CN1154272C - Receiving method based on interference elimination - Google Patents
Receiving method based on interference elimination Download PDFInfo
- Publication number
- CN1154272C CN1154272C CNB011263962A CN01126396A CN1154272C CN 1154272 C CN1154272 C CN 1154272C CN B011263962 A CNB011263962 A CN B011263962A CN 01126396 A CN01126396 A CN 01126396A CN 1154272 C CN1154272 C CN 1154272C
- Authority
- CN
- China
- Prior art keywords
- user
- symbol
- group
- signal sequence
- level
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000003379 elimination reaction Methods 0.000 title claims abstract description 32
- 230000008030 elimination Effects 0.000 title claims abstract description 31
- 108010076504 Protein Sorting Signals Proteins 0.000 claims abstract description 156
- 238000001514 detection method Methods 0.000 claims abstract description 67
- 238000011069 regeneration method Methods 0.000 claims description 25
- 230000008929 regeneration Effects 0.000 claims description 21
- 238000001228 spectrum Methods 0.000 claims description 20
- 238000001914 filtration Methods 0.000 claims description 11
- 230000001172 regenerating effect Effects 0.000 claims description 6
- 230000006978 adaptation Effects 0.000 claims description 2
- 238000012360 testing method Methods 0.000 abstract 1
- 230000003044 adaptive effect Effects 0.000 description 13
- 108010003272 Hyaluronate lyase Proteins 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005562 fading Methods 0.000 description 4
- 230000008054 signal transmission Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Images
Landscapes
- Noise Elimination (AREA)
Abstract
The present invention relates to a multiple user testing method based on interference elimination, which comprises the following steps: (1), M grade processing is carried out to all of users, in each grade, the users are divided into Gm sets according to the strength of signals, M and G are natural numbers, and m is equal to 1 to M; (2), in the first grade processing, the G1 set orderly receives baseband receiving signal sequences and residual error signal sequences after interference elimination of the prior set for carrying out the serial processing process; (3), in the second grade processing to the M grade processing, the sets in each grade orderly receive the output residual error signal sequences of the final set in the prior grade and the residual error signal sequences after the interference elimination of the prior set, and simultaneously, the sets respectively receive weighted symbol grade signal sequences outputted by users corresponding to the prior grade, carry out serial processing grade by grade and set by set for obtaining the judgment bit sequences of the users and output the judgment bit sequences of the users. The present invention can carry out multiuser detection in a self-adaptive mode under the time-variant moving channels and variable system conditions and can obviously enhance receiving performance.
Description
Technical Field
The invention relates to the field of spread spectrum communication, in particular to a method for receiving a multi-user signal at a base station end in code division multiple access spread spectrum communication.
Background
In a cell, a plurality of mobile phone users transmit signals to a base station, and the signals are received by a base station antenna and then input to a base station receiver for signal detection. In a code division multiple access spread spectrum communication system, digital information of each mobile phone user forms a baseband modulation signal through processes of source channel coding, spread spectrum modulation and the like, and then is transmitted through a high-frequency carrier. These high-frequency signals pass through an air radio propagation channel and then reach a base station receiving antenna to be received. At the receiving end of the base station, firstly, the low-pass filter removes the high-frequency carrier to obtain a baseband receiving signal, and then, the receiver detects the digital information of each user.
The basic detection principle of the conventional matched filter receiver is that the spread spectrum code information of each user is utilized, and correlation operation is performed on a received signal to find a correlation peak value, so that the signal of each user is extracted. This is based on the specific correlation property of the spreading codes, i.e. the correlation values between different spreading code words are small, and the correlation value is maximum when only the code word alignment is performed with autocorrelation. Such receivers typically include modules for matched filtering, RAKE maximum ratio combining, symbol hard decision, and the like, and of course, include functional modules for multipath searching, channel estimation, and the like to provide necessary delay information and channel parameter information for the receiving process.
The multi-user detection enhancement technology is based on the traditional receiver, further signal processing is carried out, and multiple access interference caused by the existence of other user signals is removed, so that more accurate bit estimation is obtained. A multi-user detection structure of a packet serial interference cancellation (GSIC) structure is a process for grouping all users of a transmission signal according to a certain rule, and performing interference signal regeneration, interference cancellation, re-matching filtering, RAKE combination and symbol hard judgment on the users group by group. The process of interference signal regeneration is actually the process of repeating the modulation transmission of the digital signal, and comprises the processes of spread spectrum modulation, multipath signal regeneration and the like. Since most of the interference of other users is eliminated during the re-matched filtering, more accurate bit decision can be obtained.
The multi-user detection technology of the multi-stage interference elimination structure is to further regenerate and eliminate the interference of the detection result obtained from the previous stage and to detect more and more accurate user data information step by step.
The GSIC detector structure using the packet serial interference cancellation has the disadvantage that under the fading channel condition varying with time and when other factors such as system load vary, the interference cancellation process directly subtracts the regenerated interference without considering that the regenerated interference signal may have a large estimation error, so that the interference cancellation cannot be correctly realized, and even the opposite effect of the interference amplification will be caused. This ultimately necessarily affects the detection performance, resulting in degraded or unstable receiver performance.
For further understanding of GSIC detectors, please refer to documents a. -l.johansson, l.k.rasmussen, Linear group wise temporal interference cancellation in cdma, IEEE Transactions on communications, 1998, 45 (5): pp.605-610.
Another important detection structure is a Parallel Interference Cancellation (PIC) detector for parallel processing of all users, but the parallel structure itself cancels the interference regeneration estimation signals of all users at the same time, and the interference of users with weak signals is also inaccurate in estimation, so that the interference cancellation operation of all users is seriously affected, and the detection performance is reduced. This effect can be partially removed by weighted interference cancellation, but the parallel structure treats each user equally with different strengths, and is structurally unsuitable.
Disclosure of Invention
The invention aims to provide a receiving method based on interference elimination, which can self-adaptively adjust the weight value of the interference elimination under the system condition of time-varying mobile channels and load variation, thereby obviously improving the multi-user detection performance.
The invention provides a multi-user detection method based on interference elimination, which comprises the following steps: all users are processed in M stages, and in each stage, the users are divided into G according to the signal intensitymGroup, M, GmIs a natural number, M is 1, … M; (II) in a first stage, G1And sequentially receiving the baseband receiving signal sequence and the residual signal sequence after the previous group of interference elimination, and performing the following serial processing processes: (1) carrying out symbol detection according to the group of input signal sequences, and regenerating each user signal sequence of the group of symbol levels by using the symbol detection value; (2) multiplying the symbol-level user signal sequences by corresponding user symbol-level weight signal sequences respectively to obtain symbol-level signal sequences weighted by users and outputting the symbol-level signal sequences; simultaneously, carrying out interference regeneration and interference elimination by using the symbol-level signal sequence weighted by each user, the known spreading code signal sequences of each user and the delay information of each path of each user to obtain a residual signal sequence subjected to the interference elimination of the group, and outputting the residual signal sequence; (III) in the second, …, M-stage process, each group of each stage receives the output residual signal sequence of the last group of the previous stage and the residual signal sequence of the last group after the interference is eliminated in turn, and simultaneouslyEach group respectively receives a weighted symbol-level signal sequence output by a corresponding user at the previous stage, and the serial processing step by step is as follows: (A) in each of the second, …, M-1 stages and the 1 st, …, G of the M-th stageM-1 group, performing the following serial processing: (1) carrying out symbol detection according to the two input signal sequences of the group, and regenerating each user signal sequence of the group of symbol levels of the current level by using the symbol detection value; (2) multiplying the symbol-level user signal sequence by the corresponding user symbol-level weight signal sequence of the current level to obtain a symbol-level signal sequence weighted by each user of the current level, carrying out interference regeneration and interference elimination by utilizing the symbol-level signal sequence weighted by each user and the weighted symbol-level signal sequence input by the group and the spread spectrum code signal sequence of each known user and the delay information of each path of each user to obtain a residual signal sequence subjected to the interference elimination of the group, and outputting the residual signal sequence; meanwhile, each group in the 2 … M-1 level outputs a symbol-level signal sequence weighted by each user respectively; (3) from the M-th stage GM-1 group respectively uses each user symbol detection value to obtain each user decision bit sequence and output; (B) at G of M stagemGroup, the following treatment processes are carried out: and carrying out symbol detection according to the two signal sequences input by the group, obtaining the decision bit sequence of each user of the group by using the symbol detection value of each user, and outputting the decision bit sequence.
In the above-mentioned multi-user detection method based on interference cancellation, each group of each stage in step (three) may also receive a baseband received signal sequence, and perform an interference cancellation operation.
In the above-mentioned multi-user detection method based on interference cancellation, all users in step (a) are grouped according to signal strength according to the data rate of the user.
In the above method for detecting multiple users based on interference cancellation, all users in step (one) may be grouped according to signal strength: when M is 2, G1 is G2 is 1, G1 is G2 is 2, G1 is G2 is 3, G1 is 2, G2 is 1, G1 is 3, G2 is 1.
In the above-mentioned multi-user detection method based on interference cancellation, the symbol detection method in step (two) (1) is matched filter detection.
In the above-mentioned multi-user detection method based on interference cancellation, the symbols in the symbol detection in step (ii) (1) are each user bit sequence obtained by symbol hard decision, or symbol decision sufficient statistics before each user bit decision.
In the above described multi-user detection method based on interference cancellation, the symbol detection method in step (iii) (a) (1) is: firstly, matching filtering is carried out on an input residual signal sequence, and then another input weighted symbol-level signal sequence is added to obtain a symbol detection result.
In the above-mentioned multi-user detection method based on interference cancellation, the method for regenerating each user signal sequence at symbol level in the group in step (ii) and (1) is: and obtaining a corresponding channel estimation parameter sequence and a corresponding decision bit sequence by using the symbol detection value, and multiplying to obtain a regeneration result.
In the above-mentioned multi-user detection method based on interference cancellation, the step channel estimation parameter sequence may be obtained by using the channel estimation performed by the signal of this stage, or using the channel estimation result of other stages.
In the above multi-user detection method based on interference cancellation, the selection method of the symbol-level weight signal sequence of each user in step (two) (2) is as follows: and updating the weight value by using the residual signal sequence and the symbol-level user signal sequences output by the group and the known spreading code signal sequences of the group in a certain period through a least mean square self-adaptive algorithm.
In the above multiuser detection method based on interference cancellation, under the selection scheme of M-2 and G-1-G2-1, the least mean square adaptive method is: each user only uses the symbol-level signal sequence of the current level in a certain period and the symbol-level output signal sequence after matched filtering of the residual signal sequence of the next level to update the weight.
In the above-mentioned multi-user detection method based on interference cancellation, the symbol-level weight signal sequence of each user may be a different symbol-level weight signal sequence corresponding to each path of each user.
In the above-mentioned multi-user detection method based on interference cancellation, the period is one symbol period or several symbol periods.
In the above described multi-user detection method based on interference cancellation, the interference regeneration method in step (ii) and (2) is: and spreading the symbol-level signal sequence weighted by each user by using the known spreading code sequence of the group to obtain a spreading regeneration signal sequence of the corresponding user.
In the above described multi-user detection method based on interference cancellation, the interference cancellation method in step (ii) and (2) is: and (3) performing path superposition on the spread spectrum regeneration signal sequence of the corresponding user by using the known path delay estimation information of each user in the group, and then subtracting the spread spectrum regeneration signal sequence from the input signal sequence in the group to obtain the residual signal sequence subjected to the interference elimination in the group.
In the above described multi-user detection method based on interference cancellation, the interference regeneration method in step (iii) (a) (2) is: and then, the symbol level difference value signal sequence is spread by utilizing the known spread spectrum code sequence of the group to obtain the difference value sequence of the regenerated spread spectrum signals of the current level and the previous level of each path of the group of users.
In the above described multi-user detection method based on interference cancellation, the interference cancellation method in step (iii) (a) (2) is: and carrying out path delay and superposition on the difference sequence of the regenerated spread spectrum signal by utilizing the known path delay estimation information of each user in the group, and then subtracting the difference sequence from the residual signal sequence input in the group.
The invention designs a packet Interference Cancellation (GSIC) CDMA receiving method adopting symbol-level self-adaptive weighting. In the self-adaptive GSIC detector, firstly, the strongest user group is detected and the interference of the strongest user group is estimated, the strongest user group is subjected to self-adaptive weighted elimination, and then the second strongest user group is detected and subjected to self-adaptive weighted elimination, so that the accuracy and the efficiency of the interference elimination process are greatly improved, and the performance of the whole detector can be obviously improved. The method overcomes the defects of non-weighted and fixed-weight GSIC receiving performance and stability, and fully exerts the advantages of the GSIC grouping interference elimination detection structure. And the weight value updating calculation amount of the symbol level is small, and the method is easy to realize in an actual system.
The symbol-level adaptive weighting mentioned above means that a weight is added to the symbol-level regenerated signal of each user (i.e., before re-spreading in the regeneration step), and symbol-level adaptive weight adjustment is performed. The updating of the self-adaptive weight value is independently carried out in each user group processing module of the GSIC, and the adjustment principle is to enable the energy average of the residual signals after the interference of the group is eliminated to tend to be minimum. The significance of weight adjustment under the weight optimization criterion is that different weights are given to the regenerated signals according to the reliability of the regenerated signal estimation, and the weighted partial interference elimination operation is carried out, so that the correct and effective interference elimination is realized, and the detection performance is finally improved.
The symbol-level adaptive weighting method can also be used in any other processing module adopting parallel interference elimination in the detector, so that the processing of the module is more reasonable and more efficient. Typically, we can apply this adaptive weighting method to the improved first-stage GSIC structure, i.e., within each group of first-stage GSICs, instead of the traditional matched filter detection, other detection methods such as PIC detection methods are used, for example. In addition, if the number of groups of the GSIC detector is set to one group, the detector becomes a symbol-level adaptively weighted PIC detector. If the number of packets is set as the total number of users, the GSIC detector in the present invention becomes a symbol-level adaptive weighted SIC detector. These are new technical inventions of detector structures which are not available at present.
It is particularly worth noting that the pressure sensors,
the adaptive GSIC receiving method is typically suitable for receiving the uplink base station of a variable spreading factor multi-rate CDMA system.
Drawings
FIG. 1 is a block diagram of a secondary adaptive GSIC detector overview;
FIG. 2 is a diagram of the internal structure of a first stage group 2 interference cancellation unit (GICU);
fig. 3 is a diagram of a kth user Interference Cancellation (ICU) procedure for group 2 of the first stage.
Detailed Description
In this section, we take the uplink reception of a multi-spreading factor multi-rate user in a WCDMA system as an example, and describe in detail a method for implementing symbol-level adaptive GSIC reception in a Rayleigh multipath fading channel. In order to highlight the key content of the present invention, we do not specifically describe the operations of the functional modules such as channel estimation, multipath search, etc., and only refer to the output results.
The following describes the design scheme of the receiving method in the present invention.
It is assumed that the multi-user asynchronous BPSK signal transmitted under the Rayleigh multipath fading channel is received at the base station antenna.
The system has K users. We process K users in level 2, and in each level, according to the different user data rates, divide them into 3 serially processed user groups, namely group 1, group 2, and group 3, and set G1And G2Are the number of packets in two stages, respectively, then G1＝G2Because the spreading factor of the users in group 1 is minimum, the required signal transmission power is maximumLarge, the users in group 2 have the largest spreading factor and the smallest required signal transmission power, so they are grouped and arranged in sequence according to the signal transmission power.
The baseband receiving signal sequence model is
r(n)＝∑k＝1 K∑l＝1 Lak，l(n)Akdk[i]ck(n) + z (n), (1) wherein the sequence index n is 0, 1.. denotes the nth chip, L is the number of paths of the multipath signal per user, akIs the signal amplitude of the kth user, { dk[i]}i＝0，1,.. the symbol sequence (or bit sequence) of BPSK of the kth user is { + -1 }, { ck(n)}n＝0，1,.. is the spreading code sequence of the kth user, which takes on the values { + -1 }, z (n) is additive white gaussian noise, ak，l(n) is the fading channel coefficient of the l path of the k user. In addition, the corresponding relation between the chip index n and the symbol index i is <math> <mrow> <mi>i</mi> <mo>=</mo> <mo>∉</mo> <mi>n</mi> <mo>/</mo> <msub> <mi>N</mi> <mi>k</mi> </msub> <mo>∠</mo> <mo>,</mo> <msub> <mi>N</mi> <mi>k</mi> </msub> </mrow> </math> Indicating spreading factor, sign, of the k-th user
Indicating that the integer part is taken for the fraction. (if there is a chip oversampling process, n can be understood as an index of the sample point.)
A general block diagram of a two-level symbol-level adaptive GSIC detection architecture for a receiver is shown in fig. 1. The first column represents the serial processing of the first stage and the second column represents the serial processing of the second stage. In the figure, r (n) is a baseband received signal sequence, e1 (1)(n)、e2 (1)(n)、e3 (1)(n) the output residual signal sequences of the 1 st, 2 nd and 3 rd groups in the first stage, respectively, e1 (2)(n)、e2 (2)(n) respectively represent the second in the second stageOutput residual signal sequences of group 1 and group 2, Yk，l (1)[i](K1., K, L1., L) denotes a symbol-level signal sequence that is output from each group of the first stage to each path of each user of the next stage, and d represents a symbol-level signal sequence that is weighted by each path of each user of the first stage to the next stagehigh (2)[i]、dmed (2)[i]、dlow (2)[i]And represents the output decision bit sequences of the second stage of the high rate user group (i.e., group 1), the medium rate user group (i.e., group 2), and the low rate user group (i.e., group 3), respectively, and bold d represents that a plurality of user bit sequences within a group are simultaneously output.
In the first stage, 3 groups of group 1, group 2 and group 3 sequentially receive the baseband received signal sequence and the residual signal sequence after the previous group of interference cancellation, and perform the following serial processing procedures:
(1) group Interference Cancellation Unit (GICU) of group 1 (see FIG. 2, FIG. 3)
Receiving baseband receiving signal sequence r (n), firstly using the delay information t of each path of each user in the group obtained by the multi-path searching function modulek，l(0≤tk，l＜N1，N1Is the spreading factor of the 1 st group of users) and the spreading code signal sequence c of each user known to this groupk(n) performing matched filtering symbol detection, as shown in formula (2), to obtain soft output of ith symbol of each path
yk，l (l)[i]＝∑n＝0 N1-1r(iN1+tk，l+n)ck(iN1+tk，l+n)， (2)
Then, the channel estimation value a obtained by the channel estimation function module is utilizedk，l[i](assumed here to be within one symbol period), RAKE maximum ratio combining, see equation (3), is performed to obtain sufficient statistics of symbol decisions for each user of the group
yk (1)[i]＝∑l＝1 La* k，l[i]Yk，l (1)[i]， (3) Symbol denotes complex number taking conjugate.
Obtaining the decision bit sequence of the group of users by symbol hard decision
dk (1)[i]＝sgn(yk (1)[i]) And (4) the superscript "1" in the above equation represents the decision bit sequence obtained in the first stage.
According to the group of user decision bits d obtained by hard judgmentk (1)And a channel estimation parameter sequence a obtained by using the channel estimation modulek，l[i]Then, the signal sequence of each path of each user at the symbol level of the group can be obtained
y k，l (1)[i]＝ a k，l[i]dk (1)[i]. (5)
(2) The symbol level signal sequence y of each path of each userk，l (1)[i]Respectively multiplied by the symbol-level weight signal sequences w of the respective paths of the corresponding usersk，l (1)[i](see the following equation (9) and its description), a symbol-level signal sequence weighted by each path of each user is obtained
Yk，l (1)[i]＝wk，l (1)[i] y k，l (1)[i]And (6) and outputting. Here, the symbol-level weight sequence wk，l (1)[i]Generated by a self-adaptive weight updating module, when the ith is 0 symbol, the weight sequence output by the module is an initial weight wk，l (1)(0) When i > 0, the initial weight is used to obtain the updated weight wk，l (1)[i]The specific process is shown in the following formula (9).
Symbol-level signal sequence Y weighted by each path of each userk，l (1)[i]And the known users of the groupFrequency code signal sequence ck(n) and per-user per-path delay information tk，lInterference regeneration and interference cancellation operations can be performed. The operation of interference regeneration is to weight the symbol-level signal sequence Yk，l (1)[i]Spread spectrum to obtain the chip-level regenerated signal sequence of each path of each user in the group
rk，l (1)(n)＝ck(n)Yk，l (1)[i](7) there is an obvious correspondence between the chip index n and the symbol index i, i.e., n ═ iNk+tk，l，…，(i-1)Nk+tk，l-1; the operation of interference elimination is to make the signal sequence r of each path of each user in the group after spreadingk，l (1)(n) subtracting the received signal sequence r (n) to obtain the residual signal sequence subjected to the interference elimination of the group
e1 (1)(n)＝r(n)-∑k∈K1∑l＝1 Lrk，l (1)(n)， (8)
Wherein ∑k∈K1Indicating summing the corresponding signals of all users of group 1. Similarly, hereinafter, Σk ∈K2、∑k∈K2Respectively, the summation of the corresponding signals of the users of the 2 nd group and the 3 rd group is shown. The final operation of this group is to apply the residual signal sequence e1 (1)(n) outputting.
Description of a symbol-level adaptive weight updating module: each path symbol level weight sequence w of each userk，l (1)[i]The residual signal sequence e output by the group is utilized through respective self-adaptive weight value updating modules1 (1)(n) and the respective path's own symbol-level signal sequence yk，l (1)[i]Calculated by the following formula
wk，l (1)[i+1]＝wk，l (1)[i]+β( y k，l (1)[i])* k，l (1)[i] (9)
Wherein,k，l (1)[i]＝∑n＝0 Nk-1e1 (1)(iNk+tk，l+n)ck(iNk+tk，l+ n) is the matched filtered output of the residual signal sequence, β > 0 is a positive number representing the step size parameter of the Least Mean Square (LMS) adaptive algorithm. Here, we use a period updated with one symbol period as a weight.
(3) In groups 2 and 3, the input signal sequence is changed from the baseband received signal sequence to the output residual signal sequence e of group 11 (1)(n) and 2 nd group output residual signal sequence e2 (1)(n), the other operations are completely the same as the processes of the step (1) and the step (2). In the following description of the steps, we will refer directly to the output signal sequences generated in these two groups, for example:
residual signal sequence output in group 3: e.g. of the type3 (1)(n)
The symbol-level regenerated signal sequence output by the group 3 and weighted by each path of each user in the group: y isk，l (1)[i]，k∈K3
(III) in the second stage of processing, the 1 st group, the 2 nd group and the 3 rd group sequentially receive the output residual signal sequence e of the last group, namely the 3 rd group of the previous stage3 (1)(n) and the last set of residual signal sequences after interference cancellation, i.e. e1 (2)(n)、e2 (2)(n), each group respectively receives the weighted symbol-level signal sequence Y output by the corresponding user at the previous levelk，l (1)[i]The following group-by-group serial processing is performed:
(A) in groups 1 and 2 in the second stage, the following serial processing is performed (we take group 1 as an example):
(1) firstly, matched filtering symbol detection is carried out according to the two signal sequences input in the group.
In group 1, the delay information t of each path of each user in the group is obtained by the multi-path searching function modulek，lAnd spreading code signal sequence c of each user known to this groupk(n) residual signal e of last group of previous stage inputted to this group3 (1)(n) performing matched filtering, and adding the weighted symbol-level signal sequence Y output by the corresponding user at the previous levelk，l (1)[i]To obtain the symbol soft output of the current level (K is the K1)
yk，l (2)[i]＝∑n＝0 Nk-1e3 (1)(iN1+tk，l+n)ck(iN1+tk，l+n)+Yk，l (1)[i]， (10)
Channel estimation value obtained by using channel estimation function modulea k，l[i]Performing RAKE maximum ratio combination to obtain sufficient statistics of symbol decision of each user in the group
yk (2)[i]＝∑l＝1 L a * k，l[i]yk，l (2)[i]， (11)
Obtaining the decision bit sequence of the group of users by symbol hard decision
dk (2)[i]＝sgn(yk (2)[i])， (12)
The superscript "2" in the above equation indicates the decision bit sequence obtained in the first stage.
According to the group of user decision bits d obtained by hard judgmentk (2)And a channel estimation parameter sequence obtained by using the channel estimation modulea k，l[i]Then, the signal sequence of each path of each user at the symbol level of the group can be obtained
y k，l (2)[i]＝ a k，l[i]dk (2)[i]. (13)
(2) The symbol level user signal sequencey k，l (2)[i]Multiplying the corresponding user symbol level weight signal sequence w of this level respectivelyk，l (2)[i](see equation (17) below and the description thereof) to obtain symbol-level signal sequences weighted by each user
Yk，l (2)[i]＝wk，l (2)[i] y k，l (2)[i]， (14)
And output. Here, the symbol-level weight sequence wk，l (2)[i]Generated by a self-adaptive weight updating module, when the ith is 0 symbol, the weight sequence output by the module is an initial weight wk，l (2)(0) When i > 0, the initial weight is used to obtain the updated weight wk，l (2)[i]The specific process is shown in the following weight value updating formula (17).
Symbol-level signal sequence Y weighted by each path of each userk，l (2)[i]And another input signal Y of this groupk，l (1)[i]K ∈ K1, and the set of known spreading code signal sequences c of the respective usersk(n) and per-user per-path delay information tk，lInterference regeneration and interference cancellation operations can be performed. The interference regeneration operation is to the difference signal Yk，l (2)[i]-Yk，l (1)[i]Spread spectrum to obtain the chip-level regenerated signal sequence of each path of each user in the group
rk，l (2)(n)＝ck(n)(Yk，l (2)[i]-Yk，l (1)[i])； (15)
The operation of interference elimination is to make the signal sequence r of each path of each user in the group after spreadingk，l (2)(n) input residual signals from this groupNumber sequence e3 (1)(n) subtracting to obtain the residual signal sequence subjected to the interference elimination of the group
e1 (2)(n)＝e3 (1)(n)-∑k∈K1∑l＝1 Lrk，l (2)(n), (16) wherein ∑k∈K1Indicating summing the corresponding signals of all users of group 1. The final operation of this group is to apply the residual signal sequence e1 (2)(n) outputting.
Description of a symbol-level adaptive weight updating module: each path symbol level weight sequence w of each userk，l (2)[i]The residual signal sequence e output by the group is utilized through respective self-adaptive weight value updating modules1 (2)(n) and respective path's own symbol-level signal sequencey k，l (2)[i]And a spreading code sequence ck(n) is calculated from the following formula (K. epsilon. K1)
wk，l (2)(i+1)＝wk，l (2)[i]+β( y k，l (2)[i])* k，l (2)[i]， (17)
Wherein,k，l (1)[i]＝∑n＝0 Nk-1e1 (2)(iNk+tk，l+n)ck(iNk+tk，l+ n) is the matched filtered output of the residual signal sequence, and β > 0 is the step size parameter of the Least Mean Square (LMS) adaptation algorithm.
Here, since we set the number of steps M to 2, this stage is also the last stage of the detector, and the symbol-level signal sequence Y that weights each path of each user in this group is not used any morek，l (2)[i]Outputs since there is no next stage to use them.
(3) Since the current stage is the last stage, in the first two groups of the current stage, each user decision bit sequence to be obtaineddk (2)[i]And (6) outputting.
(B) In the 3 rd group of the second stage, that is, the last group of the last stage, the following processing is performed:
first of all using the delay information tk，lAnd spreading code signal sequence c of each user in the groupk(n) residual signal e outputted from group 2 which is the previous group inputted to the present group2 (2)(n) performing matched filtering, and adding the weighted symbol-level signal sequence Y output by the corresponding user at the previous levelk，l (1)[i]To obtain the symbol soft output of the current level group (K belongs to K3)
yk，l (2)[i]＝∑n＝0 Nk-1e2 (2)(iN1+tk，l+n)ck(iN1+tk，l+n)+Yk，l (1)[i]. (18)
Channel estimation value obtained by using channel estimation function modulea k，l[i]Performing RAKE maximum ratio combination to obtain sufficient statistics of symbol decision of each user in the group
yk (2)[i]＝∑l＝1 La* k，l[i]yk，l (2)[i]. (19)
And then the symbol is judged hard to obtain the judgment bit sequence of the group of users (K belongs to K3)
dk (2)[i]＝sgn(yk (2)[i])， (20)
Finally, the decision bit sequence is still output as in the previous two groups.
Claims (17)
1. A multi-user detection method based on interference cancellation is characterized by comprising the following steps:
all users are processed in M stages, and in each stage, the users are divided into G according to the signal intensitymGroup, M, GmIs a natural number, M ═ 1.. M;
(II) in a first stage, G1And sequentially receiving the baseband receiving signal sequence and the residual signal sequence after the previous group of interference elimination, and performing the following serial processing processes:
(1) carrying out symbol detection according to the group of input signal sequences, and regenerating each user signal sequence of the group of symbol levels by using the symbol detection value;
(2) multiplying the symbol-level user signal sequences by corresponding user symbol-level weight signal sequences respectively to obtain symbol-level signal sequences weighted by users and outputting the symbol-level signal sequences; simultaneously, carrying out interference regeneration and interference elimination by using the symbol-level signal sequence weighted by each user, the known spreading code signal sequences of each user and the delay information of each path of each user to obtain a residual signal sequence subjected to the interference elimination of the group, and outputting the residual signal sequence;
(III) in the second, …, M-stage processing, each group of each stage receives the last output residual signal sequence of the previous stage and the residual signal sequence after the previous interference cancellation in turn, and each group receives the weighted symbol-stage signal sequence output by the user corresponding to the previous stage, and the processing is performed serially step by step:
(A) in each of the second, …, M-1 stages and the 1 st, …, G of the M-th stageM-1 group, performing the following serial processing:
(1) carrying out symbol detection according to the two input signal sequences of the group, and regenerating each user signal sequence of the group of symbol levels of the current level by using the symbol detection value;
(2) multiplying the symbol-level user signal sequence by the corresponding user symbol-level weight signal sequence of the current level to obtain a symbol-level signal sequence weighted by each user of the current level, carrying out interference regeneration and interference elimination by utilizing the symbol-level signal sequence weighted by each user and the weighted symbol-level signal sequence input by the group and the spread spectrum code signal sequence of each known user and the delay information of each path of each user to obtain a residual signal sequence subjected to the interference elimination of the group, and outputting the residual signal sequence; meanwhile, each group in the 2 … M-1 level outputs a symbol-level signal sequence weighted by each user respectively;
(3) from the M-th stage GM-1 group respectively uses each user symbol detection value to obtain each user decision bit sequence and output;
(B) at G of M stageMGroup, the following treatment processes are carried out:
and carrying out symbol detection according to the two signal sequences input by the group, obtaining the decision bit sequence of each user of the group by using the symbol detection value of each user, and outputting the decision bit sequence.
2. The method of claim 1, wherein each group of each stage in step (iii) further receives a baseband received signal sequence, and the interference cancellation operation is performed by subtracting the interference signal sequence from the baseband received signal sequence.
3. The method according to claim 1, wherein all users in step (a) are grouped according to signal strength according to data rate of the user from high to low.
4. The method according to claim 1, wherein all users in step (a) are grouped according to signal strength and are further characterized in that: when M is 2, G1 is G2 is 1, G1 is G2 is 2, G1 is G2 is 3, G1 is 2, G2 is 1, G1 is 3, G2 is 1.
5. The method of claim 1, wherein the symbol detection method in step (two) (1) is matched filter detection.
6. The method of claim 1, wherein the symbols in the symbol detection in step (two) (1) are each user bit sequence obtained by hard symbol decision or each user bit decision sufficient statistics before bit decision.
7. The method for multiuser detection based on interference cancellation according to claim 1, wherein the symbol detection method in step (iii) (a) (1) is: firstly, matching filtering is carried out on an input residual signal sequence, and then another input weighted symbol-level signal sequence is added to obtain a symbol detection result.
8. The method of claim 1, wherein the step (two) (1) of regenerating the symbol-level user signal sequences comprises: and obtaining a corresponding channel estimation parameter sequence of each path and a corresponding decision bit sequence by using the symbol detection value of each path of each user, and multiplying to obtain a regeneration result of each path.
9. The method of claim 8, wherein the channel estimation parameter sequence is obtained by channel estimation using signals of the current stage, or by using channel estimation results of other stages.
10. The method according to claim 1, wherein the method for selecting the symbol-level weight signal sequence of each user in step (two) (2) comprises: and updating the weight value by using the residual signal sequence and the symbol-level user signal sequences output by the group and the known spreading code signal sequences of the group in a period through a least mean square self-adaptive algorithm to obtain the weight value.
11. The method as claimed in claim 4, 5 or 10, wherein the least mean square adaptation method in the selection scheme of M2, G1, G2 and 1 is as follows: each user only uses the symbol-level signal sequence of the current level in a period and the symbol-level output signal sequence after matching filtering of the residual signal sequence of the next level to update the weight.
12. The method of claim 10, wherein the symbol-level weight signal sequences of each user are different symbol-level weight signal sequences corresponding to each path of each user.
13. The method of claim 10, wherein the weight update period is at least one symbol period.
14. The method for multiuser detection based on interference cancellation according to claim 1, wherein the interference regeneration method in the step (two) (2) is: and spreading the symbol-level signal sequence weighted by each user by using the known spreading code signal sequence of the group to obtain a spreading regeneration signal sequence of the corresponding user.
15. The method for multiuser detection based on interference cancellation according to claim 1, 12 or 14, wherein the interference cancellation method in the step (two) (2) is: and (3) performing path superposition on the spread spectrum regeneration signal sequence of the corresponding user by using the known path delay estimation information of each user in the group, and then subtracting the spread spectrum regeneration signal sequence from the input signal sequence in the group to obtain the residual signal sequence subjected to the interference elimination in the group.
16. The method for multiuser detection based on interference cancellation according to claim 1 or 12, wherein the interference regeneration method in step (iii) (a) (2) is: and then, the symbol level difference value signal sequence is spread by utilizing the known spread spectrum code sequence of the group to obtain the difference value sequence of the regenerated spread spectrum signals of the current level and the previous level of each path of the group of users.
17. The method for multiuser detection based on interference cancellation according to claim 1 or 12, wherein the interference cancellation method in the step (iii) (a) (2) is: and carrying out path delay and superposition on the difference sequence of the regenerated spread spectrum signal by utilizing the known path delay estimation information of each user in the group, and then subtracting the difference sequence from the residual signal sequence input in the group.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011263962A CN1154272C (en) | 2001-08-03 | 2001-08-03 | Receiving method based on interference elimination |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011263962A CN1154272C (en) | 2001-08-03 | 2001-08-03 | Receiving method based on interference elimination |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1400758A CN1400758A (en) | 2003-03-05 |
| CN1154272C true CN1154272C (en) | 2004-06-16 |
Family
ID=4666415
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB011263962A Expired - Fee Related CN1154272C (en) | 2001-08-03 | 2001-08-03 | Receiving method based on interference elimination |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1154272C (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7443339B2 (en) * | 2006-01-26 | 2008-10-28 | Qualcomm Incorporated | Cross-correlation suppression technique for position location receivers |
| CN1905383B (en) * | 2006-08-08 | 2010-05-12 | 北京天碁科技有限公司 | Shared frequency cell channel estimating apparatus and method |
| US8929427B2 (en) * | 2009-04-27 | 2015-01-06 | Nokia Corporation | Method, apparatus, computer program and computer program distribution medium for a communication receiver |
| CN102545958B (en) * | 2011-12-30 | 2014-09-17 | 华为技术有限公司 | Signal processing unit as well as interference cancellation method, device and system |
| CN105474549B (en) * | 2013-12-04 | 2018-05-11 | 华为技术有限公司 | Self-interference removing method, transceiver and the communication equipment of duplexer |
-
2001
- 2001-08-03 CN CNB011263962A patent/CN1154272C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN1400758A (en) | 2003-03-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1280997C (en) | Multi-user detection using an adaptive combination of joint detection and successive interference cancellation | |
| CN1926779A (en) | CPICH processing for SINR estimation in W-CDMA system | |
| CN1902834A (en) | Unified mmse equalization and multi-user detection approach for use in a CDMA system | |
| CN101036311A (en) | Use of adaptive filters in CDMA wireless system employing pilot signals | |
| CN1158803C (en) | Propagation path estimating method for interference eliminator and interference eliminator | |
| CN1777054A (en) | Array antenna channel estimating aftertreatment method | |
| CN1771671A (en) | Joint multi-code detectors in CDMA communications system | |
| CN1154272C (en) | Receiving method based on interference elimination | |
| CN1512681A (en) | Detecting method and device for training sequence of downward chain circuit in TDD/CDMA system | |
| CN1319289A (en) | Method and apparatus for radio reception | |
| CN1175606C (en) | Channel estimation method and device | |
| CN1155188C (en) | Multiple user testing method based on removal of interference | |
| CN1408148A (en) | Equalized parallel interference cancellation (EPIC) for CDMA system | |
| CN1194492C (en) | Multi-subscriber detection method of base station in mobile CDMA communication system | |
| CN1505294A (en) | A multi-user receiving device of uplink dedicated physical channel in WCDMA system | |
| CN1463100A (en) | Method and device for eliminating interference in parallel | |
| CN1155180C (en) | Double-weighing parallel interference-counteracting algorithm | |
| CN1533071A (en) | Multiple user detection method and device, and self adaptive filter detector in it | |
| CN1225855C (en) | Method and apparatus for two-level weight and parallel disturbance cancellation under MQAM modulation | |
| CN1235364C (en) | A multipath search method and apparatus using two layer filtration process | |
| CN1968030A (en) | Channel estimation method of frequency-domain receiver | |
| CN1190031C (en) | Multi-user detection device based on prior information in base station and its detection method | |
| CN1225927C (en) | Method and apparatus for two-level weight and parallel disturbance cancellation under MPSK modulation | |
| CN1816026A (en) | Channel estimation method and apparatus | |
| CN1146172C (en) | Joint detection method with decision feedback |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20040616 Termination date: 20170803 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |