CN105262708B - The soft detection method of mimo system based on pilot channel estimation - Google Patents
The soft detection method of mimo system based on pilot channel estimation Download PDFInfo
- Publication number
- CN105262708B CN105262708B CN201510725954.3A CN201510725954A CN105262708B CN 105262708 B CN105262708 B CN 105262708B CN 201510725954 A CN201510725954 A CN 201510725954A CN 105262708 B CN105262708 B CN 105262708B
- Authority
- CN
- China
- Prior art keywords
- represent
- vector
- data vector
- value
- detection
- 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.)
- Active
Links
Landscapes
- Radio Transmission System (AREA)
Abstract
The invention discloses a kind of soft detection method of the mimo system based on pilot channel estimation, mainly solves the problems, such as that message transmission rate is low in the prior art, detection complexity is very high low with Detection accuracy.Implementation step is：1st, input data vector sum pilot vector；2nd, obtain and receive signal；3rd, pre-filtering；4th, the bit of soft detection is exported；5th, the optimum allocation of data vector transimission power.The bit that the present invention carries out multiple-input, multiple-output MIMO detection models using least mean-square error MMSE signal detecting methods and log-likelihood ratio method soft detection is analyzed, message transmission rate is greatly improved in the detection performance of the bit of soft detection proposed by the present invention, detection complexity is reduced, improves Detection accuracy.
Description
Technical field
The present invention relates to wireless communication technology field, further relates to estimating based on pilot channel for multi-antenna technology field
(multiple-input multiple-output MIMO) soft detection method of the mimo systems of meter.The present invention is protecting
In the case of demonstrate,proving system detectio performance, the soft output multiple-input, multiple-output system based on pilot channel estimation can be completed with higher accuracy rate
The detection of the MIMO signal of system.
Background technology
The MIMO technology of mimo systems is the important breakthrough of wireless mobile telecommunication technology field intelligent antenna technology.Face
Speech business, data service and the Broadband Internet business growing to future mobile communications is to transmission rate, transporting
Can be with the high request of volume of business, in the case where channel capacity and frequency spectrum resource are limited, which can not increase system band
The capacity and the availability of frequency spectrum of communication system are increased exponentially in the case of width, suppresses channel fading using multiple antennas, is new
The key technology that one Generation Mobile Communication System must use.The height of one communication system communication quality is largely dependent on letter
It is accurate number to detect whether, and accurately signal detection is to ensure that MIMO technology is able to play the key of its advantage.In order to make MIMO
The superiority of technology is able to fully play out, it should uses up measured detection algorithm using performance.But the detection that performance is good
Algorithm is usually associated with higher signal processing complexity.It is good therefore, it is necessary to find one between detection performance and complexity
Good equalization point.People, which study, in recent years finds that the MIMO detection method of soft output can be in the lower complexity of mimo system
Realize preferable detection performance, and study the inspection for finding that least mean-square error MMSE detection methods can be good at balance system
Survey the relation between performance and complexity.But the MMSE detection methods in traditional mimo system are not due to accounting for letter
Influence of the channel estimation error to system detectio performance, this limits its application to a certain extent.
Paper " the Comments on Soft Decision Metric that L.Jalloul, S.Alex et al. are delivered at it
Generation for QAM with Channel Estimation Error”(IEEE Transaction on
Communications,2014,11(62):It refer to a kind of single-input single-output (single- of soft output in 4162-4163)
Input single-output SISO) system detection method.In the method, it is contemplated that the presence of channel estimation errors
Influence to system detectio performance, has obtained the soft output detection matrix in the presence of channel estimation errors in SISO systems.The party
Shortcoming is existing for method, and this method message transmission rate in the limited environment of system channel capacity and frequency spectrum resource is low,
It can not meet the demand of nowadays growing communication service, and the detection bit error rate of soft output detection matrix is higher than soft defeated
The detection of the bit signal gone out.
Patent " signal detecting method in high order modulation mimo systems " (patent of Xian Electronics Science and Technology University's application
Application number 2012101383990, publication No. CN102724160A) in disclose the letter of high-order M-QAM modulation mimo system a kind of
Number detection method.This method utilizes the tree search method of breadth-first, decomposed with the QR of sequence replace traditional QR to decompose come
Realize the trigonometric ratio to real channel matrix, it is preferred that emphasis is the ordering survivor path of same layer is divided into two groups, respectively to two
Survivor path in group expands different number of new route, and sub-optimal path is limited within limits, ensure that
The survivor path of every layer of obtained certain amount is optimal in the extensions path of respective layer.It is insufficient existing for the patent application
It is that the detection complexity of this method is very high, in the case where ensureing system detectio performance, is examined relative to the MMSE of low complex degree
Survey mode is a kind of detection algorithm of suboptimum.
Paper " the Soft-Output MIMO that Louay M.A.Jalloul, Sam P.Alex et al. is delivered at it
Detectors with Channel Estimation Error " (IEEE Signal Processing Letters, 2015,
7(22)：It refer to the detection method of signal in a kind of soft output mimo systems of nonrandom channel in 993-997).At this
Influence of the channel estimation errors to system detectio performance is take into account in method, using the estimation model of nonrandom channel, emphasis
It is to send the mode that end signal carries out Delamination Transmission, the signal of receiving terminal is obtained by being layered after reception again by filtering
The bit log likelihood ratio (log-likelihood ratio, LLR) of soft output detection in nonrandom channel.It is insufficient existing for this method
Part is, this method, which does not account for, does not account for interfering with each other and channel randomness, and receiving terminal one between receiving end signal
The detection of the signal of a node can be disturbed be subject to other detection signals, while the channel in practical application all has randomness,
Thus this method can reduce the detection performance of system.
The content of the invention
The present invention in view of the above shortcomings of the prior art, proposes the soft detection side of mimo system based on pilot channel estimation
Method, realizes the accuracy rate of higher signal detection.
To achieve the above object, the present invention includes the following steps：
(1) input data vector sum pilot vector：
(1a) in multiple-input, multiple-output MIMO detection models, input data signal is the 16QAM mapped using Gray code mode
The data vector of modulated signal composition；
(1b) arbitrarily inputs the pilot signal of a unit matrix in multiple-input, multiple-output MIMO detection models；
(1c) utilizes power normalization formula, carries out power normalization to pilot signal, obtains pilot vector；
(2) obtain and receive signal：
(2a) arbitrarily sets an original channel matrices, in original channel matrices in multiple-input, multiple-output MIMO detection models
Element to obey average be normal distribution that 0 variance is 1；
(2b) uses the channel estimation methods of pilot aided, and original channel matrices are estimated, obtain accidental channel square
Battle array；
(2c), using linearly invariant transmission method, passes data vector in multiple-input, multiple-output MIMO detection models
It is defeated, obtain the reception value of data vector；
(3) pre-filtering：
Using least mean-square error MMSE signal detecting methods, the reception value of data vector is decomposed, obtains pre-flock
Ripple weighting matrix；
(4) bit of soft detection is exported：
The row vector of pre-filtering weighting matrix is multiplied by (4a) with the reception value of data vector, obtains local signal detection
Value；
(4b) uses the probability density of Gaussian Profile, and pre-filtering weighting matrix is decomposed, obtains local signal
The conditional probability density distribution of detected value；
(4c) uses log-likelihood ratio method, and the conditional probability density distribution to local signal detected value takes likelihood logarithm,
Obtain the bit of soft detection；
(4d) exports the bit of soft detection；
(5) optimum allocation of data vector transimission power：
(5a) arbitrarily takes distribution ratio of the real number more than 0.5 as data vector transimission power in (0,1) section
Value；
Data vector in the bit and step (1a) of (5b) relatively soft detection is by bit, by all two bits
The digit of the different bit of value is added, and using it and is worth the errored bit number as multiple-input, multiple-output MIMO detection models；
Errored bit number and the total bit number of the data vector in step (1a) are divided by by (5c), obtain data vector transmission work(
The distribution ratio of rate is more than the bit error rate of the multiple-input, multiple-output MIMO detection models corresponding to 0.5；
(5d) arbitrarily takes distribution ratio of the real number less than 0.5 as data vector transimission power in (0,1) section
Value, repeat step (5b) and step (5c), obtain the distribution ratio of data vector transimission power corresponding to less than 0.5 enter it is more
Go out the bit error rate of MIMO detection models；
The distribution ratio of data vector transimission power is set to 0.5 by (5e), and repeat step (5b) and step (5c), are counted
It is equal to the bit error rate of the multiple-input, multiple-output MIMO detection models corresponding to 0.5 according to the distribution ratio of vector transmission power；
(5f) comparison step (5c), step (5d), the bit error rate of the multiple-input, multiple-output MIMO detection models in step (5e), obtains
To the minimum value of the bit error rate of multiple-input, multiple-output MIMO detection models；
(5g) passes the corresponding data vector of the minimum value of the bit error rate of multiple-input, multiple-output MIMO detection models in step (5f)
Optimum allocation ratio of the distribution ratio of defeated power as data vector transimission power.
The present invention has the following advantages compared with prior art：
First, since the present invention uses multiple-input, multiple-output MIMO detection model detection of transmitted signals, overcome in the prior art
Single-input single-output SISO detection models the shortcomings that message transmission rate is low in the limited environment of system channel capacity and frequency spectrum resource,
So that message transmission rate increases substantially in the present invention.
Second, since the present invention uses least mean-square error MMSE signal detecting methods, to the reception value of data vector into
Row pre-filtering, is directly multiplied with the reception value of data vector using pre-filtering weighting matrix, overcomes and examine in the prior art
Survey the very high shortcoming of complexity so that the present invention in detection complexity it is significantly relatively low, be better achieved detection performance and
The compromise of detection complexity.
3rd, since the present invention uses log-likelihood ratio method, the conditional probability density of local signal detected value is distributed
Take the logarithm likelihood ratio, overcome and do not account for interfering with each other between receiving end signal in the prior art and reduced with channel randomness
The shortcomings that Detection accuracy so that the Detection accuracy of the bit of soft detection is significantly lifted in the present invention.
Brief description of the drawings
Fig. 1 is the flow chart of the present invention；
Fig. 2 is relation analogous diagram of the present invention under the conditions of the bit of the soft detection of proposition；
Fig. 3 is relation analogous diagram of the present invention under the conditions of the distribution ratio of different data vector transimission powers.
Embodiment
The present invention will be further described below in conjunction with the accompanying drawings.
With reference to attached drawing 1, the specific steps to realizing the present invention are described as follows：
Step 1, input data vector sum pilot vector.
In multiple-input, multiple-output MIMO detection models, input data signal is that the 16QAM mapped using Gray code mode is modulated
The data vector of signal composition, each element of data vector are formed by 4 bits.
In multiple-input, multiple-output MIMO detection models, the pilot signal of a unit matrix is arbitrarily inputted, utilizes power normalization
Formula, carries out power normalization to pilot signal, obtains pilot vector.
The formula of power normalization is as follows：
Wherein, xtRepresent pilot vector,Represent sqrt operation, N represents multiple-input, multiple-output mimo system reception antenna
Number, × represent multiplication operations, I represents the unit matrix of N rows N row.
Step 2, obtain and receive signal.
In multiple-input, multiple-output MIMO detection models, an original channel matrices are arbitrarily set, the member in original channel matrices
White clothing is from the normal distribution that average is that 0 variance is 1.
Using the channel estimation methods of pilot aided, original channel matrices are estimated, obtain accidental channel matrix.
The channel estimation methods of pilot aided comprise the following steps that：
The first step, it is any from (0,10) section to choose a real number as more in multiple-input, multiple-output MIMO detection models
Enter to have more the general power of MIMO detection models, a real number is arbitrarily chosen from (0,1) section as data vector transimission power
Distribution ratio；
Second step, according to the following formula, calculates the transimission power of data vector：
Pd=P × r
Wherein, PdRepresent the transimission power of data vector, P represents the general power of multiple-input, multiple-output MIMO detection models, and r is represented
The distribution ratio of data vector transimission power；
3rd step, according to the following formula, calculates pilot vector transimission power：
Pt=P-Pd
Wherein, PtRepresent pilot vector transimission power, P represents the general power of multiple-input, multiple-output MIMO detection models, PdRepresent number
According to the transimission power of vector；
4th step, according to the following formula, calculates the reception value of pilot vector：
Wherein, ytRepresent the reception value of pilot vector, PtRepresent pilot transmission power,Represent sqrt operation, ×
Represent multiplication operations, HtRepresent original channel matrices, xtRepresent pilot vector, vtRepresent that it is 0 to obey average, variance is 1 normal state
The measurement noise vector of distribution；
5th step, according to the following formula, calculates the estimate of original channel matrices row vector：
Wherein, hjRepresent the estimate of original channel matrices row vector, j represents the sequence number of original channel matrices row vector, j
=1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system,Represent sqrt operation, × represent phase
Multiply operation, PtRepresent the transimission power of pilot vector, ytRepresent the reception value of pilot vector；
6th step, with the estimate h of original channel matrices row vectorjTransposition and data vector transimission power PdIt is flat
Root is multiplied, and obtains channel estimate matrix H1, the sequence number of the estimate of j expression original channel matrices row vectors, j=1,2 ...,
N, N represent the number of reception antenna in multiple-input, multiple-output mimo system；
7th step, with the evaluated error value g of original channel matrices row vectoreTransposition and data vector transimission power Pd
Square root be multiplied, obtain channel estimation errors matrix H2, e expression channel estimation errors matrix Hs2E-th of row vector sequence
Number, e=1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system, the evaluated error of channel matrix row vector
Value geIt is 0 to obey average, and variance isNormal distribution, P represent general power；
8th step, with channel estimate matrix H1With channel estimation errors matrix H2It is added, obtains the accidental channel square of M rows N row
Battle array H, wherein, M represents the number of transmission antenna in multiple-input, multiple-output mimo system, and N represents to receive day in multiple-input, multiple-output mimo system
The number of line.
Using linearly invariant transmission method, data vector is transmitted in accidental channel matrix, obtains data arrow
The reception value of amount.
The formula of linearly invariant transmission method is as follows：
Wherein, ydRepresent the reception value of data vector, PdRepresent the transimission power of data vector,Represent sqrt behaviour
Make, × representing multiplication operations, H represents accidental channel matrix, and x represents data vector, vdRepresent that it is 0 to obey average, variance is 1
The interchannel noise vector of normal distribution.
Step 3, pre-filtering.
Using least mean-square error MMSE signal detecting methods, the reception value of data vector is decomposed, obtains pre-flock
Ripple weighting matrix.
Least mean-square error MMSE signal detecting methods comprise the following steps that：
The first step, according to the following formula, calculates the consolidation noise of data vector reception value：
Wherein, z represents the consolidation noise of data vector reception value, PdRepresent the transimission power of data vector,Expression is opened
Square root functions, × represent multiplication operations, geRepresent the evaluated error value of channel matrix row vector, e represents channel estimation errors
The sequence number of e-th of row vector of matrix, e=1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system, xiTable
Show i-th of element of data vector, i represents the subscript of data vector element, i=1,2 ..., N, hjRepresent original channel matrices
The estimate of row vector, j represent the sequence number of the estimate of original channel matrices row vector, j=1,2 ..., N, and Σ represents summation behaviour
Make, vdRepresent interchannel noise vector；
Second step, according to the following formula, calculates the covariance of the consolidation noise of data vector reception value：
Wherein, σxRepresent the covariance of the consolidation noise of data vector reception value, PdRepresent the transimission power of data vector, N
Representing the number of reception antenna in multiple-input, multiple-output mimo system, P represents the general power of multiple-input, multiple-output MIMO detection models, × represent
Multiplication operations, x represent data vector, | | | |2Expression asks two norms to operate, INRepresent the unit matrix of N rows N row；
Step 3, according to the following formula, calculates pre-filtering weighting matrix：
W=(H1 T×H1+σx×IN)-1×H1 T
Wherein, W represents pre-filtering weighting matrix, H1Represent channel estimate matrix, ()TRepresent transposition operation, × represent phase
Multiply operation, σxRepresent the covariance of the consolidation noise of data vector reception value, INRepresent the unit matrix of N rows N row, N represents to enter more
Have more the number of reception antenna in mimo system, ()-1Represent inversion operation.
Step 4, the bit of soft detection is exported.
The row vector of pre-filtering weighting matrix is multiplied with the reception value of data vector, obtains local signal detected value.
Using the probability density of Gaussian Profile, pre-filtering weighting matrix is decomposed, obtains local signal detection
The conditional probability density distribution of value.
The probability density of Gaussian Profile comprises the following steps that：
The first step, according to the following formula, calculates the consolidation noise energy of data vector reception value：
Wherein, σnRepresent the consolidation noise energy of data vector reception value, n represents the consolidation noise of data vector reception value
The sequence number of energy, n=1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system, PdRepresent data vector
Transimission power, | | | |2Expression asks two norms to operate, × represent multiplication operations, wmRepresent m-th of row of pre-filtering weighting matrix
Vector, m represent the sequence number of pre-filtering weighting matrix row vector, m=1,2 ..., N, and m=n, Σ represent summation operation, hjRepresent just
The estimate of beginning channel matrix row vector, j represent the sequence number of the estimate of original channel matrices row vector, j=1,2 ..., N；
Second step, by the row vector w of pre-filtering weighting matrixmWith the estimate h of original channel matrices row vectorjIt is multiplied, obtains
To weighting coefficient ρ；
3rd step, according to the following formula, calculates the conditional probability density distribution of local detection value：
Wherein, f (yk|xi) represent local detection value conditional probability density distribution, ykRepresent local detection value, k is represented
The sequence number of local detection value element, k=1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system, xiRepresent
I-th of element of data vector x, i represent the sequence number of data vector element, and i=k, π represent pi, × represent multiplication operations,
Exp () represents to ask index operation, σnRepresent the consolidation noise energy of data vector reception value, n represents data vector reception value
Consolidation noise energy sequence number, n=k, ρ represent weighting coefficient, | |2Represent the square operation of modulus value.
Using log-likelihood ratio method, the conditional probability density distribution to local signal detected value takes likelihood logarithm, obtains
The bit of soft detection, and export the bit of soft detection.
Log-likelihood ratio method comprises the following steps that：
The first step, according to the following formula, calculates the log-likelihood ratio of the conditional probability density distribution of local detection value：
Wherein, L (bl,i) represent the log-likelihood ratio that the conditional probability density of local detection value is distributed, bl,iRepresent data arrow
L-th of bit of i-th of element is measured, i represents the sequence number of data vector element, i=1,2 ..., N, and N represents multiple-input, multiple-output
The number of reception antenna in mimo system, l represent the sequence number of data vector element bit, l=1, and 2 ..., 4, ln () represents
Right log operations are derived from, ∑ represents sum operation, S1Represent the element that l-th of bit value all in data vector x is 1
Set, xiRepresent i-th of element of data vector x, xi∈S1, ∈ represent belong to symbol, S0Represent own in data vector x
L-th of bit value be 0 element set, xpRepresent p-th of element of data vector x, p=1,2 ..., N, p ≠ i, xp
∈S0, f (yk|xi) represent i-th of element x of data vectoriLocal detection value conditional probability density distribution, ykRepresent local
Detected value, k represent the sequence number of local detection value element, k=i, f (yk|xp) represent p-th of element x of data vectorpLocal inspection
The conditional probability density distribution of measured value；
Second step, according to the following formula, simplifies log-likelihood ratio, obtains the bit of soft detection：
Wherein, Λ (bl,i) represent the bit of soft detection, bl,iRepresent l-th of ratio of i-th of element of data vector x
Spy, i represent the sequence number of data vector element, i=1,2 ..., N, and N represents the number of reception antenna in multiple-input, multiple-output mimo system,
The sequence number of the bit of l expression i-th of element of data vector, l=1,2 ..., 4, × represent multiplication operations, σnRepresent data arrow
The consolidation noise energy of the reception value of amount, the sequence number of the consolidation noise energy of the reception value of n expression data vectors, n=i, | |2
Represent the square operation of modulus value, ykRepresent local detection value, k represents the sequence number of local detection value element, k=i, S0Represent number
According to the set that l-th of bit value all in vector x is 0 element, rf∈S0, r0Represent so that expression formula | yk-rf|2Obtain
The r of minimum valuefValue, ∈ represent belong to symbol, S1Represent that l-th of bit value all in data vector x is 1 element
Set, rq∈S1, r1Represent so that expression formula | yk-rq|2Obtain the r of minimum valueqValue.
Step 5, the optimum allocation of data vector transimission power.
The first step, arbitrarily takes distribution of the real number more than 0.5 as data vector transimission power in (0,1) section
Ratio.
Second step, the bit and data vector of relatively soft detection are different by all two bit values by bit
The digit of bit be added, the errored bit number using itself and value as multiple-input, multiple-output MIMO detection models.
3rd step, the total bit number of errored bit number and data vector is divided by, and obtains the distribution of data vector transimission power
Ratio is more than the bit error rate of the multiple-input, multiple-output MIMO detection models corresponding to 0.5.
4th step, arbitrarily takes distribution of the real number less than 0.5 as data vector transimission power in (0,1) section
Ratio, repeats the second step and the 3rd step of this step, and the distribution ratio for obtaining data vector transimission power is less than corresponding to 0.5
The bit error rate of multiple-input, multiple-output MIMO detection models.
5th step, is set to 0.5 by the distribution ratio of data vector transimission power, repeats the second step and the 3rd of this step
Step, the distribution ratio for obtaining data vector transimission power are equal to the bit error rate of the multiple-input, multiple-output MIMO detection models corresponding to 0.5；
6th step, compares the 3rd step of this step, the 4th step, the error code of the multiple-input, multiple-output MIMO detection models in the 5th step
Rate, obtains the minimum value of the bit error rate of multiple-input, multiple-output MIMO detection models.
7th step, the minimum value of the bit error rate of the multiple-input, multiple-output MIMO detection models in the 6th step of this step is corresponding
Optimum allocation ratio of the distribution ratio of data vector transimission power as data vector transimission power.
Below in conjunction with the accompanying drawings 2 and attached drawing 3 effect of the present invention is described further.
1. simulated conditions：
The emulation experiment 1 and emulation experiment 2 of the present invention is based on common simulated conditions：Simulated environment is MATLAB
R2014a, the MIMO detecting systems that mimo systems are received using 2 hairs 2, the modulation system of data vector are 16QAM, channel class
Type is multiple Gauss channel, it is assumed that the channel coefficients of multiple Gauss channel type are kept not in the transmission time of adjacent input signal
Become, the cycle-index for counting bit error rate is 10000 times.
2. emulation content：
Emulation experiment 1, with the bit of the soft detection of the bit method and the prior art of soft detection proposed by the present invention
Method is compared, and emulates the relation between signal-to-noise ratio and the bit error rate.
Emulation experiment 2, ratio r is distributed with the bit method of soft detection proposed by the present invention in data vector transimission power
The simulation performance of value is made comparisons at 0.3,0.5,0.7 respectively, emulates relation between signal-to-noise ratio and the bit error rate.
3. analysis of simulation result：
Emulation experiment 1, with the method for the present invention to the bit of optimal soft detection and the bit of traditional soft detection
Make comparisons at the same time.
Attached drawing 2 is the bit of the soft detection proposed to the bit of optimal soft detection and the ratio of traditional soft detection
The analogous diagram that special position is made comparisons at the same time.Abscissa in Fig. 2 represents signal-to-noise ratio, and the unit of signal-to-noise ratio is decibel, ordinate
Represent the bit error rate.The curve identified in Fig. 2 with circle represents the bit performance simulation curve of optimal soft detection, uses asterisk
The curve of mark represents the bit performance simulation curve of the soft detection proposed, and the curve identified with square frame represents traditional soft inspection
The bit performance simulation curve of survey.In fig 2, the bit error rate changes with the change of signal-to-noise ratio, and the value of signal-to-noise ratio is got over
Greatly, the value of the bit error rate is with regard to smaller.The bit of the bit simulation curve of the soft detection of proposition and traditional soft detection is imitated
True curve is made comparisons, and when signal-to-noise ratio is identical, the corresponding bit error rate of bit of the soft detection of proposition is minimum, identical in the bit error rate
When, the corresponding signal-to-noise ratio of bit of the soft detection of proposition has the nearly gain of 4dB.By the bit bit emulator of the soft detection of proposition
Curve and the bit simulation curve of optimal soft detection are made comparisons, when signal-to-noise ratio is identical, the bit of the soft detection of proposition
The bit error rate of the bit error rate of position close to the bit of optimal soft detection.Therefore, the bit of soft detection proposed by the present invention
There is the Detection accuracy of higher in position.
Emulation experiment 2, the bit for the soft detection that the method for the present invention proposes is in data vector transimission power distribution ratio r point
Detection performance when 0.3,0.5,0.7 is not taken to make comparisons at the same time.
The bit of soft detection of the attached drawing 3 to propose distributes ratio r in data vector transimission power and takes 0.3,0.5 respectively,
The analogous diagram relatively obtained when 0.7.Abscissa in Fig. 3 represents signal-to-noise ratio, and the unit of signal-to-noise ratio is decibel, and ordinate represents to miss
Code check.The curve identified in Fig. 3 with circle represents the bit performance simulation curve of the soft detection proposed during r=0.3, uses asterisk
The curve of mark represents the bit performance simulation curve of the soft detection proposed during r=0.5, and the curve identified with square frame represents r
The bit performance simulation curve of the soft detection proposed when=0.7.In fig. 3, the bit error rate changes with the change of signal-to-noise ratio
Become, the value of signal-to-noise ratio is bigger, and the value of the bit error rate is with regard to smaller.When signal-to-noise ratio is identical, data vector transimission power distribution ratio
Bit error rate when during value r=0.5 compared to r=0.3 and r=0.7 is minimum.Therefore, the ratio for the soft detection that the method for the present invention proposes
The optimum allocation ratio of the data transmission utilization measure of special position is 0.5.
In conclusion the present invention uses least mean-square error MMSE signal detecting methods and log-likelihood ratio method to entering more
Have more the analysis that MIMO detection models carry out the bit of soft detection, tree search method and biography with prior art breadth-first
The bit of the soft detection of system is compared, and is improved the Detection accuracy of the bit of soft detection, is reduced detection complexity, is improved
Message transmission rate.
Claims (8)
1. a kind of soft detection method of mimo system based on pilot channel estimation, includes the following steps：
(1) input data vector sum pilot vector：
(1a) in multiple-input, multiple-output MIMO detection models, input data signal is that the 16QAM mapped using Gray code mode is modulated
The data vector of signal composition；
(1b) arbitrarily inputs the pilot signal of a unit matrix in multiple-input, multiple-output MIMO detection models；
(1c) utilizes power normalization formula, carries out power normalization to pilot signal, obtains pilot vector；
(2) obtain and receive signal：
(2a) arbitrarily sets an original channel matrices in multiple-input, multiple-output MIMO detection models, the member in original channel matrices
White clothing is from the normal distribution that average is that 0 variance is 1；
(2b) uses the channel estimation methods of pilot aided, and original channel matrices are estimated, obtain accidental channel matrix；
(2c) using linearly invariant transmission method, is transmitted data vector in multiple-input, multiple-output MIMO detection models, obtains
To the reception value of data vector；
(3) pre-filtering：
Using least mean-square error MMSE signal detecting methods, the reception value of data vector is decomposed, pre-filtering is obtained and adds
Weight matrix；
(4) bit of soft detection is exported：
The row vector of pre-filtering weighting matrix is multiplied by (4a) with the reception value of data vector, obtains local signal detected value；
(4b) uses the probability density of Gaussian Profile, and pre-filtering weighting matrix is decomposed, and obtains local signal detection
The conditional probability density distribution of value；
(4c) uses log-likelihood ratio method, and the conditional probability density distribution to local signal detected value takes likelihood logarithm, obtains
The bit of soft detection；
(4d) exports the bit of soft detection；
(5) optimum allocation of data vector transimission power：
(5a) arbitrarily takes distribution ratio of the real number more than 0.5 as data vector transimission power in (0,1) section；
(5b) by the bit of soft detection compared with the data vector in step (1a) is by bit, by all two bit values
The digit of different bits is added, and using it and is worth the errored bit number as multiple-input, multiple-output MIMO detection models；
Errored bit number and the total bit number of the data vector in step (1a) are divided by by (5c), obtain data vector transimission power
Distribute the bit error rate for the multiple-input, multiple-output MIMO detection models that ratio is more than corresponding to 0.5；
(5d) arbitrarily takes distribution ratio of the real number less than 0.5 as data vector transimission power, weight in (0,1) section
Multiple step (5b) and step (5c), obtain multiple-input, multiple-output of the distribution ratio of data vector transimission power corresponding to less than 0.5
The bit error rate of MIMO detection models；
The distribution ratio of data vector transimission power is set to 0.5 by (5e), repeat step (5b) and step (5c), obtains data arrow
The distribution ratio for measuring transimission power is equal to the bit error rate of the multiple-input, multiple-output MIMO detection models corresponding to 0.5；
(5f) comparison step (5c), step (5d), the bit error rate of the multiple-input, multiple-output MIMO detection models in step (5e), obtains more
Enter to have more the minimum value of the bit error rate of MIMO detection models；
The corresponding data vector of the minimum value of the bit error rate of multiple-input, multiple-output MIMO detection models in step (5f) is transmitted work(by (5g)
Optimum allocation ratio of the distribution ratio of rate as data vector transimission power.
2. the soft detection method of the mimo system according to claim 1 based on pilot channel estimation, it is characterised in that step
The data vector of the 16QAM modulated signals mapped using Gray code mode composition described in (1a) refers to that data vector x's is every
A element xiFormed by 4 bits, i represents the sequence number of i-th of element in data vector x, i=1,2 ..., N, and N is represented
The number of multiple-input, multiple-output mimo system reception antenna.
3. the soft detection method of the mimo system according to claim 1 based on pilot channel estimation, it is characterised in that step
Power normalization formula described in (1c) is as follows：
Wherein, xtRepresent pilot vector,Represent sqrt operation, N represents the number of multiple-input, multiple-output mimo system reception antenna
Mesh, × representing multiplication operations, I represents the unit matrix of N rows N row.
4. the soft detection method of the mimo system according to claim 1 based on pilot channel estimation, it is characterised in that step
The channel estimation methods of pilot aided described in (2b), estimation accidental channel matrix comprise the following steps that：
The first step, in multiple-input, multiple-output MIMO detection models, any one real number of selection more as entering more from (0,10) section
Go out the general power of MIMO detection models, point of the real number as data vector transimission power is arbitrarily chosen from (0,1) section
With ratio；
Second step, according to the following formula, calculates the transimission power of data vector：
Pd=P × r
Wherein, PdRepresent the transimission power of data vector, P represents the general power of multiple-input, multiple-output MIMO detection models, and r represents data
The distribution ratio of vector transmission power；
3rd step, according to the following formula, calculates the transimission power of pilot vector：
Pt=P-Pd
Wherein, PtRepresent the transimission power of pilot vector, P represents the general power of multiple-input, multiple-output MIMO detection models, PdRepresent data
The transimission power of vector；
4th step, according to the following formula, calculates the reception value of pilot vector：
Wherein, ytRepresent the reception value of pilot vector, PtRepresent the transimission power of pilot vector,Represent sqrt operation,
× represent multiplication operations, HtRepresent original channel matrices, xtRepresent pilot vector, vtRepresent that it is 0 to obey average, variance is 1 just
The measurement noise vector of state distribution；
5th step, according to the following formula, calculates the estimate of original channel matrices row vector：
Wherein, hjThe estimate of expression original channel matrices row vector, the sequence number of j expression original channel matrices row vectors, j=1,
2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system,Represent sqrt operation, × represent the behaviour that is multiplied
Make, PtRepresent the transimission power of pilot vector, ytRepresent the reception value of pilot vector；
6th step, with the estimate h of original channel matrices row vectorjTransposition and data vector transimission power PdSquare root
It is multiplied, obtains channel estimate matrix H1, the sequence number of the estimate of j expression original channel matrices row vectors, j=1,2 ..., N, N tables
Show the number of reception antenna in multiple-input, multiple-output mimo system；
7th step, with the evaluated error value g of original channel matrices row vectoreTransposition and data vector transimission power PdIt is flat
Root is multiplied, and obtains channel estimation errors matrix H2, e expression channel estimation errors matrix Hs2E-th of row vector sequence number, e=
1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system, the evaluated error value g of channel matrix row vectoreClothes
It is 0 from average, variance isNormal distribution, P represent multiple-input, multiple-output MIMO detection models general power；
8th step, with channel estimate matrix H1With channel estimation errors matrix H2It is added, obtains the accidental channel matrix H of M rows N row,
Wherein, M represents the number of transmission antenna in multiple-input, multiple-output mimo system, and N represents reception antenna in multiple-input, multiple-output mimo system
Number.
5. the soft detection method of the mimo system according to claim 1 based on pilot channel estimation, it is characterised in that step
The formula of the reception value for calculating data vector using linearly invariant transmission method described in (2c) is as follows：
Wherein, ydRepresent the reception value of data vector, PdRepresent the transimission power of data vector,Represent sqrt operation, ×
Represent multiplication operations, H represents accidental channel matrix, and x represents data vector, vdRepresent that it is 0 to obey average, variance is 1 normal state
The interchannel noise vector of distribution.
6. the soft detection method of the mimo system according to claim 1 based on pilot channel estimation, it is characterised in that step
(3) least mean-square error MMSE signal detecting methods are used described in, calculate comprising the following steps that for pre-filtering weighting matrix：
The first step, according to the following formula, calculates the consolidation noise of data vector reception value：
Wherein, z represents the consolidation noise of data vector reception value, PdRepresent the transimission power of data vector,Represent extraction of square root
Root operates, × represent multiplication operations, geRepresent the evaluated error value of channel matrix row vector, e represents channel estimation errors matrix
The sequence number of e-th of row vector, e=1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system, xiRepresent number
According to i-th of element of vector, i=1,2 ..., N, hjRepresent the estimate of original channel matrices row vector, j represents initial channel
The sequence number of the estimate of matrix row vector, j=1,2 ..., N, ∑ represent sum operation, vdRepresent interchannel noise vector；
Second step, according to the following formula, calculates the covariance of the consolidation noise of data vector reception value：
Wherein, σxRepresent the covariance of the consolidation noise of data vector reception value, PdRepresent the transimission power of data vector, N is represented
The number of reception antenna in multiple-input, multiple-output mimo system, P represent the general power of multiple-input, multiple-output MIMO detection models, × represent to be multiplied
Operation, x represent data vector, | | | |2Expression asks two norms to operate, INRepresent the unit matrix of N rows N row；
Step 3, according to the following formula, calculates pre-filtering weighting matrix：
W=(H1 T×H1+σx×IN)-1×H1 T
Wherein, W represents pre-filtering weighting matrix, H1Represent channel estimate matrix, ()TRepresent transposition operation, × represent the behaviour that is multiplied
Make, σxRepresent the covariance of the consolidation noise of data vector reception value, INRepresent the unit matrix of N rows N row, N represents multiple-input, multiple-output
The number of reception antenna, () in mimo system-1Represent inversion operation.
7. the soft detection method of the mimo system according to claim 1 based on pilot channel estimation, it is characterised in that step
The tool of the conditional probability density distribution that local signal detected value is calculated using Gaussian Profile probability density described in (4b)
Body step is as follows：
The first step, according to the following formula, calculates the consolidation noise energy of data vector reception value：
Wherein, σnRepresent the consolidation noise energy of data vector reception value, n represents the consolidation noise energy of data vector reception value
Sequence number, n=1,2 ..., N, N represent multiple-input, multiple-output mimo system in reception antenna number, PdRepresent the transmission of data vector
Power, | | | |2Expression asks two norms to operate, × represent multiplication operations, wmRepresent m-th of row vector of pre-filtering weighting matrix,
M represents the sequence number of pre-filtering weighting matrix row vector, m=1,2 ..., N, m=n, and ∑ represents summation operation, hjRepresent initial letter
The estimate of road matrix row vector, j represent the sequence number of the estimate of original channel matrices row vector, j=1,2 ..., N；
Second step, by the row vector w of pre-filtering weighting matrixmWith the estimate h of original channel matrices row vectorjIt is multiplied, is added
Weight coefficient ρ；
3rd step, according to the following formula, calculates the distribution of the conditional probability density of local detection value：
Wherein, f (yk|xi) represent local detection value conditional probability density distribution, ykRepresent local detection value, k represents local
The sequence number of detected value element, k=1,2 ..., N, N represent the number of reception antenna in multiple-input, multiple-output mimo system, xiRepresent data
I-th of element of vector x, i represent the sequence number of data vector element, and i=k, π represent pi, × represent multiplication operations, exp
() represents to ask index operation, σnRepresent the consolidation noise energy of data vector reception value, n represents the conjunction of data vector reception value
And the sequence number of noise energy, n=k, ρ represent weighting coefficient, | |2Represent the square operation of modulus value.
8. the soft detection method of the mimo system according to claim 1 based on pilot channel estimation, it is characterised in that step
The bit that soft detection is calculated using log-likelihood ratio method described in (4c) is comprised the following steps that：
The first step, according to the following formula, calculates the log-likelihood ratio of the conditional probability density distribution of local detection value：
Wherein, L (bl,i) represent the log-likelihood ratio that the conditional probability density of local detection value is distributed, bl,iRepresent data vector the
L-th of bit of i element, i represent the sequence number of data vector element, i=1,2 ..., N, and N represents multiple-input, multiple-output MIMO systems
The number of reception antenna in system, l represent the sequence number of data vector element bit, l=1, and 2 ..., 4, ln () represents to be derived from so
Log operations, ∑ represent sum operation, S1Represent the collection for the element that l-th of bit value all in data vector x is 1
Close, xiRepresent i-th of element of data vector x, xi∈S1, ∈ represent belong to symbol, S0Represent l all in data vector x
A bit value be 0 element set, xpRepresent p-th of element of data vector x, p=1,2 ..., N, p ≠ i, xp∈S0, f
(yk|xi) represent i-th of element x of data vectoriLocal detection value conditional probability density distribution, ykRepresent local detection value,
K represents the sequence number of local detection value element, k=i, f (yk|xp) represent p-th of element x of data vectorpLocal detection value bar
Part probability density distribution；
Second step, according to the following formula, simplifies log-likelihood ratio, obtains the bit of soft detection：
Wherein, Λ (bl,i) represent the bit of soft detection, bl,iRepresent l-th of bit of i-th of element of data vector x, i tables
Registration represents the number of reception antenna in multiple-input, multiple-output mimo system according to the sequence number of vector element, i=1,2 ..., N, N, and l is represented
The sequence number of the bit of i-th of element of data vector, l=1,2 ..., 4, × represent multiplication operations, σnRepresent connecing for data vector
The consolidation noise energy of receipts value, the sequence number of the consolidation noise energy of the reception value of n expression data vectors, n=i, | |2Expression is asked
The square operation of modulus value, ykRepresent local detection value, k represents the sequence number of local detection value element, k=i, S0Represent data vector x
In all l-th of bit value be 0 element set, rf∈S0, r0Represent so that expression formula | yk-rf|2Obtain minimum value
RfValue, ∈ represent belong to symbol, S1Represent the set that l-th of bit value all in data vector x is 1 element,
rq∈S1, r1Represent so that expression formula | yk-rq|2Obtain the r of minimum valueqValue.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510725954.3A CN105262708B (en) | 2015-10-30 | 2015-10-30 | The soft detection method of mimo system based on pilot channel estimation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510725954.3A CN105262708B (en) | 2015-10-30 | 2015-10-30 | The soft detection method of mimo system based on pilot channel estimation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105262708A CN105262708A (en) | 2016-01-20 |
| CN105262708B true CN105262708B (en) | 2018-05-04 |
Family
ID=55102216
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510725954.3A Active CN105262708B (en) | 2015-10-30 | 2015-10-30 | The soft detection method of mimo system based on pilot channel estimation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105262708B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111628848B (en) * | 2020-05-12 | 2023-02-03 | 鹏城实验室 | Method and apparatus for detecting communication symbol, and computer-readable storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101233680A (en) * | 2005-07-25 | 2008-07-30 | 纳维尼网络公司 | Low complexity soft detection in multiple transmit and receive antenna systems with M-QAM modulations |
| CN103002480A (en) * | 2011-09-09 | 2013-03-27 | 上海贝尔股份有限公司 | Distributed type collaborative detecting method and distributed type collaborative detecting equipment for uplink baseband signals |
| CN104079302A (en) * | 2013-03-25 | 2014-10-01 | 华为技术有限公司 | Channel decoding method and channel decoding device for double sending signals |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FI20040196A0 (en) * | 2004-02-09 | 2004-02-09 | Nokia Corp | Signal detection using sphere decoding |
-
2015
- 2015-10-30 CN CN201510725954.3A patent/CN105262708B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101233680A (en) * | 2005-07-25 | 2008-07-30 | 纳维尼网络公司 | Low complexity soft detection in multiple transmit and receive antenna systems with M-QAM modulations |
| CN103002480A (en) * | 2011-09-09 | 2013-03-27 | 上海贝尔股份有限公司 | Distributed type collaborative detecting method and distributed type collaborative detecting equipment for uplink baseband signals |
| CN104079302A (en) * | 2013-03-25 | 2014-10-01 | 华为技术有限公司 | Channel decoding method and channel decoding device for double sending signals |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105262708A (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104618061B (en) | The detection method of multiple user signals in a kind of large-scale multi-antenna system | |
| RU2303330C1 (en) | Method for receiving signal in communication system with several channels for transmitting and receiving | |
| CN109951214B (en) | Signal detection method suitable for large-scale MIMO system | |
| CN105915477B (en) | Extensive MIMO detection method and hardware structure based on GS method | |
| CN104954056B (en) | The hardware architecture and method of matrix inversion in extensive MIMO linearity tests | |
| CN103685090A (en) | Apparatus for MIMO channel performance prediction | |
| WO2010021767A1 (en) | Mimo receiver using ml depth-first and k-best detectors for snr higher and lower than a threshold | |
| CN102006148B (en) | Multiple-input multiple-output (MIMO) signal detection method based on breadth-first tree search | |
| TWI591973B (en) | A signal detection method and device | |
| CN104243377B (en) | A kind of disturbance restraining method and device | |
| CN103188703A (en) | Survival constellation point choosing method and QRM-maximum likehood detection (QRM-MLD) signal detection method | |
| CN110430150A (en) | A kind of cell mobile communication systems receiver design method neural network based | |
| CN107276934B (en) | A kind of extensive mimo system multi-user uplink Robust Detection Method | |
| CN112215335A (en) | System detection method based on deep learning | |
| Hua et al. | Signal detection in uplink pilot-assisted multi-user MIMO systems with deep learning | |
| CN106160832B (en) | Low-complexity MIMO interference channel receiver based on mean square error | |
| CN105262708B (en) | The soft detection method of mimo system based on pilot channel estimation | |
| CN112054832B (en) | Deep learning multiple-input multiple-output detection method | |
| CN103607234A (en) | Relay combination receiving method and system thereof | |
| Kulkarni et al. | Performance analysis comparison of transmit antenna selection with maximal ratio combining and orthogonal space time block codes in equicorrelated Rayleigh fading multiple input multiple output channels | |
| CN109818891B (en) | Lattice reduction assisted low-complexity greedy sphere decoding detection method | |
| CN112039568A (en) | Large-scale MIMO system cross-layer design scheme based on incomplete channel state information | |
| ElMossallamy et al. | Noncoherent MIMO codes construction using autoencoders | |
| CN110474663A (en) | One kind is neural network based can iterative intelligent signal detecting method | |
| Seyman | Symbol Detection Based on Back Tracking Search Algorithm in MIMO-NOMA Systems. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |