WO2020259838A1

WO2020259838A1 - Techniques for user equipment beamforming

Info

Publication number: WO2020259838A1
Application number: PCT/EP2019/067135
Authority: WO
Inventors: Andre Janssen; Yeong-Sun Hwang; Holger Neuhaus; Bertram Gunzelmann; Bernhard Raaf; Qing Xu; Zhibin Yu
Original assignee: Apple Inc.
Priority date: 2019-06-27
Filing date: 2019-06-27
Publication date: 2020-12-30

Abstract

The disclosure relates to methods and devices for improving TX beam forming. A method according to a first approach is for determining an RX beam pattern for a user equipment, UE. A method according to a second approach is for polarization control in TX beam forming of a user equipment, UE. A method according to a third approach is for user equipment, UE, TX beam forming based on UE beam correspondence. A method according to a fourth approach is for re-acquiring a beam pair between a user equipment, UE, and a base station based on grouping information.

Description

TECHNIQUES FOR USER EQUIPMENT BEAMFORMING

FIELD

[0001] The disclosure relates to techniques for user equipment (UE) beamforming.

BACKGROUND

[0002] Compared with 4G LTE, beam management (BM) is one of the major new features in 5G NR, especially in FR2 (frequency range 2 in mmWave band) where antenna array based analog beamforming is supported not only on gNB but also on UE side. The target of downlink (DL) beam management is to identify the matched pair between a gNB beam and a UE beam, as fast as possible and as accurate as possible, so as to have robust DL reception and uplink (UL) transmission even in high UE mobility scenarios.

[0003] The disclosure deals with the question how to implement beamforming on UE side, in particular by exploiting the advantages of new 5G NR standards.

BRIEF DESCRIPTION OF THE DRAWINGS

[0004] The accompanying drawings are included to provide a further understanding of embodiments and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments and together with the description serve to explain principles of embodiments. Other embodiments and many of the intended advantages of embodiments will be readily appreciated as they become better understood by reference to the following detailed description. [0005] Fig. 1 is a schematic diagram of a communication system 100 illustrating a user equipment (UE) 110 communicating with a base station 120 such as a gNB by using one or more beams 111a, 111b, 111c.

[0006] Fig. 2 is a schematic diagram illustrating a method 200 of conditionally triggering on-demand SIB and use it for extra UE RX beam measurements according to the disclosure.

[0007] Fig. 3 is a schematic diagram illustrating an example of UE behavior 300 of using SIB PDSCH redundancy for measuring other UE RX beam candidates according to the disclosure .

[0008] Fig. 4 is an example procedure 400 of conditionally transmitting on-demand SIB requests and use it for extra UE RX beam measurements according to the disclosure.

[0009] Fig. 5 is an example of UL interference reduction 500 by UE TX polarization control according to the disclosure.

[0010] Fig. 6 is an example of DL interference nulling 600 by UE RX polarization control according to the disclosure.

[0011] Fig. 7 is an exemplary signal flow block diagram 700 to realize the two TX beam sweeping strategies according to the disclosure.

[0012] Fig. 8 is an example UE procedure 800 for dynamic UE TX beam sweeping strategy selection for 5G NR UL interference reduction according to the disclosure. [0013] Fig. 9 is an example of superposed gNB TX polarization angle estimation 900 on UE side according to the disclosure.

[0014] Fig. 10 is a schematic diagram illustrating three different examples 1001, 1002, 1003 of different variability in polarization domain and spatial domain according to the disclosure .

[0015] Fig. 11 is a schematic diagram of a communication system 1100 illustrating SRS based UE TX beam sweeping and SRS beam indication used for improving UL performance for partial beam correspondence case.

[0016] Fig. 12 is an example procedure 1200 of UE online beam correspondence accuracy learning based on TX beam sweeping and SRS beam indication statistics according to the disclosure .

[0017] Fig. 13 is an example procedure 1300 of SRS overhead reduction by adaptively modifying the supported number of BM SRS resources through UE capability indications according to the disclosure.

[0018] Fig. 14 is an example state diagram 1400 illustrating transition between enabling and disabling for online beam calibration & learning according to the disclosure.

[0019] Fig. 15 is a schematic diagram of a communication system 1500 illustrating SSB TX beam neighborhood relationship detection by using a wide UE RX beam reception of different SSBs according to the disclosure. [0020] Fig. 16 is a schematic diagram of an exemplary initial RSRP-Angle profile 1600 by UE RX beam sweeping for different two SSB indices according to the disclosure.

[0021] Fig. 17 is a schematic diagram of an exemplary interpolated RSRP-Angle profile 1700, peak RSRP selection and relative AoA estimation according to the disclosure.

[0022] Fig. 18 is a schematic diagram of a communication system 1800 illustrating an example of SSB grouping based UE RX beam acquisition according to the disclosure.

[0023] Fig. 19 is a schematic diagram illustrating an exemplary method 1900 for determining an RX beam for a UE according to the disclosure.

[0024] Fig. 20 is a schematic diagram illustrating an exemplary method 2000 for polarization control in TX beamforming of a UE according to the disclosure.

[0025] Fig. 21 is a schematic diagram illustrating an exemplary method 2100 for UE beamforming based on UE beam correspondence according to the disclosure.

[0026] Fig. 22 is a schematic diagram illustrating an exemplary method 2200 for re-acquiring a beam pair between a UE and a radio cell based on grouping information according to the disclosure.

[0027] Fig. 23 is a block diagram illustrating a UE circuitry 2300 according to the disclosure. DETAILED DESCRIPTION

[0028] In the following detailed description, reference is made to the accompanying drawings, which form a part thereof, and in which is shown by way of illustration specific aspects in which the invention may be practiced. It is understood that other aspects may be utilized and structural or logical changes may be made without departing from the scope of the present invention. The following detailed description, therefore, is not to be taken in a limiting sense, and the scope of the present invention is defined by the appended claims .

[0029] The following terms, abbreviations and notations will be used herein:

5G NR: 3GPP fifth generation new radio specifications QoS : Quality of Service

UE: User Equipment

gNB : gNodeB, base station in 5G

LTE : Long Term Evolution

RF: Radio Frequency

BM: Beamforming management

UL: uplink

DL: downlink

RX: receive

TX: transmit

UL-SCH uplink shared channel

DL-SCH downlink shared channel

SCH: shared channel

PDSCH : physical downlink shared channel

PDCCH : physical downlink control channel

PUSCH: physical uplink shared channel

FR2 : frequency range 2 according to 5G NR HARQ : Hybrid Automatic Repeat Request

ACK: Acknowledge

NACK: Non-Acknowledge

CSI-RS channel state information reference signal

S SB : synchronization signal block

SRS : sounding reference signals

DMRS : demodulate reference signals

SIB: system information block

BLER: block error rate

SRS : sounding reference signal

MAC : media access control (OSI-Layer)

RRC : radio resource control

DRX : discontinuous reception

CDRX : connected mode DRX

RSRP : reference signal received power

RSSI : received signal strength indication, can be the measurement metric for extra RX beam measurement using SIB PDSCH data symbols

AoA : angle of arrival

PL: path loss

TA : timing advance

EIRP: equivalent isotropically radiated power

[0030] It is understood that comments made in connection with a described method may also hold true for a corresponding device configured to perform the method and vice versa. For example, if a specific method step is described, a corresponding device may include a unit to perform the described method step, even if such a unit is not explicitly described or illustrated in the figures. Further, it is understood that the features of the various exemplary aspects described herein may be combined with each other, unless specifically noted otherwise.

[0031] The techniques described herein may be implemented in wireless communication networks, in particular communication networks based on mobile communication standards such as 5G new radio (NR), in particular for millimeter-wave data rate. The techniques may also be applied in LTE networks, in particular LTE-A and/or OFDM and successor standards. The methods are also applicable for high speed communication standards from the 802.11 family according to the WiFi alliance, e.g. 802. Had and successor standards. The methods and devices described below may be implemented in electronic devices such as cellular handsets and mobile or wireless devices or User Equipment communicating with radio cells such as access points, base stations, gNodeBs and/or eNodeBs . The described devices may include integrated circuits (ICs) and/or passives and may be manufactured according to various technologies. For example, the circuits may be designed as logic integrated circuits, ASICs, analog integrated circuits, mixed signal integrated circuits, optical circuits, memory circuits and/or integrated passives.

[0032] Fig. 1 is a schematic diagram of a communication system 100 illustrating a user equipment (UE) 110 communicating with a base station 120 such as a gNB by using one or more beams 111a, 111b, 111c.

[0033] In the following, four approaches are presented for improving the beamformed communication between UE 110 and base station, gNB 120 as shown in Fig. 1. The first approach is related to improving determining an RX beam for a user equipment. The second approach is related to improve beamformed communication by polarization control in TX beamforming of the UE . The third approach is related to improve UE TX beamforming by applying on-line UE beam correspondence accuracy learning. The fourth approach is related to re-acquiring a beam pair between a UE and a radio cell based on grouping information.

[0034] The first approach, i.e. improving determining an RX beam for a user equipment, as described in the following, is further described below with respect to Figures 2, 3, 4 and 19.

[0035] Compared with 4G LTE, beam management (BM) is one major new features in 5G NR, especially in FR2 (frequency range 2 in mmWave band) where antenna array based analog beamforming is supported not only on gNB but also on UE side. The target of downlink (DL) beam management is to identify the matched beam pair between gNB TX and UE RX as fast as possible and as accurate as possible, so as to have robust DL reception even in high UE mobility scenarios. Within 3GPP defined BM framework, UE RX beam sweeping is one key sub functionality to ensure robust DL reception in mmWave bands.

[0036] For example, in RRC_CONNTECTED mode, due to fast UE rotation, an early selected gNB TX beam can still be globally optimal for the same UE, but the previously acquired optimal UE RX beam can be out-of-date with respect to the same gNB TX beam. In this scenario, UE needs to test (by DL measurements) as many UE RX beam candidates as possible and as fast as possible, so that UE can quickly find a new best UE RX beam which is optimal for the most recent rotated angle of arrival (AoA) and to always ensure robust DL PDSCH/PDCCH reception. [0037] However, based on current 3GPP NR defined BM framework in release 15, the capability of RX beam candidate testing on UE side is restricted by the DL measurement opportunities which are determined by the time allocation of DL reference signals (e.g. CSI-RS or SSB) . The DL reference signals allocation is totally driven by the gNB . It means, the allocation of those DL reference signals used for UE RX beam sweeping can NOT be triggered or even proposed by the UE . Furthermore, there is no possibility for UE to indicate the gNB about UE rotation status by 3GPP NR release 15 procedures. As a result, gNB cannot accordingly adapt the time density of DL RS allocations to improve UE RX beam sweeping capability when UE is in fast rotation scenarios!

[0038] The first approach as described in this disclosure provides a solution to the following problem, in particular related to 5G mmWave band communications: When the UE is in fast rotation scenarios (e.g. gaming), how can UE test more UE RX beam candidates as fast as possible, i.e., how can UE opportunistically increase the RX beam measurement capability based on the current 3GPP defined BM framework.

[0039] System Information Block (SIB) in NR contains cell specific information which can be repeatedly broadcasted by the gNB and carried by DL PDCCH and PDSCH. In NR, two types of SIB are defined: on-demand SIB (newly introduced in NR than LTE, with massive payloads) and periodic SIB (this is similar like SIB in LTE but can be a longer periodicity) . In order to improve the spectrum efficiency, on-demand SIB be only be allocated upon UE request. Since a SIB contains repeated cell specific information (static) , once UE has successfully decoded a SIB repetition, the other SIB repetitions become redundant to a same UE . Hereby, UE could then explore such redundancy and use the received DL signals from the SIB repetitions as the extra UE RX beam measurement opportunities to test other candidate UE RX beams. This results in improved UE RX beam sweeping capability.

[0040] In particular, for on-demand SIB which can be conditionally allocated by a gNB upon the request from a UE, UE could adapt the on-demand SIB request based on the detection of UE rotation event, so that such extra measurement opportunity by on-demand SIBs can only be activated based on UE needs. This results in optimized trade off between the radio spectrum usage (used for allocating the massive payload of on-demand SIBs) and UE mobility performance (increased measurement opportunities by making use of the redundant resources, e.g. PDSCH DMRS or even PDSCH data symbols, from the allocated on-demand SIBs) .

[0041] As one extension, since an on-demand SIB could be shared by multiple UEs, when UE has detected the fast rotation event (e.g. by DL measurements based on the current activated UE RX beam or based on motion sensors) , instead of immediately transmitting on-demand SIB request (by sending an PRACH preamble) during a UL slot, UE can instead open the RX within the same UL slot and monitor the availability of on- demand SIB requests sent by other UEs. If PRACH preamble signals, which is associated to the same on-demand SIB has already been sent by other UEs, the UE could bypass sending the same PRACH signals, but instead directly wait for the allocation of such on-demand SIB and then use it for extra UE RX beam measurements. This results in reduced UL interferences and also reduces the UE power consumption (TX power consumption is usually higher than RX) . [0042] As another extension, on top of on-demand SIB, periodic SIBs can also be explored to increase the UE RX beam sweeping capability. That can be done by periodically testing other UE RX beam candidates in a round-robin manner for SIB repetitions. The RSRP or RSSI measurements based on PDSCH DMRS or PDSCH data symbols from the periodic SIBs can then be used for RX beam candidate beam selection.

[0043] Thus, the first approach as described in this disclosure provides the technical advantage of improved UE RX beam sweeping capability and robust UE DL reception

especially in UE fast rotation scenarios, particularly for 5G NR mmWave bands .

[0044] The second approach, i.e. improving beamformed communication by polarization control in TX beamforming of the UE as described in the following, is further described below with respect to Figures 5 to 10 and 20.

[0045] Antenna array based analog beamforming has been introduced for UE device implementation for 5G NR mmWave band communications. In 5G cellular networks, when UE is located at the beam-edge of different gNBs, the UE beam is usually not narrow enough to isolate among different neighbour (adjacent) gNB beams. That is because, for a mobile device, the size of the antenna array circuitry is very limited, which results in limited number of antenna elements that can be placed within the antenna array circuity. Typical array size for a phone device for 5G NR mmWave bands is usually 1X2 or 1X4.

[0046] As a result, DL/UL interferences can exist when UE is located at the beam-edge of different gNBs . For UL, the gNB which receives the UL signals from its connected UEs can also suffer the interferences of UL signals from other co-located UEs which is connected to a different neighbour gNB . For DL, the UE which receives the DL signals from the serving gNB can also suffer the interferences of DL signals from a different co-located neighbour gNB. When the neighbour gNBs are not coordinated (e.g. at network edges), the introduced UL/DL interference cannot be well controlled.

[0047] The second approach as described in this disclosure provides a solution to the following problem, in particular related to 5G mmWave band communications: How to reduce the UL/DL interferences when UE is located at beam-edge of multiple gNBs.

[0048] In mmWave band communications, the transmitter diversity and the receiver diversity are usually achieved by dual-polarizations within the same antenna panel. For single layer transmission mode, by aligning superposed RX polarization and the superposed TX polarization which transmits the useful signals, or by orthogonalizing the RX polarization and the superposed TX polarization which transmits interference signals, the UL/DL interferences from adjacent spatial direction (adjacent beams) can be significantly reduced or even nulled.

[0049] In particular, by following 5G NR procedures, the UL interferences can be reduced by aligning the superposed UE TX polarization with the serving gNB RX polarization. This can be achieved by adding UE intelligence into the SRS based TX beam sweeping mechanism supported by 5G NR UL BM framework. In this proposal, UE can adaptively select from two different TX beam sweeping strategies, referred to as beam sweeping strategy 1 and Beam sweeping strategy 2 as described in the following.

[0050] Beam Sweeping Strategy 1: Sweep the TX beam candidates with the same superposed TX polarization (same common phase delta among two TX polarizations) but with different spatial directions (different relative phase shifter settings of the antenna elements within the antenna array) .

[0051] Beam Sweeping Strategy 2: Sweep the TX beam candidates with the different superposed TX polarizations (different common phase deltas among two TX polarizations) but with the same spatial direction (the same relative phase shifter settings of the antenna elements within the antenna array) .

[0052] In particular, by following 5G NR procedures, the DL interferences can also be reduced by controlling the superposed UE RX polarization based on the detected serving gNB TX polarization angle as well as the interference gNB TX polarization angle. Hereby the polarization detection can be based on SSB based RSRP measurements on the two orthogonal RX polarizations in UE side.

[0053] Thus, the second approach as described in this disclosure provides the technical advantage of reduced UL/DL interferences without significantly increasing UE or gNB complexity .

[0054] The third approach, i.e. improving UE TX beamforming by using UE beam correspondence as described in the following, is further described below with respect to Figures 11 to 14 and 21.

[0055] UE Beam correspondence (i.e. UE TX beam pattern is spatially overlapping with UE RX beam pattern) was the initial 5G NR assumption for mmWave bands, which aims at exploring the high channel reciprocity (due to highly beam- formed channels) in mmWave bands, so that the best UL TX beam can be directly derived if the best DL RX beam has already been acquired by the UE . UE beam correspondence can avoid the extra radio resources allocation in network side for UL beam or UE TX beam training and thus it is efficient for radio resource allocation. However, in reality, due to the non idealism of TX/RX circuitries (e.g. different implementations of phase shifters for TX and RX) and phone structures, it is very challenging for UE vendors to have a perfect UE RF design which can support full beam correspondence in 5G mmWave bands, for each of the manufactured phone device.

[0056] As a result, the TX beam pattern that UE is using for transmitting the UL channels may not be totally overlapping with the best RX beam pattern which is acquired from DL beam acquisition procedures. As a further result, the EIRP (equivalent isotropically radiated power) that the UE is delivering from its TX antenna panel, is NOT 100% focusing on the desired direction, which leads to sub-optimal UL performance. To solve this issue, it has been finally agreed in 3GPP RAN plenary meeting in December 2018 that, in FR2 (frequency range 2 for mmWave bands) , UE can choose to support either full beam correspondence capability or partial beam correspondence capability, through UE capability indication to the network. When UE indicates the capability of supporting partial beam correspondence, it means UE can meet the beam correspondence accuracy requirements with assistance of UE TX beam sweeping as described below in the example of Figure 11.

[0057] According to 3GPP 5G NR specification, UE TX beam sweeping is based on transmitting of bursts of BM SRS resources. The maximal number of BM SRS resources that can be allocated for UE to apply TX beam sweeping is determined through UE capability indication. Each BM SRS resource consists of at least one SRS symbol. All SRS symbols within a BM SRS resource need to be transmitted by using a same TX beam. The target of SRS TX beam sweeping is to refine the UE TX beam pattern by sweeping multiple UE TX beam candidates, which are spatially adjacent to current activated UE RX beam. gNB can then measure the received SRS resources and indicates the UE the best SRS resource ID. Based on the gNB indicated (selected) SRS resource ID, UE can then figure out the best TX beam candidate that have been swept and can apply such TX beam candidate for UL transmissions (PUSCH/PUCCH) . By such procedure, the TX/RX beam pattern mismatch (beam correspondence accuracy) can be compensated back by SRS based UE TX beam sweeping and SRS beam indication. One example of SRS based UE TX beam sweeping and SRS beam indication is shown in Figure 11 described below.

[0058] The third approach as described in this disclosure provides a solution to the following problem, in particular related to 5G mmWave band communications: SRS TX sweeping can help UE to gain back the beam correspondence accuracy but introduces high SRS overhead, especially when the number of adjacent TX beam candidates with respect to a reference RX beam is high. The high SRS overhead further introduces power consumption overhead in UE side and reduced radio resource allocation efficiency in the network side.

[0059] According to the third approach, the SRS TX beam sweeping process during field operations can be viewed as "on-line RF calibration". Based on that, UE can iteratively apply on-line learning of the beam correspondence accuracy and then accordingly reduce the SRS overhead used for TX beam sweeping. Hereby, the on-line beam correspondence accuracy learning can be achieved by monitoring the SRS beam indications from gNB side, which is associated to the previous TX beam sweeping burst; Furthermore, UE can estimate a quality metric of the swept TX beams, and then accordingly narrow down the size of the TX beam candidate set for each reference RX beam. Based on such on-line learning iterations after a time period of field deployment and field operations, UE can accordingly reduce the supported number of BM SRS resources, which is defined by 5G NR as one UE capability. This results in SRS overhead reduction.

[0060] Thus, the third approach as described in this disclosure provides the technical advantage of achieving the optimized trade-off between UE UL performance and SRS

overhead, without further introducing UE cost.

[0061] The fourth approach, i.e. re-acquiring a beam pair between a UE and a base station based on grouping information as described in the following, is further described below with respect to Figures 15 to 18 and 22.

[0062] Beamforming techniques can significantly improve the network resource efficiency in 5G NR. In 5G NR FR1 (sub6GHz bands) , antenna array based analogue beamforming can be deployed in gNB side. In 5G NR FR2 (mmWave bands), antenna based beamforming is deployed in both gNB and UE side. For initial access, in order to find the best gNB TX beam (in FR1) or the best beam pair between gNB TX beam and UE RX beam (in FR2), beam acquisition needs to be done in UE side. In 5G NR, beam acquisition is based on sweeping the measurements on all Synchronization Signal Blocks (SSBs) in UE side. According to 3GPP definition, a different SSB index can be associated to one different gNB TX beam. Further, a number of SSBs with different indexes form a SSB burst. The number of SSBs within a SSB burst can be up to 64 (meaning up to 64 gNB TX beams) . The SSB burst is repeated with a pre-defined periodicity (e.g. 20ms).

[0063] Furthermore, in DRX (Dis-continuous Reception) operations (for both IDLE mode DRX and Connected mode DRX) , after UE is waken up from sleep for a long period, UE side beam re-acquisition is also needed to re-identify the best beam (pair) between a new best SSB TX beam from the gNB and a new best UE RX beam. That is because, due to UE mobility (e.g. UE moving or UE rotation), the previously identified best beam pair (in the previous DRX on-duration) can be out- of-date. The UE side beam re-acquisition is also based on SSB measurements .

[0064] UE side beam (re-) acquisition in FR1 is 1-D search (identify the best gNB SSB TX beam only) , while in FR2 it is 2-D search (identify the best gNB SSB TX beam and UE RX beam combination) . Thus, the time overhead for beam (re- ) acquisition in FR2 (mmWave bands) can be much longer than that in FR1, especially when UE has a lot of RX beam candidates. In particular, when UE is re-entering from DRX off duration into DRX on duration, the long latency due to beam re-acquisition can also have significant negative impacts for UE power consumption. That is because, to secure the 2-D beam search for beam re-acquisition in FR2, UE needs to be waken-up much earlier.

[0065] The fourth approach as described in this disclosure provides a solution to the following problem, in particular related to 5G mmWave band communications, in particular in FR2 : How can UE speed up beam re-acquisition e.g. after DRX waken up.

[0066] SSBs are cell-specific, not UE-specific. Therefore, the mapping between a gNB TX beam and a SSB index is fixed (aiming for the maximal gNB coverage) . It also means, the SSB beam neighborhood relationship information (for example SSB1 beam is spatially co-located with to SSB4 beam, etc.) are also fixed for a same gNB. UE could pre-detect SSB beam neighborhood relationships (e.g. by historical SSB measurements when UE was in non-DRX operations) and then use this information to speed up the UE RX beam sweeping in the later-on beam re-acquisition scenarios (e.g. DRX or CDRX waken-up) . It is very important to note that: comparing with existing known approach which uses pre-detected historical SSB strength information, the proposed SSB beam neighborhood relationship information is irrelevant to UE mobility and is constant for a same gNB.

[0067] According to the fourth approach, since SSB TX beam neighborhood relationship information is cell specific and robust against UE movements, the detection can be opportunistically activated when UE was in non-DRX scenarios. In particular, two sub-methods how to detect such information in UE side are introduced as described in the following. [0068] According to a first sub-method, UE can use a same wider RX beam to receive different SSBs, and then compare the channel (parameter) estimation from different SSBs. For SSBs whose TX beams are spatially correlated, the channel parameters are also correlated (e.g. similar RSRP measurements, similar delay spread measurements, or even correlated channel transfer functions) , those information could be jointly explored to derive a spatial correlation metric between SSBs.

[0069] According to a second sub-method, SSB TX beam neighborhood relationship information can be detected by comparing the estimated relative Angle of Arrivals (relative AoAs) from different SSB beams. As for one example, the relative AoA among different SSB beams can be estimated by first scanning narrow UE RX beams with a-prior UE RX beam angle information (pre-known by UE RF characterizations) so as to derive a sub-sampled RSRP-Angle profile for different SSBs. And then, in the offline processing, we can optionally interpolate such sub-sampled RSRP-Angle profile (interpolation is only needed if the swept UE RX beam has lower resolution with higher angle stepping) , and identify the best receiver angles (angle mapped to the peak RSRPs) for different SSB indexes. And finally we can compute the angle difference which is associated to the peak RSRPs of different SSB indexes within the derived RSRP-Angle profile. Such angle difference is the relative AOAs among different SSBs.

[0070] After having detected SSB beam neighborhood relationship information, UE could later-on use this information to speed up the beam re-acquisition when waking up from CDRX/DRX. For example, instead of sweeping all RX beam candidates per SSB index in a round-robin manner (2-D search) , UE could group the SSBs which are pre-detected to be spatially correlated (SSBs with co-located gNB TX beams) and apply the wide UE RX beam sweeping within a co-located SSB group. Afterwards, UE can down-select the best SSB group, and further apply narrow RX beam sweeping within the down- selected best SSB group. The procedure can terminate until a good reception quality on a SSB is reached. Note that, the innovation is NOT about hierarchical RX beam acquisition, but rather to make use of pre-detected SSB gNB beam neighborhood relationship information to further speed up the hierarchical RX beam re-acquisition. Furthermore, how UE makes use of such pre-detected SSB TX beam neighborhood relationship information to further speed up RX beam re-acquisition is not restricted to the mentioned example.

[0071] Thus, the fourth approach as described in this disclosure provides the technical advantage of reduced UE early waken-up time overhead and reduced latency for RX beam re-acquisition, in particular in DRX or C-DRX operations.

[0072] Fig. 2 is a schematic diagram illustrating a method 200 of conditionally triggering on-demand SIB and use it for extra UE RX beam measurements according to the disclosure.

[0073] In New Radio (NR) Specification, the System Information Block (SIB) contains cell specific information which can be repeatedly transmitted by the gNB 120 and is carried by DL PDCCH and PDSCH. In NR, two types of SIB are defined: on-demand SIB (newly introduced in NR with massive payloads) and periodic SIB (this is similar like SIB in LTE but can have a longer periodicity) . In order to improve the spectrum efficiency, on-demand SIB can only be allocated upon UE request. Since a SIB contains repeated cell specific static information, once UE has successfully decoded a SIB repetition, the other SIB repetitions become redundant to a same UE . Hereby, UE can then explore such redundancy and use the received DL signals from the SIB repetitions as the extra UE RX beam measurement opportunities to test other candidate UE RX beams. This results in improved UE RX beam sweeping capability .

[0074] Hereby, on-demand SIB can be conditionally allocated by a gNB upon the request from a UE . For on-demand SIB UE can adapt the on-demand SIB request based on the detection of a UE rotation event, so that such extra measurement opportunity using on-demand SIBs can only be activated based on UE needs. This results in optimized trade-off between the radio spectrum usage (used for allocating the massive payload carried by on-demand SIBs) and UE RX beam sweeping performance. This further results in increased measurement opportunities by making use of the redundant resources, e.g. PDSCH DMRS symbols or even PDSCH data symbols, from the allocated on-demand SIBs. Note that a UE fast rotation event can be detected by UE DL measurements based on current activated RX beam or based on sensing information from motion sensors within the UE device. The key idea of conditionally triggering on-demand SIB and using it for extra UE RX beam measurements is shown in Figure 2.

[0075] In a first action 201, UE 110 detects a fast rotation event, e.g. by detecting a change of UE antenna radiation pattern. In a second action 202, UE requests gNB 120 to allocate on-demand SIB. When gNB 120 receives this request, in a third action 203 gNB 120 allocates on-demand SIB (including DMRS and PDSCH data symbols) . Here, UE 110 can for example just monitor the PDCCH on the expected time/frequency location of a potential on-demand SIB occasion. In a fourth action 204, UE makes use of the allocated on-demand SIB (e.g. DMRS and PDSCH data symbols) to measure extra RX beam candidates .

[0076] After getting the allocation of an on-demand SIB, UE 110 first needs to decode the PDCCH associated to the SIB. The SIB PDCCH decoding is preferred to use a stable UE RX beam, e.g. the current activated and the best UE RX beam. After having decoded SIB PDCCH, UE knows the time and frequency allocation of the SIB PDSCH region. Hereby, since UE is NOT interested to decode the SIB PDSCH (redundant information) again, but rather to use it to measure other UE RX beam candidates, UE can sweep multiple UE RX beam candidates during the SIB PDSCH region and measure those UE RX beam candidates. The above described UE behavior is shown in Figure 3.

[0077] Fig. 3 is a schematic diagram illustrating an example of UE behavior 300 using SIB PDSCH redundancy for measuring other UE RX beam candidates according to the disclosure.

[0078] Not only the PDSCH DMRS symbols within a SIB can be used as the extra UE RX beam measurement opportunity, the PDSCH data sub-carriers (SCs) within the SIB can also be explored. This can be done, for example in a simpler approach by RSSI measuring, which reflects the total RX beamformed energy on PDSCH data symbols. In another approach, this can be done by re-encoding and re-modulation based on already decoded SIB information bits, which were decoded from historical SIB repetitions, and then by reproducing the channel transfer function (CTF) on the received PDSCH data subcarriers (SCs) within the SIB repetition and measuring the reference signal received power (RSRP) based on the reproduced CTFs . The measured data SC RSRP can then be used for selecting the best UE RX beam candidates.

[0079] The procedure is shown in Fig. 3: When SIB PDCCH OFDMs 310 are received, SIB PDCCH is decoded 311 using the current best activated UE RX beam 301. Thereafter, when SIB PDSCH DMRS and data symbols 320 are received, UE RX beam candidates

302, 303, 304 are measured 321 during the SIB PDSCH region, e.g. using the PDSCH DMRS or even the PDSCH data symbols. In Fig. 3, an exemplary number of three candidate RX beams 302,

303, 304 are measured based on the received PDSCH DMRS or PDSCH data symbols.

[0080] As one extension, UE can monitor the on-demand SIB requests sent by other UEs. Since an on-demand SIB can be shared by multiple UEs, when UE has detected a fast rotation event, instead of immediately transmitting the on-demand SIB request (by sending an PRACH preamble) with pre-configured UL slot, UE can instead open the RX within the same UL slot and monitor the on-demand SIB requests sent by other UEs. The fast rotation event may for example be detected by DL measurements of the current activated UE RX beam or based on sensors .

[0081] If the PRACH signals which are associated to the same on-demand SIB have already been sent by other UEs, the UE can bypass sending the same on-demand SIB request, but instead directly wait for the allocation of such on-demand SIB from the base station and then use it for extra UE RX beam measurements. This results in reduced UL interferences and also reduces the UE power consumption. Note that the TX power consumption is usually higher than the power consumption of RX. Figure 4 shows an example procedure for this extension.

[0082] Fig. 4 is an example procedure 400 of conditionally transmitting on-demand SIB requests and use it for extra UE RX beam measurements according to the disclosure.

[0083] The procedure 400 comprises a first block 401 including detecting a UE fast rotation event, e.g. based on a change of UE antenna radiation pattern. Then, a second block 402 includes identifying the candidate UL slots (based on PRACH resource configuration) in which the PRACH signal which is associated to the on-demand SIB request is allowed to be transmitted. A third block 403 includes randomly selecting one PRACH candidate UL slot and receiving the IQ samples during the selected UL slot. A fourth block 404 includes detecting the same PRACH signal which can be transmitted by the other UEs, e.g. by correlations, based on the received IQ samples within the selected PRACH candidate UL slot. Then, a fifth block 405 includes a condition to proceed: If the same PRACH signal is detected a seventh block 407 is performed which includes waiting for on-demand SIB to be allocated by the gNB; and setting up a waiting timer. Otherwise, a sixth block 406 is performed which includes selecting a second PRACH candidate UL slot and transmitting the PRACH signal in the selected slot to request on-demand SIB. After the sixth block 406, the seventh block 407 is performed. After the seventh block 407 an eight block 408 includes a condition to proceed: If on-demand SIB is allocated before time-out, a ninth block 409 is performed which includes measuring other UE RX beam candidates based on the allocated on-demand SIB signals. Otherwise, the sixth block 406 is performed again. [0084] As another further extension, besides of on-demand SIB, other periodic SIBs can also be explored to increase the UE RX beam sweeping capability, in the background. Hereby, after having successfully decoded the information bits from a periodic SIB repetition based on the current activated UE RX beam, UE can then apply other UE RX beam candidates (e.g. in a periodically round-robin manner) to receive the SIB repetitions and measure the reception quality (e.g. by DMRS RSRP measurements) .

[0085] As another further extension, besides of SIB redundancy, UE can also make use of DL PDSCH retransmission redundancies from normal DL traffic receptions to further increase the RX beam measurement opportunities. This can be done by enforcing the DL HARQ feedback bits associated to a received DL-SCH channel to be false ( NACK"), even though the PDSCH which carries the DL-SCH has already been successfully decoded (CRC pass) . The enforced DL NACK bits are transmitted by the UE in normal PUCCH or PUSCH. After the gNB has received enforced NACK bits, it triggers the PDSCH retransmissions which now become redundant information to the same UE . UE can then make use of the DMRS or even data symbols within the PDSCH retransmissions to test other UE RX beam candidates to speed up the UE RX beam sweeping.

[0086] Fig. 5 is an example of UL interference reduction 500 by UE TX polarization control according to the disclosure.

[0087] In mmWave band communications, the transmit diversity and receiver diversity are usually achieved by dual polarizations within the same antenna panel. In 5G mmWave communication scenarios, in order to reduce the UL/DL interferences when UE is located at the beam-edges of neighboring gNBs, the second approach as described above with respect to Fig. 1 presents a solution to adaptively control the angle difference between the superposed TX polarization (see 512, 513, 514 in Fig. 5) and the superposed RX polarization (see 511 in Fig. 5) to reduce the UL/DL interferences. The presented second approach can be applied, for example, for single layer transmission mode (e.g. PUCCH, PDCCH, single layer PUSCH/PDSCH) for both UL and DL in 5G NR. In this case, by aligning the angle between superposed RX polarization 511 and the superposed TX polarization 512, 513, 514 which transmits the useful signals, or by orthogonalizing the angle between the superposed RX polarization 521 and the superposed TX polarization 522, 523 which transmits interference signals, UL/DL interferences from adjacent spatial direction (adjacent beams) can be significantly reduced. Figure 5 shows an example for UL interference reduction while Figure 6 gives another example for DL interference reduction.

[0088] For the UL interference reduction in Fig. 5, a polarization 511 of gNB RX beam of gNBl (a first base station, e.g. serving base station) is determined. Polarizations 512, 513, 514 of UE TX beams (for intra-gNB UEs, time-division multiplexed) may be provided to be included within an angle range 510 of the polarization 511 of gNB RX beam of gNBl. Similarly, a polarization 521 of gNB RX beam of gNB2 (a second base station, e.g. interfering base station) is determined. Polarizations 522, 523 of UE TX beams (for intra-gNB UEs, time-division multiplexed) may be provided to be included within an angle range 520 of the polarization 521 of gNB RX beam of gNB2. [0089] By following 5G NR procedures, the UL interferences can be reduced by aligning the superposed UE TX polarization 512, 513, 514 with the serving gNB RX polarization 511. This can be achieved by adding UE intelligence into the existing SRS based TX beam sweeping mechanism supported by 5G NR. UE can adaptively choose between two TX beam sweeping strategies as described in the following.

[0090] Strategy 1 (TX beam pattern sweeping) : Sweep the TX beam candidates with the constant superposed TX polarization angle (same common phase delta among two TX polarizations) but with different spatial directions (different relative phase shifter settings for the antenna elements within the antenna array) . Such strategy aims at finding a TX beam pattern which maximizes the TX power delivery to optimal spatial direction of the serving gNB receiver.

[0091] Strategy 2 (TX polarization angle sweeping) : Sweep the TX beam candidates with the different superposed TX polarization angles (different common phase deltas among two TX polarizations) but with the same spatial direction (constant relative phase shifter settings for the antenna elements within the antenna array) . Such strategy aims at finding the optimal superposed TX polarization angle which minimizes the UL interference at the gNB receiver side. The signal flow block diagram which realizes of the two TX beam sweeping strategies is shown in Figure 7.

[0092] Fig. 6 is an example of DL interference nulling 600 by UE RX polarization control according to the disclosure.

[0093] For the DL interference reduction or nulling in Fig. 6, an old superposed polarization 602 of UE RX beam of gNB (a serving base station) can be rotated by UE polarization rotation 604 around the superposed polarization 603 of the serving gNB TX beam to a new superposed polarization 605 of UE RX beam. The new superposed polarization 605 of UE RX beam can have an orthogonal angle with respect to a superposed polarization 601 of an interference gNB TX beam. By this orthogonal angle, interference nulling or at least reduction can be achieved.

[0094] Fig. 7 is an exemplary signal flow block diagram 700 to realize the two TX beam sweeping strategies according to the disclosure.

[0095] Time domain SRS (sounding reference signal) IQ samples

701 are provided via two TX polarization paths 717, 727, i.e. via a vertical polarization path 717 and a horizontal polarization path 727.

[0096] In the vertical polarization path 717, the SRS IQ samples 701 are phase-shifted by TX polarization angle setting 702, passed through a digital-to-analog converter (DAC) 713 to an analog signal S (t) 714. The analog signal S (t) 714 passes a power splitter 715 to split the signal S (t) 714 into a plurality of analog sub-signals which are passed through a TX beam pattern setting module 706 to set a respective beam pattern of the analog sub-signals by a respective phase shifting. At the outputs of the TX beam pattern setting module 706, the different vertical components 707 of the analog TX beam patterns are provided.

[0097] In the horizontal polarization path 727, the SRS IQ samples 701 are passed without TX polarization angle setting

702 through a second digital-to-analog converter (DAC) 723 to a second analog signal S (t) 724. The second analog signal S (t) 724 passes a second power splitter 725 to split the signal S (t) 724 into a plurality of analog sub-signals which are passed through the TX beam pattern setting module 706 to set a respective beam pattern of the analog sub-signals by a respective phase shifting. At the outputs of the TX beam pattern setting module 706, the different horizontal components 707 of the analog TX beam patterns are provided.

[0098] In Figure 7, within a same antenna array, for each TX polarization 717, 727 there are N (typical values for phone device are N=2 or N=4) antenna elements. For both sweeping strategies described above, within a SRS sweeping burst, for a single SRS resource (index p) the two TX polarizations 717, 727 are applied with the same beam pattern for transmitting such SRS resource (controlled by relative phase shifting vectorO_p , 706) . On top, a common phase shifting 702 is applied on one TX polarization stream (controlled by common phase shifting scalar f ), which controls the superposed TX polarization angle in UE side. For strategy 1 (TX beam pattern sweeping), (P_pis constant for all transmitted SRS resources while q_r is varying for different SRS resources within the sweeping burst. For strategy 2 (TX polarization angle sweeping), (P_pis varying for different SRS resources while q_r is constant for all transmitted SRS resources within the sweeping burst.

[0099] Further note that, UE implementation wise, the common phase shifting f_r , which is used for controlling the superposed TX polarization angle on UE side, can be implemented either in digital domain (e.g. in baseband processor) or in analog domain (e.g. as additional common phase which is offsetting the phase shifting vector q_r within the antenna array) .

[0100] The two presented SRS TX beam sweeping strategies can be dynamically and adaptively selected by the UE : As one example, for a dynamic SRS sweeping event which is dynamically triggered by the gNB (e.g. an aperiodic SRS burst within one UL slot, which is triggered by DCI carried by DL PDCCH) , UE can dynamically decide which sweeping strategy to use for each single event. In another example, for pre defined SRS sweeping burst pattern (e.g. periodic SRS or semi-persistent SRS pattern which is pre-configured by gNB to UE) , UE can dynamically adjust the occupation ratio between the two sweeping strategies. For both examples, the selection (or occupation ratio) among sweeping strategy 1 and sweeping strategy 2 on UE side can be based on three factors: A first factor is communication range between the UE and the serving gNB. This can be reflected by path-loss measurement or timing-advance measurement. A second factor is UE mobility status. This can be reflected by UE mobility measurement or speed sensors within the UE . A third factor is the detection of gNB polarization control behaviors, i.e. whether gNB is also dynamically adapting its RX polarization for receiving UL signals or not. These three factors can impact the sweeping strategy selection on UE side as described in the following .

[0101] First, the sweeping strategy 2 (TX polarization angle sweeping) may be prioritized over sweeping strategy 1 (TX beam pattern sweeping) , if the communication range between the UE and the serving gNB is shorter. That is because for short range communications, the TX power delivery from UE to gNB can be easily guaranteed even if the TX beam pattern is not optimal. The communication range can be determined based on the path-loss (PL) measurement (e.g. using DL SSB or CSI- RS) on UE side. The communication range can also be determined based on timing advance (TA) measurement (e.g. using UL PRACH) on gNB side which is later-on indicated to UE through TA commands .

[0102] Second, the sweeping strategy 1 (TX beam pattern sweeping) may be prioritized over sweeping strategy 2 (TX polarization angle sweeping) , if the UE is detected to be at high mobility. That is because angle delta between the superposed UE TX polarization and superposed gNB RX polarization is sensitive to propagation delay fluctuations which becomes significant when UE is in high mobility status. The UE mobility status detection can be based on observation of the fluctuation levels of UE DL signal parameter measurements (e.g. RSRP/CQI etc.), or based on the sensing information collected by the mobility sensors from UE device.

[0103] Third, the sweeping strategy 1 (TX beam pattern sweeping) may be prioritized over sweeping strategy 2 (TX polarization angle sweeping) , if the serving gNB has been detected to also apply dynamic RX polarization control for UL signal reception. That is because when the gNB applies dynamic RX polarization control also on the received SRS resources, it may confuse the gNB if the UE is also applying dynamic TX polarization for the same training signals (SRS) . The gNB RX polarization control behavior can be detected by two sub-methods as described in the following.

[0104] According to Sub-method 1 gNB RX polarization control behavior can be detected by measuring the superposed polarization angle of the DL RS, which is indicated by the gNB to be spatially associated SRS resources which are allocated for TX beam sweeping. In this case, gNB is supposed to use the same beam configuration for receiving the SRS resources and for transmitting the spatially associated DL RS ( SSB or CSI-RS) .

[0105] According to Sub-method 2 gNB RX polarization control behavior can be detected by observing the statistics of SRS beam indications from gNB to UE, which are the outcomes of the SRS beam sweeping events. When multiple repeated SRS sweeping bursts are all based on sweeping strategy 2 (TX polarization angle sweeping) , and when UE observes that gNB indicates back fluctuated SRS IDs with the different UE TX polarization angle settings, it means that: either the gNB does not consider interference measurement on their side for UE TX beam selection, or the gNB also applies the RX polarization sweeping during UE SRS sweeping phase, which got confused. In both cases, UE should disable sweeping strategy 2 but only activate the sweeping strategy 1.

[0106] The overall UE procedure for adaptive TX beam sweeping strategy selection used for UL interference reduction is shown in Figure 8 described below.

[0107] Fig. 8 is an example UE procedure 800 for dynamic UE TX beam sweeping strategy selection for 5G NR UL interference reduction according to the disclosure.

[0108] The procedure 800 comprises a first block 801 including: waiting for a scheduled SRS burst for UE TX beam sweeping. Then, a second block 802 includes: collecting the information which can impact the selection of TX beam sweeping strategies (e.g. PL, TA, UE mobility status, historical received SRS beam indications, etc.) . A third block 803 includes: selecting the beam sweeping strategy for the upcoming SRS burst. A fourth block 804 includes: waiting for the selected SRS ID through SRS beam indication from gNB . Then, a fifth block 805 includes a condition: Was sweeping strategy 1 selected? If no, a sixth block 806 is performed including: apply the same phase angle between two TX polarizations used by the same SRS ID during sweeping to tranmit PUSCH/PUCCH. If the condition in block 805 is yes, a seventh block 807 is performed including: apply the same TX beam pattern used by the same SRS ID during sweeping to transmit PUSCH/PUCCH.

[0109] Fig. 9 is an example of superposed gNB TX polarization angle estimation 900 on UE side according to the disclosure. The gNB TX polarization angle 930 can be estimated from DL RS (CSI-RS/SSB) measurements from two parallel UE RX polarizations 910, 920 (set to be 90 degree apart) . The first UE RX polarization 910 is measured by RSRP1 (received signal received power) 911, the second UE RX polarization 920 is measured by RSRP2 921. Superposition of RSRP1 911 and RSRP2 921 gives the angle of the superposed gNB TX polarization 930.

[0110] Similar like UL, in 5G NR mmWave bands, for single layer transmission mode, DL interference can also be reduced by polarization angle control, as shown in Figure 6 and described above. In particular, when targeting at 3GPP 5G NR UE implementation, this can be realized by the following two schemes depending on the currently activated DL beam training phases defined by 5G NR beam management framework: [0111] In gNB TX beam refinement phase (P2), UE needs to measure and report one best gNB TX beam candidates to the gNB. The gNB TX beam reporting can be jointly decided based on the RSRP measurement of a gNB TX beam candidate, but also based on the measured gNB TX polarization angle difference between a serving gNB TX beam candidate and a detected strong neighboring gNB TX beam (DL interferences) .

[0112] In UE RX beam refinement phase (P3) , UE needs to acquire an optimal UE RX beam candidate with respect to a fixed serving gNB TX beam. The UE RX beam pattern is first selected based on the maximal measured RSRP with respect to the already fixed gNB TX beam. And then, the polarization angle of the selected RX beam pattern is then rotated based on the measured polarization angle from the fixed serving gNB TX beam and the measured polarization angles from the detected strong neighboring gNB TX beams (DL interferences) . As one example, the new UE RX beam polarization angle can be rotated by maximizing the receiving signal to noise and interference ratio (SINR) for a MRC demodulator.

[0113] As another example, the new UE RX beam polarization angle can be derived by nulling the strongest DL interference (by making the UE RX polarization angle to be orthogonal with respect to the detected TX polarization angle of the strongest neighboring TX beam) .

[0114] In particular, as for one example, the gNB TX polarization angle can be estimated from DL RS (CSI-RS/SSB) measurements from two parallel UE RX polarizations (set to be 90 degree apart) . Figure 9 shows one example for such measurement . [0115] Fig. 10 is a schematic diagram illustrating three different examples 1001, 1002, 1003 of different variability in polarization domain and spatial domain according to the disclosure .

[0116] For simplicity the diagrams plot only polarization on y axis and angle of arrival/departure (as a representative spatial parameter) on the x axis. The second diagram 1002 shows only variation on angle and no variability in polarization. The first diagram 1001 shows variations from two polarizations and three angular positions, i.e., still more variability in spatial than in polarization domain. The third diagram 1003 shows similar variability as the first diagram 1001 but a different distribution: for the first polarization three lower angles are used while for the second polarization three higher angles are used.

[0117] The constellations from diagrams 1001, 1002, 1003 can be mirrored at the 45 degree line to get similar constellations prioritizing polarization over spatial to a more or less degree.

[0118] Fig. 11 is a schematic diagram of a communication system 1100 illustrating SRS based UE TX beam sweeping and SRS beam indication used for improving UL performance for partial beam correspondence case.

Fig. 11 is an example of the existing mechanism (also working assumption) for TX beam refinement by 3GPP. The third approach according to this disclosure is built on top of this mechanism by introducing the further items: on-line beam correspondence accuracy learning based on the statistics of SRS beam indications, then updating candidate TX beam set, and in the long term modifying BM SRS capabilities as described in the following.

[0119] When a UE 110 does not support full beam correspondence capability, in order to reduce SRS overhead used for UE TX beam sweeping while still to achieve enough beam correspondence accuracy, this third approach introduces a scheme where, after being deployed in the field, UE 110 can iteratively apply on-line UE beam correspondence accuracy learning, and then accordingly reduce the SRS overhead based on the learning results. The core concept is to view the SRS TX beam sweeping process in the field as "on-line RF calibration". Hereby, the on-line beam correspondence accuracy learning can be achieved by monitoring the SRS beam indications 1120 from gNB 120 side, which is associated to the previous TX beam sweeping burst. Furthermore, UE 110 can estimate a quality metric of the swept TX beams, and then accordingly narrow down the size of the TX beam candidate set 1101, 1102, 1103 for each reference RX beam 1110. Based on such on-line learning after a time period of the field deployment and field operations, UE 110 can accordingly reduce the supported number of BM SRS resources, which is defined by 5G NR as one UE capability. This results in adaptive SRS overhead reduction.

[0120] In this third approach, the UE procedures can be divided into two types of time behaviors, after having been deployed in the field: short-term behavior (dynamic behavior) and long-term behavior (semi-statistic behavior) .

[0121] The short-term behavior is mainly targeting at online beam correspondence accuracy learning on UE side 110, which can be applied in an iterative manner. Figure 11 shows one example UE procedure for these new methods.

[0122] Fig. 12 is an example procedure 1200 of UE online beam correspondence accuracy learning based on TX beam sweeping and SRS beam indication statistics according to the disclosure .

[0123] The procedure 1200 comprises a first block 1201 including: Identify the current activated UE RX beam as the reference RX beam. Then, a second block 1202 includes: Extract M TX beam candidates adjacent to the reference beam. M is any integer number. Then a third block 1203 includes: Transmit M SRX resources, each associated to one selected TX beam candidate. A fourth block 1203 includes: Receive the selected SRS ID from SRS beam indication from gNB (MAC CE or RRC reconfiguration) . A fifth block 1205 includes: Apply the TX beam candidate associated to received SRS ID for PUSCH/PUCCH transmission. A sixth block 1206 includes: Update the beam correspondence accuracy metric (also refered to as beam correspondence metric or beam quality metric) for the selected and applied TX beam candidate, e.g. based on UL quality estimations. Then, a seventh block 1207 includes a condition: Is the previous metric greater than a threshold (Th) ? If no, the procedure 1200 jumps back to the first block 1201. If yes, an eighth block 1208 is performed including: Update the size of the TX beam candidate set M, e.g. M--, i.e. by decreasing M by 1. Thus, the number of TX beam candidate set is reduced.

[0124] Within the example of Figure 12, each reference RX beam is associated with a set of spatially adjacent TX beam candidates, each TX beam candidate is further associated with a beam correspondence accuracy metric which reflects spatial overlapping level of its beam pattern with respect to the reference UE RX beam. The initial size of the TX beam candidate set, as well as the beam correspondence accuracy metric values for the TX beam candidates can be pre-generated through RF Lab characterization based on a few sample phone devices, before the UEs are massively deployed into the field. After the massive field deployment, for each UE, at each SRS sweeping burst and based on the SRS beam indication associated to the same SRS sweeping burst, UE can then iteratively narrow-down the size of the TX beam candidate set as well as the associated beam correspondence accuracy metrics for the TX beam candidates within the set.

[0125] In particular, the beam correspondence accuracy metric can be determined by the received SRS ID through SRS beam indication (can be received SRS quality measurements indicated by gNB through MAC CE or RRC re-configuration) . The beam correspondence accuracy metric can further be determined by the UL quality estimation after the selected TX beam candidate has been applied for normal UL transmissions (e.g. UL BLER and UL throughput) , which can be further conditioned by the actual path-loss (PL) values. Furthermore, the size of the TX beam candidate set, which is associated to the same reference RX beam, can be iteratively reduced based on the updated beam correspondence accuracy metric.

[0126] In this third approach, the long-term behavior is mainly executed by UE to regularly check, for each of the RX beam within the codebook, the current number of associated TX beam candidates which have been has been learnt and updated during the field operations. When the maximal number of TX beam candidates among all reference RX beams is below the currently supported number of BM SRS resources, UE can then reduce the SRS overhead by reducing the number of supported BM SRS resources in UE capability indications. Figure 13 shows one example UE procedure of it.

[0127] Fig. 13 is an example procedure 1300 of SRS overhead reduction by adaptively modifying the supported number of BM SRS resources through UE capability indications according to the disclosure.

[0128] The procedure 1300 starts with a first block 1301 including: Loop over all RX beams in the pre-defined code book with UE . Then, a second block 1302 includes: Identify the maximal number (Mmax) of adjacent TX beam candidates for all RX beams. Then a third block 1303 includes a condition: Mmax less than number of SRS (numOfSRS) ? If no, the procedure 1300 jumps back to the first block 1301. If yes, a fourth block 1304 is performed including: numOfSRS : =Mmax, i.e. the maximal number of UE supported BM SRS resources is set to the maximal number Mmax of adjacent TX beam candidates for all RX beams. Then, a fifth block 1305 is performed that includes: Reducing the supported number of BM SRS resources (numOfSRS) to gNB .

[0129] Fig. 14 is an example state diagram 1400 illustrating transition between enabling and disabling for online beam calibration & learning according to the disclosure.

[0130] The state diagram 1400 includes two states: a first state 1401 is On-line beam correspondence calibration and learning enabled; a second state 1402 is On-line beam correspondence calibration and learning disabled. In the first state 1401 UE actions 1403 include operations as shown in Fig. 12. In the second state 1402 UE actions 1404 include minimal number of supported BM SRS resources to base station, as UE capability.

[0131] A first transition 1405 from state 1401 to state 1402 is performed upon trigger events 1405: e.g. more than 80% of RX beams have been calibrated, each for its associated TX beam candidate set. A second transition 1406 from state 1402 to state 1401 is performed upon trigger events 1406: e.g. extreme temperatures; e.g. measured poor UL qualities even though DL beam quality is good (line of sight channel, low path-loss, etc.).

[0132] According to 3GPP signaling, following message fields within UE capability message (TS 38.331 standard) indicates the SRS overhead for TX beam sweeping:

MIMO-ParametersPerBand ::=SEQUENCE { beamCorrespondence ENUMERATED {supported} OPTIONAL, uplinkBeamManagement SEQUENCE {

maxNumberSRS-ResourcePerSet-BM ENUMERATED {n2, n4, n8, nl6}, maxNumberSRS-ResourceSet INTEGER (1..8)

}

[0133] For example, when UE needs the assistance of SRS TX beam sweeping, it will leave "beamCorrespondence" field un configured. When it is un-configured, gNB will trigger SRS TX beam sweeping. The supported number of BM SRS resources used for sweeping is indicated as another UE capability by the following 2 fields. This is the trade-off between UL

beamforming accuracy and SRS overhead. maxNumberSRS-ResourcePerSet-BM

maxNumberSRS-ResourceSet

[0134] After initial deployment, UE will indicate a high number of BM SRS resources as the UE capability. That is to speed up the RX-TX beam pattern on-line calibration. After long-term learning and adaptively narrowing down the

candidate TX beam candidate set for each RX beam, UE can gradually reduce the capability number of BM SRS resources.

[0135] In the extreme case, when UE is confident enough about the on-line calibration, UE can explicitly configure the beamCorrespondence field as a UE capability, so that SRS TX beam sweeping is no longer required, which can totally remove the SRS overhead.

[0136] The transition between enabling 1401 and disabling 1402 for online beam calibration and learning is shown in Fig. 14.

[0137] Fig. 15 is a schematic diagram of a communication system 1500 illustrating SSB TX beam neighborhood relationship detection by using a wide UE RX beam reception of different SSBs according to the disclosure.

[0138] The base station 120 generates multiple SSB TX beams 1520 from which a first beam 1521, a third beam 1523 and a fourth beam 1524 are spatially adjacent and directed towards the UE 110 and a second beam 1522 and a fifth beam 1525 are directed reverse to the UE 110. The UE 110 generates a wide UE RX beam 1510 towards the gNB 120 which at least partially overlaps with the SSB TX beams 1521, 1523 and 1524 of gNB 120.

[0139] In order to reduce the UE RX beam re-acquisition latency (e.g. DRX or C-DRX waken up in FR2) in mmWave bands, the fourth approach as described above with respect to Fig. 1 introduces a concept to pre-detect the SSB TX beam spatial neighborhood relationship information, which is cell specific and unchanged with respect to UE mobility, and later-on apply this pre-detected information to further speed up the later- on UE RX beam re-acquisition. Note that, SSB TX beam neighborhood relationship does not necessarily follow the time order of SSB allocations and it is up to gNB implementation. With the detection sub-methods as described below, UE is robust against any time order allocation of SSB TX beams .

[0140] In the following, methods of SSB TX beam neighborhood relationship information blind detection are described. Such SSB TX beam neighborhood relationship information can be pre detected by a UE when it was in RRC-CONNECTED mode without non-DRX operations for a same serving cell. Since such information is cell specific and unchanged with respect to UE mobility, the detection can be done in an opportunistic manner (e.g. sweep RX beams on different serving cell SSBs only when serving cell beam with respect to the selected SSB is already on-track and when there is no FDMed PDSCH/PDCCH scheduled in parallel with those SSBs) . Furthermore, since this is static information, the detection can be long time averaged to get reliable results. The following two sub methods may be applied. [0141] Sub-method 1: UE can use a same wider RX beam to receive different SSB TX beams, and then compare the channel parameter estimation from different SSBs. For SSBs whose TX beams are spatially correlated, the channel parameters are also correlated (e.g. similar RSRP measurements, similar delay spread measurements, or even high correlated channel transfer functions), those information can be jointly explored to derive a spatial correlation metric between SSB beams. One example of this method is shown in Figure 15. Hereby, the SSB TX beam neighborhood configuration is the same as the example in Figure 15. UE can setup a wider RX beam 1510 and measure the channel parameters for each SSB index. Since SSB1, SSB3 and SSB4 according to beams 1521,

1523 and 1524 are co-located gNB TX beams 1520, the measurement should show similar channel parameters (E.g. RSRPs, delay spreads) . In this example, SSB2 and SSB5 according to beams 1522, 1525 will show significant different channel parameters (e.g. much weaker RSRP due to mismatched angle of departure and angle of arrival, and much longer delay spread due to reflections) .

[0142] Sub-method 2: SSB TX beam neighborhood relationship information can be detected by UE 110 by estimating the relative Angle of Arrivals (relative AoAs) for different SSBs. The relative AoA of different SSBs can be estimated by first scanning narrow UE RX beams with a-priori UE RX beam angle information (pre-known by UE RF characterizations) so as to derive a (can be sub-sampled) RSRP-Angle profile for different SSBs. And then, in the offline processing, the sub sampled RSRP-Angle profile can be optionally interpolated (interpolation is only needed if the swept UE RX beam angle has high steps) , the best receiver angles can be identified (angle mapped to the peak RSRPs) for different SSB indexes, and then the delta angle which is associated to the peak RSRPs of different SSB indexes within the derived RSRP-Angle profile can be computed. This method is conceptually illustrated by the example in Figure 16 and Figure 17 together .

[0143] Fig. 16 is a schematic diagram of an exemplary initial RSRP-Angle profile 1600 by UE RX beam sweeping for different two SSB indices according to the disclosure.

[0144] For this sub-method, as shown by the conceptual example in Figure 16, UE scans narrow RX beams for different SSB indexes with a pre-defined angle step. The angle step is determined by angle resolution of UE RX beam patterns (pre known to UE by RF characterizations, which is further pre determined by the UE code-word design and the number of antenna elements within a UE antenna panel) . In this example, the stepping size is 30°.

[0145] Hereby, as a further option, as a trade-off between the resolution and the scanning time, UE can also generate higher angle stepping size and later-on apply RSRP interpolations to gain back the resolution. By scanning narrow UE RX beams, UE can derive a (sub-sampled) RSRP-Angle profile for different SSB indexes (In this example SSB index 1 and 2 ) .

[0146] In Figure 16, the measured RSRP 1601 for SSB index 1 is shown for each swept UE RX beam; and the measured RSRP 1602 for SSB index 2 is shown for each swept UE RX beam. [0147] Fig. 17 is a schematic diagram of an exemplary interpolated RSRP-Angle profile 1700, peak RSRP selection and relative AoA estimation according to the disclosure.

[0148] And then, as shown by Figure 17, UE can optionally apply RSRP interpolations on the initially derived RSRP-Angle profile to improve the resolution. UE further identifies the peak RSRP values for different SSB indexes, and then estimates the relative AoA 1703 of different SSB indexes by taking the difference of the angles which are associated to the peak RSRP values.

[0149] In the example of Figure 17, the measured RSRP 1701 for SSB index 1 is shown for each swept UE RX beam; and the measured RSRP 1702 for SSB index 2 is shown for each swept UE RX beam. Additionally, the interpolated RSRP 1711 for SSB index 1 and the interpolated RSRP 1712 for SSB index 2 is shown. The two peaks are from the measured RSRP 2 and the interpolated RSRP for SSB index 1 resulting in a relative AoA of 52.5° .

[0150] Furthermore, for this sub-method, the UE RX narrow beam scanning can be activated in an opportunistic manner (e.g. in non-DRX operations and the target SSB is not FDMed with DL PDCCH/PDSCH) . Furthermore, it can happen that at a certain point of time, only a sub-set of SSB indexes can be detected because UE may not be under the coverage by some other SSB TX beams, but the concept still works because UE can later-on detect neighborhood relationships of those SSB TX beams when opportunistically being under the coverage. [0151] Fig. 18 is a schematic diagram of a communication system 1800 illustrating an example of SSB grouping based UE RX beam acquisition according to the disclosure.

[0152] After having detected SSB beam neighborhood relationship information, UE 110 can later-on use this information to speed up the beam re-acquisition when being waken up from CDRX/DRX operations. For example, instead of sweeping all RX beam candidates per SSB index in a round- robin manner (2-D search) , UE can group SSBs which are pre detected to be spatially correlated (SSBs whose gNB TX beams in the neighborhood of each other) and apply the wide UE RX beam sweeping per SSB group. Afterwards, UE 110 can down- select the best SSB group, and further apply narrow RX beam sweeping within the down-selected best SSB group. The procedure can terminate until a good reception quality on a SSB is reached.

[0153] Fig. 18 shows one example to illustrate the gain of the presented fourth approach. A 2-level of hierarchical beam structure is assumed which is a normal UE beam configuration: 4 wide UE beams and within each wide UE beam there are 4 narrow UE beams (in total 16 narrow UE beams) .

[0154] I.e., UE 110 produces four wide beams WB(1) referred to as 1810, WB(2) referred to as 1820, WB(3) referred to as 1830 and WB(4) referred to as 1840 which are directed in four different sectors. In each wide beam four narrow beams are included. In particular WB(1) includes NB(1,1) referred to as 1811, NB(1,2) referred to as 1812, NB(1,3) referred to as 1813 and NB(1,4) referred to as 1814. WB(2) includes NB(2,1) referred to as 1821, NB(2,2) referred to as 1822, NB(2,3) referred to as 1823 and NB(2,4) referred to as 1824. WB(3) includes NB(3,1) referred to as 1831, NB(3,2) referred to as 1832, NB(3,3) referred to as 1833 and NB(3,4) referred to as 1834. WB(4) includes NB(4,1) referred to as 1841, NB(4,2) referred to as 1842, NB(4,3) referred to as 1843 and NB(4,4) referred to as 1844. In this example, gNB 120 produces two SSBs, i.e. SSB1 corresponding to narrow beam 1801 and SSB2 corresponding to narrow beam 1802 that are directed towards UE 110.

[0155] Furthermore, in this example SSB1 (according to beam 1801) and SSB2 (according to beam 1802) are from neighboring gNB TX beams. Such SSB TX beam neighborhood relationship information may have been pre-detected by the UE 110. In this example, UE can acquire narrow UE beam NB (4, 2), 1842 for SSB1, 1801 as the best beam pair.

[0156] In the following, a comparison for UE beam acquisition time between a hierarchical UE beam acquisition without exploring the SSB TX beam neighboring information and a hierarchical UE RX beam acquisition by exploring the SSB TX beam neighboring information (which corresponds to the fourth approach as described above with respect to Figs. 15 to 18) is performed.

[0157] For the hierarchical UE RX beam acquisition approach without exploring the SSB TX beam neighboring information, the search procedure is abstracted in Table 1, in which the acquisition of SSB index 1 and SSB index 2 are applied independently by ignoring the SSB TX beam neighborhood relationship information. Hereby, as shown in Figure 18, after the measurement in the 4^th SSB burst, wide UE beam WB(4) 1840 shall be detected as the best UE wide beam. And then, starting from 5^th SSB burst, UE sweeps the narrow beams WO 2020/259838 PCT/EP2019/067135

within wide beam WB(4) 1840 for SSB1 (1801) and SSB2 (1802) independently. It will need in total 8 SSB bursts to finish the beam acquisition. [0158] Note that, by comparing the SSB1 and SSB2 measurements based on a same wide UE beam within the 4^th SSB burst, if the measurement quality of SSB2 is much worse than that of SSB1, UE can skip the narrow beams measurement for SSB2 in the follow up steps. However, the UE beam acquisition time cannot be reduced in this approach because the acquisition time is bounded by the repetition period of SSB burst, not SSB indexes.

Table 1: Procedure of hierarchical UE beam acquisition without exploring SSB beam neighborhood info

[0159] Hierarchical UE RX beam acquisition by exploring the SSB TX beam neighboring information corresponds to the fourth approach as described above with respect to Figs. 15 to 18. The search procedure is abstracted in Table 2. Hereby, SSB1 (1801) and SSB2 (1802) within a same SSB burst are from neighboring gNB TX beams. Such information is assumed to be pre-detected by the UE 110. In this approach, instead of always applying RX beam sweeping over SSB1 and SSB2 independently, UE groups SSB1 and SSB2 during wide beam sweeping steps by alternating the UE wide RX beam candidates for SSB1 and SSB2 within a same SSB burst. This results in two wide RX beam candidates tested within a single SSB burst, which speeds up the wide beam acquisition. Since SSB1 and SSB2 are co-located, UE is supposed to find the best wide UE beam after measuring SSB2 using WB(4), after the 2^nd SSB burst .

[0160] Note that, since SSB1 and SSB2 are measured using different wide beam candidates, after 2^nd SSB burst, UE does not know which SSB beam is the best. So it has to test all narrow beam candidates within WB(4) for both SSB1 and SSB2. However, since SSB1 and SSB2 are available in every SSB burst, it does not extend the UE acquisition time. According to table 2, the UE can finish beam re-acquisition by using only 6 SSB bursts.

Table 2: Procedure of hierarchical UE beam acquisition exploring SSB beam neighborhood info

[0161] By comparing the above two approaches (i.e. without or with using neighborhood information) for this particular example, by exploring the pre-detected SSB TX beam neighborhood relationship information, the UE beam re acquisition latency can be reduced by 25%.

[0162] In a further embodiment, gNB SSB TX beam spatial adjacent information can be shared from one UE to another UE (e.g. through application layer information exchanges), which are associated with multiple detected base stations within a geographical area.

[0163] gNB SSB TX beam spatial adjacent information sharing can be explicitly configured from base station to a connected UE through RRC messages.

[0164] Fig. 19 is a schematic diagram illustrating an exemplary method 1900 for determining an RX beam pattern for a UE according to the disclosure. The blocks described in Fig. 19 follow the order of the procedures described above with respect to Fig. 2. In a first block 1901, UE detects a change of an antenna radiation condition. In a second block 1902, UE receives downlink signals from base station for UE RX beam sweeping. In a third block 1903, UE determines a UE RX beam pattern based on a selection from candidate beams formed on the basis of the downlink signals. In a fourth block 1904, UE determines the UE RX beam pattern based on a selection from a set of UE RX candidate beam patterns formed on the basis of further downlink signals carrying pre-known information, wherein the further downlink signals are received from the base station upon request of the UE or another UE .

[0165] Thus, the method 1900 comprises: receiving 1902 downlink signals from a base station for UE RX beam sweeping; determining 1903 a UE RX beam pattern based on a selection from candidate beams formed on the basis of the downlink signals; and upon detection 1901 of a change of an antenna radiation condition, determining 1904 the UE RX beam pattern based on a selection from a set of UE RX candidate beam patterns formed on the basis of further downlink signals carrying pre-known information, wherein the further downlink signals are received from the base station upon request of the UE or another UE . The method 1900 provides improved UE beamforming according to the first approach as described above with respect to Fig. 1. Examples of the method 1900 are described above with respect to Figures 2 to 4.

[0166] The change of antenna radiation condition may comprise a change/dynamics of propagation status at the UE antenna array circuity or a change of a mobility status of the UE . Note that not only the UE can be moving, but also the UE environment can move, for example a door could be

opened/closed or a lorry changing the environment, which also changes the optimal beams.

[0167] The downlink signals may include DL reference and/or DL data signals. A SIB that may be included in the DL signals contains DMRS, PDCCH and PDSCH, so not only reference

signals, but also data signals.

[0168] Pre-known information may include redundancy

information .

[0169] The further downlink signals may comprise a Physical Downlink Control Channel, PDCCH, and an associated Physical Downlink Shared Channel, PDSCH, wherein the PDSCH comprises: a System Information Block, SIB, which is associated to an on-demand SIB request from the UE or further UEs; or a re- transmitted Downlink Shared Channel, DL-SCH, data which is associated to a Downlink, DL, re-transmission request by the UE, e.g. as described above with respect to Figures 2 to 4.

[0170] The SIB is associated to an on-demand SIB request from the UE or further UEs, even though the UE may have already successfully decoded the SIB.

[0171] The re-transmitted UL-SCH data is associated to a DL re-transmission request by the UE, even though the UE may have already successfully decoded the first transmission of the same data.

[0172] The method 1900 may further comprise: sending a request for a further downlink signal to the base station; or monitoring a request for a further downlink signal sent by other UEs without sending an own request to the base station. The request comprises the on-demand SIB request even though the UE has already decoded the same SIB. Alternatively the request comprises the DL re-transmission request even though the UE has already successfully decoded a first transmission of the same DL-SCH, data.

[0173] The method 1900 may further comprise: determining the UE RX beam based on measuring UE RX candidate beam patterns using PDSCH Demodulate Reference Signals, DMRS, and/or PDSCH data symbols of the further downlink signals, e.g. as

described above with respect to Figures 1 to 4. The method 1900 may further comprise: measuring the UE RX candidate beam patterns by reproducing PDSCH I/Q phase information through re-encoding and re-modulating of pre-known information bits; and based on the reproduced PDSCH I/Q phase information, estimating a reference signal received power, RSRP, of the received PDSCH data symbols which are received using a candidate RX beam pattern.

[0174] The method 1900 may further comprise: decoding a

Physical Downlink Control Channel, PDCCH, within the further downlink signals based on an activated UE RX beam to obtain a time and frequency allocation of a PDSCH region of a SIB or re-transmitted DL data; and measuring the set of UE RX candidate beam patterns during the PDSCH region of the further downlink signals, e.g. as described above with respect to Figures 1 to 4.

[0175] The method 1900 may further comprise: detecting the change of antenna radiation condition based on detection of a UE antenna position change event, e.g. as described above with respect to Figures 1 to 4.

[0176] The method 1900 may further comprise: determining the UE RX beam based on on-demand SIBs received from the base station upon request of the UE or another UE and/or based on periodic SIBs initiated by the base station, e.g. as

described above with respect to Figures 1 to 4.

[0177] Fig. 20 is a schematic diagram illustrating an exemplary method 2000 for polarization control in TX beamforming of a UE according to the disclosure. The method 2000 may include the two main blocks 2001 and 2002. In the first main block 2001, a TX beam sweeping strategy is determined. In the second main block 2002, the TX beam sweeping strategy is applied by transmitting TX beam sweeping signal bursts. These two main blocks 2001 and 2002 can be described by the following procedures of the method 2000. [0178] Generating 2001 one or more TX beam sweeping signal bursts, each TX beam sweeping signal burst comprising a plurality of uplink beam sweeping reference signals, wherein a polarization pattern and a spatial beam pattern of the uplink beam sweeping reference signals is varied over the one or more beam sweeping signal bursts, and wherein a variability of the polarization pattern and the spatial beam pattern of the uplink beam sweeping reference signals depends on propagation parameters.

[0179] The method 2000 provides improved UE beamforming according to the second approach as described above with respect to Fig. 1. Examples of the method 2000 are described above with respect to Figures 5 to 9.

[0180] A core idea behind this method 2000 is to either offer more options on polarization patterns or special beam

patterns, depending on what is more beneficial and that may depend on some criteria of the environment and mobility status .

[0181] The variation may be based on TX beam pattern sweeping or based on TX polarization pattern sweeping.

[0182] TX beam pattern sweeping may comprise: varying the uplink beam sweeping reference signals based on the same polarization pattern but different spatial beam patterns.

[0183] TX polarization pattern sweeping may comprise: varying the uplink beam sweeping reference signals based on different polarization patterns but with the same spatial beam pattern. [0184] The method 2000 may further comprise: determining the propagation parameters based on path-loss measurements on UE side, e.g. as described above with respect to Figures 5 to 9.

[0185] The path-loss measurement on UE side are reflecting radio propagation distance; for example for low path-loss, a wide spatial beam pattern may be used but the focus lies more on polarization pattern sweeping, e.g. as described above with respect to Figures 5 to 9.

[0186] The method 2000 may further comprise: determining the propagation parameters based on polarization dynamics of the base station, e.g. as described above with respect to Figures 5 to 9.

[0187] When the base station is also sweeping its RX

polarization pattern at the same time, a same TX polarization pattern may be used but with the focus lying on more spatial beam pattern sweeping.

[0188] The method 2000 may further comprise: determining the propagation parameters based on a mobility status of the UE, e.g. as described above with respect to Figures 5 to 9.

[0189] When UE is moving fast the focus will lie on more spatial beam pattern sweeping because polarization pattern optimization is less robust in high mobility.

[0190] The method 2000 may further comprise: using the polarization control for single layer transmission mode in TX beamforming of the UE, e.g. as described above with respect to Figures 5 to 9. [0191] The method 2000 may further comprise: aligning an angle between a polarization of the one or more TX beam sweeping signal bursts from the UE and a polarization of an RX beam from a serving base station, e.g. as described above with respect to Figures 5 to 9.

[0192] The method 2000 may further comprise: orthogonalizing an angle between a polarization of the one or more TX beam sweeping signal bursts from the UE and a polarization of an RX beam from a neighboring base station, e.g. as described above with respect to Figures 5 to 9.

[0193] The method 2000 may further comprise: prioritizing TX polarization pattern sweeping over TX beam pattern sweeping upon a communication range to a serving base station going below a threshold, e.g. as described above with respect to Figures 5 to 9.

[0194] The method 2000 may further comprise: determining the communication range based on a path-loss or based on a timing advance in communications with the serving base station, e.g. as described above with respect to Figures 5 to 9.

[0195] The method 2000 may further comprise: prioritizing TX beam pattern sweeping over TX polarization pattern sweeping upon a mobility status metric of the UE being higher than a pre-defined threshold, e.g. as described above with respect to Figures 5 to 9.

[0196] The method 2000 may further comprise: prioritizing TX beam pattern sweeping over TX polarization pattern sweeping upon detecting that the serving base station also applies polarization control for uplink signal reception, e.g. as described above with respect to Figures 5 to 9.

[0197] The method 2000 may further comprise: detecting polarization control of the serving base station based on determining polarization angle of downlink reference signals from the serving base station or based on determining

statistics of sounding reference signal, SRS, beam

indications from the serving base station, e.g. as described above with respect to Figures 5 to 9.

[0198] Fig. 21 is a schematic diagram illustrating an exemplary method 2100 for UE beamforming based on UE beam correspondence according to the disclosure. The method 2100 may include the following four main blocks 2101, 2102, 2103 and 2104. The first main block 2101 includes generating a set of candidate TX beam patterns. The second main block 2102 includes sweeping TX beam patterns by transmitting beam sweeping reference signal bursts. The third main block 2103 includes collecting the beam correspondence accuracy metrics of swept TX beam patterns. The fourth main block 2104 includes updating a size of the set of candidate TX beam patterns. These four blocks 2101, 2102, 2103 and 2104 can be described by the following procedures of the method 2100.

[0199] Generating 2101 a set of candidate TX beam patterns for transmission to a base station; updating 2102 a size of the set of candidate TX beam patterns based on a quality metric with respect to a correspondence criterion indicating a beam correspondence of an RX beam pattern steered by the same UE, which receives downlink signals transmitted from the base station, with any of the candidate TX beam patterns from previous transmissions; and based on the updated size of the set of candidate TX beam patterns, updating 2103 a UE

capability message indicating a number of TX beam sweeping resources supported by the UE .

[0200] The method 2100 provides improved UE beamforming according to the third approach as described above with respect to Fig. 1. Examples of the method 2100 are described above with respect to Figures 10 to 13.

[0201] Beam correspondence reflects the level of spatial overlapping between a TX beam pattern and an associated RX beam pattern steered by the same UE, not with a RX beam pattern from the base station. That is because the UE is receiving signals from the base station through UE's RX beam pattern. So, the pattern is not transmitted by the base station, only the signal that is received using the RX beam pattern was transmitted by the base station.

[0202] The method 2100 may further comprise: updating the size of the set of candidate TX beam patterns based on learning of a beam correspondence metric, which is associated with each candidate TX beam pattern, e.g. as described above with respect to Figures 10 to 13. A beam correspondence metric reflects a level of spatial overlapping between a TX beam pattern and an associated RX beam pattern steered by the same UE .

[0203] The method 2100 may further comprise: updating the size of the set of candidate TX beam patterns to include a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is higher than a pre-defined threshold; and updating the size of the set of candidate TX beam patterns to exclude a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is lower than a pre-defined threshold, e.g. as described above with respect to Figures 10 to 13.

[0204] The method 2100 may further comprise: selecting a number of candidate TX beam patterns based on a number of beamforming management, BM, sounding reference signal, SRS, resources within a triggered BM SRS sweeping burst;

transmitting SRS resources, each associated to one of the candidate TX beam patterns from the set; receiving an SRS resource ID for a SRS resource selected by the base station; updating the beam correspondence metric of the candidate beam pattern associated to the SRS resource ID; and updating the set of candidate TX beam patterns upon the beam

correspondence metric passing a threshold, e.g. as described above with respect to Figures 10 to 13.

[0205] The beam correspondence metric can be updated, for example as follows: it can be based on the counting of the statistics that a same candidate TX beam pattern has been selected after historical SRS sweeping bursts. It can further be determined by the UL quality measurement after the

selected candidate TX beam candidate is used for actual PUSCH transmission .

[0206] During SRS sweeping, UE transmits multiple SRS

resources with multiple SRS IDs. Then gNB down-selects the best received SRS resource and indicates back to UE the only one selected SRS ID.

[0207] The method 2100 may further comprise: updating the beam correspondence metric based on statistics of a same candidate TX beam pattern whose associated SRS ID has been selected by the base station through historical SRS sweeping events, e.g. as described above with respect to Figures 10 to 13.

[0208] An initial size of the set of candidate TX beam patterns and initial values of the quality metric can be pre determined based on RF Lab characterizations.

[0209] The method 2100 may further comprise: updating the beam correspondence metric based on a SRS quality metric which is associated to a UE transmitted SRS resource and indicated back from the base station to the UE through higher layer messages, e.g. as described above with respect to

Figures 10 to 13.

[0210] As for an example, the beam correspondence metric can be updated based on a SRS quality metric, e.g. SRS SINR, which is associated to a UE transmitted SRS resource and indicated back from the base station to the UE through higher layer messages, e.g. media access control channel estimation (MAC CE) or radio resource control (RRC) .

[0211] The method 2100 may further comprise: updating the beam correspondence metric based on based on the UL quality measurement in UE side (UL BLER and/or UL throughput) after the candidate TX beam pattern has been selected by the base station and applied by UE for UL data transmission, e.g. as described above with respect to Figures 10 to 13.

[0212] The method 2100 may further comprise: reducing a number of supported BM SRS resources upon detecting that a number of candidate TX beam patterns among all reference RX beam patterns is below the number of supported BM SRS resources, e.g. as described above with respect to Figures 10 to 13.

[0213] The method 2100 may further comprise: disabling the updating of the size of the set of candidate TX beam patterns upon detecting that a threshold number of RX beam patterns have been calibrated, each for its associated set of

candidate TX beam patterns, e.g. as described above with respect to Figures 10 to 13.

[0214] The method 2100 may further comprise: enabling the updating of the size of the set of candidate TX beam patterns upon detecting uplink qualities falling below a threshold, e.g. as described above with respect to Figures 10 to 13.

[0215] Fig. 22 is a schematic diagram illustrating an exemplary method 2200 for re-acquiring a beam pair between a UE and a base station based on grouping information according to the disclosure.

[0216] The method 2200 comprises: determining 2201 a TX beam configuration of the base station based on a selection of a beam pair between the UE and the base station for each of a plurality of UE positions towards the base station, each beam pair comprising an RX beam pattern of the UE and a TX beam pattern of the base station; determining 2202 grouping information of downlink reference signals, wherein downlink reference signals within a group are associated to a subset of at least two TX beam patterns of the base station; and re acquiring 2203 a beam pair between the UE and the base station based on the determined grouping information. [0217] The method 2200 provides improved UE beamforming according to the fourth approach as described above with respect to Fig. 1. Examples of the method 2200 are described above with respect to Figures 14 to 17.

[0218] The method 2200 may further comprise: determining the grouping information based on determining a spatial adjacent metric among different TX beam patterns of the base station over time, e.g. as described above with respect to Figures 14 to 17.

[0219] The grouping information indicates a neighborhood relationship between the TX beam patterns of the base

station. The grouping information is cell specific and unchanged with respect to UE mobility.

[0220] The method 2200 may further comprise: determining the grouping information based on a database comprising grouping information of multiple base stations, wherein each grouping information is associated with a base station identity and a geographical region identity, e.g. as described above with respect to Figures 14 to 17.

[0221] The method 2200 may further comprise: determining the grouping information based on messages from a serving base station, e.g. as described above with respect to Figures 14 to 17.

[0222] The method 2200 may further comprise: re-acquiring the beam pair based on measuring different downlink reference signals within a group by using different UE RX beam patterns for fast acquisition, e.g. as described above with respect to Figures 14 to 17. [0223] The method 2200 may further comprise: re-acquiring the beam pair based on measuring different downlink reference signals within a group by using a single UE RX beam pattern; and combining results of the measurements for high accuracy, e.g. as described above with respect to Figures 14 to 17.

[0224] The method 2200 may further comprise: re-acquiring the beam pair based on SNR conditions, UE mobility and/or UE battery status, e.g. as described above with respect to

Figures 14 to 17.

[0225] The method 2200 may further comprise: determining the spatial adjacent metric based on a difference of estimated channel parameters from different TX beam patterns of the base station, received using a same UE RX beam pattern, e.g. as described above with respect to Figures 14 to 17.

[0226] The method 2200 may further comprise: determining the estimated channel parameters based on Reference Signal

Received Power, RSRP, measurements, delay spread measurements and/or channel transfer functions of the different TX beam patterns, e.g. as described above with respect to Figures 14 to 17.

[0227] The method 2200 may further comprise: determining the grouping information based on determining relative angle of arrivals, AoAs, of the different TX beam patterns, e.g. as described above with respect to Figures 14 to 17.

[0228] The method 2200 may further comprise: determining the relative AoAs based on determining a sub-sampled angle profile for the different TX beam patterns by scanning narrow UE RX beams with pre-known radiation patterns, e.g. as described above with respect to Figures 14 to 17.

[0229] The method 2200 may further comprise: interpolating the sub-sampled angle profile and identifying for each of the different TX beam patterns an angle having a peak RSRP, e.g. as described above with respect to Figures 14 to 17.

[0230] The method 2200 may further comprise: determining the relative AoAs of the different TX beam patterns based on a difference of angles which are associated with the peak

RSRPs, e.g. as described above with respect to Figures 14 to 17.

[0231] The downlink reference signals may be arranged in Synchronization Signal Blocks, SSBs, having SSB indices, each SSB index associated to a different TX beam pattern of the base station, wherein a number of SSBs with different SSB indices form an SSB burst.

[0232] Fig. 23 is a block diagram illustrating a UE circuitry 2300 according to the disclosure. The UE circuitry 2300 includes an RX/TX (receive/transmit) circuitry 2301 for receiving downlink signals from a serving base station, e.g. the base station 120 shown in Fig. 1 and transmitting uplink signals to the serving base station or to other base stations. The UE circuitry 2300 further includes a beamforming management (BM) circuitry 2302 for performing beamforming in uplink and downlink direction. BM circuitry

2302 and RX/TX circuitry 2301 are coupled by an interface

2303 for exchanging data and/or control signals. The BM circuitry 2303 may be implemented in a baseband processor or in any other processor of the UE circuitry 2300 or the UE . The RX/TX circuitry 2301 may be implemented in a RF (Radio Frequency) circuitry of the UE .

[0233] The UE circuitry 2300 provides improved UE beamforming by implementing any one or any combination of the first, second, third and fourth approaches as described above with respect to Fig. 1. Thus, the UE circuitry 2300 can implement any of the methods 1900, 2000, 2100, 2200 as described above with respect to Figures 19 to 22. In particular, the UE circuitry 2300 is configured to implement any one or any combination of the techniques described above with respect to Figures 1 to 18.

[0234] With respect to the first approach, the RX circuitry is configured to receive downlink signals from a base station for UE RX beam sweeping; and the BM circuitry is configured to determine a UE RX beam pattern based on a selection from candidate beams formed on the basis of the downlink signals. The BM circuitry 2302 is configured, upon detection of a change of antenna radiation condition, to determine the UE RX beam pattern based on a selection from a set of UE RX candidate beam patterns formed on the basis of further downlink signals carrying pre-known information. The further downlink signals are received from the base station upon request of the UE or another UE .

[0235] The further downlink signals may comprise a Physical Downlink Control Channel, PDCCH, and an associated Physical Downlink Shared Channel, PDSCH. The PDSCH may comprise: a System Information Block, SIB, which is associated to an on- demand SIB request from the UE or further UEs; or a re transmitted Uplink Synchronization Channel, UL-SCH, data which is associated to a Downlink, DL, re-transmission request by the UE .

[0236] The RX/TX circuitry 2301 may be configured to send a request for a further downlink signal to the base station. Alternatively the UE circuitry may comprise a monitoring circuitry, configured to monitor a request for a further downlink signal sent by other UEs without sending an own request to the base station. The request may comprise the on- demand SIB request even though the UE has already decoded the same SIB. Alternatively, the request may comprise the DL re transmission request even though the UE has already

successfully decoded a first transmission of the same UL-SCH, data .

[0237] The BM circuitry 2302 may be configured to determine the UE RX beam based on measuring UE RX candidate beam patterns using PDSCH Demodulate Reference Signals, DMRS, and/or PDSCH data symbols of the further downlink signals.

[0238] The BM circuitry 2302 may further be configured to measure the UE RX candidate beam patterns by reproducing PDSCH I/Q phase information through re-encoding and re modulating of pre-known information bits; and based on the reproduced PDSCH I/Q phase information, estimating a

reference signal received power, RSRP, of the received PDSCH data symbols which are received using a candidate RX beam pattern .

[0239] The BM circuitry 2302 may be configured to detect the change of antenna radiation condition based on detection of a UE antenna position change event. [0240] The BM circuitry 2302 may be configured to: decode a Physical Downlink Control Channel, PDCCH, within the further downlink signals based on an activated UE RX beam to obtain a time and frequency allocation of a SIB PDSCH region; and measure the set of UE RX candidate beam patterns during the PDSCH region of the further downlink signals.

[0241] With respect to the second approach, the BM circuitry 2302 is configured to generate one or more TX beam sweeping signal bursts, each TX beam sweeping signal burst comprising a plurality of uplink beam sweeping reference signals, wherein a polarization pattern and a spatial beam pattern of the uplink beam sweeping reference signals is varied over the one or more beam sweeping signal bursts, wherein a

variability of the polarization pattern and the spatial beam pattern of the uplink beam sweeping reference signals depends on propagation parameters.

[0242] The variation may be based on TX beam pattern sweeping or based on TX polarization pattern sweeping.

[0243] The TX beam pattern sweeping may comprise: varying the uplink beam sweeping reference signals based on the same polarization pattern but different spatial beam patterns.

[0244] The TX polarization pattern sweeping may comprise: varying the uplink beam sweeping reference signals based on different polarization patterns but with the same spatial beam pattern.

[0245] The UE circuitry 2300 may be configured to determine the propagation parameters based on path-loss measurements on UE side. The UE circuitry 2300 may be configured to determine the propagation parameters based on a mobility status of the UE . The UE circuitry 2300 may be configured to determine the propagation parameters based on polarization dynamics of the base station.

[0246] With respect to the third approach, the BM circuitry 2302 is configured to generate a set of candidate TX beam patterns for transmission to a base station. The BM circuitry 2302 is further configured to update a size of the set of candidate TX beam patterns based on a quality metric with respect to a correspondence criterion indicating a

correspondence of an RX beam pattern steered by the same UE, which receives downlink signals transmitted from the base station, with any of the candidate TX beam patterns from previous transmissions. The BM circuitry 2302 is further configured, based on the updated size of the set of candidate TX beam patterns, to update a UE capability message

indicating a number of TX beam sweeping resources supported by the UE .

[0247] The BM circuitry 2302 may be configured to update the size of the set of candidate TX beam patterns based on learning of a beam correspondence metric, which is associated with each candidate TX beam pattern. A beam correspondence metric reflects a level of spatial overlapping between a TX beam pattern and an associated RX beam pattern steered by the same UE .

[0248] The BM circuitry 2302 may be configured to update the size of the set of candidate TX beam patterns to include a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is higher than a pre-defined threshold. The BM circuitry 2302 may be configured to update the size of the set of candidate TX beam patterns to exclude a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is lower than a pre-defined threshold.

[0249] The BM circuitry 2302 may be configured to select a number of candidate TX beam patterns based on a number of beamforming management, BM, sounding reference signal, SRS, resources within a triggered BM SRS sweeping burst; transmit SRS resources, each associated to one of the candidate TX beam patterns from the set; receive an SRS resource ID for a SRS resource selected by the base station; update the beam correspondence metric of the candidate beam pattern

associated to the SRS resource ID; and update the set of candidate TX beam patterns upon the beam correspondence metric passing a threshold.

[0250] With respect to the fourth approach, the BM circuitry 2302 is configured to determine a TX beam configuration of a base station based on a selection of a beam pair between the UE and the base station for each of a plurality of UE positions towards the base station, each beam pair comprising an RX beam pattern of the UE and a TX beam pattern of the base station. The BM circuitry 2302 is further configured to determine grouping information of downlink reference signals, wherein downlink reference signals within a group are associated to a subset of at least two TX beam patterns of the base station. The BM circuitry 2300 is further configured to re-acquire a beam pair between the UE and the base station based on the determined grouping information.

[0251] The BM circuitry 2300 may be configured to determine the grouping information based on determining a spatial adjacent metric among different TX beam patterns of the base station over time.

[0252] The BM circuitry 2300 may be configured to determine the grouping information based on a database comprising grouping information of multiple base stations, wherein each grouping information is associated with a base station identity and a geographical region identity.

[0253] The BM circuitry 2300 may be configured to determine the grouping information based on messages from a serving base station.

[0254] The BM circuitry 2300 may be configured to re-acquire the beam pair based on measuring different downlink reference signals within a group by using a single UE RX beam pattern and to combine results of the measurements for high accuracy.

[0255] The BM circuitry 2300 may be configured to re-acquire the beam pair based on measuring different downlink reference signals within a group by using different UE RX beam patterns for fast acquisition.

[0256] The BM circuitry 2300 may be configured to re-acquire the beam pair based on SNR conditions, UE mobility and/or UE battery status.

EXAMPLES

[0257] The following examples pertain to further embodiments. Example 1 is a method for determining an RX beam pattern for a user equipment, UE, the method comprising: receiving downlink signals from a base station for UE RX beam sweeping; determining a UE RX beam pattern based on a selection from candidate beams formed on the basis of the downlink signals; and upon detection of a change of an antenna radiation condition, determining the UE RX beam pattern based on a selection from a set of UE RX candidate beam patterns formed on the basis of further downlink signals carrying pre-known information, wherein the further downlink signals are received from the base station upon request of the UE or another UE .

[0258] The change of antenna radiation condition may comprise a change/dynamics of propagation status at the UE antenna array circuity or a change of a mobility status of the UE . Note that not only the UE can be moving, but also the UE environment can move, for example a door could be

[0259] The downlink signals may include DL reference and/or DL data signals. A SIB that may be included in the DL signals contains DMRS, PDCCH and PDSCH, so not only reference

signals, but also data signals. Pre-known information may include redundancy information.

[0260] In Example 2, the subject matter of Example 1 can optionally include that the further downlink signals comprise a Physical Downlink Control Channel, PDCCH, and an associated Physical Downlink Shared Channel, PDSCH, wherein the PDSCH comprises: a System Information Block, SIB, which is associated to an on-demand SIB request from the UE or further UEs; or a re-transmitted Downlink Shared Channel, DL-SCH, data which is associated to a Downlink, DL, re-transmission request by the UE . [0261] The SIB is associated to an on-demand SIB request from the UE or further UEs, even though the UE may have already successfully decoded the SIB.

[0262] The re-transmitted UL-SCH data is associated to a DL re-transmission request by the UE, even though the UE may have already successfully decoded the first transmission of the same data.

[0263] In Example 3, the subject matter of Example 2 can optionally include: sending a request for a further downlink signal to the base station; or monitoring a request for a further downlink signal sent by other UEs without sending an own request to the base station, wherein the request comprises the on-demand SIB request even though the UE has already decoded the same SIB; or wherein the request comprises the DL re-transmission request even though the UE has already successfully decoded a first transmission of the same DL-SCH, data.

[0264] In Example 4, the subject matter of Example 1 or

Example 2 can optionally include: determining the UE RX beam based on measuring UE RX candidate beam patterns using PDSCH Demodulate Reference Signals, DMRS, and/or PDSCH data symbols of the further downlink signals. The subject matter of Example 4 can optionally further include: measuring the UE RX candidate beam patterns by reproducing PDSCH I/Q phase information through re-encoding and re-modulating of pre known information bits; and based on the reproduced PDSCH I/Q phase information, estimating a reference signal received power, RSRP, of the received PDSCH data symbols which are received using a candidate RX beam pattern. [0265] In Example 5, the subject matter of Example 1 or

Example 2 can optionally include: decoding a Physical

Downlink Control Channel, PDCCH, within the further downlink signals based on an activated UE RX beam to obtain a time and frequency allocation of a PDSCH region of a SIB or re

transmitted DL data; and measuring the set of UE RX candidate beam patterns during the PDSCH region of the further downlink signals .

[0266] In Example 6, the subject matter of Example 1 or

Example 2 can optionally include: detecting the change of antenna radiation condition based on detection of a UE antenna position change event.

[0267] In Example 7, the subject matter of Example 1 or

Example 2 can optionally include: determining the UE RX beam based on on-demand SIBs received from the base station upon request of the UE or another UE and/or based on periodic SIBs initiated by the base station.

[0268] Example 8 is a User Equipment, UE, circuitry,

comprising: a receive, RX, circuitry, configured to receive downlink signals from a base station for UE RX beam sweeping; and a Beamforming Management, BM, circuitry, configured to determine a UE RX beam pattern based on a selection from candidate beams formed on the basis of the downlink signals, wherein the BM circuitry is configured, upon detection of a change of antenna radiation condition, to determine the UE RX beam pattern based on a selection from a set of UE RX

candidate beam patterns formed on the basis of further downlink signals carrying pre-known information, wherein the further downlink signals are received from the base station upon request of the UE or another UE .

[0269] In Example 9, the subject matter of Example 8 can optionally include that the further downlink signals comprise a Physical Downlink Control Channel, PDCCH, and an associated Physical Downlink Shared Channel, PDSCH, wherein the PDSCH comprises: a System Information Block, SIB, which is

associated to an on-demand SIB request from the UE or further UEs; or a re-transmitted Uplink Synchronization Channel, UL- SCH, data which is associated to a Downlink, DL, re

transmission request by the UE .

[0270] In Example 10, the subject matter of Example 9 can optionally include: a transmit, TX, circuitry, configured to send a request for a further downlink signal to the base station; or a monitoring circuitry, configured to monitor a request for a further downlink signal sent by other UEs without sending an own request to the base station, wherein the request comprises the on-demand SIB request even though the UE has already decoded the same SIB; or wherein the request comprises the DL re-transmission request even though the UE has already successfully decoded a first transmission of the same UL-SCH, data.

[0271] In Example 11, the subject matter of Example 8 or Example 9 can optionally include that the BM circuitry is configured to determine the UE RX beam based on measuring UE RX candidate beam patterns using PDSCH Demodulate Reference Signals, DMRS, and/or PDSCH data symbols of the further downlink signals. [0272] In Example 12, the subject matter of Example 11 can optionally include that the BM circuitry is configured to measure the UE RX candidate beam patterns by reproducing PDSCH I/Q phase information through re-encoding and re modulating of pre-known information bits; and based on the reproduced PDSCH I/Q phase information, estimating a

[0273] In Example 13, the subject matter of Example 8 or Example 9 can optionally include that the BM circuitry is configured to detect the change of antenna radiation

condition based on detection of a UE antenna position change event .

[0274] In Example 14, the subject matter of Example 8 or Example 9 can optionally include that the BM circuitry is configured to: decode a Physical Downlink Control Channel, PDCCH, within the further downlink signals based on an activated UE RX beam to obtain a time and frequency

allocation of a SIB PDSCH region; and measure the set of UE RX candidate beam patterns during the PDSCH region of the further downlink signals.

[0275] Example 15 is a device for determining an RX beam pattern for a user equipment, UE, the device comprising: means for receiving downlink signals from a base station for UE RX beam sweeping; means for determining a UE RX beam pattern based on a selection from candidate beams formed on the basis of the downlink signals; means for detecting a change of antenna radiation condition; and upon detection of a change of antenna radiation condition, means for determining the UE RX beam pattern based on a selection from a set of UE RX candidate beam patterns formed on the basis of further downlink signals carrying pre-known information, wherein the further downlink signals are received from the base station upon request of the UE or another UE .

[0276] In Example 16, the subject matter of Example 15 can optionally include that the further downlink signals comprise a Physical Downlink Control Channel, PDCCH, and an associated Physical Downlink Shared Channel, PDSCH, wherein the PDSCH comprises: a System Information Block, SIB, which is

transmission request by the UE .

[0277] Example 17 is a computer readable non-transitory medium on which computer instructions are stored which when executed by a computer cause the computer to perform the method of any one of Examples 1 to 7.

[0278] Example 18 is a method for polarization control in TX beamforming of a user equipment, UE, the method comprising: generating one or more TX beam sweeping signal bursts, each TX beam sweeping signal burst comprising a plurality of uplink beam sweeping reference signals, wherein a

polarization pattern and a spatial beam pattern of the uplink beam sweeping reference signals is varied over the one or more beam sweeping signal bursts, and wherein a variability of the polarization pattern and the spatial beam pattern of the uplink beam sweeping reference signals depends on

propagation parameters. [0279] A core idea behind this method is to either offer more options on polarization patterns or special beam patterns, depending on what is more beneficial and that may depend on some criteria of the environment and mobility status.

[0280] In Example 19, the subject matter of Example 18 can optionally include that the variation is based on TX beam pattern sweeping or based on TX polarization pattern

sweeping .

[0281] In Example 20, the subject matter of Example 19 can optionally include that TX beam pattern sweeping comprises: varying the uplink beam sweeping reference signals based on the same polarization pattern but different spatial beam patterns .

[0282] In Example 21, the subject matter of Example 19 can optionally include: varying the uplink beam sweeping

reference signals based on different polarization patterns but with the same spatial beam pattern.

[0283] In Example 22, the subject matter of Example 18 or Example 19 can optionally include: determining the

propagation parameters based on path-loss measurements on UE side .

[0284] The path-loss measurement on UE side are reflecting radio propagation distance; for example for low path-loss, a wide spatial beam pattern may be used but the focus lies more on polarization pattern sweeping.

[0285] In Example 23, the subject matter of Example 18 or Example 19 can optionally include: determining the propagation parameters based on polarization dynamics of the base station.

[0286] When base station, e.g. base station is also sweeping its RX polarization pattern at the same time, a same TX polarization pattern may be used but with the focus lying on more spatial beam pattern sweeping.

[0287] In Example 24, the subject matter of Example 18 or Example 19 can optionally include: determining the

propagation parameters based on a mobility status of the UE .

[0288] When UE is moving fast the focus will lie on more spatial beam pattern sweeping because polarization pattern optimization is less robust in high mobility.

[0289] In Example 25, the subject matter of Example 18 or Example 19 can optionally include: using the polarization control for single layer transmission mode in TX beamforming of the UE .

[0290] In Example 26, the subject matter of Example 18 or Example 19 can optionally include: aligning an angle between a polarization of the one or more TX beam sweeping signal bursts from the UE and a polarization of an RX beam from a serving base station.

[0291] In Example 27, the subject matter of Example 18 or Example 19 can optionally include: orthogonalizing an angle between a polarization of the one or more TX beam sweeping signal bursts from the UE and a polarization of an RX beam from a neighboring base station. [0292] In Example 28, the subject matter of Example 19 can optionally include: prioritizing TX polarization pattern sweeping over TX beam pattern sweeping upon a communication range to a serving base station going below a threshold.

[0293] In Example 29, the subject matter of Example 28 can optionally include: determining the communication range based on a path-loss or based on a timing advance in communications with the serving base station.

[0294] In Example 30, the subject matter of Example 19 can optionally include: prioritizing TX beam pattern sweeping over TX polarization pattern sweeping upon a mobility status metric of the UE being higher than a pre-defined threshold.

[0295] In Example 31, the subject matter of Example 19 can optionally include: prioritizing TX beam pattern sweeping over TX polarization pattern sweeping upon detecting that the serving base station also applies polarization control for uplink signal reception.

[0296] In Example 32, the subject matter of Example 31 can optionally include: detecting polarization control of the serving base station based on determining polarization angle of downlink reference signals from the serving base station or based on determining statistics of sounding reference signal, SRS, beam indications from the serving base station.

[0297] Example 33 is a User Equipment, UE, circuitry, comprising: a Beamforming Management, BM, circuitry,

configured to generate one or more TX beam sweeping signal bursts, each TX beam sweeping signal burst comprising a plurality of uplink beam sweeping reference signals, wherein a polarization pattern and a spatial beam pattern of the uplink beam sweeping reference signals is varied over the one or more beam sweeping signal bursts, wherein a variability of the polarization pattern and the spatial beam pattern of the uplink beam sweeping reference signals depends on propagation parameters .

[0298] In Example 34, the subject matter of Example 33 can optionally include that the variation is based on TX beam pattern sweeping or based on TX polarization pattern

sweeping .

[0299] In Example 35, the subject matter of Example 34 can optionally include: varying the uplink beam sweeping

reference signals based on the same polarization pattern but different spatial beam patterns.

[0300] In Example 36, the subject matter of Example 34 can optionally include: varying the uplink beam sweeping

[0301] In Example 37, the subject matter of Example 34 or Example 35 can optionally include that the UE circuitry is configured to determine the propagation parameters based on path-loss measurements on UE side.

[0302] In Example 38, the subject matter of Example 34 or Example 35 can optionally include that the UE circuitry is configured to determine the propagation parameters based on a mobility status of the UE . [0303] In Example 39, the subject matter of Example 34 or Example 35 can optionally include that the UE circuitry is configured to determine the propagation parameters based on polarization dynamics of the base station.

[0304] Example 40 is a device for polarization control in TX beamforming of a user equipment, UE, the device comprising: means for generating one or more TX beam sweeping signal bursts, each TX beam sweeping signal burst comprising a plurality of uplink beam sweeping reference signals, wherein a polarization pattern and a spatial beam pattern of the uplink beam sweeping reference signals is varied over the one or more beam sweeping signal bursts, and wherein a

[0305] In Example 41, the subject matter of Example 40 can optionally include that the variation is based on TX beam pattern sweeping or based on TX polarization pattern

sweeping .

[0306] Example 42 is a computer readable non-transitory medium on which computer instructions are stored which when executed by a computer cause the computer to perform the method of any one of Examples 18 to 32.

[0307] Example 43 is a method for user equipment, UE, TX beamforming based on UE beam correspondence, the method comprising: generating a set of candidate TX beam patterns for transmission to a base station; updating a size of the set of candidate TX beam patterns based on a quality metric with respect to a correspondence criterion indicating a beam correspondence of an RX beam pattern steered by the same UE, which receives downlink signals transmitted from the base station, with any of the candidate TX beam patterns from previous transmissions; and based on the updated size of the set of candidate TX beam patterns, updating a UE capability message indicating a number of TX beam sweeping resources supported by the UE .

[0308] Beam correspondence reflects the level of spatial overlapping between a TX beam pattern and an associated RX beam pattern steered by the same UE, not with a RX beam pattern from the base station. That is because the UE is receiving signals from the base station through UE's RX beam pattern. So, the pattern is not transmitted by the base station, only the signal that is received using the RX beam pattern was transmitted by the base station.

[0309] In Example 44, the subject matter of Example 43 can optionally include: updating the size of the set of candidate TX beam patterns based on learning of a beam correspondence metric, which is associated with each candidate TX beam pattern, wherein a beam correspondence metric reflects a level of spatial overlapping between a TX beam pattern and an associated RX beam pattern steered by the same UE .

[0310] In Example 45, the subject matter of Example 43 or Example 44 can optionally include: updating the size of the set of candidate TX beam patterns to include a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is higher than a pre-defined threshold; and updating the size of the set of candidate TX beam patterns to exclude a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is lower than a pre-defined threshold.

[0311] In Example 46, the subject matter of Example 44 can optionally include: selecting a number of candidate TX beam patterns based on a number of beamforming management, BM, sounding reference signal, SRS, resources within a triggered BM SRS sweeping burst; transmitting SRS resources, each associated to one of the candidate TX beam patterns from the set; receiving an SRS resource ID for a SRS resource selected by the base station; updating the beam correspondence metric of the candidate beam pattern associated to the SRS resource ID; and updating the set of candidate TX beam patterns upon the beam correspondence metric passing a threshold.

[0312] The beam correspondence metric can be updated, for example as follows: it can be just a counter ++ which counts the statistics that a same candidate TX beam pattern is selected after SRS sweeping. It can further be determined by the UL quality measurement after the selected candidate TX beam candidate is used for actual PUSCH transmission.

[0313] During SRS sweeping, UE transmits multiple SRS resources with mulitple SRS IDs. Then gNB down-selects the best received SRS resource and indicates back to UE the only one seleced SRS ID.

[0314] In Example 47, the subject matter of Example 46 can optionally include: updating the beam correspondence metric based on statistics of a same candidate TX beam pattern whose associated SRS ID has been selected by the base station through historical SRS sweeping events. [0315] In Example 48, the subject matter of Example 43 or Example 44 can optionally that an initial size of the set of candidate TX beam patterns and initial values of the quality metric are pre-determined based on RF Lab characterizations.

[0316] In Example 49, the subject matter of Example 46 can optionally include: updating the beam correspondence metric based on a SRS quality metric which is associated to a UE transmitted SRS resource and indicated back from the base station to the UE through higher layer messages.

[0317] As for an example, the beam correspondence metric can be updated based on a SRS quality metric, e.g. SRS SINR, which is associated to a UE transmitted SRS resource and indicated back from the base statoin to the UE through higher layer messages, e.g. media access control channel estimation (MAC CE) or radio resource control (RRC) .

[0318] In Example 50, the subject matter of Example 46 can optionally include: updating the beam correspondence metric based on based on the UL quality measurement in UE side (UL BLER and/or UL throughput) after the candidate TX beam pattern has been selected by the base statoin and applied by UE for UL data transmission.

[0319] In Example 51, the subject matter of Example 43 or Example 44 can optionally include: reducing a number of supported BM SRS resources upon detecting that a number of candidate TX beam patterns among all reference RX beam patterns is below the number of supported BM SRS resources.

[0320] In Example 52, the subject matter of Example 43 or Example 44 can optionally include: disabling the updating of the size of the set of candidate TX beam patterns upon detecting that a threshold number of RX beam patterns have been calibrated, each for its associated set of candidate TX beam patterns .

[0321] In Example 53, the subject matter of Example 52 can optionally include: enabling the updating of the size of the set of candidate TX beam patterns upon detecting uplink qualities falling below a threshold.

[0322] Example 54 is a User Equipment, UE, circuitry, comprising: a Beamforming Management, BM, circuitry,

configured to generate a set of candidate TX beam patterns for transmission to a base station, wherein the BM circuitry is configured to update a size of the set of candidate TX beam patterns based on a quality metric with respect to a correspondence criterion indicating a correspondence of an RX beam pattern steered by the same UE, which receives downlink signals transmitted from the base station, with any of the candidate TX beam patterns from previous transmissions, wherein the BM circuitry is configured, based on the updated size of the set of candidate TX beam patterns, to update a UE capability message indicating a number of TX beam sweeping resources supported by the UE .

[0323] In Example 55, the subject matter of Example 54 can optionally include that the BM circuitry is configured to update the size of the set of candidate TX beam patterns based on learning of a beam correspondence metric, which is associated with each candidate TX beam pattern, wherein a beam correspondence metric reflects a level of spatial overlapping between a TX beam pattern and an associated RX beam pattern steered by the same UE . [0324] In Example 56, the subject matter of Example 55 can optionally include that the BM circuitry is configured to update the size of the set of candidate TX beam patterns to include a candidate TX beam pattern if the beam

correspondence metric associated with the candidate beam pattern is higher than a pre-defined threshold; and that the BM circuitry is configured to update the size of the set of candidate TX beam patterns to exclude a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is lower than a pre-defined threshold.

[0325] In Example 57, the subject matter of Example 56 can optionally include that the BM circuitry is configured to: select a number of candidate TX beam patterns based on a number of beamforming management, BM, sounding reference signal, SRS, resources within a triggered BM SRS sweeping burst; transmit SRS resources, each associated to one of the candidate TX beam patterns from the set; receive an SRS resource ID for a SRS resource selected by the base station; update the beam correspondence metric of the candidate beam pattern associated to the SRS resource ID; and update the set of candidate TX beam patterns upon the beam correspondence metric passing a threshold.

[0326] Example 58 is a device for user equipment, UE, TX beamforming based on UE beam correspondence, the device comprising: means for generating a set of candidate TX beam patterns for transmission to a base station; means for updating a size of the set of candidate TX beam patterns based on a quality metric with respect to a correspondence criterion indicating a beam correspondence of an RX beam pattern steered by the same UE, which receives downlink signals transmitted from the base station, with any of the candidate TX beam patterns from previous transmissions; and means for updating a UE capability message indicating a number of TX beam sweeping resources supported by the UE, based on the updated size of the set of candidate TX beam patterns .

[0327] In Example 59, the subject matter of Example 58 can optionally include: means for updating the size of the set of candidate TX beam patterns based on learning of a beam correspondence metric, which is associated with each

candidate TX beam pattern, wherein a beam correspondence metric reflects a level of spatial overlapping between a TX beam pattern and an associated RX beam pattern steered by the same UE .

[0328] Example 60 is a computer readable non-transitory medium on which computer instructions are stored which when executed by a computer cause the computer to perform the method of any one of Examples 43 to 53.

[0329] Example 61 is a method for re-acquiring a beam pair between a user equipment, UE, and a base station based on grouping information, the method comprising: determining a TX beam configuration of the base station based on a selection of a beam pair between the UE and the base station for each of a plurality of UE positions towards the base station, each beam pair comprising an RX beam pattern of the UE and a TX beam pattern of the base station; determining grouping information of downlink reference signals, wherein downlink reference signals within a group are associated to a subset of at least two TX beam patterns of the base station; and re- acquiring a beam pair between the UE and the base station based on the determined grouping information.

[0330] In Example 62, the subject matter of Example 61 can optionally include: determining the grouping information based on determining a spatial adjacent metric among

different TX beam patterns of the base station over time.

[0331] The grouping information indicates a neighborhood relationship between the TX beam patterns of the base station. The grouping information is cell specific and unchanged with respect to UE mobility.

[0332] In Example 63, the subject matter of Example 61 or Example 62 can optionally include: determining the grouping information based on a database comprising grouping

information of multiple base stations, wherein each grouping information is associated with a base station identity and a geographical region identity.

[0333] In Example 64, the subject matter of Example 61 or Example 62 can optionally include: determining the grouping information based on messages from a serving base station.

[0334] In Example 65, the subject matter of Example 61 can optionally include: re-acquiring the beam pair based on measuring different downlink reference signals within a group by using different UE RX beam patterns for fast acquisition.

[0335] In Example 66, the subject matter of Example 61 can optionally include: re-acquiring the beam pair based on measuring different downlink reference signals within a group by using a single UE RX beam pattern; and combining results of the measurements for high accuracy.

[0336] In Example 67, the subject matter of Example 65 or Example 66 can optionally include: selecting a re-acquiring mode of the beam pair based on SNR conditions, UE mobility and/or UE battery status.

[0337] In Example 68, the subject matter of Example 62 can optionally include: determining the spatial adjacent metric based on a difference of estimated channel parameters from different TX beam patterns of the base station, received using a same UE RX beam pattern.

[0338] In Example 69, the subject matter of Example 68 can optionally include: determining the estimated channel parameters based on Reference Signal Received Power, RSRP, measurements, delay spread measurements and/or channel transfer functions of the different TX beam patterns.

[0339] In Example 70, the subject matter of Example 61 or Example 62 can optionally include: determining the grouping information based on determining relative angle of arrivals, AoAs, of the different TX beam patterns.

[0340] In Example 71, the subject matter of Example 70 can optionally include: determining the relative AoAs based on determining a sub-sampled angle profile for the different TX beam patterns by scanning narrow UE RX beams with pre-known radiation patterns.

[0341] In Example 72, the subject matter of Example 71 can optionally include: interpolating the sub-sampled angle profile and identifying for each of the different TX beam patterns an angle having a peak RSRP.

[0342] In Example 73, the subject matter of Example 72 can optionally include: determining the relative AoAs of the different TX beam patterns based on a difference of angles which are associated with the peak RSRPs .

[0343] In Example 74, the subject matter of Example 61 or Example 62 can optionally include that the downlink reference signals are arranged in Synchronization Signal Blocks, SSBs, having SSB indices, each SSB index associated to a different TX beam pattern of the base station, wherein a number of SSBs with different SSB indices form an SSB burst.

[0344] Example 75 is a User Equipment, UE, circuitry, comprising: a Beamforming Management, BM, circuitry,

configured to determine a TX beam configuration of a base station based on a selection of a beam pair between the UE and the base station for each of a plurality of UE positions towards the base station, each beam pair comprising an RX beam pattern of the UE and a TX beam pattern of the base station, wherein the BM circuitry is configured to determine grouping information of downlink reference signals, wherein downlink reference signals within a group are associated to a subset of at least two TX beam patterns of the base station, wherein the BM circuitry is configured to re-acquire a beam pair between the UE and the base station based on the

determined grouping information.

[0345] In Example 76, the subject matter of Example 75 can optionally include that the BM circuitry is configured to determine the grouping information based on determining a spatial adjacent metric among different TX beam patterns of the base station over time.

[0346] In Example 77, the subject matter of Example 75 or Example 76 can optionally include that the BM circuitry is configured to determine the grouping information based on a database comprising grouping information of multiple base stations, wherein each grouping information is associated with a base station identity and a geographical region identity .

[0347] In Example 78, the subject matter of Example 75 or Example 76 can optionally include that the BM circuitry is configured to determine the grouping information based on messages from a serving base station.

[0348] In Example 79, the subject matter of Example 75 or Example 76 can optionally include that the BM circuitry is configured to re-acquire the beam pair based on measuring different downlink reference signals within a group by using a single UE RX beam pattern and to combine results of the measurements for high accuracy.

[0349] In Example 80, the subject matter of Example 75 or Example 76 can optionally include that the BM circuitry is configured to re-acquire the beam pair based on measuring different downlink reference signals within a group by using different UE RX beam patterns for fast acquisition.

[0350] In Example 81, the subject matter of Example 75 or Example 76 can optionally include that the BM circuitry is configured to re-acquire the beam pair based on SNR

conditions, UE mobility and/or UE battery status. [0351] Example 82 is a device for re-acquiring a beam pair between a user equipment, UE, and a base station based on grouping information, the device comprising: means for determining a TX beam configuration of the base station based on a selection of a beam pair between the UE and the base station for each of a plurality of UE positions towards the base station, each beam pair comprising an RX beam pattern of the UE and a TX beam pattern of the base station; means for determining grouping information of downlink reference signals, wherein downlink reference signals within a group are associated to a subset of at least two TX beam patterns of the base station; and means for re-acquiring a beam pair between the UE and the base station based on the determined grouping information.

[0352] In Example 83, the subject matter of Example 82 can optionally include: means for determining the grouping information based on determining a spatial adjacent metric among different TX beam patterns of the base station over time .

[0353] Example 84 is a computer readable non-transitory medium on which computer instructions are stored which when executed by a computer cause the computer to perform the method of any one of Examples 61 to 74.

[0354] In addition, while a particular feature or aspect of the disclosure may have been disclosed with respect to only one of several implementations, such feature or aspect may be combined with one or more other features or aspects of the other implementations as may be desired and advantageous for any given or particular application. Furthermore, to the extent that the terms "include", "have", "with", or other variants thereof are used in either the detailed description or the claims, such terms are intended to be inclusive in a manner similar to the term "comprise". Furthermore, it is understood that aspects of the disclosure may be implemented in discrete circuits, partially integrated circuits or fully integrated circuits or programming means. Also, the terms "exemplary", "for example" and "e.g." are merely meant as an example, rather than the best or optimal.

[0355] Although specific aspects have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that a variety of alternate and/or equivalent implementations may be substituted for the specific aspects shown and described without departing from the scope of the present disclosure. This application is intended to cover any adaptations or variations of the specific aspects discussed herein.

[0356] Although the elements in the following claims are recited in a particular sequence with corresponding labeling, unless the claim recitations otherwise imply a particular sequence for implementing some or all of those elements, those elements are not necessarily intended to be limited to being implemented in that particular sequence.

Claims

1. A method for determining an RX beam pattern for a user equipment, UE, the method comprising:

receiving downlink signals from a base station for UE RX beam sweeping;

determining a UE RX beam pattern based on a selection from candidate beams formed on the basis of the downlink signals; and

upon detection of a change of an antenna radiation condition, determining the UE RX beam pattern based on a selection from a set of UE RX candidate beam patterns formed on the basis of further downlink signals carrying pre-known information,

wherein the further downlink signals are received from the base station upon request of the UE or another UE .

2. The method of claim 1,

wherein the further downlink signals comprise a Physical Downlink Control Channel, PDCCH, and an associated Physical Downlink Shared Channel, PDSCH, wherein the PDSCH comprises: a System Information Block, SIB, which is associated to an on-demand SIB request from the UE or further UEs; or

a re-transmitted Downlink Shared Channel, DL-SCH, data which is associated to a Downlink, DL, re-transmission request by the UE .

3. The method of claim 2, comprising:

sending a request for a further downlink signal to the base station; or

monitoring a request for a further downlink signal sent by other UEs without sending an own request to the base station, wherein the request comprises the on-demand SIB request even though the UE has already decoded the same SIB; or

wherein the request comprises the DL re-transmission request even though the UE has already successfully decoded a first transmission of the same DL-SCH, data.

4. The method of claim 1 or 2, comprising:

determining the UE RX beam based on measuring UE RX candidate beam patterns using PDSCH Demodulate Reference Signals, DMRS, and/or PDSCH data symbols of the further downlink signals.

5. The method of claim 1 or 2, comprising:

decoding a Physical Downlink Control Channel, PDCCH, within the further downlink signals based on an activated UE RX beam to obtain a time and frequency allocation of a PDSCH region of a SIB or re-transmitted DL data; and

measuring the set of UE RX candidate beam patterns during the PDSCH region of the further downlink signals.

6. The method of claim 1 or 2, comprising:

detecting the change of antenna radiation condition based on detection of a UE antenna position change event.

7. A method for polarization control in TX beamforming of a user equipment, UE, the method comprising:

generating one or more TX beam sweeping signal bursts, each TX beam sweeping signal burst comprising a plurality of uplink beam sweeping reference signals,

wherein a polarization pattern and a spatial beam pattern of the uplink beam sweeping reference signals is varied over the one or more beam sweeping signal bursts, and wherein a variability of the polarization pattern and the spatial beam pattern of the uplink beam sweeping

reference signals depends on propagation parameters.

8. The method of claim 7, wherein the variation is based on TX beam pattern sweeping or based on TX polarization pattern sweeping .

9. The method of claim 8, wherein TX beam pattern sweeping comprises :

varying the uplink beam sweeping reference signals based on the same polarization pattern but different spatial beam patterns .

10. The method of claim 8, wherein TX polarization pattern sweeping comprises:

varying the uplink beam sweeping reference signals based on different polarization patterns but with the same spatial beam pattern.

11. The method of claim 7 or 8, comprising:

determining the propagation parameters based on path- loss measurements on UE side.

12. The method of claim 7 or 8, comprising:

determining the propagation parameters based on

polarization dynamics of the base station.

13. A method for user equipment, UE, TX beamforming based on UE beam correspondence, the method comprising:

generating a set of candidate TX beam patterns for transmission to a base station; updating a size of the set of candidate TX beam patterns based on a quality metric with respect to a correspondence criterion indicating a beam correspondence of an RX beam pattern steered by the same UE, which receives downlink signals transmitted from the base station, with any of the candidate TX beam patterns from previous transmissions; and based on the updated size of the set of candidate TX beam patterns, updating a UE capability message indicating a number of TX beam sweeping resources supported by the UE .

14. The method of claim 13, comprising:

updating the size of the set of candidate TX beam patterns based on learning of a beam correspondence metric, which is associated with each candidate TX beam pattern,

wherein a beam correspondence metric reflects a level of spatial overlapping between a TX beam pattern and an

associated RX beam pattern steered by the same UE .

15. The method of claim 14, comprising:

updating the size of the set of candidate TX beam patterns to include a candidate TX beam pattern if the beam

correspondence metric associated with the candidate beam pattern is higher than a pre-defined threshold; and

updating the size of the set of candidate TX beam patterns to exclude a candidate TX beam pattern if the beam correspondence metric associated with the candidate beam pattern is lower than a pre-defined threshold.

16. The method of claim 14, comprising:

selecting a number of candidate TX beam patterns based on a number of beamforming management, BM, sounding reference signal, SRS, resources within a triggered BM SRS sweeping burst ; transmitting SRS resources, each associated to one of the candidate TX beam patterns from the set;

receiving an SRS resource ID for a SRS resource selected by the base station; and

updating the beam correspondence metric of the candidate beam pattern associated to the SRS resource ID; and

updating the set of candidate TX beam patterns upon the beam correspondence metric passing a threshold.

17. The method of claim 16, comprising:

updating the beam correspondence metric based on

statistics of a same candidate TX beam pattern whose

associated SRS ID has been selected by the base station through historical SRS sweeping events.

18. The method of claim 13 or 14,

wherein an initial size of the set of candidate TX beam patterns and initial values of the quality metric are pre determined based on RF Lab characterizations.

19. A method for re-acquiring a beam pair between a user equipment, UE, and a base station based on grouping

information, the method comprising:

determining a TX beam configuration of the base station based on a selection of a beam pair between the UE and the base station for each of a plurality of UE positions towards the base station, each beam pair comprising an RX beam pattern of the UE and a TX beam pattern of the base station; determining grouping information of downlink reference signals, wherein downlink reference signals within a group are associated to a subset of at least two TX beam patterns of the base station; and re-acquiring a beam pair between the UE and the base station based on the determined grouping information.

20. The method of claim 19, comprising:

determining the grouping information based on

determining a spatial adjacent metric among different TX beam patterns of the base station over time.

21. The method of claim 19 or 20, comprising:

determining the grouping information based on a database comprising grouping information of multiple base stations, wherein each grouping information is associated with a base station identity and a geographical region identity.

22. The method of claim 19 or 20, comprising:

determining the grouping information based on messages from a serving base station.

23. The method of claim 19, comprising:

re-acquiring the beam pair based on measuring different downlink reference signals within a group by using different UE RX beam patterns for fast acquisition.

24. The method of claim 19, comprising:

re-acquiring the beam pair based on measuring different downlink reference signals within a group by using a single UE RX beam pattern; and

combining results of the measurements for high accuracy.

25. A computer readable non-transitory medium on which computer instructions are stored which when executed by a computer cause the computer to perform the method of any one of claims 1 to 24.