CN109787637A - A kind of integer finite field compression sensing method - Google Patents
A kind of integer finite field compression sensing method Download PDFInfo
- Publication number
- CN109787637A CN109787637A CN201910019107.3A CN201910019107A CN109787637A CN 109787637 A CN109787637 A CN 109787637A CN 201910019107 A CN201910019107 A CN 201910019107A CN 109787637 A CN109787637 A CN 109787637A
- Authority
- CN
- China
- Prior art keywords
- transmission signal
- signal
- matrix
- integer
- compression
- 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.)
- Granted
Links
Abstract
The invention discloses a kind of integer finite field compression sensing methods, comprising: the grouping of length is fixed to the information transmitted of needs, the projective transformation on integer field, Modulo division, modulation obtain transmission signal, transmission signal is sent；Receiving end receives the transmission signal, and transmission signal is screened and demodulated；Compressed sensing model is obtained according to signal integer mold compression observation corresponding with its after demodulation, and based on compression observation；Using compressed sensing model and restructing algorithm, original signal is found out.The present invention, wherein being directly to handle digitized source symbol in information source end, omits the process for compressing information redundance in message sink coding, reduces the complexity at source device end by directly carrying out compressed sensing processing to the signal of numeric field.Due to directly being handled in numeric field, algorithm is conducive to use digital circuit, while having the function similar to channel error correction coding, and the anti-interference ability of communication system can be improved.
Description
Technical field
The present invention relates to Information & Communication Technology field more particularly to a kind of integer finite field compression sensing methods.
Background technique
The reliable transmission wirelessly communicated under interference or noise circumstance can largely rely on forward error correction coding skill
Art is realized.But Error Correction of Coding needs to increase the information redundance for resisting channel disturbance, applies in systems in practice
When, in order to efficiently transmit, it is necessary first to first carry out message sink coding to information source, compress wherein information source bring information redundancy itself
Degree, in the complexity for objectively increasing source device.
On the other hand, compressed sensing technology directly utilizes the information redundance of information source itself, can reduce and sample in information source end
Rate, to reduce the complexity of information source end equipment.Currently, being transmitted across for wireless communication is not actually used in compressed sensing technology
Journey is largely due to the calculating that current compressed sensing technology is mainly a kind of real number or complex field, is not easy to count
The signal processing in word domain, when being directly used in Digital Signal Processing, there are quantization error, finite word length effects etc. to reduce accuracy
The problem of.
Furthermore in terms of the compressed sensing technology of integer or finite field, have some similar researchs at present.Document
" Deterministic construction of compressed sensing matrices ", (Devore R
A..Journal of Complexity, 2007,23 (4-6): 918-925) it proposes to make using the polynomial matrix in finite field
Hereafter there is more document proposition error correcting code since multinomial finite field is the common tool of channel coding for observing matrix
Coding or check matrix as observing matrix, such as document " the compressed sensing measurement based on quasi-cyclic low-density parity check codes
Matrix " (Jiang little Yan thanks to positive light etc., " computer application ", o. 11th in 2014) is mentioned with GF (26) in multinomial finite field
The check matrix of LDPC code is all made of binary element 0 and 1 and constructs observing matrix, institute as observing matrix, but in these methods
Finite field be GF (2q), and the recovery algorithms for the compression perceptual system for using encoder matrix are decoded using linear programming
Algorithm.However entire compression perceptual system is set to be suitable for relatively simple conventional compact sensory perceptual system, example using such method
Such as most-often used image perception field, and can not be changeable applied to modulation system, in kind and plural number communication can be used in mixed way
Field, and such method to used coding or check matrix there are also limitation, such as code ring length therein limits
System.
Summary of the invention
An object of the present invention at least that, for how to overcome the above-mentioned problems of the prior art, provide one kind
Integer finite field compression sensing method directly can carry out compressed sensing processing to digital signal, including first to digitized
Source signal is converted using the INTEGER MATRICES for meeting compressed sensing observing matrix requirement, then to obtained signal using mutually
The transformation base of matter carries out modular arithmetic, will obtain the digital signal sequences of multiple limited sizes in this way, convenient for utilizing digital modulation
Carry out communications.
To achieve the goals above, the technical solution adopted by the present invention includes following aspects.
A kind of compression sensing method based on integer finite field, it is described to include:
The grouping of length is fixed to the information that needs transmit, obtains the first transmission signal；To first transmission signal into
Projective transformation on row integer field obtains the second transmission signal；Multiple relatively prime integer moulds two-by-two are carried out to the second transmission signal
The Modulo division of value obtains multiple remainder sequences, and the multiple remainder sequence is third transmission signal；Signal is transmitted to third
The 4th transmission signal is obtained after being modulated, and the 4th transmission signal is sent to receiving end by carrier wave；
Receiving end receives the 4th transmission signal, and is carried out based on what demodulation threshold required to the 4th transmission signal
Screening selects carrier-to-noise ratio to be higher than the modulation symbol of demodulation threshold, carries out demodulation process to the signal after screening and obtains the 5th transmission
Signal；Observing matrix is obtained according to integer field projective transformation matrix and modulation symbol selection matrix；According to the 5th transmission signal and
Its corresponding integer modulus value solves the compression observation of the second transmission signal；Rarefaction representation is carried out to the first transmission signal, according to
The second transmission compression observation of signal, observing matrix, the first transmission signal rarefaction representation, acquisition compressed sensing model；Benefit
With compressed sensing model and restructing algorithm, the estimated value of the first transmission signal is found out.
Preferably, a kind of compression sensing method based on integer finite field, the projective transformation are to transmit signal for first
It is multiplied with the random number matrix on integer field.
Preferably, a kind of compression sensing method based on integer finite field, the Modulo division are to transmit signal for second
With multiple relatively prime integral divides two-by-two, to obtain the multiple remainder sequence.
Preferably, a kind of compression sensing method based on integer finite field, the columns etc. of the modulation symbol selection matrix
In the dimension of the first transmission signal, the carrier-to-noise ratio that line number is equal in the 4th transmission signal is higher than of the modulation symbol of demodulation threshold
Number, and it is 1 that every a line of symbol selection matrix, which is only 1 element, the vector that remaining element is 0, the position pair of element 1
The 4th is answered to transmit the position that carrier-to-noise ratio in signal is higher than the modulation symbol of demodulation threshold.
Preferably, a kind of compression sensing method based on integer finite field, by modulation symbol selection matrix and projective transformation
Matrix multiple obtains the observing matrix, and the observing matrix is by the part row of modulation symbol selection square and projective transformation matrix
What vector was constituted.
Preferably, a kind of compression sensing method based on integer finite field, the compression observation of the second transmission signal,
It is to solve congruence equations according to the integer mould of the 5th transmission signal, construction third transmission signal and obtain.
Preferably, a kind of compression sensing method based on integer finite field, the compression according to the second transmission signal are seen
Measured value, observing matrix, first transmission signal rarefaction representation, obtain compressed sensing model；Utilize compressed sensing model and reconstruct
Algorithm, the estimated value for finding out the first transmission signal include:
With the compression observation of the second transmission signal, observing matrix, sparse transformation matrix, as the defeated of sparse restructing algorithm
Enter parameter, by orthogonal matching pursuit algorithm, is degree of rarefication as the optimization target values of algorithm using 1, obtains sparse signal representation
Estimation coefficient；According to the sparse transformation matrix and sparse signal representation estimation coefficient, the first signal is obtained by signal reconstruction
Estimated value.
In conclusion by adopting the above-described technical solution, the present invention at least has the advantages that
1. compressed sensing technology whole process is placed on integer number domain to handle, the compression for being currently based on floating number is overcome
Perception handles the difficulty for being not easy to digital circuit, convenient for being applied to wireless communication field.
2. the information redundance that source signal itself is had directly is used in channel error correction, legacy communications system is simplified
The middle message sink coding that first carries out is to reduce information redundancy, then carries out channel coding to increase the process of information redundancy, simplifies system
Processing step realizes a kind of novel Joint Source/channel Coding Design form, increases Transmission system in noise and interference environment
Under robustness.
Detailed description of the invention
Fig. 1 is a kind of integer finite field compression sensing method flow chart according to an exemplary embodiment of the present invention.
Fig. 2 is a kind of integer finite field compression perceptual system structural schematic diagram according to an exemplary embodiment of the present invention.
Specific embodiment
With reference to the accompanying drawings and embodiments, the present invention will be described in further detail, so that the purpose of the present invention, technology
Scheme and advantage are more clearly understood.It should be appreciated that described herein, specific examples are only used to explain the present invention, and does not have to
It is of the invention in limiting.
Embodiment 1
Fig. 1 shows a kind of integer finite field compression sensing method according to an exemplary embodiment of the present invention.The embodiment
Method specifically include that
Step 101, the grouping of length is fixed in the information to be transmitted by transmitting terminal, obtains the first transmission signal；
Specifically, assuming that needing the digital signal transmitted is by obtaining after information source direct digitization, without message sink coding
In Information Compression process.Terminal (transmitting terminal) is divided the information code current for the integer form that needs transmit by certain code element number
Group obtains the first transmission signal x；
Step 102, the projective transformation on integer field is carried out to the first transmission signal, obtains the second transmission signal；
Specifically, precoding is carried out to the first transmission signal x using the projective transformation matrix Σ on integer field, after making coding
Packet signal y in each element can carry the information of all symbols in x, this conversion process is referred to as projective transformation.It throws
Shadow conversion process can be write as matrix form:
Σ is that projective transformation matrix (can also be written as Σ in above formulaN×N, expression Σ is square matrix, dimension N), the institute in matrix
There is element σijIt is all integer, which meets the requirement in compressed sensing to integer observing matrix；Y is by the transformed letter of Σ
Number grouping, y be referred to as second transmit signal.
Step 103, the Modulo division for carrying out multiple relatively prime integer modulus value two-by-two to the second transmission signal obtains multiple remaining
Number Sequence obtains third transmission signal with this；
Specifically, using the second transmission signal y p (p >=2), relatively prime integer is as mould two-by-two, progress Modulo division,
P transmission packet signal z1, z2 ..., zp are obtained, z1 is claimed, z2 ..., zp are that third transmits signal, i.e., the multiple remainder sequence
As third transmits signal.
And each grouping zi (i=1,2 ..., p) is that y is carried out mould with i-th in p two-by-two relatively prime integer for mould to remove
The remainder sequence obtained afterwards.
Step 104, the 4th transmission signal is obtained after being modulated to third transmission signal, is sent out by carrier wave to receiving end
It send；
Specifically, third transmission signal is obtained the 4th transmission signal after ovennodulation, sent out by radio frequency transmitting terminal
It is fed into channel.In the channel, the signal after modulated will be by natural or artificial interference.
Step 105, receiving end receives the 4th transmission signal, and is transmitted based on what demodulation threshold required to the described 4th
Signal is screened, and carrier-to-noise ratio is selected to be higher than the modulation symbol of demodulation threshold, is carried out demodulation process to the signal after screening and is obtained
5th transmission signal；
Specifically, the signal received is carried out demodulation process by radio frequency reception end, the reception signal is judged before demodulation
Carrier-to-noise ratio abandons the modulated signal if carrier-to-noise ratio is lower than a certain thresholding.It will be obtained after the signal for having enough carrier-to-noise ratios demodulation
5th transmission signal s1, s2 ..., sp thinks that be grouped zi obtains grouping si after transmission and demodulation (i represents 1,2 ..., p).
The modulation symbol that those discardings are recorded in demodulation transmits the position in grouping zi at place.
Step 106, observing matrix is obtained according to integer field projective transformation matrix and modulation symbol selection matrix；
Specifically, each reception grouping si (i=1,2 ..., p) the 5th transmitted in signal expands into vector form si
=[si1 si2 … siM]T.It is abandoned since previous step has carried out part, when the element in each grouping si is relative to transmission
The element being grouped in zi is few.It is assumed that the element number in each si is M, without loss of generality, it is assumed that M < < N, then si can be write as
Following form:
Wherein []piIt indicates to carry out Modulo division to all elements in bracket with mould pi.
(i=1,2 ... p) be exactly the integer mould relatively prime two-by-two selected in step 3 to pi；Matrix BM×NFor modulation symbol
Selection matrix obtains by the following method: designing a unit matrix first, the line number of matrix is equal in the first transmission signal x
The number N of element；The line number of the unit matrix is equal to again where the modulation symbol being dropped in the vector of the 5th transmission signal
Position corresponding to row delete, final line number be M.Therefore final matrix ΦM×NIt is to be sweared by the part row of projection matrix Σ
What amount was constituted, which is referred to as observing matrix.
Step 107, the rarefaction representation of the first transmission signal is found out；
Specifically, since the first transmission signal x is directly obtained by information source, without Information Compression process, therefore can be with
Rarefaction representation is carried out, therefore can further be write (3) formula as following formula:
Si=[ΦM×N·x]pi=[ΦM×N·ΨN×P·βP×1]pi (4)
If ignoring the dimension of above-mentioned matrix and vector, above formula also be can be written as:
Si=[Φ Ψ β]pi (5)
Receiving end is measurement vector from the signal phasor si that carrier-to-noise ratio thresholding demodulates is higher than in above formula, and Φ is observation square
Battle array, Ψ are sparse transformation matrixes, and dimension is N × P.Since signal x is directly obtained by information source, rarefaction representation can be carried out
For x=Ψ β, x=Ψ can also be written asN×P·βP×1, Ψ thereinN×PRepresenting matrix Ψ has dimension N × P, βP×1Indicate vector beta
It is P × 1 with dimension, is sparse transformation coefficient vector of the x on Ψ.
Step 108, it is observed according to the compression that the 5th transmission signal integer modulus value corresponding with its solves the second transmission signal
Value；
Specifically, setting θ=Φ Ψ β, then congruence equations can be listed according to (5):
Si=[θ]pi, (i=1,2 ..., p) (6)
The value of θ is found out using Chinese remainder theorem to above formula, which is actually the compression observation of the second transmission signal
Value.
Step 109, the rarefaction representation of signal is transmitted according to the compression observation of the second transmission signal, observing matrix, first,
Obtain compressed sensing model；
Specifically, obtaining compressed sensing model according to required by step 107, step 108:
θ=Φ Ψ β
Wherein, θ is the compression observation for the second transmission signal that step 8 solves, and it is the first biography that Φ, which is observing matrix, Ψ,
The sparse transformation matrix of defeated signal, is given value.
Step 110, using compressed sensing model and restructing algorithm, the estimated value of the first transmission signal is found out.
Specifically, finally finding out sparse system using common compressed sensing restructing algorithms such as orthogonal matching pursuit algorithm (OMP)
Number vector beta, and finally according to x=Ψ β, find out transmitted original signal (the i.e. first transmission signal) x.
Embodiment 2
Fig. 2 shows a kind of integer finite field compression perceptual system structural representations according to an exemplary embodiment of the present invention
Figure.Integer finite field compression perceptual system of the invention specifically includes: first communication device, the first communication device include:
Quantify grouping module, projective transformation module, modulation module, radiofrequency launcher；Secondary communication device (transmitting terminal) includes: that radio frequency connects
Receive device, carrier-to-noise ratio judgment module, symbol selecting module, demodulation module, equation solution module, compressed sensing reconstructed module.
Embodiment 3
The compressed sensing algorithm provided by the invention based on integer field is described in detail below with reference to Fig. 1, Fig. 2.Root
According to aforementioned thought, we can design a kind of specific embodiment.
It is assumed that needing the initial data transmitted that can be grouped into x=[x1 x2 …… xN]T, wherein N=15.Due to signal
X is obtained by information source direct digitization, can carry out rarefaction expression.A redundant dictionary Ψ can be constructed, signal x can in the domain Ψ
To be expressed as the sparse coefficient vector beta of degree of rarefication K.Code word x is carried out at projective transformation using INTEGER MATRICES Σ (15 × 15 dimension)
Reason after removing using two relatively prime integers such as (2,3) mould, obtains the packet signal y1 and y2 of two 15 dimensions, then be modulated
After be transmitted into channel.Receiving end first passes around carrier-to-noise ratio judgement, it is assumed that has 7 tune for having same position on each packet signal
The carrier-to-noise ratio of symbol processed is less than thresholding, then then selecting 2 groupings for modulation symbol of these carrier-to-noise ratios less than thresholding is abandoned
Upper respectively remaining 8 modulation symbols are demodulated, and then use Chinese Remainder to totally 16 solution adjusting datas in this 2 groupings
Theorem carries out congruence solution, obtains 8 integer datas transmitted by reality.Use OMP (orthogonal matching pursuit) again accordingly
Algorithm obtains the estimated value of coefficient vector β according to M=8 integer data is obtained to recovery and rebuilding is carried outFurther according to sparse change
The relationship of changing finds out the estimation of coding codeword xSpecific implementation process is as follows:
Step 1: being grouped initial data x by certain amount, obtains the first transmission signal；
The user data information that terminal is sent can be expressed as initial data vector x=[x1 x2 …… xN]T, wherein respectively
Element is integer.For example, initial data is the integer sequence for including integer 0 to 7, block length number N=15 is selected, due to these
Data are directly obtained from quantizing of information source, it can be considered that there is information redundancy between data symbols, can carry out sparse table
Show, if x=Ψ β, Ψ are sparse transformation matrix, β is transformation coefficient.
Step 2: precoding is carried out to the first transmission signal x using the projective transformation matrix Σ on integer field, after making coding
Packet signal y in each element can carry the information of all symbols in x, this conversion process is referred to as projective transformation.It throws
Shadow conversion process can be write as matrix form:
Σ is that projective transformation matrix (can also be written as Σ in above formulaN×N, expression Σ is square matrix, dimension N), the institute in matrix
There is element σijAll it is an integer, can choose the part Hadamard matrix for 15 ranks.Y is by the transformed signal of Σ point
Group, y are referred to as the second transmission signal.
Step 3: selecting relatively prime integer 2,3 is mould, is remmed to y, obtains the remainder sequence z1 that two sizes are 15,
Z2, referred to as third transmit signal.
Step 4: again third transmission signal is obtained into the 4th transmission signal after QPSK is modulated, is sent to enter letter
Road.In the channel, the signal after modulated will be by natural or artificial interference.
Step 5: the signal received is carried out demodulation process by receiving end, judges that the load of the reception signal is made an uproar before demodulation
Than abandoning the modulated signal if carrier-to-noise ratio is lower than the demodulation threshold of QPSK.Carrier-to-noise ratio is higher than to the letter of QPSK demodulation threshold
Number demodulation after obtain the 5th transmission signal s1, s2, that is, think be grouped z1, z2 obtained after transmission and demodulation grouping s1, s2.It is solving
The modulation symbol of those discardings is recorded in tune respectively in the position in transmission grouping z1, z2.
Step 6: each reception grouping si (i=1,2) the 5th transmitted in signal expands into vector form si=[si1
si2 … siM]T.Due to losing, each element being grouped in si is few relative to the element in grouping zi when sending.
It is assumed that the element number in each si is M=7, it is clear that 7 < < N=15, then the form that si can be written as follow:
Wherein []piIt indicates to carry out the remainder that Modulo division obtains to all elements in bracket with mould pi (i=1,2), and
And p1=2, p2=3；Matrix B7×15It for modulation symbol selection matrix, obtains by the following method: designing a unit square first
Battle array, the line number of matrix are equal to the number 15 of element in the first transmission signal x；The line number of unit matrix is equal to the 5th transmission again to believe
Number vector in row corresponding to position where the modulation symbol that is dropped delete.Therefore final matrix Φ7×15Be by
What the part row vector of projection matrix Σ was constituted, which is referred to as observing matrix.
Step 7: since the first transmission signal x is directly obtained by information source, rarefaction representation can be carried out, therefore can
Further to be write (3) formula as following formula:
Si=[Φ7×15·x]pi=[Φ7×15·Ψ15×128·β128×1]pi (4)
If ignoring the dimension of above-mentioned matrix and vector, above formula also be can be written as:
Si=[Φ Ψ β]pi (5)
Receiving end is measurement vector from the signal phasor si that carrier-to-noise ratio thresholding demodulates is higher than in above formula, and Φ is observation square
Battle array, Ψ are sparse transformation matrixes, and dimension is 15 × 128.Since signal x is directly obtained by information source, can carry out sparse
Being expressed as x=Ψ β, β is sparse transformation coefficient vector of the x on Ψ, and dimension is 128 × 1.
Step 8: setting θ=Φ Ψ β, then can list congruence equations according to (5):
Si=[θ]pi, (i=1,2) (6)
The value of θ is found out using Chinese remainder theorem to above formula, which is actually the compression observation of the second transmission signal
Value.
Step 9: after step 8, compressed sensing model can be write out:
θ=Φ Ψ β
Wherein, θ, Φ, Ψ become given value, then are reconstructed using the common compressed sensing such as orthogonal matching pursuit algorithm (OMP)
Algorithm finds out sparse coefficient vector beta, and finally according to x=Ψ β, finds out transmitted original signal x.
Further, the process that the present invention handles signal is actually similar to the congruence transformation of vector signal.When
The congruence transformation signal of different moulds part signal after having the transmission of noise or interference will be lost, and the present invention is subsequent
Signal recovery is carried out using Chinese remainder theorem and compressed sensing restructing algorithm in receiving end.This process ensures entire compression
Sampling process, observation process and recovery process are all handled in integer number domain, so as to avoid finite word length effect sum number
Quantization error bring in modular transformation influences.The scheme proposed through the invention can make the compressed sensing of domain digital signal
Method is easier to utilize digital circuit, to reduce sending ending equipment complexity, and realizes a kind of different from conventional channel
The antijam communication method of coding techniques.
In above-described embodiment, compressed sensing processing is carried out by the signal directly to numeric field, wherein being straight in information source end
It connects and digitized source symbol (also including the digital signal before modulation) is handled, it is superfluous to omit compression information in message sink coding
The process of remaining reduces the complexity at source device end.Due to directly being handled in numeric field, algorithm is conducive to use digital circuit
It realizes, while there is the function similar to channel error correction coding, the anti-interference ability of communication system can be improved.
The above, the only detailed description of the specific embodiment of the invention, rather than limitation of the present invention.The relevant technologies
The technical staff in field is not in the case where departing from principle and range of the invention, various replacements, modification and the improvement made
It should all be included in the protection scope of the present invention.
Claims (7)
1. a kind of compression sensing method based on integer finite field, which is characterized in that described to include:
The grouping of length is fixed to the information that needs transmit, obtains the first transmission signal；First transmission signal is carried out whole
Projective transformation on number field obtains the second transmission signal；Multiple relatively prime integer modulus value two-by-two are carried out to the second transmission signal
Modulo division obtains multiple remainder sequences, and the multiple remainder sequence is third transmission signal；Third transmission signal is carried out
The 4th transmission signal is obtained after modulation, and the 4th transmission signal is sent to receiving end by carrier wave；
Receiving end receives the 4th transmission signal, and is sieved based on what demodulation threshold required to the 4th transmission signal
Choosing selects carrier-to-noise ratio to be higher than the modulation symbol of demodulation threshold, obtains the 5th transmission to the signal progress demodulation process after screening and believes
Number；Observing matrix is obtained according to integer field projective transformation matrix and modulation symbol selection matrix；According to the 5th transmission signal and its
Corresponding integer modulus value solves the compression observation of the second transmission signal；Rarefaction representation is carried out to the first transmission signal, according to the
The two transmission compression observations of signals, observing matrix, the first transmission signal rarefaction representation, acquisition compressed sensing model；It utilizes
Compressed sensing model and restructing algorithm find out the estimated value of the first transmission signal.
2. the method according to claim 1, wherein the projective transformation is by the first transmission signal and integer field
On random number matrix be multiplied.
3. the method according to claim 1, wherein the Modulo division is by the second transmission signal and multiple two
Two relatively prime integral divides, to obtain the multiple remainder sequence.
4. being passed the method according to claim 1, wherein the columns of the modulation symbol selection matrix is equal to first
The dimension of defeated signal, the carrier-to-noise ratio that line number is equal in the 4th transmission signal are higher than the number of the modulation symbol of demodulation threshold, and accord with
Every a line of number selection matrix is that only 1 element is 1, the vector that remaining element is 0, the position the corresponding 4th of element 1
Transmit the position that carrier-to-noise ratio in signal is higher than the modulation symbol of demodulation threshold.
5. the method according to claim 1, wherein modulation symbol selection matrix is multiplied with projective transformation matrix
The observing matrix is obtained, the observing matrix is to select the part row vector of square and projective transformation matrix to constitute by modulation symbol
's.
6. the method according to claim 1, wherein the compression observation of the second transmission signal, is basis
The integer mould of 5th transmission signal, construction third transmission signal, solves congruence equations and obtains.
7. method according to any one of claim 1 to 6, which is characterized in that the pressure according to the second transmission signal
Contracting observation, observing matrix, first transmission signal rarefaction representation, obtain compressed sensing model；Using compressed sensing model and
Restructing algorithm, the estimated value for finding out the first transmission signal include:
With the compression observation of the second transmission signal, observing matrix, sparse transformation matrix, the input as sparse restructing algorithm is joined
Number is degree of rarefication as the optimization target values of algorithm using 1, obtains sparse signal representation estimation by orthogonal matching pursuit algorithm
Coefficient；According to the sparse transformation matrix and sparse signal representation estimation coefficient, estimating for the first signal is obtained by signal reconstruction
Evaluation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910019107.3A CN109787637B (en) | 2019-01-09 | 2019-01-09 | Integer finite field compressed sensing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910019107.3A CN109787637B (en) | 2019-01-09 | 2019-01-09 | Integer finite field compressed sensing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109787637A true CN109787637A (en) | 2019-05-21 |
| CN109787637B CN109787637B (en) | 2020-07-07 |
Family
ID=66500050
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910019107.3A Active CN109787637B (en) | 2019-01-09 | 2019-01-09 | Integer finite field compressed sensing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109787637B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102104396A (en) * | 2011-03-15 | 2011-06-22 | 清华大学深圳研究生院 | Pulse UWB (Ultra Wide Band) communication system based on CS (Compressed Sensing) theory |
| US20110260036A1 (en) * | 2010-02-22 | 2011-10-27 | Baraniuk Richard G | Temporally- And Spatially-Resolved Single Photon Counting Using Compressive Sensing For Debug Of Integrated Circuits, Lidar And Other Applications |
| WO2012128897A2 (en) * | 2011-02-25 | 2012-09-27 | California Insitute Of Technology | Systems and methods for acquiring and decoding signals using compressed sensing |
| CN105359419A (en) * | 2014-01-07 | 2016-02-24 | 国立大学法人东京大学 | Transmission device, receiving device, and transmission/receiving system |
| CN106027061A (en) * | 2016-05-06 | 2016-10-12 | 南京信息工程大学 | Lamb wave compression sensing method based on adaptive observation matrix |
| CN106209703A (en) * | 2016-07-08 | 2016-12-07 | 中国人民解放军信息工程大学 | A kind of Frequency Hopping Signal Blind Parameter Estimation and device |
| CN106302293A (en) * | 2016-08-26 | 2017-01-04 | 电子科技大学 | A kind of broadband based on compressed sensing antijam communication method and system |
| CN106452626A (en) * | 2016-10-11 | 2017-02-22 | 北京邮电大学 | Broadband spectrum compression sensing based on multi-group relatively-prime sampling |
-
2019
- 2019-01-09 CN CN201910019107.3A patent/CN109787637B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110260036A1 (en) * | 2010-02-22 | 2011-10-27 | Baraniuk Richard G | Temporally- And Spatially-Resolved Single Photon Counting Using Compressive Sensing For Debug Of Integrated Circuits, Lidar And Other Applications |
| WO2012128897A2 (en) * | 2011-02-25 | 2012-09-27 | California Insitute Of Technology | Systems and methods for acquiring and decoding signals using compressed sensing |
| CN102104396A (en) * | 2011-03-15 | 2011-06-22 | 清华大学深圳研究生院 | Pulse UWB (Ultra Wide Band) communication system based on CS (Compressed Sensing) theory |
| CN105359419A (en) * | 2014-01-07 | 2016-02-24 | 国立大学法人东京大学 | Transmission device, receiving device, and transmission/receiving system |
| CN106027061A (en) * | 2016-05-06 | 2016-10-12 | 南京信息工程大学 | Lamb wave compression sensing method based on adaptive observation matrix |
| CN106209703A (en) * | 2016-07-08 | 2016-12-07 | 中国人民解放军信息工程大学 | A kind of Frequency Hopping Signal Blind Parameter Estimation and device |
| CN106302293A (en) * | 2016-08-26 | 2017-01-04 | 电子科技大学 | A kind of broadband based on compressed sensing antijam communication method and system |
| CN106452626A (en) * | 2016-10-11 | 2017-02-22 | 北京邮电大学 | Broadband spectrum compression sensing based on multi-group relatively-prime sampling |
Non-Patent Citations (2)
| Title |
|---|
| VINH NGUYEN XUAN等: "Combined based on minimum distance orthogonal matching pursuit method for support recovery improvement in super-resolution compressed sensing", 《WORKSHOP ON COMPRESSED SENSING ON RADAR,SONAR,AND REMOTE SENSING》 * |
| 罗晖等: "基于DCS的WMSN多视角视频编解码", 《计算机工程与设计》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109787637B (en) | 2020-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020035683A1 (en) | Joint source channel coding for noisy channels using neural networks | |
| WO2020035684A1 (en) | Joint source channel coding of information sources using neural networks | |
| CN102017498A (en) | Recovery of transmission errors | |
| EP3418821A1 (en) | Method and device for configuring a data transmission system | |
| GB2330501A (en) | Puncturing and recovering code in a spread spectrum communication system | |
| KR101298745B1 (en) | Methods and devices for decoding and encoding data | |
| CN114301573B (en) | Federal learning model parameter transmission method and system | |
| CN103125090A (en) | Method and device for transmission and reception in a multi-input multi-output channel, distributing a code word between multiple mapping matrices, and corresponding computer program | |
| US8538329B2 (en) | System with sub-nyquist signal acquisition and transmission and associated methods | |
| EP3291449B1 (en) | Methods and devices for generating optimized coded modulations | |
| CN110476357A (en) | Polarize decoding transmission method and device | |
| CN107911152B (en) | Space coding modulation system and method suitable for any number of transmitting antennas | |
| CN109787637A (en) | A kind of integer finite field compression sensing method | |
| JPH0214649A (en) | Code error correction device | |
| CN109150397B (en) | A kind of method and device constructing polarization code sequence | |
| KR101405457B1 (en) | Method of block-coded group modulation, and transmitter using the same | |
| CN104683814A (en) | Visual-quality-oriented image transmission method and device | |
| CN117203902A (en) | Method and system for channel state information feedback using sub-codebook based trellis coded quantization | |
| CN104540147B (en) | One kind cooperation compressed sensing quantifies retransmission method | |
| JP2006333055A (en) | System and method for observation device distributed observation | |
| CN103036646A (en) | ECC (error correction code) decoding controlling method used for data communication | |
| KR102328268B1 (en) | Apparatus and method of encoding and decoding of short length block codes in wireless communication system | |
| KR20200110520A (en) | Human body communication transmitter and human body communication receiver | |
| KR101661676B1 (en) | A method of dirty paper coding by using nested lattice codes | |
| WO2020083492A1 (en) | A channel encoder and a method for encoding an information word |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |