CN1242629C - A multi-user self-adaptive packet layering spacetime signal sending-receiving system - Google Patents

Abstract

The multi-user adaptive grouping layered space-time signal transceiving system of the present invention is characterized in that the base station uses the total system theoretical capacity or the condition number of the channel as the optimization target according to the estimated value of the channel response matrix from all users to the base station, and adaptively The antennas of each user are grouped; the grouping information is fed back to each user through the downlink channel; the data carried by the antennas of each user is multiplied by the orthogonal code corresponding to the group; the base station receiver decodes the orthogonal code, and the data sent by the antenna The signal is separated into several groups; the multiple transmitting antennas in each group and the receiving antennas of the base station constitute a layered space-time system, and the interference cancellation demodulation algorithm of the layered space-time system is used to demodulate the data of each antenna, and after reordering, all user data. The invention combines the traditional uplink multi-user code division multiple access mode, utilizes the degree of freedom of the space channel, improves the system performance; especially under the bad space channel condition, it can avoid the deterioration of the system performance.

Description

A multi-user adaptive grouping layered spatio-temporal signal transceiving system

Technical field:

The invention belongs to the field of mobile communication multiple-input multiple-output (MIMO) antenna signal processing technology and mobile communication uplink multiple access technology, and particularly relates to a multi-user adaptive grouping layered spatio-temporal signal transceiving system for improving system performance under non-ideal spatial channels and Its grouping optimization method.

Background technique:

Multiple-input multiple-output (MIMO) antenna signal processing technology can increase system capacity and improve system performance by utilizing the degree of freedom of spatial channels. MIMO technology has become one of the research hotspots in recent years, and has begun to be applied in practical systems.

According to the introduction of "Signals, Systems, and Electronics, International Symposium on, 29 Sep-20ct 1998; Page(s): 295-300), proposed by Bell Laboratories, USA A multi-antenna technology called Vertica-Belaboratories Layered Space-Time (V-BLAST technology for short), uses the degree of freedom of the spatial channel to increase the capacity of the system. The multi-channel antennas at the transmitting end transmit multiple independent data concurrently, and the receiving end jointly processes all the signals received by multiple antennas, and demodulates all the transmitted data by using the multi-user interference cancellation algorithm. Since all the data are sent simultaneously in the same frequency band, no additional communication bandwidth resources are occupied, the system using the V-BLAST technology significantly improves the spectrum utilization efficiency. However, V-BLAST technology is a point-to-point communication technology and does not support simultaneous communication of multiple users.

"International Institute of Electronics and Electrical Engineers International Conference on Communications 2001" (Communications, IEEE International Conference on, 2001; Page (s): 565-569 vol.2.) gives a multi User synchronous V-BLAST system, each user in the system has one or more transmit antennas, the multi-user synchronous V-BLAST system uses orthogonal codes to divide users into multiple groups, and the antennas of all users in each group and Multiple receiving antennas of the base station constitute an equivalent V-BLAST system; since the multi-user synchronous V-BLAST system utilizes the degree of freedom of the spatial channel, the capacity of the system is improved to a certain extent.

However, orthogonal codes group all the antennas, and there are many possible combinations of groups. Under the same channel conditions, various grouping combinations lead to different system performance; especially in the case of non-ideal spatial channels, such as The performance difference caused by different groups is particularly obvious when the channel correlation between some transceiver antenna pairs is strong, or when the "keyhole phenomenon" (KEYHOLE) occurs in the spatial channel. However, under certain channel conditions, there are grouping combinations with better performance. How to select grouping combinations with better performance is an unsolved problem in the prior art; in addition, the channels in the actual system are constantly changing, so the grouping combination of channels The way needs to change with the change of the channel in order to ensure that the performance of the system is always optimized. But so far there is no report on how to obtain a grouping combination method that is beneficial to system performance improvement.

Technical content:

Aiming at the above-mentioned deficiencies of the prior art, the present invention proposes an adaptive multi-user packet layered space-time signal transceiving system applied to the uplink synchronization link of a mobile communication system. The base station selects this One of the two grouping optimization methods provided by the invention can adaptively group the antennas of each user to effectively improve system performance.

The present invention is applied to a multi-user adaptive packet layered space-time signal transceiving system for an uplink synchronous link of a mobile communication system, comprising a mobile user sending module and a base station receiver (10);

Assume that the system includes K users numbered U_1,..., U_K; the kth mobile user transmission module has m _k transmission antennas, and the K users in the system have M transmission antennas in total, denoted as [A ₁ , A ₂ , ..., A _M ];

The kth mobile user sending module modulates the data (1) to be sent by the modulation module (2), and the serial-to-parallel conversion module (3) takes out m _k N _d continuous modulated data as a data block , be set as the nth block, it is converted into a data stream with m _k road length as N _d , and each road data stream corresponds to a transmitting antenna; the multiplexing module (5) inserts the data stream in each road data stream with time division multiplexing Based on the training sequence (4) with a length of N _p estimated by the channel, a data stream (7) with a training sequence of length N _B =N _d +N _p is formed; let T be the duration of each data symbol, the nth The time length of block data block transmission is T _B =N _B T, in the time period (n-1)T _B ~nT _B , each data stream (7) with training sequence is transmitted within the duration of each data symbol Multiply the orthogonal code sequence (6) assigned to it by the base station to become a transmitted signal, and this step is called a channelization process; the above-mentioned signals sent by all transmitting antennas (8) simultaneously arrive at the base station receiver (10) through the uplink channel (9) );

The antenna (12) of the base station receiver (10) sends the received signals of all users to the de-orthogonal code module (13); the orthogonal code sequences (6) used by the antennas of all users have G different sequence, the de-orthogonal code module (13) utilizes the de-orthogonal code technology to decompose the signals of all users into corresponding G grouped signals according to G different orthogonal code sequences, wherein each grouped signal includes the same The received signal (15) corresponding to the training sequence and the received signal (14) corresponding to the transmission data; for any signal grouped, the channel estimation module (16) uses the received signal (15) corresponding to the training sequence as input, and the grouped Each transmit antenna (8) to each base station receive antenna (12) within the time period (n-1) T _B ~ _nTB within the channel H _p channel estimation, obtain channel estimate value (17); V-BLAST data solution The modulation module (19) uses the channel estimation value (17) and the above-mentioned received signal (14) corresponding to the transmission data, and invokes the multi-user interference cancellation demodulation algorithm to demodulate the time period (n-1) T _B ~ nT Data (20) transmitted on all transmitting antennas in _B ;

It is characterized by:

The base station receiver (10) also inputs the channel estimation value (17) in the time period (n-1) _TB ～ _nTB obtained by the channel estimation module (16) to the optimization grouping module (18), and the optimization grouping module (18) According to the channel conditions, choose one of the grouping optimization methods (24) given below to optimize the grouping of all antennas in the time period _nTB ~ (n+1) _TB , and divide the M antennas of K users The transmitting antennas [A ₁ , A ₂ ,..., A _M ] are divided into G groups, and the grouping vector F=[F ₁ , F ₂ ,..., F _M ](25), where F _i represents The i antennas are divided into the F _i group, F _i ∈ {1, 2, ..., G}; the base station informs each user through the downlink feedback channel (11) that in the time period nT _B ~ (n+1)T _B In the group number of each antenna, it is assumed that the mobile user knows that the i-th antenna is divided into the p=F _i group, and the unique orthogonal code sequence corresponding to this group is W _p , and its length is L, Taken from a set containing G Walsh (WALSH) codes, W=[W ₁ , W ₂ ,..., W _G ], the channelization process in the time period nT _B ~ (n+1)T _B Among them, the orthogonal code sequence (6) multiplied by the corresponding data stream (7) with the training sequence on the first antenna is S _i =W _p ; the optimization grouping module (18) also divides the grouping vector Send to recombination and parallel-to-serial conversion module (21), recombination and parallel-to-serial conversion module (21) utilizes the grouping vector (25) of the user corresponding to sending data time period, above-mentioned V-BLAST data demodulation module (19) The demodulated data (20) is sequenced and reassembled, and restored to corresponding data (22, 23) of each user after parallel-to-serial conversion.

One of the grouping optimization methods that can be selected by the optimization grouping module (18) proposed by the present invention is to use the total system theoretical capacity as the optimization target, calculate the sum of the channel theoretical capacity weighted under various grouping modes, and select to maximize it The grouping vector F _cap of

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; cap</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}}\underset{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; \}</span>$

Parameters in the formula: grouping vector F _cap = [F ₁ , F ₂ , ..., F _i , ..., F _M ], F _i ∈ {1, 2, ..., G}; G is the number of groups ; M is the total number of transmit antennas in the system; the optimization objective is the weighted sum of G capacity values C _p ,

${\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; det</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}}{{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$

is the capacity of all the M _p transmitting antennas in the p-th group and the N receiving antennas of the base station to form an equivalent M _p ×N V-BLAST subsystem, where I _N is the unit matrix of N×N; M _p is the number of transmitting antennas of the pth group; $h_p=[{\mathrm{<figure-callout\; id="1"\; label="h\; p"\; filenames="C0313196400111.png,C0313196400121.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}{\mathrm{<figure-callout\; id="2"\; label="h\; p"\; filenames="C0313196400111.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}...,h_{p_{m_p}}]$ is the channel response matrix of the pth group, its elements are also vectors, h _i =[h _{1, i} h _{2, i} ,..., h _{N, i} ] ^T , [v] ^T represents the transposition of vector v, h _{j, i} is the channel response coefficient from sending element i to receiving element j; (p ₁ , p ₂ ,...,p _Mp ) is their corresponding positions in the sending sequence; ρ _p is the Signal-to-noise ratio of the V-BLAST subsystem at the receive antenna.

The second grouping optimization method that the present invention proposes that can be selected for use by the optimization grouping module (18) uses the condition number of the channel as the optimization target, calculates the sum of the channel matrix condition number weighted under various grouping modes, and selects to minimize it The grouping vector F _cond of

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; cond</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; Cond</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}}\underset{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; Cond</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$

Parameters in the formula: grouping vector F _cond =[F ₁ , F ₂ ,..., F _i ,..., F _M ], F _i ∈ {1, 2,..., G}; G is the number of groups ; Cond(H _p ) is the condition number of H _p ; where $h_p=[{\mathrm{<figure-callout\; id="1"\; label="h\; p"\; filenames="C0313196400111.png,C0313196400121.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}{\mathrm{<figure-callout\; id="2"\; label="h\; p"\; filenames="C0313196400111.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}...,h_{p_{m_p}}]$ is the channel response matrix of the pth group, its elements are also vectors, h _i =[h _{1, i} h _{2, i} ,..., h _{N, i} ] ^T , h _i =[h _{1, i} h _{2, i} ,..., h _{N, i} ] ^T = [v] ^T , [v] ^T represents the transposition of vector v, h _{j, i} is the channel response coefficient from sending element i to receiving element j; (p ₁ , p ₂ ,..., p _Mp ) are their corresponding positions in the transmission sequence; H _p is decomposed by SVD singular value to obtain H _p =UΛV, U, V are unitary matrix, diagonal matrix Λ=diag{Λ ₁ , Λ ₂ , ..., Λ _r , 0, ...} are singular values of H _p , and the condition number of the channel is defined as the ratio of the largest singular value to the smallest singular value.

The optimization grouping module (18) can also select other grouping optimization methods according to channel conditions.

Compared with the prior art, since the multi-user grouping layered spatio-temporal signal transceiving system of the present invention adaptively utilizes the degree of freedom of the spatial channel, the performance of the multi-user system is significantly improved, especially in the case of non-ideal spatial channels Under this condition, the deterioration of multi-user system performance can be effectively avoided. The wireless communication channel is complex and changeable. Under the same channel, various group combinations of multi-user packet layered space-time transceiver system may lead to different system performance. In the case of non-ideal spatial channels, for example, the channel correlation between some transceiver antenna pairs is strong, or when the KEYHOLE phenomenon occurs, the performance difference of different groups is particularly obvious, and improper grouping will cause the system to fail to communicate effectively . If only according to the traditional multiple access method, each user fixedly uses an orthogonal code, the system adopts multi-antenna technology and cannot guarantee the system performance to be improved, and when the spatial channel is not ideal, the system performance will suffer serious deterioration . The two grouping optimization methods provided by the present invention can ensure that the grouping made by the base station is optimized every time, and the grouping method based on one of the two methods will try to divide the antennas with strong correlation into different groups, and divide the affected antennas into different groups. The antennas seriously affected by KEYHOLE are divided into different groups; in each equivalent sub-V-BLAST system, the equivalent channel response matrix has a good structure, which is beneficial to reduce the bit error rate. The channel in the actual system changes, and the way of combining packets also needs to change accordingly. The self-adaptive grouping layered space-time transceiving system proposed by the present invention provides a self-adaptive grouping platform, which can ensure correct and timely execution of the optimized grouping results and ensure that the performance of the system is always optimized. Compared with fixed grouping or random grouping, the multi-user self-adaptive grouping V-BLAST system of the present invention can effectively improve the performance of the system and improve the reliability of wireless communication.

Description of drawings:

Fig. 1 is a structural diagram of the adaptive packet V-BLAST system of the multi-user sending part;

Fig. 2 is a structural diagram of the adaptive packet V-BLAST system in the receiving part of the base station.

Fig. 3 is a performance comparison chart of the adaptive grouping V-BLAST system.

Detailed ways:

Embodiments of the present invention are described below in conjunction with the accompanying drawings.

Example 1:

This embodiment takes a synchronous uplink system with K=4 users as an example to describe a multi-user adaptive grouping layered spatio-temporal signal transceiving system applied to the uplink synchronous link of a mobile communication system.

Assume that each user has 3 transmit antennas, _and K=4 users have M=12 transmit antennas in total, denoted as [A ₁ , A ₂ , . The base station has N=4 receiving antennas.

The transmission module of the kth user modulates the data (1) to be transmitted by the modulation module (2), and the serial-to-parallel conversion module (3) takes 3 N _d continuous modulated data as a data block, which is set as In the nth block, it is converted into m _k =3 road length data streams of N _d , each road data stream corresponds to a transmitting antenna, and the multiplexing module (5) is inserted in each road data stream in time division multiplexing for The length of channel estimation is the training sequence (4) of N _p , forms the data stream (7) with training sequence, and length is N _B =N _d +N _p ; Let T be the duration of each data symbol, the nth block data The time length of block transmission is T _B = N _B T, N _p , N _d needs to be set reasonably according to the speed of channel change. It is required that N _p is greater than the number of transmitting antennas and T _B is less than the channel correlation time. N _d /N _p is as large as possible to improve the system channel utilization efficiency. In the time period (n-1)T _B ~nT _B , each data stream with training sequence (7) is multiplied by the orthogonal code sequence (6) assigned to it by the base station within the duration of each data symbol, This step is called channelization process; all transmitting antennas (8) transmit simultaneously, and the signal reaches the base station receiver (10) through the uplink channel (9).

The antennas of all users are arranged in order, and the channel response matrix of uplink multi-antennas in the time period (n-1) _TB ~ _nTB is defined: H=[h ₁ , h ₂ , h ₃ , h ₄ ,..., h ₁₂ ], whose elements are also vectors, h _i =[h _{1, i} h _{2, i} ..., h _{4, i} ] ^T , h _{j, i} is the channel response coefficient from sending element i to receiving element j. The data streams (7) with training sequences corresponding to each multi-antenna channel are arranged as B=[b ₁ , b ₂ , b ₃ , . . . , b ₁₂ ]. Due to the role of ideal power control in a multi-user system, the difference in the average fading of each user can be ignored, which is equivalent to sending data with the same power for each user. Let the matrix T=diag(S ₁ , S ₂ ,...,S _i ,...,S ₁₂ ) ^T composed of orthogonal code sequences (6) used by all antennas of all users, where S _i represents the i-th antenna Orthogonal code sequence to use. The signal vector r=(r ₁ , r ₂ ,..., r ₄ ) ^T received on all antennas of the base station (10) is:

r=HTB+n

n is Gaussian white noise.

可以提取第p组的信号：The antenna (12) of the base station receiver (10) sends the received signals of all users to the de-orthogonal code module (13). The orthogonal code sequences (6) used by the antennas of all users have a total of G=4 different sequences, and antennas using different orthogonal code sequences are considered to be in different groups, and all members of the p-th group are assumed to use orthogonal codes W _p , the p-th group has M _p antennas in total, and their corresponding positions in the transmission sequence are (p ₁ , p ₂ ,...,p _Mp ), and the orthogonal code solution module (13) uses the orthogonal code W _p Correlation calculation with the received signal (12) on the antenna (12): r=(r ₁ , r ₂ ,..., r ₄ ) ^T

The signal of group p can be extracted:

r _p = (r _p1 , r _p2 , . . . , r _p4 ) ^T

The data streams of the M _p transmitting antennas and the r _p =(r _p1 ,r _p2 ,...,r _p4 ) ^T received by the N=4 receiving antennas of the base station constitute an equivalent M _p ×4 V -BLAST subsystem:

r _p =H _p B _p +n _p

where n _p is Gaussian white noise, $B_p={[{\mathrm{<figure-callout\; id="1"\; label="b\; p"\; filenames="C0313196400111.png,C0313196400121.png"\; state="\{\{state\}\}">b</figure-callout>}}_{p_{}}{\mathrm{<figure-callout\; id="2"\; label="b\; p"\; filenames="C0313196400111.png"\; state="\{\{state\}\}">b</figure-callout>}}_{p_{}}...,b_{p_{m_p}}]}^T$ is the data to be sent in the subsystem, $h_p=[{\mathrm{<figure-callout\; id="1"\; label="h\; p"\; filenames="C0313196400111.png,C0313196400121.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}{\mathrm{<figure-callout\; id="2"\; label="h\; p"\; filenames="C0313196400111.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}...,h_{p_{m_p}}]$ is the channel response matrix of the pth group. There are totally G=4 such equivalent V-BLAST subsystems.

For any grouping, set the p-th grouping, the above-mentioned separated group signal includes two parts: the received signal (15) corresponding to the training sequence and the received signal (14) corresponding to the transmitted data; the channel estimation module (16) With the received signal (15) corresponding to the training sequence as input, carry out channel estimation to the channel _Hp of each transmitting antenna (8) to each base station receiving antenna (12) in the pth grouping, obtain the channel estimated value (17); The V-BLAST data demodulation module (19) utilizes the channel estimation value (17) and the above-mentioned received signal (14) corresponding to the transmission data, and invokes the multi-user interference to eliminate the demodulation algorithm to demodulate the time period (n-1 ) Data on all antennas of the p-th group in ) _TB ~ _nTB , carry out similar above-mentioned demodulation process to all groups, obtain the data on all antennas in the system (20); recombination and parallel-to-serial conversion module (21) utilize Obtain the user grouping vector (25) corresponding to the time period of sending data from the following optimization grouping module (18), the data (20) demodulated by the above-mentioned V-BLAST data demodulation module (19) is ordered, and serialized After the conversion, restore the data (22, 23) of each user within the corresponding time period (n-1) _TB to _nTB ;

The above-mentioned channel estimation module (16) of this embodiment obtains channel estimation values (17) from all transmitting antennas (8) to each base station receiving antenna (12) within the time period (n-1) _TB ~ _nTB , except as The input of the V-BLAST data demodulation module (19) is also input to the optimization grouping module (18), and the optimization grouping module (18) selects one of the following two grouping optimization methods (24) for the next time period according to channel conditions All antennas within _nTB to (n+1) _TB are optimized into groups, and the optimization group is performed every _TB . _The optimal grouping module (18) divides M=12 transmitting antennas [A ₁ , _{A 2} , . F ₂ ,..., F _i ,..., F ₁₂ ], where F _i ∈ {1, 2, ..., 4}, F _i means that the i-th antenna is divided into the F _i- th group;

When the total system theoretical capacity is selected as the optimization target, the grouping vector

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; cap</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}}\underset{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; \}</span>$

in, $C_p={\log }_2\left\{\det (I_N+\frac{{\mathrm{\&rho;}}_p}{m_p}{h}_p{h}_p^{*})\right\};$ Parameters in the formula: grouping vector F _cap = [F ₁ , F ₂ , ..., F _i , ..., F _M ], F _i ∈ {1, 2, ..., G},; G=4 is the number of groups; ρ _p is the signal-to-noise ratio of the V-BLAST subsystem formed by the p-th group at the receiving antenna; M _p is the number of sending antennas in the p-th group; $h_p=[{\mathrm{<figure-callout\; id="1"\; label="h\; p"\; filenames="C0313196400111.png,C0313196400121.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}{\mathrm{<figure-callout\; id="2"\; label="h\; p"\; filenames="C0313196400111.png"\; state="\{\{state\}\}">h</figure-callout>}}_{p_{}}...,h_{p_{m_p}}]$ is the channel response matrix of the pth group, h _i =[h _{1, i} h _{2, i} ,..., h _{N, i} ] ^T , h _{j, i} is the channel response coefficient from sending element i to receiving element j ; _IN is a 4×4 unit matrix; M=12 is the total number of transmitting antennas in the system; F=[F ₁ , F ₂ , . . . , F ₁₂ ] is a grouping vector. This method is to calculate the sum of the weighted channel theoretical capacities under various grouping modes, and select the grouping mode F which maximizes it.

When the condition number of the channel is chosen as the optimization target, the grouping vector

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; cond</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; Cond</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}}\underset{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; Cond</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$

Parameters in the formula: grouping vector F _cond = [F ₁ , F ₂ , ..., F _i , ..., F _M ], F _i ∈ {1, 2, ..., G}; G=4 is The number of groups; Cond(H _p ) is the condition number of H _p . The definition of the condition number is: perform singular value decomposition (SVD) on H _p to get H _p = UΛV, U, V are unitary matrix, diagonal matrix Λ = diag{Λ ₁ , Λ ₂ ,..., Λ _r , 0,...}'s diagonal elements are the singular values of H _p , and the condition number is defined as the ratio of the largest singular value to the smallest singular value. The method is to calculate the sum of the condition numbers of the weighted channel matrix in various grouping modes, and select the grouping mode F which minimizes it.

The optimization grouping module (18) can also select other grouping optimization methods according to channel conditions.

The optimization grouping module (19) outputs the grouping vector (25), and the base station notifies each user of the group number where each antenna is located in the next time period nTB _~ (n+1) _TB through the downlink feedback channel (11). It is known that the i-th antenna is divided into the p=F _i-th group, and the unique orthogonal code sequence corresponding to this group is W _p , whose length is L=4, taken from a set containing G=4 WALSH codes, W=[W ₁ , W ₂ ,...,W ₄ ], in the above channelization process in the next time period _nTB ~ (n+1) _TB , the corresponding data with training sequence on the i-th antenna The above-mentioned orthogonal code sequence (6) multiplied by the stream (7) is S _i =W _p ; at the same time, the grouping vector (25) of the time period nT _B ~ (n+1) _TB will also be input to the recombination and parallel string The conversion module (21) enables the module to correctly restore the data of each user within the time period _nTB to (n+1) _TB .

The key of this system is that the base station optimizes the combination of packets. The goal of optimization is: under the given power and modulation mode, the capacity of the system is given, so that the bit error rate of the system is as small as possible. Because the V-BLAST demodulation algorithm has a nonlinear interference cancellation process, the bit error rate of the system cannot be expressed quantitatively, and the optimal grouping scheme cannot be accurately found in the grouped V-BLAST. The total theoretical capacity of the system or the condition number of the channel are two numbers that are indirectly related to the bit error rate. In this embodiment, the two numbers are optimized by using one of the two grouping methods mentioned above to obtain a practically feasible approximation to the optimal The optimal grouping scheme significantly improves the performance of the system.

In order to evaluate the improvement of multi-user system performance by the present invention, a computer numerical simulation is carried out on a synchronous uplink system with K=4 users. In the simulation, all users adopt 16QAM digital modulation mode, the wireless channel model adopted is quasi-static flat Rayleigh fading model, and the spatial channel model from M=12 transmitting antennas to receiving antennas is a mixed model, including independent channel conditions, correlated channel Situation, KEYHOLE channel situation.

H = [H ₁ , H ₂ , H ₃ , H ₄ , ..., H ₁₂ ] = [A, B, C, D], H _i = [H _{1, i} H _{2 , i} ..., H _{4, i} ] ^T , H _{j, i} is the channel response coefficient from sending element i to receiving element j. Each element in A=[H ₁ , H ₂ , H ₃ ] and B=[H ₄ , H ₅ , H ₆ ] is a complex Gaussian random variable with independent and identical distribution. C=[H ₇ , H ₈ , H ₉ ] is a channel with spatial correlation, and D=[H ₁₀ , H ₁₁ , H ₁₂ ] is a channel model when the KEYHOLE phenomenon exists. If the user is grouped and each antenna of the same user uses the same orthogonal code, the corresponding grouping vector is F=[1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4] , which is exactly the same as in the case of traditional multi-user multiple access, where orthogonal codes are used to distinguish users. The channel matrices of each equivalent V-BLAST subsystem are exactly A, B, C, D. This method of grouping users is used as a reference comparison use case for evaluation. Similarly, randomly grouping M=12 transmit antennas is also a reference comparison case. Figure 3 shows a performance comparison diagram of the adaptive grouping V-BLAST system. The abscissa in the figure is the average signal-to-noise ratio expressed in dB, and the ordinate is the average bit error rate of the system. Curve 1 shows the system performance of grouping by method 1, curve 2 shows the system performance of grouping by method 2, curve 3 shows the system performance of random grouping, and curve 4 shows the system performance of grouping by users. It can be seen from the figure that the performance is extremely deteriorated under such non-ideal channel conditions according to the fixed way of grouping users. The performance of the random grouping method is also not ideal. Compared with the fixed grouping method or the random grouping method, the multi-user adaptive grouping V-BLAST system can effectively improve system performance and improve the reliability of wireless communication under the two optimization methods proposed by the present invention.

Claims (3)

1. A multi-user adaptive grouping layered space-time signal transceiving system applied to the uplink synchronous link of a mobile communication system, comprising a mobile user sending module and a base station receiver (10); Assume that the system includes K users numbered U_1,..., U_K; the kth mobile user transmission module has m _k transmission antennas, and the K users in the system have M transmission antennas in total, denoted as [A ₁ , A ₂ ,...,A _M ]; The kth mobile user sending module modulates the data (1) to be sent by the modulation module (2), and the serial-to-parallel conversion module (3) takes out m _k N _d continuous modulated data as a data block , be set as the nth block, it is converted into a data stream with m _k road length as N _d , and each road data stream corresponds to a transmitting antenna; the multiplexing module (5) inserts the data stream in each road data stream with time division multiplexing Based on the training sequence (4) of length _Np estimated by the channel, a data stream (7) with a training sequence of length N _B =N _d +N _p is formed; let T be the duration of each data symbol, the nth The time length of block data block transmission is T _B =N _B T, in the time period (n-1)T _B ~nT _B , each data stream (7) with training sequence is transmitted within the duration of each data symbol Multiply the orthogonal code sequence (6) assigned to it by the base station to become a transmitted signal, and this step is called a channelization process; the above-mentioned signals sent by all transmitting antennas (8) simultaneously arrive at the base station receiver (10) through the uplink channel (9) ); The antenna (12) of the base station receiver (10) sends the received signals of all users to the de-orthogonal code module (13); the orthogonal code sequences (6) used by the antennas of all users have G different sequence, the de-orthogonal code module (13) decomposes the signals of all users into corresponding G grouped signals according to G different orthogonal code sequences, wherein each grouped signal includes a received signal corresponding to the training sequence ( 15) and the received signal (14) corresponding to the transmission data; for the signal of any one grouping, the channel estimation module (16) takes the received signal (15) corresponding to the training sequence as input, and each transmitting antenna (8) in the grouping The channel H _p in the time period (n-1) _TB ～ _nTB of each base station receiving antenna (12) performs channel estimation to obtain the channel estimation value (17); the layered space-time data demodulation module (19) utilizes this The channel estimation value (17) and the above-mentioned received signal (14 ₎ corresponding to the transmission data, and call the multi-user interference cancellation demodulation algorithm to demodulate and _get data(20); It is characterized by: The base station receiver (10) also inputs the channel estimation value (17) in the time period (n-1) _TB ～ _nTB obtained by the channel estimation module (16) to the optimization grouping module (18), and the optimization grouping module (18) Select the grouping optimization method (24) to optimize the grouping of all antennas in the time period nT _B ~ (n+1) _TB , and put the M transmitting antennas of K users [A ₁ , A ₂ ,..., A _M ] is divided into G groups, record the group vector F=[F ₁ , F ₂ ,..., F _M ] (25), where F _i represents that the i-th antenna is divided into the first group F _i , F _i ∈{1,2,...,G}; the base station notifies each user of the group number of each antenna in the time period nT _B ~ (n+1) _TB through the downlink feedback channel (11), assuming that the mobile user knows The i-th antenna is divided into the p=F _i group, and the unique orthogonal code sequence corresponding to this group is W _p , whose length is L, taken from a set containing G Walsh codes, W =[W ₁ , W ₂ ,...,W _G ], during the channelization process within the time period _nTB ~ (n+1) _TB , the corresponding data stream with training sequence on the i-th antenna ( 7) The multiplied orthogonal code sequence (6) is S ₁ =W _p ; the optimization grouping module (18) also sends the grouping vector in each time period to the reorganization and parallel-serial conversion module (21), and the reorganization and parallel-serial conversion module (21) The conversion module (21) reorganizes the data (20) demodulated by the layered spatio-temporal data demodulation module (19) by using the user's grouping vector (25) corresponding to the time period for sending data, and restores them after parallel-to-serial conversion is the data (22, 23) of each corresponding user. 2. The multi-user adaptive grouping layered space-time signal transceiver system as claimed in claim 1 is characterized in that the grouping optimization method (24) selected by the optimization grouping module (18) is to calculate weighted channels under various grouping modes The sum of theoretical capacities, choose the grouping vector F _cap that maximizes it, that is ${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; acp</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}}\underset{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; \}</span>$ Parameters in the formula: grouping vector F _cap = [F ₁ , F ₂ , ..., F _i , ..., F _M ], F _i ∈ {1, 2, ..., G}; G is the number of groups ; M is the total number of transmit antennas in the system; the optimization objective is the weighted sum of G capacity values C _p , ${\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; det</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}}{{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$ is the capacity of all M _p transmitting antennas in the p-th group and the N receiving antennas of the base station to form an equivalent M _p ×N hierarchical space-time subsystem capacity, where I _N is the unit matrix of N×N; M _p is the number of transmitting antennas of the pth group; $h_p=[h_{p1},h_{p2},\&CenterDot;\&CenterDot;\&CenterDot;,{h_{\mathrm{pM}}}_p]$ is the channel response matrix of the pth group, its elements are also vectors, h _i =[h _{1, i} h _{2, i} ,..., h _{N, i} ] ^T = [v] ^T , [v] ^T represents the vector The transpose of v, h _{j, i} is the channel response coefficient from sending element i to receiving element j; (p ₁ , p ₂ ,...,p _Mp ) is their corresponding position in the sending sequence; ρ _p is the The signal-to-noise ratio of the layered spatio-temporal subsystem composed of p groups at the receiving antenna. 3. The multi-user adaptive grouping layered space-time signal transceiving system as claimed in claim 1, characterized in that the grouping optimization method (24) selected by the optimization grouping module (18) is to calculate weighted channels under various grouping modes The sum of matrix condition numbers, select the grouping vector F _cond that minimizes it, ie ${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; cond</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arg</span>}\underset{\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; Cond</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}}\underset{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; Cond</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \}</span>$ Parameters in the formula: grouping vector F _cond =[F ₁ , F ₂ ,..., F _i ,..., F _M ], F _i ∈ {1, 2,..., G}; G is the number of groups ; Cond(H _p ) is the condition number of H _p ; where $h_p=[h_{p1},h_{p2},\&Center\; Dot;\&Center\; Dot;\&Center\; Dot;,h_{{\mathrm{pM}}_p}]$ is the channel response matrix of the pth group, its elements are also vectors, h _i =[h _{1, i} h _{2, i} ,..., h _{N, i} ] ^T , h _i =[h _{1, i} h _{2, i} ,..., h _{N, i} ] ^T = [v] ^T , [v] ^T represents the transposition of vector v, h _{j, i} is the channel response coefficient from sending element i to receiving element j; (p ₁ , p ₂ ,...,p _Mp ) are their corresponding positions in the sending sequence; Singular value decomposition is performed on H _p to obtain H _p =UΛV, U, V are unitary matrix, diagonal matrix Λ=diag{Λ ₁ , Λ ₂ ,..., Λ _r , 0,...} are the singular values of H _p , and the condition number of the channel is defined as the ratio of the largest singular value to the smallest singular value.

Images