CN101494467B - Multiuser serial interference delete receiver and method with improved constringency performance - Google Patents

Abstract

The invention discloses a multi-user serial interference deletion receiver with improved convergence performance and a method thereof. In the receiver, an interference deletion unit adds reconstructing signals corresponding to users in a former iterated order with residual signals which are connected with all the front interference deletion units in series and are obtained after the interference is eliminated from receiving signals; added data signals are detected and the detected data is used for generating an estimation value of the user data; and the reconstructed signals generated by the estimation value of the user data are subtracted from the added data signals so as to obtain the residual signals which are then transmitted to a latter interference deletion unit. In addition, stepped iteration is carried out to the reconstructing signals generated by the estimation value of the user data, and the estimation value of the user data obtained by the last step iteration is input into a judging unit for judgment so as to obtain a restoring signal of the user. The interference deletion unit comprises a nonlinear processing module, and the estimation value of the user data is determined according to the credibility of the user data detected currently so as to lead the receiver to be converged fast.

Description

Multiuser serial interference delete receiver and method with improved constringency performance

Technical field

The present invention relates to the signal reception technique in CDMA (code division multiple access) communication system, relate in particular to a kind of multiuser serial interference delete receiver and method with improved constringency performance.

Background technology

Cdma system is an interference-limited system, and multiple access disturbs all the time and exists, therefore need to seek suitable receiver technology and eliminate interference, and be anti-multipath fading, the effective way of eliminating interference and process based on the time domain that Rake (RAKE) receives.In the CDMA spread spectrum system, channel width is far longer than the flat fading bandwidth of channel.Being different from traditional modulation technique need to eliminate intersymbol interference between adjacent-symbol with equalization algorithm, and the CDMA spreading code just requires it that good autocorrelation performance is arranged when selecting.Like this, the time delay expansion that occurs in the wireless channel can be regarded as just by the again transmission of the number of delivering a letter.Owing to containing utilizable information in multipath signal, cdma receiver can improve the signal to noise ratio that receives signal by merging multipath signal.What RAKE receiver was done is exactly: by each the road signal in a plurality of correlation detector receiving multipath signals, and they are combined, the multipath structure that therefore can take full advantage of signal in time domain improves the output signal-to-noise ratio of system, thereby becomes the Technology of Anti Multipath Fading that extensively is employed in the present cdma system.

Serial interference delete receiver is to estimate first between multiple access interference and footpath to disturb, and then deletes successively in order the interference that estimates.Multistage serial interference delete device is estimated successively and is deleted in multistage middle interference exactly.

And the grouping serial interference delete is exactly on the basis of serial interference delete, and the user is divided into groups according to the quality of condition of acceptance, and the group that condition of acceptance is good at first detects.Carry out the parallel interference deletion in the group, carry out serial interference delete between group, compare like this serial interference delete, can reduce Time Delay of Systems.

To utilizing RAKE receiver to carry out our discovery in the research of (grouping) serial interference delete, how to recover rightly to disturb, make it close to original interference signal, be the key that determines (grouping) serial interference delete receiver performance.Owing in the multi-user system, all being to disturb for other users of some users, recovering so rightly to disturb is exactly will be in the iterative process of interference delete, exactly each user's of reconstruct the information of trying one's best.In traditional multi-user interference delete procedure, to directly carry out hard decision to user's data, obviously such processing is not quite reasonable, particularly smaller in noise or disturbing the larger iteration initial stage, direct hard decision diffusion phenomena that can make the mistake, final so that whole interference delete iterative algorithm can not be restrained.

Summary of the invention

In view of above situation, the technical problem to be solved in the present invention is exactly to propose a kind of multiuser serial interference delete receiver and method with improved constringency performance, so that the user's reconstruct data in the iterative process is more reasonable, thereby has accelerated the convergence of receiver.

In order to address the above problem, the invention provides a kind of multiuser serial interference delete receiver with improved constringency performance, the decision unit that the interference delete unit that comprises the series winding with at least one iteration stages links to each other with output with afterbody series winding interference delete unit, wherein

The interference delete unit, be used for the reconstruction signal of a upper iteration stages respective user is removed the residual signals addition that obtains after the interference with each the interference delete unit that is connected in the front from receive signal, data-signal after the addition is detected, the data that utilization detects generate the estimated value of described user data, and from the data-signal after the described addition, deduct the reconstruction signal that is generated by described user data estimated value, obtain residual signals and send a rear interference delete unit to

And, interference delete unit on each iteration stages of corresponding same subscriber signal, be used for the described reconstruction signal that is generated by described user data estimated value is carried out step by step iteration, and described decision unit corresponding to the estimated value of the described user data that the afterbody iteration is obtained input adjudicate, and obtains described user's restoring signal;

Described interference delete unit comprises the Nonlinear Processing module, is used for determining according to the confidence level of the current user data that detects the estimated value of described user data.

Described Nonlinear Processing module is the signal to noise ratio according to iteration sum of series respective user corresponding to current interference delete unit, determines through judgement or direct estimated value by the described described user data of data acquisition that detects.

The tupe that described Nonlinear Processing module is configured is as follows:

When the user data sequence that detects is z _{G, p}={ z _{G, p, i}, i=0,1 ..., during N-1, for permanent width of cloth modulation system, through the user data sequence after the Nonlinear Processing ${\tilde{Z}}_{g,p}=\left\{{\tilde{z}}_{g,p,i}\right\},$ I=0,1 ..., the expression formula of N-1 is:

Wherein, DEC () represents hard decision, arg () ∈ (π, π] expression calculated complex the main value argument, η _k∈ (π, π] (and k=0,1 ..., 2 ^M-1) be that size is the main value argument of the permanent width of cloth modulation constellation points of M, ξ is the threshold value by the signal to noise ratio decision of current iteration sum of series respective user.

The algorithm of described threshold xi is as follows:

$\mathrm{\ξ}=\frac{1}{14}+\frac{3}{{\mathrm{\η}}_2}\·\[g+{\log }_{10}(\mathrm{SNR}+{\mathrm{\η}}_1)\]$

Wherein, g represents the rank of interference delete iteration, and SNR represents the signal to noise ratio of respective user, η ₁And η ₂Be two and be used for controlling the constant that ξ satisfies 1/14＜ξ＜1/2.

Described interference delete unit is that the data-signal to unique user or one group of user detects.

Described testing process comprises that Rake RAKE receives and despreading.

Corresponding to above-mentioned receiver, the present invention and then a kind of multiuser serial interference delete method of reseptance with improved constringency performance is provided, after serial interference delete receiver was received the upward signal that the user sends, the concrete steps of carrying out multiuser serial interference delete were as follows:

Steps A, for a user on the iteration stages, the reconstruction signal of respective user in the upper iteration stages is come the residual signals addition that obtains after each user's of front the interference with the removal of reception signal, data-signal after the addition is detected, the data that detect are carried out generating after the Nonlinear Processing estimated value of described user data, and from the data-signal after the described addition, deduct the reconstruction signal that is generated by described user data estimated value, obtain residual signals;

Step B, repeating step A, until all users' signal all carried out detection on all iteration stages, the estimated value of each user data that the afterbody iteration is obtained is adjudicated respectively, obtains the restoring signal of respective user;

In the described steps A data that detect being carried out Nonlinear Processing, is the estimated value of determining user data according to the confidence level of the current user data that detects.

The estimated value of the data generation user data that described utilization detects is the signal to noise ratio according to current iteration sum of series respective user, determines through judgement or direct estimated value by the described described user data of data acquisition that detects.

The algorithm of described Nonlinear Processing is as follows:

When the user data sequence that detects is z _{G, p}=(z _{G, p, i}, i=0,1 ..., during N-1, for permanent width of cloth modulation system, through the user data sequence after the Nonlinear Processing ${\tilde{Z}}_{g,p}=\left\{{\tilde{z}}_{g,p,i}\right\},$ I=0,1 ..., the expression formula of N-1 is:

Wherein, DEC () represents hard decision, arg () ∈ (π, π] expression calculated complex the main value argument, η _k∈ (π, π] (and k=0,1 ..., 2 ^M-1) be that size is the main value argument of the permanent width of cloth modulation constellation points of M, ξ is the threshold value by the signal to noise ratio decision of current iteration sum of series respective user.

The algorithm of described threshold xi is as follows:

$\mathrm{\ξ}=\frac{1}{14}+\frac{3}{{\mathrm{\η}}_2}\·\[g+{\log }_{10}(\mathrm{SNR}+{\mathrm{\η}}_1)\]$

Wherein, g represents the rank of interference delete iteration, and SNR represents the signal to noise ratio of respective user, η ₁And η ₂Be two and be used for controlling the constant that ξ satisfies 1/14＜ξ＜1/2.

Described steps A is that the data-signal to unique user or one group of user detects.

Described testing process comprises that Rake RAKE receives and despreading.

Advantage and characteristics that the present invention is main are as follows:

1. the present invention is based on existing serial interference delete structure, be convenient to realize, this structure can reduce strong user to greatest extent to weak user's interference, can realize accurately and effectively multi-user's interference delete;

2. the recovered part of multipath channel adopted the lower RAKE receiver of complexity, and the equalization methods that uses in the existing cdma system is the RAKE reception, has kept the existing certain compatibility of cdma system;

3. the Nonlinear Processing module of setting up is used for the user data that will offer reconstructed module is carried out Nonlinear Processing, has reduced the generation of error diffusion phenomenon, has obtained good systematic function.

Description of drawings

Be described in detail in conjunction with the following drawings and understand principle of the present invention, step, characteristics and advantage with being conducive to, in the accompanying drawing:

Fig. 1 is the system configuration schematic diagram with multiuser serial interference delete receiver of improved constringency performance of the present invention;

Fig. 2 is interference delete unit ICU among Fig. 1 of the present invention _{(g, p)}The composition structure chart;

Fig. 3 is the flow chart with multiuser serial interference delete method of reseptance of improved constringency performance of the present invention;

Fig. 4 illustrates the bit error rate performance of technical scheme of the present invention under simulation scenarios.

Embodiment

In (grouping) serial interference delete algorithm of time domain, user's detection data confidence level is different in each (grouping) serial interference delete iteration rank.Under the condition of (grouping) serial interference delete algorithmic statement, the rank of iteration is higher, and signal to noise ratio is higher, and user's detection data confidence level is just higher so.Therefore, main thought of the present invention is to set up a nonlinear function, the confidence level of the judgement mode of user's detection data and data is closely connected is in the same place.Like this, so that our serial (grouping serial) interference delete algorithm is restrained rapidly, receiver can be detected performance preferably with less iterations.

Below, with reference to Fig. 1 the structure of the multi-user with improved constringency performance (grouping) serial interference delete receiver of the present invention's proposition is described.

The difference of serial interference delete and grouping serial interference delete only is, in the grouping interference delete, the user will divide into groups according to signal to noise ratio, at first the strongest user's group of signal to noise ratio is detected, and then successively each group is detected according to the ordering of signal to noise ratio size.What all users in the group used is the parallel interference deletion, and when certain group detected, delete all users' of group reconstruct data, so what carry out between the group is serial interference delete.Parallel interference deletion in the group can reduce Time Delay of Systems, and the serial interference delete between group can reduce strong user to greatest extent to weak user's interference.If in the grouping serial interference delete algorithm, only have a user in each grouping, so just become the serial interference delete reception.So serial interference delete is a special case of grouping serial interference delete.

Because adopt the structure of (grouping) serial interference delete, so define the interference delete unit (Interference Cancellation Unit, ICU) of a unified structure.Because serial interference delete and grouping serial interference delete receiver structure are closely similar, principle is also basic identical, so each interference delete unit in the up serial interference delete receiver of CDMA shown in Figure 1, both can be regarded as unique user had been processed, and also can be regarded as one group of user is processed.For some module I CU among Fig. 1 _{(g, p)}, two inputs are arranged: residual signals E _{G, p}With in the upper level iteration to this user's (group) reconstruction signal I _{G-1, p}ICU _{(g, p)}Three output: I are arranged _{G, p}Provide the reconstruction signal to this user of next stage iteration (group), E _{G, p+1}Provide the residual signals to the next user of iteration at the corresponding levels (group), if the detection data of iteration at the corresponding levels are adjudicated, just use the 3rd output

Here, residual signals E _{G, p}As an input signal of ICU at the corresponding levels unit, can be understood as it is that total reception signal has deducted the corresponding reconstruction signal of the user who has detected, and the difference that obtains.Because detecting the user data that recovers each time is not entirely accurate certainly, so still keeping the portion of energy of useful user data in this difference.For first ICU unit of the first order of (grouping) serial interference delete, residual error E _1,1Just equal the reception signal R of system, for remaining ICU unit, residual error E _{G, p+1}=E _{G, p}+ I _{G-1, p}-I _{G, p}In addition, the one-time detection of finishing all users is exactly the one-level iteration, and namely each row ICU unit has all been carried out one-time detection and carried out exactly the one-level iteration among Fig. 1.As for reconstruction signal above-mentioned, in the interference delete iterative process, some users are through (Nonlinear Processing of the present invention belongs to a kind of method of part judgement) after hard decision or the Nonlinear Processing, user data obtains all or the recovery of part, then these data will be passed through spread spectrum again, and again carry out convolution with channel, so just obtained reconstruction signal.This reconstruction signal is the part of this user in receiving signal, also is the interference to other users.

Fig. 2 is unit ICU among Fig. 1 _{(g, p)}The detailed structure view of an embodiment.In this embodiment, when detecting, this ICU unit carries out RAKE (Rake) reception ﹠ disposal to subscriber signal first to user's (group).

At first, the residual signals E of input _{G, p}Will with the upper level iteration in to the present user's that will detect (group) reconstruction signal I _{G-1, p}Addition.Carry out channel equalization by the RAKE technology, then through obtaining detected burst z after the despreading resume module _{G, p}Before user's data are carried out spread spectrum, multipath reconstruct, use a nonlinear function (Nonlinear Function) user's data to be processed the burst after obtaining processing

Processing by nonlinear function interrelates the user's that feeds back data and the confidence level of detection, thereby the algorithm of whole serial iteration is restrained at faster speed.

Can be used for carrying out the hard decision of data, perhaps be used for producing user's's (group) reconstruction signal I _{G, p}Residual signals E _{G, p+1}Calculation expression be: E _{G, p+1}=E _{G, p}+ I _{G-1, p}-I _{G, p}

ICU shown in Figure 2 _{(g, p)}The Nonlinear Processing module that the unit uses is to interrelate for the user data that will feed back and the confidence level of the current user data that detects.Obviously, guarantee that at algorithm the progression of serial interference delete is higher under the condition of convergence, signal to noise ratio is higher, and the user data confidence level after the despreading is just higher in the ICU module so.When user's data were adjudicated, the threshold value that at first obtains according to calculated in advance obtained decision region and non-decision region.In fact this decision region is exactly the subregion on the signal constellation (in digital modulation), and data drop in the decision region, just thinks that these data of user are credible, can carry out hard decision; If drop on non-decision region, just think that data are insincere, do not keep it constant with regard to not carrying out hard decision.Threshold value can adaptive adjustment decision region and the size of non-decision region.And higher according to the progression of serial interference delete, signal to noise ratio is higher, and the user data confidence level after the despreading is with regard to higher principle in the ICU module so, and in this case, threshold value can be so that decision region enlarges; Otherwise threshold value can so that decision region dwindle.

And the progression of this credibility and iteration is the non-linear relation of a complexity.We set up a nonlinear function and simulate this effect, at first suppose among Fig. 2 through RAKE receive and despreading after detected user data sequence be z _{G, p}={ z _{G, p, i}, i=0,1 ..., N-1 is so for the permanent width of cloth modulation such as BPSK, QPSK and 8PSK, through the user data sequence after the nonlinear function processing ${\tilde{Z}}_{g,p}=\left\{{\tilde{z}}_{g,p,i}\right\},$ I=0,1 ..., the expression formula of N-1 is:

Wherein, DEC () represents hard decision, and arg () ∈ (π, π] be the main value argument of calculating a plural number.η _k∈ (π, π] (and k=0,1 ..., 2 ^MDoes-1) being size, (may I ask: M represent to receive signal amplitude for the main value argument of the permanent width of cloth modulation constellation points of M?), such as η in BPSK _k∈ { 0, π }.ξ is a threshold value that is determined by iteration sum of series signal to noise ratio, and it is modeled as following form:

ξ＝f(g，SNR) (2)

Wherein, g is the rank of interference delete iteration, and SNR is this user's signal to noise ratio.

As an example, functional expression (2) specific design is become following form:

$\mathrm{\ξ}=\frac{1}{14}+\frac{3}{{\mathrm{\η}}_2}\·[g+{\log }_{10}(\mathrm{SNR}+{\mathrm{\η}}_1)]---\left(3\right)$

Wherein, η ₁And η ₂Be two constants, they are used to control ξ and satisfy condition: 1/14＜ξ＜1/2.Easily see, along with the rising of iterations and signal to noise ratio, the decision region that is determined by ξ is also reasonably enlarging.

Based on above-mentioned receiver structure, the present invention provides again multi-user's (grouping) serial interference delete that has accordingly improved constringency performance method of reseptance, see also Fig. 3, this figure is the flow chart that has multi-user's (grouping) serial interference delete method of reseptance of improved constringency performance among the present invention, and its specific implementation process is as follows:

After step 1, receiver are received the upward signal that the user sends, receive signal and be admitted to interference delete receiver.Because this receiver is (grouping) serial structure, so on each iteration stages g, the user of front (grouping) ICU _{(g, p)}User's (grouping) ICU with the back _{(g, p+1)}Be cascaded structure, and offer an input of user's (grouping) the residual signals E of back _{G, p+1}, its calculation expression is: E _{G, p+1}=E _{G, p}+ I _{G-1, p}-I _{G, p}

Step 2, each interference delete unit ICU _{(g, p)}Input be residual signals E _{G, p}With in the upper level iteration to this user's (group) reconstruction signal I _{G-1, p}, after these two the input additions, be input to the deconvolution of carrying out multipath channel in the RAKE receiver and process.

Step 3, RAKE receiver will be carried out despreading to the signal chip after processing.Data after the despreading are carried out the processing of part judgement by the Nonlinear Processing module, through after this module, a part of believable data have obtained the primary signal estimated value through hard decision, and another part is considered to insecure data and remains unchanged.

Step 4, the output of nonlinear function module

Can input to two modules.If the interference delete iterative process is proceeded, this output will be sent into the signal reconstruction module so, and will be namely right

Carry out again spread spectrum, and carry out convolution with multipath channel, reconstruct the contribution I of this user (group) in the signal that receiver receives _{G, p}

Step 5, according to the unified definition of interference delete, all users are detected a circulation of finishing, and just are called the one-level iteration of interference delete.The last residual signals that produces of upper level iteration is directly sent into the next stage iteration, continues on for the interference delete algorithm.If according to system's setting, the number of times of interference delete is enough, and iterative process just finishes so,

Just input to judging module, all data are carried out hard decision, obtain this user's recovery data.

Below, specify the performance of multi-user interference delete receiver of the present invention by an example.The analogue system of setting up is as follows:

The common cdma system backward channel of additional CP or ZP is not encoded, and radio frequency carrier frequency is 2GHz, and channel adopts the M.1225 vehicle-mounted A channel model of regulation, the BPSK modulation.Chip transmits according to time slot, 2048 chips of each time slot, and spreading rate is 1.2288Mchip/s.Adopt ideal communication channel to estimate in the emulation.Number of users is 16, spread spectrum coefficient SF=16, and we simulate the user distinguishes in the HRPD system with scrambler the practice, and using length is 2 ⁴²The truncated code of-1 m sequence is as the sign of distinguishing the user.Two coefficients in the above-mentioned formula (3) are set as respectively: η ₁=2, η ₂=98.Adopt the system configuration of serial interference delete.

16 users' average error rate performance as shown in Figure 4, in two criterion situations, the iteration before the 5th is vastly improved the performance of system, to the 7th iteration, algorithm approaches convergence.

Theory analysis and emulation all proves, our receiver performance of design is better, and complexity is low, and preferably industrial applicibility is arranged.

Claims (10)

1. multiuser serial interference delete receiver with improved constringency performance, it is characterized in that, the interference delete unit that comprises the series winding with at least one iteration stages and the decision unit that links to each other with the output of the interference delete unit of the series winding of last iteration stages, wherein

The interference delete unit, be used for the reconstruction signal of a upper iteration stages respective user is removed the residual signals addition that obtains after the interference with each the interference delete unit that is connected in the front from receive signal, data-signal after the addition is detected, the data that utilization detects generate the estimated value of described user's data, and from the data-signal after the described addition, deduct the reconstruction signal that the data estimation value by described user generates, obtain residual signals and send a rear interference delete unit to

And, interference delete unit on each iteration stages of corresponding same subscriber signal, be used for the described reconstruction signal that is generated by described user data estimated value is carried out step by step iteration, and described decision unit corresponding to the estimated value of the described user data that the afterbody iteration is obtained input adjudicate, and obtains described user's restoring signal;

Described interference delete unit comprises the Nonlinear Processing module, is used for determining according to the confidence level of the current user data that detects the estimated value of described user's data; Wherein, the Nonlinear Processing module is determined the estimated value of described user data according to the confidence level of the current user data that detects, comprising:

When the user data sequence that detects is Z _{G, p}={ z _{G, p, i}, i=0,1 ..., during N-1, N represents a certain positive integer, for permanent width of cloth modulation system, through the user data sequence after the Nonlinear Processing ${\tilde{Z}}_{g,p}=\left\{{\tilde{z}}_{g,p,i}\right\},i=\mathrm{0,1},...,N-1$ Expression formula be:

Wherein, DEC () represents hard decision, arg () ∈ (π, π] expression calculated complex the main value argument, η _k∈ (π, π] be that size is the main value argument of the permanent width of cloth modulation constellation points of M, k=0,1 ..., 2 ^M-1, ξ is that g is the rank of interference delete iteration by the threshold value of the signal to noise ratio decision of current iteration sum of series respective user, and p is user's sequence.

2. receiver as claimed in claim 1, it is characterized in that, described Nonlinear Processing module is the signal to noise ratio according to iteration sum of series respective user corresponding to current interference delete unit, determines through judgement or direct estimated value by the described described user data of data acquisition that detects.

3. receiver as claimed in claim 1 is characterized in that, the algorithm of described threshold xi is as follows:

$\mathrm{\ξ}=\frac{1}{14}+\frac{3}{{\mathrm{\η}}_2}\·[g+{\log }_{10}(\mathrm{SNR}+{\mathrm{\η}}_1)]$

Wherein, g represents the rank of interference delete iteration, and SNR represents the signal to noise ratio of respective user, η ₁And η ₂Be two and be used for controlling the constant that ξ satisfies 1/14＜ξ＜1/2.

4. receiver as claimed in claim 1 is characterized in that, described interference delete unit is that the data-signal to unique user or one group of user detects.

5. receiver as claimed in claim 1 is characterized in that, described testing process comprises that Rake receives and despreading.

6. the multiuser serial interference delete method of reseptance with improved constringency performance is characterized in that, after serial interference delete receiver was received the upward signal that the user sends, the concrete steps of carrying out multiuser serial interference delete were as follows:

Steps A, for a user on the iteration stages, the reconstruction signal of respective user in the upper iteration stages is come the residual signals addition that obtains after each user's of front the interference with the removal of reception signal, data-signal after the addition is detected, the data that detect are carried out generating after the Nonlinear Processing estimated value of described user data, and from the data-signal after the described addition, deduct the reconstruction signal that the data estimation value by described user generates, obtain residual signals;

Step B, repeating step A, until all users' signal all carried out detection on all iteration stages, the estimated value of each user data that the afterbody iteration is obtained is adjudicated respectively, obtains the restoring signal of respective user;

In the described steps A data that detect being carried out Nonlinear Processing, is the estimated value of determining user data according to the confidence level of the current user data that detects, and the algorithm of described Nonlinear Processing is as follows:

When the user data sequence that detects is Z _{G, p}={ z _{G, p, i}, i=0,1 ..., during N-1, N represents a certain positive integer, for permanent width of cloth modulation system, through the user data sequence after the Nonlinear Processing ${\tilde{Z}}_{g,p}=\left\{{\tilde{z}}_{g,p,i}\right\},i=\mathrm{0,1},...,N-1$ Expression formula be:

Wherein, DEC () represents hard decision, arg () ∈ (π, π] expression calculated complex the main value argument, η _k∈ (π, π] be that size is the main value argument of the permanent width of cloth modulation constellation points of M, k=0,1 ..., 2 ^M-1, ξ is that g is the rank of interference delete iteration by the threshold value of the signal to noise ratio decision of current iteration sum of series respective user, and p is user's sequence.

7. method as claimed in claim 6, it is characterized in that, the estimated value that data that described utilization detects generate user data is the signal to noise ratio according to current iteration sum of series respective user, determines through judgement or directly by the estimated value of the described described user's of data acquisition who detects data.

8. method as claimed in claim 6 is characterized in that, the algorithm of described threshold xi is as follows:

$\mathrm{\ξ}=\frac{1}{14}+\frac{3}{{\mathrm{\η}}_2}\·[g+{\log }_{10}(\mathrm{SNR}+{\mathrm{\η}}_1)]$

Wherein, g represents the rank of interference delete iteration, and SNR represents the signal to noise ratio of respective user, η ₁And η ₂Be two and be used for controlling the constant that ξ satisfies 1/14＜ξ＜1/2.

9. method as claimed in claim 6 is characterized in that, described steps A is that the data-signal to unique user or one group of user detects.

10. method as claimed in claim 6 is characterized in that, described testing process comprises that Rake receives and despreading.

Images