WO2007000079A1

WO2007000079A1 - A method for configuring and demodulating multiple scrambled codes in cdma system

Info

Publication number: WO2007000079A1
Application number: PCT/CN2005/000939
Authority: WO
Inventors: Zhaohua Zeng; Ping Li; Hai Jiang
Original assignee: Zte Corporation
Priority date: 2005-06-28
Filing date: 2005-06-28
Publication date: 2007-01-04
Also published as: CN101133581B; CN101133581A

Abstract

A method for configuring and demodulating multiple scrambled codes in CDMA system, increases the system capacity by configuring multiple scrambled codes in multiple antennas system, the method comprises the following steps: configures multiple scrambled codes for each cell, the generating method and amount of spread codes for each cell is consistent with that in conventional way; the generating method for subscriber training sequence is consistent with the provision of current TD protocol, only different in shortening the width of the channel pulse response window for each subscriber when probably causing multiple scrambled codes. The method according to the invention takes full advantage of any multiple antennas system, especially makes use of the space-division (SDMA) character resulting from multiple antennas in TD-SCDMA system, adopts a particular method for configuring wireless resource and the demodulating method thereof, reallocates the scrambled code resource, uses multiple scrambled codes in each cell, thereby greatly improves the system capacity, increases the availability of frequency band, and saves the frequency resource. .

Description

Multi-scrambling code configuration in CDMA system and related demodulation method

The present invention relates to a method for configuring a plurality of scrambling codes in a cell to increase system capacity in a system, and is particularly suitable for a multi-scrambling code configuration method in a TD-SCDMA system in the field of mobile communications.

The method of the present invention can also be applied to all CDMA systems in the field of mobile communications, and is particularly suitable for TDD-CDMA systems and all SCDMA systems. Background technique

In all existing mobile communication systems, a multiple access system is usually constructed by using one or several combinations of frequency division (FDMA), time division (TOMA) or code division (CDMA) characteristics of radio resources, for a large number of mobile users. Provide services. For a 3G mobile system with CDMA as its main feature, the same frequency network coverage is usually achieved by configuring one scrambling code for each cell (different adjacent cell scrambling codes). In the network, the wireless resources occupied by different users are different, which greatly limits the system bandwidth utilization and limits the user capacity in the unit frequency band. At the same time, it can be seen that any introduction of wireless resource independent features will bring about a significant increase in frequency band utilization. For example, TD-SCDMA allocates resources in FDMA/TDMA/CDMA mode, and its frequency band utilization is far from Other systems. - In the TD-SCDMA system, after the data on the transmission channel is mapped by the physical channel, the data on the channel is subjected to spreading and scrambling. Document TS 25. 221 Physical channels and mapping of transport channels onto physical channels (TDD), V4. 7. 0, 3 scrambling codes are defined in 3GPP, and scrambling codes are used to identify the cell attributes of only i data, each cell Only one scrambling sequence is configured, and different neighboring cells will use different scrambling codes. In this way, the maximum number of users per cell in a certain time slot is determined by the number of channelization codes. According to the current TD-SC A protocol, a 12.2k voice user will occupy 2 SF=16 code channels, so that there are only 8 users in a single cell and one time slot.

However, increasing system capacity (or system data throughput) has always been one of the focus issues of third-generation mobile communications (3G). As the number of users (or user traffic data rates) continues to increase, frequency resources are becoming increasingly scarce. More advanced technologies are needed to make full use of frequency resources. How to serve as many users as possible under limited resource conditions is the goal of 3G. Summary of the invention

The object of the present invention is to provide a multi-scrambling code configuration in a CDMA system and its associated demodulation method, which is fully beneficial. With any multi-antenna system, especially the spatial division (SOMA) feature brought by multiple antennas in the TD-SCDMA system, multiple scrambling codes and their corresponding channel estimation and joint detection methods are simultaneously configured in the same cell, and full consideration is given. Under the multi-antenna SDMA feature, a specific radio resource configuration method and corresponding demodulation technology are provided, and the user channelization code resources are fully multiplexed, thereby multiplying the system capacity and improving the frequency band utilization.

The technical solution of the present invention is as follows:

Multi-scrambling code configuration in a CDMA system and its associated demodulation method. In a multi-antenna system, the system capacity is increased by configuring a multi-scrambling code, which includes - a), configuring multiple scrambling codes per cell, each scrambling code The method and quantity for generating the spread spectrum code are consistent with the conventional method;

b) The method for generating the user training sequence is consistent with the current TD protocol, except that when the multi-scrambling code is likely to occur, the channel impulse response window width of each user becomes shorter.

The method further includes:

Al), the system groups the users served by the cell, the number of groups is equal to the number of scrambling codes, and each group of users allocates a channelization code according to a conventional orthogonal variable spreading factor code method;

Bl), according to the code group number of the downlink pilot of the current cell, select N scrambling codes of the plurality of scrambling codes corresponding to the code group as the local cell scrambling code, where N is greater than or equal to 1 and less than or equal to the scrambling code of the code group. Number

Cl ), processing the data at the transmitting end, spreading, scrambling, modulating and raising the cosine filtering, transmitting through the air interface;

Dl), separating the training sequence received signal from the received data, completing channel estimation and channel post-processing for each user, and the received training sequence can be expressed as

e = Gh + n (1) where G matrix is a cyclic right shift matrix consisting of a basic Midamble code, h is the user's channel impulse response, and n is Gaussian white noise.

The method further includes: 信道 using the following method to perform channel estimation: using a basic Midamble code, a total of NK voice users, K is the number of users using one scrambling code, for counting natural numbers, each user's The channel estimation window length W is reduced to ^.

NK

The method, wherein the number of scrambling codes configured per cell is N=2.

The method, wherein the method is applied to a TD-SCDMA system, and further includes the step of assigning a plurality of scrambling codes:

All ), a total of 128 scrambling codes are defined, and are divided into 32 groups, each of which has 4 scrambling codes. 4 scrambling codes in the group are used for cross correlation;

Bl l ) , preferentially select the scrambling code with small cross-correlation as the scrambling code used by the cell, and the value of the number of scrambling codes N is 2≤N≤4.

The method further includes: a channel estimation step for multiple scrambling codes:

Cl l ) , generating a G matrix, a cyclic right shift matrix G consisting of a basic Midamble code, assuming that N scrambling codes are simultaneously used in one cell, the number of users is increased by N times, a total of NK voice users, each user's The channel estimation window length W is reduced to ^, where K is the number of users using one scrambling code;

NK

Dl l), the training sequence of the received data is proposed for channel estimation, the maximum likelihood estimation of the channel impulse response, h = G; setting a noise threshold Γ, performing channel post-processing on the estimated £, retaining greater than the decision threshold , less than the decision threshold is set to zero;

Ell), obtain the channel impulse response of each user, and take NK values from ϋ in the length of W to form the channel estimation of each user. The channel response window length of each user is W.

In the method, the method further includes a joint detection step of multiple scrambling codes - demodulating multi-user joint detection data of multiple scrambling codes, specifically, using respective spreading spread scrambling codes of the user and after passing through the channel The processed channel impulse response fi generates a matrix A, which demodulates the user data: wherein the matrix A is a matrix composed of a convolution vector of each user spread spectrum scrambling code sequence and a channel impulse response using a plurality of scrambling codes, To ensure normal demodulation, it needs to be satisfied: The number of antennas is Ka≥N.

The multi-scrambling code configuration and related demodulation method in a CDMA system provided by the present invention utilizes any multi-antenna system, especially the spatial division (SDMA) characteristic brought by multiple antennas in the TD-SCDMA system, A specific radio resource configuration method and corresponding demodulation technology are adopted to re-allocate the scrambling code resources, and multiple scrambling codes are used in one cell, compared with a scrambling code corresponding to an existing cell. The method can greatly improve the user capacity of the system, improve the frequency band utilization, and save frequency resources. DRAWINGS

1 is a structural diagram of a system transmission matrix A of the method of the present invention, A is composed of the same V block; FIG. 2 is a structural diagram of the block V shown in FIG. 1, and V is a composite spread scrambling code volume of each user. The b-vector obtained by the accumulative channel impulse response, the different blocks in the picture represent the b-vectors obtained by using the same spreading code for different scrambling codes;

3(a) and (b) are respectively schematic diagrams of system signal transmission and signal reception demodulation corresponding to the present invention. detailed description

The preferred embodiments of the technical solution are further described in detail below with reference to the accompanying drawings:

In the TD-SCDMA system, 128 scrambling codes are defined and divided into 32 groups, each of which has 4 scrambling codes. However, the prior art protocol only supports only one scrambling code per cell. .

However, in the TD-SCDMA system, in fact, the SDMA characteristics brought by the use of multiple antennas can be fully utilized, and the system capacity can be improved by configuring a multi-scrambling code, specifically: each cell is configured with multiple scrambling codes; The method and quantity of the spreading code generation under each scrambling code are consistent with the conventional method; the generation method of the user training sequence is consistent with the current TD protocol, and the only difference is that the channel impulse response of each user may be caused by the multi-scrambling code. The window width becomes shorter, but this has little effect in the urban environment.

Considering the complexity of the base station algorithm and the wireless channel characteristics of the urban environment, and referring to the method for generating the user training sequence in the TD-SCDMA system, the present invention recommends the number of scrambling codes per cell configuration N=2, and the single time slot for each scrambling code. Supports up to 8 12. 2k voice users, so that each cell single slot can support 16 12.2k voice users, and each user has a channel impulse response window width of 8 chips on average, but the present invention The patent is not limited to N=2, and N is equal to other values in the scope of the present invention. At the same time, the method of user channel impact window width dynamic allocation can also be adopted to solve the problem that the window width may be insufficient, and the technical content of this aspect is adopted. It is also within the scope of protection of this patent.

In order to clearly describe the patent of the present invention, the detailed description is as follows:

The users served by the cell are grouped, the number of groups is equal to the number of scrambling codes, and each group of users allocates a channelization code according to a conventional orthogonal variable spreading factor of 0 VSF code.

According to the code group number of the downlink pilot DwPTS of the current cell, the N scrambling codes of the four scrambling codes corresponding to the code group are selected as the local cell scrambling code, 1 < N ≤ 4, generally N=2 or N= l.

The data is processed at the transmitting end, spread spectrum, scrambling (packet scrambling), modulation and raised cosine filtering, and transmitted through the air interface, equivalent to over the wireless channel.

The training sequence received signal is separated from the received data, and channel estimation and channel post-processing for each user are completed. The received training sequence can be expressed as

e = G + n (1) where G matrix is a cyclic right shift matrix consisting of a basic Midamble code, h is the user's channel impulse response, and n is Gaussian white noise.

Assuming that N scrambling codes are simultaneously used in one cell, the number of users is increased by N times. To this end, the present invention can use the following method for channel estimation: Using a basic Midamble code, a total of NK voice users, K To use the number of users of one scrambling code, the channel estimation window length w of each user is reduced to ^.

NK

Complete the maximum likelihood estimation of the channel impulse response h using equation (2), ie

h = (G* ^r G)- ^I G* ^r e (2) Since the G matrix is a cyclic right shift matrix, there is

h = G ^_1 e (3) After the user data is scrambled by the spread spectrum, after reaching the receiver, the received signal e can be expressed as - e = Ad+n (4) d is the column vector composed of the symbols of all users. The user's respective spread spectrum scrambling code and post-processed channel impulse response generation system transmission matrix A,

A = [A ^kl A ^k2 LA ^Ka ]', Ka is the number of antennas, Ka ≥ N (5) Demodulation of user data,

d = (A ^ff -A)- ¹ A ⁱ e (6)

A ^to is a block diagonal matrix, the block ^to can be expressed as

=[ 3⁄4 L 3⁄4 3⁄4 3⁄4 L 3⁄4 3⁄4 3⁄4 L 3⁄4„] (7) where is the channel-excitation response of the k-th user's spread-spectrum scrambling sequence using the 1st |3⁄4 code and its corresponding user/ 3⁄4 convolution; & £ is the convolution of the k-th user's spread spectrum scrambling code composite sequence using the second scrambling code with the channel impulse response / ^ of its corresponding user, ie k is the count natural number, similar Expressed as a matrix

= ® , 3⁄4=3⁄4®3⁄4, ······, _N =c _N ® _N (8) 3⁄4 is a matrix consisting of the k-th user's spread spectrum scrambling code composite sequence of the first scrambling code, _w is Use the first

The matrix consisting of the k-th users of the k-th users of the spread spectrum of the scrambling code is the channel impulse response of the kth user using the Nth scrambling code.

The method of the present invention is a scheme of using multiple scrambling codes in one cell, and a method of channel estimation and joint detection in this case, as shown in Figs. 3(a) and (b), which includes the following Steps:

How to assign multiple scrambling codes:

The first step: In the TD-SCDMA system, a total of 128 scrambling codes are defined and divided into 32 groups, each of which has 4 scrambling codes, and the four scrambling codes in the selected group are mutually Correlation.

Step 2: The scrambling code with small cross-correlation is preferentially selected as the scrambling code used by the cell, and the value of the number of scrambling codes N is 2 ≤ N ≤ 4.

Channel estimation method for multiple scrambling codes: Step 1: Generate a G matrix.

A cyclic right shift matrix G consisting of a basic Midamble code, assuming that N scrambling codes are simultaneously used in one cell, the number of users is increased by N times, a total of NK voice users, and the channel estimation window length W of each user is reduced to

^, K is the number of users using a scrambling code.

N

Step 2: Present the training sequence of the received data for channel estimation.

1 maximum likelihood estimation of channel impulse response, fi - G^e ;

2 - noise threshold Γ, the estimated channel post-processing, the retention greater than the decision threshold, is less than the decision threshold is set to zero.

Step 3: Obtain the channel impulse response for each user.

Taking the value of W from the length of each user constitutes the channel estimation of each user, and the channel response window length of each user is w.

Joint detection method for multiple scrambling codes:

Multi-user joint detection data demodulation of multiple scrambling codes, specifically using user's respective spread spectrum scrambling code and channel post-processing channel impulse response generation matrix A to demodulate user data:

0 = (Α ^Η 'Α)- H

The matrix A is a matrix composed of a convolution vector of each user spread spectrum scrambling code and a channel impulse response using a plurality of scrambling codes. To ensure normal demodulation, it is necessary to satisfy: the number of antennas Ka ≥ N.

In the following, a cell of the method of the present invention is further illustrated by using two scrambling codes as an example. For convenience of description, it is assumed that the spreading factors SF of all users are equal.

First, the data is processed at the transmitting end, spread spectrum, scrambling, modulation, and raised cosine filtering, and transmitted through the antenna. The training sequence for receiving data is proposed for channel estimation, and the data portion is subjected to joint detection and demodulation by multiple users to complete data separation.

Complete the maximum likelihood estimation of the channel impulse response h with equation (1), ie

h = (G* ^r G)- ¹ G* ^r e (2) Since the G matrix is a cyclic right shift matrix, there is h = G ^! e (3) Therefore, the above channel estimation can be implemented by the FFT/IFFT fast method, which significantly improves the operation speed, as follows - = ifft( t(e)/fft(m)) (12) The power of each tap of the impulse response is judged. If it is greater than or equal to the set threshold, the position of the tap is retained; otherwise, the response of the tap is removed.

The user data arrives at the receiver through the spread spectrum scrambling and air channel, and the received signal e can be expressed as:

(13)

d is the column vector composed of the symbols of all users, the user's respective spread spectrum scrambling code and the post-processed channel impulse response £ generation system transmission matrix A, the structure of the A matrix is as shown in Fig. 1, where Ns is the user symbol The length of the number, K is the number of users using a scrambling code, and Q is the spreading factor, as shown in the following equation -

Demodulate user data, Α = (Α" · Α)- .

A = [A ^kl A ^kz LA ^Ka ]', Ka is the number of antennas, Ka ≥ N (14)

A ^to is a block diagonal array, the block ^" can be expressed as

=[ 3⁄4 I 3⁄4 3⁄4 3⁄4 L 3⁄4 3⁄4 3⁄4 I 3⁄4J (15) User data demodulation with matrix A, you can use whitening matched filter algorithm (Li F), zero-forcing linear block equalizer algorithm (ZF), minimum mean A variety of joint detection algorithms such as variance algorithm (MMSE). Here, the minimum mean square error algorithm is used as an example for data estimation.

d = (A ^H R _n - ^l A + R _d - ^l T ^{l H} R -'e (16) Thereby, demodulated data of each user is obtained.

Compared with the prior art, the method disclosed in the method of the present invention uses multiple scrambling codes and their corresponding channel estimation and joint detection to increase system capacity in one cell, which greatly improves the capacity of the TD-SCDMA system. , improve the frequency band utilization, or equivalently improve the data throughput rate of a single cell.

The present invention is applicable to all applications to all CDMA systems, particularly TDD-CDMA systems and SCDMA systems, and improvements to other systems can also be made with reference to the above-described concepts of the method of the present invention. Any engineer having a background in signal processing, communication, etc., can design corresponding channel estimation and joint detection devices in accordance with the present invention, which are all included in the spirit and scope of the present invention, and thus are protected by the appended claims. range.

Claims

Rights request

A multi-scrambling code configuration and associated demodulation method in a CDMA system. In a multi-antenna system, system capacity is increased by configuring a multi-scrambling code, which includes:

a), each cell is configured with multiple scrambling codes, and the method and quantity for generating the spreading code under each scrambling code are consistent with the conventional methods;

2. The method according to claim 1, wherein the method further comprises: -), the system groups the users served by the cell, the number of groups is equal to the number of scrambling codes, and each group of users is orthogonal according to a conventional a method of variable spreading factor code to allocate a channelization code;

3. The method according to claim 2, wherein the method further comprises: performing channel estimation by using: a basic Midamble code, a total of NK voice users, and K is a user using a scrambling code For the number of natural numbers, the channel estimation window length W of each user is reduced to ^.

NK

The method according to claim 3, wherein the number of scrambling codes configured per cell in the method is N=2.

5. The method according to claim 1, wherein the method is applied to a TD-SCDMA system, and further includes the step of allocating a plurality of scrambling codes:

All), a total of 128 scrambling codes are defined, and are divided into 32 groups, each of which has 4 scrambling codes, and cross-correlation is performed on 4 scrambling codes in the selected group;

Bll), preferentially select a scrambling code with small cross-correlation as the scrambling code used by the cell, and the value of the number of scrambling codes N The range is 2 ≤ N ≤ 4.

6. The method according to claim 5, wherein the method further comprises a channel estimation step - cl l for the plurality of scrambling codes, a G matrix, and a cyclic right shift matrix G composed of a basic Midamble code. Assuming that N scrambling codes are simultaneously used in a cell, the number of users is increased by N times, and a total of NK voice users are used. The channel estimation window length W of each user is reduced to ^, and K is the number of users using one scrambling code;

NK

Dl l), the training sequence for receiving data is proposed for channel estimation, the maximum likelihood estimate of the channel impulse response, = G; setting a noise threshold Γ, performing channel post-processing on the estimated £, greater than the threshold threshold retention, Less than the decision threshold is set to zero;

Ell), obtain the channel impulse response of each user, and take NK values from & in the length of W to form each user channel estimation, and the channel response window length of each user is w.

7. The method of claim 6 wherein the method further comprises the step of jointly detecting the plurality of scrambling codes:

Multi-user joint detection data demodulation of multiple scrambling codes, specifically, using user's respective spread spectrum scrambling code and channel impulse response after channel processing to generate matrix A, demodulating user data -