JP4077830B2 - Wireless device and adaptive array processing method - Google Patents

Description

  The present invention relates to a configuration of a radio apparatus used for a base station in radio communication mainly for a mobile body such as a mobile phone.

  2. Description of the Related Art In recent years, mobile communication systems (for example, Personal Handyphone System: hereinafter referred to as PHS), which are rapidly developing, have an array antenna composed of a plurality of antennas in order to suppress the influence of interference waves and obtain good communication quality. An adaptive array base station that separates and extracts a signal of a desired wave by performing known adaptive array processing on a received signal has been put into practical use.

  Furthermore, if such an adaptive array base station is used, in order to improve the frequency utilization efficiency of radio waves, the same time slot of the same frequency is spatially divided so that a mobile terminal device of multiple users is path-multiplexed to the radio base system. It is possible to realize a PDMA (Path Division Multiple Access) system that can be connected. The PDMA system is also called an SDMA system (Space Division Multiple Access).

  FIG. 11 shows channel configurations in various communication systems such as frequency division multiple access (FDMA), time division multiple access (TDMA), and space division multiple access (SDMA). FIG.

  First, FDMA, TDMA, and SDMA will be briefly described with reference to FIG. FIG. 11A is a diagram showing FDMA, in which analog signals of users 1 to 4 are transmitted by being divided by radio waves of different frequencies f1 to f4, and the signals of users 1 to 4 are separated by a frequency filter. The

  In the TDMA shown in FIG. 11 (b), the digitized signal of each user is transmitted in radio waves of different frequencies f1 to f4 and time-divisionally transmitted at fixed time intervals (time slots). The signal is separated by the frequency filter and time synchronization between the base station and each user mobile terminal device.

  On the other hand, in the SDMA system, as shown in FIG. 11C, one time slot at the same frequency is spatially divided to transmit data of a plurality of users. In this SDMA, the signal of each user is separated using a frequency filter, time synchronization between the base station and each user mobile terminal device, and a mutual interference canceller such as an adaptive array.

  FIG. 12 is a schematic block diagram showing a configuration of a transmission / reception system 2000 of a conventional SDMA base station.

  In the configuration shown in FIG. 12, for example, n antennas # 1 to #n (n: natural number) are provided to identify users PS1 and PS2.

  In the reception operation, the output of the antenna is given to the RF circuit 2101. In the RF circuit 2101, after being amplified by the reception amplifier and frequency-converted by the local oscillation signal, unnecessary frequency signals are removed by the filter, and A / D-converted and supplied to the digital signal processor 2102 as a digital signal.

  The digital signal processor 2102 is provided with a channel assignment reference calculator 2103, a channel assigner 2104, and an adaptive array 2100. The channel assignment criterion calculator 2103 pre-calculates whether the signals from the two users can be separated by the adaptive array. In response to the calculation result, the channel allocator 2104 provides the adaptive array 2100 with channel allocation information including user information for selecting a frequency and time. Adaptive array 2100 separates only a specific user's signal by performing a weighting operation in real time on the signals from antennas # 1 to #n based on the channel assignment information.

[Configuration of adaptive array antenna]
FIG. 13 is a block diagram illustrating a configuration of the transmission / reception unit 2100a corresponding to one user in the adaptive array 2100. In the example shown in FIG. 13, n input ports 2020-1 to 2020-receiving signals from antennas # 1 to #n in order to extract a desired user signal from input signals including a plurality of user signals. n is provided.

  Signals input to the input ports 2020-1 to 2020-n are given to the weight vector calculator 2011 and the multipliers 2012-1 to 2012-n via the switch circuits 2010-1 to 2010-n.

The weight vector calculator 2011 calculates the weight vectors w _{1i to} w _ni using the input signal, the unique word signal that is a reference signal stored in advance in the memory 2014, and the output of the adder 2013. Here, the subscript i indicates a weight vector used for transmission / reception with the i-th user. Therefore, the unique word signal is a training signal for adaptive array processing.

Multipliers 2012-1 to 2012-n multiply the input signals from the input ports 2020-1 to 2020-n and the weight vectors w _{1i to} w _ni , respectively, and supply the result to the adder 2013. The adder 2013 adds the output signals of the multipliers 2012-1 to 2012-n and outputs the result as a received signal S _RX (t). The received signal S _RX (t) is also given to the weight vector calculator 2011. .

Further, the transceiver unit 2100a receives the output signal S _TX (t) from the adaptive array radio base station, multiplies the weight vectors w _{1i to} w _ni given by the weight vector calculator 2011, and outputs the multipliers 2015- 1-2015-n. Outputs of the multipliers 2015-1 to 2015-n are respectively supplied to switch circuits 2010-1 to 2010-n. That is, when the switch circuits 2010-1 to 2010-n receive a signal, the signals given from the input ports 2020-1 to 2020-n are given to the signal receiving unit 1R and the signals are transmitted. The signal from the signal transmission unit 1T is given to the input / output ports 2020-1 to 2020-n.

[Operation principle of adaptive array]
Next, the operation principle of the transmission / reception unit 2100a shown in FIG. 13 will be briefly described.

  In the following description, in order to simplify the description using mathematical formulas, the number of antenna elements is four, and the number of users PS simultaneously communicating is two. At this time, a signal given from each antenna to the receiving unit 1R is expressed by the following equation.

Here, the signal RX _j (t) indicates the received signal of the j-th (j = 1, 2, 3, 4) antenna, and the signal Srx _i (t) is the i-th (i = 1, 2). Indicates a signal transmitted by the user.

Further, the coefficient h _ji indicates a complex coefficient of the signal from the i th user received by the j th antenna, and n _j (t) indicates noise included in the j th received signal.

  The above equations (1) to (4) are expressed in the vector format as follows.

In the formulas (6) to (8), [...] ^T represents transposition of [...].
Here, X (t) represents an input signal vector, _Hi represents a reception response vector of the i-th user, and N (t) represents a noise vector.

As shown in FIG. 13, the adaptive array antenna outputs a signal synthesized by multiplying input signals from the respective antennas by weighting factors w _{1i to} w _4i as a received signal S _RX (t).

With the above preparation, for example, the operation of the adaptive array when extracting the signal Srx ₁ (t) transmitted by the _first user is as follows.

The output signal y1 of the adaptive array 2100 (t) is the multiplication of the input signal vector X (t) and weight vector W ₁ vector can be represented by the following formula.

That is, the weight vector W ₁ is a vector whose elements are weighting factors w _j1 (j = 1, 2, 3, 4) multiplied by the _jth input signal RX _j (t).

  Here, when the input signal vector X (t) expressed by the equation (5) is substituted for y1 (t) expressed by the equation (9), the result is as follows.

Here, if the adaptive array 2100 is operating ideally, by well known methods, weight vector W ₁ is sequentially controlled by weight vector calculator 2011 so as to satisfy the following simultaneous equations.

When the weight vector W ₁ is completely controlled so as to satisfy the expressions (12) and (13), the output signal y1 (t) from the adaptive array 2100 is eventually expressed as the following expression.

That is, the signal Srx ₁ (t) transmitted by the first user of the two users is obtained as the output signal y1 (t).

On the other hand, in FIG. 13, an input signal S _TX (t) to the adaptive array 2100 is given to the transmission unit 1T in the adaptive array 2100, and multipliers 2015-1, 2015-2, 2015-3,. To one input. The weight vectors w _1i , w _2i , w _3i ,..., W _ni calculated based on the received signal as described above by the weight vector calculator 2011 are copied and applied to the other inputs of these multipliers. Is done.

  The input signals weighted by these multipliers correspond to the corresponding antennas # 1, # 2, # 3,..., Through the corresponding switches 2010-1, 2010-2, 2010-3,. Sent to #n and transmitted.

  Here, the identification of the users PS1 and PS2 is performed as described below. That is, the radio signal of the mobile phone is transmitted in a frame configuration. The radio signal of a mobile phone is mainly composed of a preamble made up of a signal sequence known to the radio base station and data (speech etc.) made up of a signal sequence unknown to the radio base station.

  The preamble signal sequence includes a signal sequence (unique word signal) of information for identifying whether or not the user is a desired user to communicate with the radio base station. The weight vector calculator 2011 of the adaptive array radio base station 1 compares the unique word signal extracted from the memory 2014 with the received signal sequence, and extracts a signal that seems to include a signal sequence corresponding to the user PS1. Thus, the weight vector control (determination of the weighting factor) is performed.

  In the above description, at the time of signal transmission, the weight vector at the time of reception is copied to form the directivity of the transmission signal. At the time of transmission, the moving speed of the terminal device is taken into consideration at the time of reception. Alternatively, the weight vector corrected may be used as a transmission weight vector.

  On the other hand, an orthogonal frequency division multiplexing (OFDM) system is known as a communication system with high frequency utilization efficiency.

  The OFDM system is a type of multi-carrier modulation in which modulation is performed by dispersing 1-channel data over a plurality of carriers. In the OFDM system, the frequency spectrum of a signal used for communication has a shape close to a rectangle.

  FIG. 14 is a diagram showing three carriers extracted from the frequency spectrum of a plurality of carriers (carrier waves) used in such an OFDM system.

  In the OFDM system, as shown in FIG. 14, when attention is paid to the spectrum of one carrier, the frequency interval of a plurality of carriers is set so that the zero point of the spectrum of this carrier coincides with the frequency of an adjacent carrier. Yes. In other words, the carriers are arranged at frequencies that do not interfere with each other, and the carriers are orthogonal to each other.

  Here, the frequency interval Δf of each carrier wave is given by the following equation, where Ts is the duration of one symbol of data to be transmitted.

Δf = 1 / Ts × n (n: natural number)
FIG. 15 is a diagram showing a waveform of a transmission symbol transmitted by such an OFDM method.

  As a result of synthesizing the waveforms of N carriers from i = 1 to i = N, a signal as shown by the bottom waveform in FIG. 15 is used as an OFDM transmission symbol.

  In OFDM modulation, inverse discrete Fourier transform is performed on a baseband signal in order to obtain each carrier component. Correspondingly, in the demodulation process of the received wave, the discrete Fourier transform is performed on the received signal by a so-called fast Fourier transform (FFT) algorithm.

  Here, in FIG. 15, a “guard interval” is provided before the effective symbol period in the OFDM signal waveform. As such a guard interval, a part of an effective symbol waveform, for example, a signal of a predetermined period Tg at the end of the effective symbol waveform is copied and added.

  Such a guard interval is provided as a countermeasure against an interference wave signal generated by multipath interference.

  When a desired wave and an interference wave that arrives with a time delay are combined into a received signal, if the delay time of the interference wave is within the time set as the guard interval, the influence of the interference wave is limited within the guard interval period. . If the guard interval period is set longer than the expected delay time of the interference wave, it is possible to demodulate by eliminating the influence of the interference wave as described below.

  FIG. 16 is a conceptual diagram for explaining a demodulation operation when such a desired wave and an interference wave are received.

  In the demodulation of the OFDM system, as shown in FIG. 16, a time window called an FFT window is provided in each symbol period. This time window indicates a section in which only the effective symbol section is cut out of the received OFDM transmission symbols. Here, such an FFT window is made equal to the effective symbol period length Ts. Further, as described above, the guard interval period is set longer than the delay time of the interference wave. In this way, even when an interference wave is present, the signals present in the guard interval period are signals within the same effective symbol, so that the orthogonality of each carrier wave of the received wave can be maintained. It becomes possible. Therefore, the receiving side can perform demodulation without the influence of such interference waves.

  Therefore, combining the above-described adaptive array system and such an OFDM system is expected to realize higher communication quality and a reception system with higher frequency utilization efficiency.

  However, simply combining these two methods has the following problems.

[Problems of configuring different adaptive arrays for each carrier]
Hereinafter, a first configuration example for performing OFDM transmission using an adaptive array will be described.

  By using such a configuration, it is also possible to perform the SDMA multiple connection as described above by applying the adaptive array technology.

  FIG. 17 is a schematic block diagram for explaining the configuration of such an adaptive array base station 3000.

  Referring to FIG. 17, adaptive array base station 3000 is assumed to perform transmission / reception using an adaptive array antenna having four antennas # 1 to # 4 for the sake of simplicity. In FIG. 17, the configuration for performing reception is extracted from the configuration of the adaptive array base station.

  Referring to FIG. 17, adaptive array base station 3000 receives signals from adaptive array antennas # 1 to # 4, and performs A / D conversion unit 3010 for performing detection and analog / digital conversion, and A / D An FFT unit 3020 is provided for receiving a digital signal output from the conversion unit 3010, performing a fast Fourier transform, and separating a signal for each carrier wave.

  Here, among the signals output from the FFT unit 3020, the signal from the i-th antenna for the l-th carrier is hereinafter expressed as a signal fl, i (l, i: natural number).

  Adaptive array base station 3000 is further provided for each carrier, and each receives a corresponding carrier component obtained by subjecting signals from antennas # 1 to # 4 to Fourier transform by FFT section 3020. , N (N: total number of carriers) adaptive array blocks 3030.1 to 3030. N is provided.

  However, in FIG. 17, the adaptive array block 3030. Only l is shown.

  Adaptive array block 3030. In the same manner as the adaptive array base station shown in FIG. 13, l is a reception weight vector calculator 3041 for receiving signals fl, 1 to fl, 4 and calculating reception weight vectors, and signals fl, 1 to fl, 4 are respectively received at one input, and the outputs of the multipliers 3042-1 to 3042-4 and the multipliers 3042-1 to 3042-4 that receive the reception weight vectors from the reception weight vector calculator 3041 are respectively received at the other inputs. An adder 3043 for combining, and a memory 3044 for storing in advance a unique word signal (reference signal) used in the calculation of adaptive array processing in the reception weight vector calculator 3041 are provided. The adder 3043 outputs a desired signal S1 (t) for the carrier l, and this desired signal S1 (t) is also given to the reception weight vector calculator 3041.

  With such a configuration, it is possible to receive a signal transmitted from the desired user by the adaptive array base station by using the OFDM transmission scheme, and to separate and receive the signal for each carrier by adaptive array processing.

  However, the configuration of the adaptive array base station 3000 has the following problems.

  As described above, in the OFDM system, a signal for one channel is distributed and transmitted over many carriers.

  Therefore, in general, in a signal transmitted by the OFDM method, the number of reference signal symbols included in each carrier is often not sufficient. For example, in “Multimedia Mobile Access Communication systems (MMAC)” promoted by the Ministry of Internal Affairs and Communications, etc., a reference signal for each OFDM carrier (subcarrier) is defined as two symbols.

  In such a case, with the configuration of the adaptive array base station 3000 as shown in FIG. 17, there is a problem that it is difficult to converge the weights, and it is impossible to form highly accurate directivity.

  Furthermore, the configuration of the adaptive array base station 3000 shown in FIG. 17 has the following problems.

  FIG. 18 is a conceptual diagram showing timing of signals received by adaptive array base station 3000 shown in FIG.

  In FIG. 18, the portion indicated by “G” in the received signal represents the guard interval period as described above.

  Further, the original desired wave is generally a signal that first arrives at the base station, and the signal that arrives first is hereinafter referred to as a “first arrival signal”.

  Also, a signal that arrives at this head arrival signal with a delay time within the guard interval period due to the influence of the multipath is called a “short delay signal”. A signal that arrives with a delay as described above will be referred to as a “long delay signal”. Further, such a path through which the head arrival signal, the short delay signal, and the long delay signal are transmitted is referred to as a “path”.

  Also, in FIG. 18, adaptive array block 3030. In l, the timing for sampling a signal is indicated by an arrow.

  In adaptive array base station 3000, adaptive array processing is performed on the signal after being divided for each carrier, so that the sampling timing is also sufficient to extract the signal waveform for each carrier. Any interval may be used.

  By performing adaptive array processing, it is possible to remove a long delay signal as shown in FIG.

  On the other hand, since the bandwidth of the band-divided carrier is so narrow that the short delay signal cannot be separated, the head arrival signal and the short delay signal are processed in the adaptive array processing as the same signal.

  FIG. 19 is a diagram showing a signal intensity distribution corresponding to each carrier after passing through such an adaptive array.

  In FIG. 19, at each of the frequencies f1 to fN of each carrier, the spectrum of the head arrival signal (“head wave”) and the spectrum of the short delay signal (“short delay wave”) are adaptive array processing as described above. Later, it looks like the same signal. However, since the bandwidth for all carriers is very wide, for example, in the carrier indicated by the arrow in FIG. 19, there may be a case where the leading wave and the short delay wave are in opposite phases.

  FIG. 20 is a diagram showing an intensity distribution when signals for each carrier are combined in the case as shown in FIG.

  When adaptive array reception is performed using a reference signal whose timing is aligned with the leading wave, only a signal with a low level can be extracted for a carrier having a frequency in which the leading signal and the short delay signal are in opposite phases. . In other words, when adaptive array reception is performed for each carrier, as shown in FIG. 19, only signals with extremely low signal levels are extracted for carriers having frequencies in which the leading wave and the short delay wave are in opposite phases. Will not be able to.

  For this reason, since the carrier indicated by the arrow in FIG. 19 cannot perform sufficient signal transmission, it is necessary to perform control such as using a redundant code or performing communication without using this carrier. End up. In the latter case, this is equivalent to removing the signal originally arriving at the base station as a short delay signal as an unnecessary signal, resulting in a decrease in reception sensitivity.

  Therefore, in summary, in the case of a configuration in which a different adaptive array is operated for each carrier as shown in FIG. 17, first, it is difficult to secure a sufficient reference signal for accurate directivity control. There is a problem that.

  Furthermore, there is a problem that the reception sensitivity is lowered because the maximum ratio combining of multipath signals within the guard interval cannot be performed.

  In other words, a signal with a delay time within the guard interval (short delay component) has a high correlation with the head signal. Therefore, when adaptive array synthesis is performed using a reference signal whose timing is aligned with the head signal, a short delay component is present in the array synthesis output. Will be included. However, in the OFDM transmission method, when a plurality of carriers used for communication are distributed in a very wide band, the phase of the leading wave and the short delay wave may be opposite depending on the carrier. In this case, there arises a problem that the maximum ratio synthesis is not performed in the entire carrier.

[Problems of Adaptive Array Operation with Common Weight for All Carriers]
Since the configuration of adaptive array base station 3000 has the above-described problems, another configuration is also possible in which adaptive array processing is performed on a signal before being subjected to band division by FFT processing.

  FIG. 21 is a schematic block diagram for explaining the configuration of an adaptive array base station 4000 that operates the adaptive array by calculating weights common to all such carriers.

Referring to FIG. 21, adaptive array base station 4000 receives signals from four antennas # 1 to # 4 in the same manner as adaptive array base station 3000 shown in FIG. An A / D conversion unit 4010 for receiving the signal, a reception weight vector calculation unit 4041 for calculating a reception weight vector for the signal for each antenna in response to the output of the A / D conversion unit 4010, Receiving signals from the array antenna and receiving outputs from the multipliers 4042-1 to 4042-4 and the multipliers 4042-1 to 4042-4 receiving the weight vectors from the reception weight vector calculator 4041 at the other input, and combining them When the adder 4043 and the reception weight vector calculator 4041 calculate the weight vector A memory 4044 for storing a reference signal used in advance, performs fast Fourier transform processing by receiving the output of the adder 4043, a desired wave signal _{S 1 (t) ~S N (} t) of each carrier And an FFT unit 4050 for separation. The output of the adder 4043 is given to the reception weight vector calculator 4041 and used for calculation of the reception weight vector.

  FIG. 22 is a conceptual diagram for explaining the operation of adaptive array base station 4000 shown in FIG.

  Also in FIG. 22, “G” indicates a guard interval period. Further, in order to perform adaptive array processing on the signal before band division, in the adaptive array, for example, the sampling timing of the reception weight vector calculator 4041 is compared with that for the signal after band division as shown in FIG. It is necessary to perform sampling in a shorter period.

  In this case as well, the long delay signal can be removed by adaptive array processing by the adaptive array block.

  On the other hand, since the signal input to the adaptive array block is not band-divided, it becomes a signal having a very wide band. That is, the reception weight vector calculator 4041 recognizes the head arrival signal and the short delay signal as completely different signals. Therefore, such a short delay signal is also removed by the adaptive array processing.

  When such an operation is performed, the short delay signal itself is actually a desired wave, and if it is effectively used, its characteristics are expected to be improved. Since it is removed by adaptive array processing, there is a problem in that communication quality deteriorates.

  Furthermore, since the short delay signal is also regarded as an interference signal, when viewed from the adaptive array base station 4000, it appears that a large number of interference waves have arrived. If directivity is formed by an adaptive array to remove these signals, the degree of freedom of the antenna may be used up.

  As described above, when the degree of freedom of the antenna has been used up, there is a problem that the gain in the desired wave direction is reduced, or interference exceeding the degree of freedom of the antenna is seen, so that all interference cannot be removed. .

  The present invention has been made to solve the above-described problems, and its purpose is to provide a multipath signal within a guard interval even when adaptive array reception is performed for an OFDM transmission system. Is to provide an adaptive array base station capable of improving the reception sensitivity by combining the maximum ratio.

  Another object of the present invention is to provide an adaptive array base station capable of maintaining interference suppression performance without consuming antenna degrees of freedom when combining multipath signals within a guard interval period. It is to be.

  In summary, the present invention provides a radio apparatus for transmitting / receiving a signal transmitted by an orthogonal frequency division communication system using a plurality of carriers, comprising: an array antenna having a plurality of antennas; and a first signal for a desired wave signal. Received response vector estimating means for estimating a response vector, first Fourier transform means for extracting a component for each of a plurality of carriers by Fourier transforming the first response vector, and Fourier transforming a received signal from the array antenna A second Fourier transform means for extracting a component for each carrier of the received signal for each antenna and a plurality of carriers, each of which corresponds to a corresponding carrier among the components for the carrier of the received signal for each antenna. The component is received from the second Fourier transform means, and the corresponding carrier component in the desired wave is extracted. Adaptive array processing means, wherein the adaptive array processing means extracts a corresponding carrier component based on a component of the first response vector from the first Fourier transform means for the corresponding carrier. Is derived.

  Preferably, the radio apparatus detects the arrival timing of the desired wave from the signal received by the array antenna, and corresponds to the reception signal and the plurality of carriers before being Fourier-transformed by the second Fourier transform unit for each antenna. It further includes arrival timing detection means for detecting a desired wave when the cross-correlation with the reference signal including the training signal component exceeds a predetermined threshold.

  Preferably, the radio apparatus further includes arrival timing detection means for detecting arrival timing of a desired wave from a signal received by the array antenna, and the reception response vector estimation means is the arrival timing detected by the arrival timing detection means. The level of the response in the first response vector at a time other than is 0.

  Preferably, the radio apparatus further includes arrival timing detection means for detecting an arrival timing of a desired wave from a signal received by the array antenna, and the radio apparatus receives a signal in which a guard interval section is added to an effective symbol section. The reception response vector estimation means sets the response level in the first response vector to 0 at the time after the guard interval.

  Preferably, the wireless device further includes arrival timing detection means for detecting arrival timing of a desired wave from a signal received by the array antenna, wherein the adaptive array processing means is a component for the corresponding carrier of the first response vector. A weight vector for extracting a desired wave for the corresponding carrier is derived from the correlation matrix for each carrier derived based on.

  Preferably, the arrival timing detection means further detects the arrival timing of n interference waves (n: natural number, n ≧ 1) from the signal received by the array antenna, and the reception response vector estimation means For each of the interference waves, the second to (n + 1) th response vectors for each signal are estimated, and the first Fourier transform means further Fourier transforms the second to (n + 1) th response vectors to obtain a plurality of response vectors. And the adaptive array processing means extracts the corresponding carrier components based on the corresponding carrier components of the first to (n + 1) th response vectors from the first Fourier transform means. A weight vector for extraction is derived.

  Furthermore, it is preferable that the arrival timing detection means has a predetermined cross-correlation between the reception signal before being Fourier transformed by the second Fourier transform means and a reference signal including training signal components corresponding to a plurality of carriers for each antenna. The desired wave and the interference wave are detected in response to exceeding the threshold value.

  Preferably, the reception response vector estimation means sets the response level in the first to (n + 1) th response vectors at time other than the arrival timing detected by the arrival timing detection means to be zero.

  Preferably, the adaptive array processing means extracts a desired wave for the corresponding carrier based on a correlation matrix for each carrier derived based on components for the corresponding carrier of the first to (n + 1) th response vectors. The weight vector of is derived.

  Preferably, the adaptive array processing means derives a weight vector for extracting an interference wave for the corresponding carrier from the correlation matrix for each carrier.

  Preferably, the reception response vector estimation means estimates the first to (n + 1) th response vectors by the MMSE method.

  According to another aspect of the present invention, an adaptive array for receiving a signal transmitted by an orthogonal frequency division communication system using a plurality of carriers with an array antenna and extracting each component corresponding to the carrier by adaptive array processing A processing method for estimating a first response vector for a desired wave signal, Fourier transforming the first response vector to extract a component for each of a plurality of carriers, and a first response vector Deriving a weight vector for separating the component corresponding to the carrier for the desired wave by adaptive array processing based on the component for each carrier, and Fourier-transforming the received signal from the array antenna, Extracting the carrier component of the received signal and for each antenna The carrier component of the signal signal, by multiplying the weight vector, and a step of extracting the corresponding component of the carrier for the desired wave.

  Preferably, the adaptive array processing method detects at least the arrival timing of a desired wave from signals received by a plurality of antennas, detects the arrival timing of at least one interference wave, and estimates a response vector for the interference wave And a step of Fourier-transforming a response vector for the interference wave to extract a component for each of the plurality of carriers, and the step of deriving a weight vector includes converting the response vector for the interference wave to a component for each carrier. Based on this, a weight vector is derived.

[Embodiment 1]
FIG. 1 is a schematic block diagram showing a configuration of an adaptive array base station 1000 according to the embodiment of this invention. The adaptive array base station 1000 of the present invention transmits and receives a signal having directivity to and from a mobile station such as a user terminal by adaptive array processing. However, as will be apparent from the following description, adaptive array base station 1000 can also transmit and receive signals to and from mobile stations using the space division multiplexing method.

  Referring to FIG. 1, adaptive array base station 1000 receives an array antenna composed of n (n: natural number) antennas and signals from array antennas # 1 to #n, and performs detection and analog / digital conversion. A / D conversion unit 1010 for performing and FFT provided for each of the n antennas, receiving an output from the A / D conversion unit 1010, and separating and extracting a signal for each carrier for the corresponding antenna Part 1020.1-1020. n and a correlator 1030 for detecting the arrival timing of the desired wave and the interference wave, and a correlator 1030 for detecting the desired wave and the interference wave, as will be described later, upon receiving the signal from the A / D converter 1010 In order to detect the arrival timing, a memory 1040 for holding a reference signal corresponding to each of the desired wave and the interference wave, and a signal before fast Fourier transform, which is a signal before the fast Fourier transform, from the A / D converter 1010 to the correlator 1030 And a reception response vector for estimating a response vector for a desired wave and an interference wave by a procedure to be described later by receiving from the correlator 1030 information on the arrival timing of the signal detected by the correlator 1030 Each estimator 1050 is provided corresponding to each antenna and estimated by the reception response vector estimator 1050. Receiving a reception response for each burner, FFT unit for extracting a response vector for each carrier by performing a fast Fourier transform from 1060.1 to 1060. n, provided for each carrier, and FFT units 1060.1 to 1060. n, adaptive array blocks 1070.1-1070 for performing adaptive array processing in response to response vectors of corresponding carriers for antennas # 1 to #n. N (N is the total number of carriers).

  In FIG. 1, adaptive array block 1070. Only k is extracted and shown.

  Adaptive array block 1070. k is a received weight calculator 1072. k and FFT sections 1020.1 to 1020. n receives a signal for the kth carrier from n at one input node, and the other input node receives a receive weight calculator 1072. The signals from the multipliers 1080-1 to 1080-n receiving the weight vector from k and the signals from the multipliers 1080-1 to 1080-n are added and output, and the desired signal Sk (t) for the kth carrier is output. And an adder 1090.

  FIG. 2 is a conceptual diagram for explaining a received signal of adaptive array base station 1000 shown in FIG.

Adaptive array base station 1000 receives desired wave S _d (t), desired wave delay wave S _d (t−τ _S ), interference wave S _u (t), and interference wave delay as received waves. There is a wave S _u (t−τ _i ). Here, the times τ _S and τ _i are delay times. Here, the subscript d of the signal means that the signal is a desired wave. In addition, the signal of the interference wave is represented with a suffix u.

FIG. 3 is a conceptual diagram for explaining the configuration of the desired wave S _d (t) and the interference wave S _u (t).

The desired wave S _d (t) is not particularly limited. For example, the desired wave S _d (t) has a reference signal section (training signal section) d (t) for two symbols at the head and a data signal section that follows.

  Here, the reference signal d (t) is a signal obtained by performing inverse Fourier transform on training symbols of input signals arranged in the frequency domain, and is a signal in the time domain.

Similarly, the interference wave S _u (t) also has, for example, a reference signal section u (t) for two symbols at the head and a data signal section that follows.

  Here, it is assumed that the reference signal section d (t) of the desired wave is a signal different from the reference signal section u (t) of the interference wave without losing generality.

  Therefore, the adaptive array base station 1000 can identify a mobile station such as a user terminal by using such different reference signals (training signals).

[Correlator operation]
FIG. 4 is a conceptual diagram for explaining the operation of correlator 1030 in adaptive array base station 1000 shown in FIG.

The signals input to the correlator 1030 are the first arrival signal S _d (t) and the short delay signal S _d (t−τ ₂ ) that arrives with a delay time τ ₂ smaller than the guard interval period for the desired wave. It is assumed that there is a long delay signal S _d (t−τ ₃ ) that arrives with a delay time τ ₃ longer than the guard interval period. Similarly for the interference wave, the head arrival signal S _u (t), the short delay signal S _u (t−τ ₂ ) that arrives with a delay time τ ₂ smaller than the guard interval period, and the delay time longer than the guard interval period It is assumed that there is a long delay signal S _u (t−τ ₃ ) that arrives at τ ₃ .

  Also in adaptive array base station 1000, when correlator 1030 operates, it is necessary to process a signal before performing FFT processing. Therefore, a sampling timing sufficiently short to perform such signal processing before FFT is performed. Correlator 1030 samples the received signal.

The signal X _n (t) input from the antenna #n to the correlator 1030 can be expressed by the following equation (16).

In Expression (16), h _{n, 1} represents the response (element of response vector) of the leading wave of the desired wave received by the nth antenna #n, and _{pn, 1} represents the nth antenna #n. The response of the leading wave of the interference wave (multiple partner in SDMA) received at 1 is shown.

Similarly, the coefficients h _{n, 2} and h _{n, 3} indicate the response (element of response vector) of the delayed wave of the desired wave received by the n-th antenna #n, and the coefficients p _{n, 2} and p _{n , 3} indicate delayed wave responses of interference waves (multiple counterparts in SDMA) received by the nth antenna #n.

Further, as described above, the signal s _d (t) is a signal of a desired wave, and the signal s _u (t) is a signal of an interference wave (multiple partner in SDMA).

  However, in the case where interference or multiple users further exist, the term of the interference wave increases in Equation (16).

Correlation function ρ _{n, d} (t) between received signal X _n (t) of antenna #n and reference signal s _d (t) of a desired wave (t is a reference signal section), and received signal X of antenna #n _When the correlation function ρ _{n, u} (t) between _n (t) and the reference signal s _u (t) (t is the reference signal section) of the interference wave is calculated, the following is obtained.

In the correlation function ρ _{n, d} (t) between the reference signal of the desired wave and the received signal of the antenna #n, the correlation component between the desired wave and the reference signal remains, but a slight correlation component I between the interference wave and noise. _d (t) also remains.

Similarly, in the correlation function ρ _{n, u} (t) between the received signal of antenna #n and the reference signal of the interference wave, not only the correlation component of the interference wave and the reference signal remains but also a slight amount of interference wave and noise. The correlation component I _u (t) remains.

Such a correlation function ρ _{n, d} (t) or correlation function ρ _{n, u} (t) is also referred to as “sliding correlation”.

FIG. 5 is a diagram showing the time dependence of such a correlation function ρ _{n, d} (t).
Since the correlation function ρ _{n, d} (t) is actually a complex number, it is a signal whose absolute value and phase change on the complex plane as time passes. It is assumed that only components in a predetermined direction on the complex plane are represented.

Referring to FIG. 5, in correlation function ρ _{n, d} (t), first, peak P1 exists as the head arrival signal component. Immediately after this peak P1, there is a peak P2 corresponding to the short delay signal component. Furthermore, a peak P3 corresponding to the long delay signal component exists at a time delayed from the peak P2 of the short delay signal component.

The same applies to the correlation function ρ _{n, u} (t) for the interference wave.
FIG. 6 is a diagram showing the time dependency of the absolute value component of the correlation function ρ _{n, d} (t) shown in FIG. 5. In FIG. 6, the value Vt is used for performing processing as described later. Indicates a threshold.

[Reception Response Vector Estimator Operation]
The above-described correlation function ρ _{n, d} (t) corresponds to the reception response (complex number) of the desired wave signal of the n-th antenna, but the complex response obtained by such correlation has non-orthogonal noise and interference. The component remains and the error is large.

However, with this correlation function ρ _{n, d} (t), the arrival time of the head arrival signal and the delay time itself of the arrival time of the delay signal can be accurately obtained.

Therefore, the reception response of the desired wave user terminal and the reception response of the interference wave user terminal are obtained more accurately by using the delay time obtained by the correlation function ρ _{n, d} (t) or the correlation function ρ _{n, u} (t). It is assumed that the procedure is performed by the reception response vector estimator 1050 according to the procedure described below.

(Step 1)
First, a threshold value Vt is set in advance for the absolute values | ρ _{n, d} (t) | and | ρ _{n, u} (t) | of the correlation function as shown in FIG. Pick up the signal that is above the value. Here, as such a threshold value Vt, a predetermined value is set in advance, or a standard such as extraction from a maximum signal level to a signal lower by a predetermined value is used.

(Step 2)
For the signal picked up in this way, as will be described below, the complex response is accurately estimated by a so-called MMSE (Minimum Mean Square Error) method that minimizes the mean square error between the array output and the reference signal. .

  Here, attention is paid to one of the antennas, and a signal sequence obtained by sampling the received signal at this antenna is set as a vector X, which is expressed as follows.

  Although not particularly limited, the number of elements of such a vector can be, for example, 64 samples or 128 samples.

Further, signals obtained by performing inverse Fourier transform on the reference signal of the desired wave are denoted by sd ₁ , sd ₂ , sd ₃ ,. The delay signal of the desired wave is assumed to arrive through the path k, and when the delay time is τ _k , the reference signal for such a desired wave corresponds to the vector composed of the sampling value of the received signal described above. Is expressed as follows.

In the above equation (20), there are a number of elements represented by □ corresponding to the delay time τ _k . In addition, as a value of the element represented by □, for example, when a guard interval exists before the reference signal, an inverse Fourier transform component of a signal existing in the guard interval exists in this portion.

Further, in the following, a case where there is one interference signal is considered for the sake of simplicity.
At this time, it is assumed that the time series of elements obtained by performing the inverse Fourier transform on the reference signal of the interference signal in the same manner is represented as su ₁ , su ₂ ,. Assuming that the delay signal of the interference wave arrives through the path k ′ and the delay time is τ _{k ′} , the interference signal reference is made corresponding to the vector X formed by the sampling elements of the received signal. The time series of signals is expressed as follows.

  In the following, when estimating the response vector by the MMSE method, there are a plurality of, for example, three paths k having different delay times for the desired wave, and a plurality of, for example, three paths k ′ for the interference wave. And

Under such conditions, for example, the processing for obtaining the response vector for the received signal of the n-th antenna is a response h _k for the desired wave so as to minimize the evaluation function J ₁ expressed by the following equation (22). This corresponds to _obtaining the response p _{k ′} for the interference wave.

Here, when the matrix Q and the vector a are defined as in the equations (23) and (24), the evaluation function J ₁ is expressed as in (25).

Furthermore, from the condition that the evaluation function J ₁ or the vector a is minimal, the vector a can be obtained as shown in Expression (26) by the following procedure.

  As described above, the complex amplitude of each path can be obtained for the desired signal and the interference signal.

  The above procedure is, for example, derivation for the n-th antenna. Such processing is similarly performed for the other antennas, and the response of the desired wave and the interference wave is obtained for each antenna.

(Step 3)
As described above, all signal levels other than the signal that is picked up as being above the threshold and whose complex response is estimated are set to 0.

By this processing, residual noise and interference components can be removed.
(Step 4)
Next, for a delay time longer than the guard interval time, the complex response is set to 0, and a complex response signal consisting only of a component corresponding to the delayed wave within the guard interval time is obtained _{. It is} assumed that _dd (t) and the interference wave is also set as a reception response (correlation function) ρ _{n, ud} (t). Here, the subscript dd indicates that the signal is within the guard interval for the desired wave, and the subscript ud indicates that the signal is within the guard interval for the interference wave.

(Step 5)
Similarly, for delayed waves that are shorter than the guard interval time, all the complex responses are set to 0, and other responses that leave the level of the complex response are newly set as responses (correlation functions) ρ _{n, It is} assumed that _du (t) and the interference wave is a response (correlation function) ρ _{n, uu} (t).

  Here, the subscript du indicates that the desired wave is a delay wave longer than the guard interval time, and the subscript uu indicates that the interference wave is a delay wave having a delay time longer than the guard interval time. .

  The response for the interference wave thus obtained corresponds to a complex response for multiple users in SDMA.

FIG. 7 is a diagram showing temporal changes in the response ρ _{n, dd} (t) and the response ρ _{n, du} (t) calculated as described above.

  In FIG. 7, there are three paths, and the signals from the path 1 and the path 2 for the first incoming wave and the first delayed wave have arrived at the base station within the guard interval length and correspond to the path 3 The delay wave arrives at the adaptive array base station 1000 after a delay time longer than the guard interval length has elapsed from the arrival time of the leading wave.

Accordingly, the response ρ _{n, dd} (t) includes two peaks, and the response ρ _{n, du} (t) includes one peak.

With such a procedure, for the desired wave, the first response vector consisting of the response of each antenna corresponding to the response ρ _{n, dd} (t) and the response of each antenna corresponding to the response ρ _{n, du} (t) The second response vector consisting of is derived.

Similarly, the response of the interference wave includes a third response vector composed of responses of the respective antennas corresponding to the response ρ _{n, ud} (t) and responses of the respective antennas corresponding to the response ρ _{n, uu} (t). A fourth response vector will be derived.

  If there are m interference waves (m ≧ 2), each antenna corresponding to a complex response consisting of only a component corresponding to a delayed wave within the guard interval time is similarly applied to the m-th interference wave. And a (2m + 1) th response vector consisting of responses of antennas corresponding to complex responses leaving only components corresponding to delayed waves after the guard interval time. It will be.

[FFT section 1060.1-1060. Operation of n]
As described above, reception response vector estimator 1050 obtains a response for each antenna.

Next, FFT section 1060.1-1060. In n, the following processing is performed.
The complex response ρ _{n, dd} (t) of the signal transmitted from the desired terminal and arrived within the guard interval is converted into a complex response ξ _{n, dd} (k) for each carrier by performing a fast Fourier transform. Here, k is a carrier number.

FIG. 8 is a diagram showing the complex response ρ _{n, dd} (t) and the complex response ξ _{n, dd} (k) for each carrier obtained by the fast Fourier transform.

The same operation is performed for all antennas, and the complex response for each carrier of all antennas is calculated. A response vector having a complex response as an element is calculated for each carrier. Further, a complex response ξ _{n, du} (k) for each carrier is obtained by performing a fast Fourier transform on the complex response ρ _{n, du} (t).

If such processing is performed, for example, a response vector d _d (k) for a signal that has arrived within the guard interval period of the k-th carrier is expressed as shown in Expression (27). Similarly, the response vector d _u (k) of the k-th carrier from the complex response ρ _{n, du} (t) of the signal transmitted from the desired terminal and arriving with a delay time _equal to or longer than the guard interval is given by Equation (28). Is calculated.

Similarly, by performing a fast Fourier transform on the complex response ρ _{n, ud} (t) of a signal transmitted from an interfering user (in SDMA, all connected users other than the desired user) and arriving within the guard interval, a carrier is obtained. Each response ξ _{n, ud} (k) is converted, and the response vector i _d (k) of the k-th carrier of the interference wave is calculated as in equation (29).

Similarly, with respect to the interference wave, the complex response ρ _{n, uu} (t) of the signal arriving after the guard interval period is converted into a response ξ _{n, uu} (k) for each carrier by _{performing a} fast Fourier transform _, and the signal exceeds the guard interval. The response vector i _u (k) for the interference wave arriving at a delay time of is expressed as in equation (30).

  When there are a plurality of interference or multiple users, such a response vector is calculated for each interference wave or multiple users.

[Operation of Receive Weight Calculator]
Based on the response vector of the k-th carrier for each antenna obtained by the fast Fourier transform as described above, the reception weight calculator 1070. k calculates the reception weight vector for the k-th carrier as follows.

Response vector d _d (k) of desired wave with delay time within guard interval of k-th carrier, response vector d _u (k) of desired wave with delay time exceeding guard interval, interference wave with delay time within guard interval Response vector i _d (k) and the response vector i _u (k) of the interference wave with a delay time exceeding the guard interval are as follows. It is calculated | required by n.

From these signals, the correlation matrix R _xx ^(k) of the ^kth carrier is calculated as shown in Expression (32), and based on this, the reception weight vector of the desired signal is calculated as shown in Expression (33). .

  Further, the reception weight vector for the counterpart of the multiplexed signal in the case of SDMA is calculated as shown in Equation (34).

In Expression (32), σ ² is a positive real number, and this value may be a value empirically determined so that the correlation matrix does not become singular or may be a value of thermal noise power of the system. . Further, I represents an n × n unit matrix.

  When there is no interference wave, the term corresponding to the interference wave in equation (32) is 0, and even when only the desired wave is present, the reception weight vector of the desired signal is calculated as in equation (33). Will be.

  9 and 10 are flowcharts for explaining the overall operation of adaptive array base station 1000 described above.

  Referring to FIG. 9, when the processing is started (step S100), the correlator 1030 performs a sliding correlation between the received signal and the reference signal of the desired wave for each antenna of the array antenna (step S102).

  Further, the correlator 1030 performs a sliding correlation between the received signal and the reference signal of the interference wave for each antenna of the array antenna (step S104).

  Subsequently, reception response vector estimator 1050 includes a signal having an absolute value of a correlation value exceeding a predetermined threshold for a desired wave, and a signal having an absolute value of a correlation value exceeding a predetermined threshold for an interference wave Is picked up (steps S106, S108).

  Further, reception response vector estimator 1050 estimates a complex response to the picked-up signal for the desired wave and interference wave by the MMSE method or the like (step S110). Then, the level of signals other than the picked-up signal is set to 0 (step S112).

Then, the reception response vector estimator 1050 sets the signal component having a delay time longer than the guard interval to 0, and receives the reception response ρ _{n, dd} (t) for the desired wave and the reception response ρ _{n, ud} for the interference wave. (T) is obtained (step S114). Further, reception response vector estimator 1050 sets the signal component having a delay time shorter than the guard interval to 0, and receives reception response ρ _{n, du} (t) for the desired wave and reception response ρ _{n, uu} (t for the interference wave. ) Is obtained (step S116).

Next, referring to FIG. 10, FFT units 1060.1 to 1060. In n, each of the received responses ρ _{n, dd} (t), received response ρ _{n, ud} (t), received response ρ _{n, du} (t) and received response ρ _{n, uu} (t) is _subjected to Fourier transform. For the antenna, a complex response for each carrier is obtained (step S118).

Accordingly, 1) a complex response vector d _d (k) for each carrier with respect to a signal arriving with a delay within the guard interval among the desired waves, and 2) for each carrier with respect to a signal arriving with a delay exceeding the guard interval among the desired waves. Complex response vector d _u (k), 3) Complex response vector i _d (k) for each carrier with respect to a signal arriving with a delay within the guard interval of the interference wave, 4) Arrival with a delay exceeding the guard interval of the interference wave A complex response vector i _u (k) for each carrier with respect to the received signal is derived (step S120).

Further, the reception weight calculator 1072. At k, a correlation matrix R _XX ^(k) of the kth carrier is derived based on the derived complex response vector, and a weight vector for the kth carrier is calculated for the desired wave (step S122).

  Multipliers 1080-1 to 1080-n and adder 1090 multiply the signal for each carrier obtained by Fourier transform of the received signal from each antenna of the array antenna by a weight vector to obtain the kth A desired signal for the carrier is extracted (step S124). In the SDMA system, if necessary, a weight vector is also obtained for an interference wave, and the interference wave is extracted.

  Furthermore, if the components for each carrier are combined, the signal transmitted by the OFDM method can be demodulated. The process ends here (step S130).

The reason why the short delay signal is synthesized without being attenuated by the phase difference from the head arrival wave by the above method is that the response vector d _d (k) for each carrier includes both the head signal and the short delay signal. Because the component is included, the adaptive array operation is performed so that the beam is directed to each of the head signal and the short delay signal for each carrier.

For this reason, even if the frequency of the carrier is different, it is not synthesized in the opposite phase. In addition, since there are four significant signal component terms (in the case of one interference) from the expression of the correlation matrix R _XX (k), the degree of freedom consumed is 3, for example, a four-element antenna. It is possible to perform complete null direction control.

  On the other hand, regardless of the method as described above, when performing null control on all signals, it is necessary to perform control that directs null even for a short delay. In this case, the four-element antenna has insufficient flexibility and cannot obtain sufficient characteristics.

[Embodiment 2]
In the first embodiment, as an operation of reception response vector estimator 1050, a complex response of a desired signal and a complex response of an interference signal are obtained according to the method described in equations (22) to (26).

  However, when there is no overlap between the reference signal section of the desired signal and the reference signal section of the interference signal, the methods described in the equations (22) to (26) cannot be applied as they are.

  In the second embodiment, a method for deriving a complex response of a desired signal and a complex response of an interference signal that can be applied even in such a case will be described.

(Estimation of desired signal response)
In obtaining the complex response of the desired signal, the evaluation function J ₂ given by the following equation (35) is used. In addition, the notation of the following formulas is the same as the formulas (22) to (26) unless otherwise specified.

  Here, a matrix Q ′ and a vector h defined by the following equations (36) and (37) are used.

  At this time, Expression (35) can be rewritten as follows.

With respect to the vector h, the complex response h _k with respect to the path k of the desired wave is _obtained as in the following equation (39), as in the equation (26), under the condition that it is minimal.

(Estimation of interference signal response)
In obtaining the complex response of the interference signal, the evaluation function J ₃ given by the following equation (40) is used.

  Here, a matrix Q ″ and a vector p defined by the following equations (41) and (42) are used.

  At this time, Expression (40) can be rewritten as follows.

With respect to the vector p, the complex response p _k ′ with respect to the path k ′ of the interference wave is obtained as in the following equation (44), as in the equation (39), under the condition that it is minimal.

  The reception response vector estimator 1050 performs the complex response estimation method as described above, and the same effects as those of the first embodiment can be obtained.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  As described above, by using the configuration of the adaptive array base station according to the present invention, multipath signals within a guard interval can be completely combined at the maximum ratio to maximize reception sensitivity. Furthermore, when combining multipath signals within the guard interval, it is possible to maintain the interference suppression performance without consuming antenna freedom.

It is a schematic block diagram which shows the structure of the adaptive array base station 1000 of embodiment of this invention. It is a conceptual diagram for demonstrating the received signal of the adaptive array base station 1000 shown in FIG. It is a conceptual diagram for demonstrating the structure of desired wave _Sd (t) and interference wave _Su (t). It is a conceptual diagram for demonstrating operation | movement of the correlator 1030 among the adaptive array base stations 1000 shown in FIG. It is a figure which shows the time dependence of correlation function (rho) _{n, d} (t). It is a figure which shows the time dependence of the absolute value component of the correlation function (rho) _{n, d} (t) shown in FIG. It is a figure which shows the time change of response (rho) _{n, dd} (t) and response (rho) _{n, du} (t). It is a figure which shows complex response (p) _{n, dd} (t) and the complex response (xi) _{n, dd} (k) for every carrier obtained by the fast Fourier transformation with respect to this. 3 is a first flowchart for explaining the operation of adaptive array base station 1000 as a whole. 6 is a second flowchart for explaining the operation of adaptive array base station 1000 as a whole. FIG. 2 is a channel layout diagram in various communication systems of frequency division multiplex connection, time division multiplex connection, and spatial multiplex division connection. It is a schematic block diagram which shows the structure of the transmission / reception system 2000 of the conventional SDMA base station. It is a block diagram which shows the structure of the transmission / reception part 2100a corresponding to one user among the adaptive arrays 2100. FIG. It is a figure which extracts and shows three carriers out of the frequency spectrum of the some carrier (carrier wave) used in an OFDM system. It is a figure which shows the waveform of the transmission symbol transmitted by OFDM system. It is a conceptual diagram for demonstrating the demodulation operation | movement when a desired wave and an interference wave are received. 4 is a schematic block diagram for explaining a configuration of an adaptive array base station 3000. FIG. It is a conceptual diagram which shows the timing of the signal which the adaptive array base station 3000 shown in FIG. 17 receives. It is a figure which shows intensity distribution of the signal corresponding to each carrier after passing an adaptive array. It is a figure which shows intensity distribution when the signal for every carrier is synthesize | combined in the case as shown in FIG. 2 is a schematic block diagram for explaining a configuration of an adaptive array base station 4000. FIG. It is a conceptual diagram for demonstrating operation | movement of the adaptive array base station 4000 shown in FIG.

Explanation of symbols

  1000 Adaptive array base station, # 1 to #n antenna, 1010 A / D converter, 1020.1 to 1020. n FFT unit, 1030 correlator, 1040 memory, 1050 reception response vector estimator, 1060.1-1060. n FFT section, 1070.1-1070. N adaptive array block, 1072. k Receive weight calculator, 1080-1 to 1080-n multiplier, 1090 adder.

Claims (10)

A wireless device for transmitting and receiving signals transmitted by an orthogonal frequency division communication system using a plurality of carriers,
An array antenna having a plurality of antennas;
Reception response vector estimation means for estimating a first response vector for a desired wave signal;
First Fourier transform means for Fourier transforming the first response vector to extract a component for each of the plurality of carriers;
Second Fourier transform means for performing Fourier transform on the received signal from the array antenna and extracting a component for each carrier of the received signal for each antenna;
Provided for each of the plurality of carriers, each of which receives a corresponding carrier component from the second Fourier transform means among the components of the received signal carrier for each antenna, and corresponding carrier component in the desired wave Adaptive array processing means for extracting
The adaptive array processing means derives a weight vector for extracting the component of the corresponding carrier based on the component of the first response vector from the first Fourier transform means for the corresponding carrier. apparatus.   From the signals received by the array antenna, the arrival timing of a desired wave is detected, and for each antenna, the received signals before being Fourier transformed by the second Fourier transform means and the training signals corresponding to the plurality of carriers The radio apparatus according to claim 1, further comprising arrival timing detection means for detecting the desired wave in response to a cross-correlation with a reference signal including a component exceeding a predetermined threshold. The wireless device further includes arrival timing detection means for detecting arrival timing of a desired wave from a signal received by the array antenna,
The radio apparatus according to claim 1, wherein the reception response vector estimation unit sets a response level in the first response vector to 0 at a time other than the arrival timing detected by the arrival timing detection unit. The wireless device further includes arrival timing detection means for detecting arrival timing of a desired wave from a signal received by the array antenna,
The radio apparatus receives a signal in which a guard interval period is added to an effective symbol period, and the reception response vector estimation means sets a response level in the first response vector to 0 at a time after the guard interval period. The wireless device according to claim 1. The wireless device further includes arrival timing detection means for detecting arrival timing of a desired wave from a signal received by the array antenna,
The arrival timing detection means further detects the arrival timing of n interference waves (n: natural number, n ≧ 1) from the signal received by the array antenna,
The reception response vector estimation means estimates second to (n + 1) th response vectors for each signal for each of the n interference waves,
The first Fourier transform means further performs Fourier transform on the second to (n + 1) th response vectors to extract a component for each of the plurality of carriers,
The adaptive array processing means is a weight vector for extracting the component of the corresponding carrier based on the component for the corresponding carrier of the first to (n + 1) th response vectors from the first Fourier transform means. The wireless device according to claim 1, wherein The arrival timing detection means has a predetermined cross-correlation between the received signal before being Fourier transformed by the second Fourier transform means and a reference signal including training signal components corresponding to the plurality of carriers for each antenna. The wireless device according to claim 5 , wherein the desired wave and the interference wave are detected in response to exceeding a threshold value. The reception response vector estimating means, and zero level of the response in the response vector of the from the first at a time other than said detected arrival timing (n + 1) that by the arrival timing detecting means of claim 5, wherein Wireless device. 6. The radio apparatus according to claim 5 , wherein the reception response vector estimation means estimates the first to (n + 1) th response vectors by MMSE method. An adaptive array processing method for receiving a signal transmitted by an orthogonal frequency division communication system using a plurality of carriers with an array antenna and extracting each component corresponding to the carrier by adaptive array processing,
Estimating a first response vector for the desired wave signal;
Fourier transforming the first response vector to extract a component for each of the plurality of carriers;
Deriving a weight vector for separating a component corresponding to the carrier for a desired wave by adaptive array processing based on a component for each carrier of the first response vector;
Fourier transforming the received signal from the array antenna to extract a carrier component of the received signal for each antenna;
An adaptive array processing method comprising: extracting the corresponding carrier component of the desired wave by multiplying the carrier component of the received signal for each antenna by the weight vector. The adaptive array processing method includes:
Detecting at least the arrival timing of a desired wave from a signal received by an array antenna having a plurality of antennas, and detecting the arrival timing of at least one interference wave;
Estimating a response vector for the interference wave;
Further comprising: Fourier transforming a response vector for the interference wave to extract a component for each of the plurality of carriers,
The adaptive array processing method according to claim 9 , wherein the step of deriving the weight vector derives the weight vector based on a component for each carrier of a response vector with respect to the interference wave.

Images