CN109889187A - Signal processing method, device and electronic equipment based on sef-adapting filter - Google Patents
Signal processing method, device and electronic equipment based on sef-adapting filter Download PDFInfo
- Publication number
- CN109889187A CN109889187A CN201910034406.4A CN201910034406A CN109889187A CN 109889187 A CN109889187 A CN 109889187A CN 201910034406 A CN201910034406 A CN 201910034406A CN 109889187 A CN109889187 A CN 109889187A
- Authority
- CN
- China
- Prior art keywords
- weight
- sub
- adaptive
- weight vector
- weight vectors
- 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
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/40—Arrangements for reducing harmonics
Landscapes
- Radio Transmission System (AREA)
Abstract
This application involves a kind of signal processing method based on sef-adapting filter, device and electronic equipments.The above method, comprising: construct the weight vectors of sef-adapting filter, the weight vectors include Adaptive Weight vector sum fixed weight vector；The Adaptive Weight vector is split according to selected sub- weight segmentation is several, obtains the sub- weight vectors with the sub- weight segmentation number corresponding number；Each sub- weight vectors are calculated, and obtain the Adaptive Weight vector according to each sub- weight vectors being calculated；The input signal of the sef-adapting filter is handled according to the Adaptive Weight vector sum fixed weight vector, obtains output signal.The above-mentioned signal processing method based on sef-adapting filter, device and electronic equipment can be improved the performance of wireless communication, reduce the interference in signal, and improve calculating speed.
Description
Technical field
This application involves field of communication technology, more particularly to a kind of signal processing method based on sef-adapting filter,
Device and electronic equipment.
Background technique
Minimum variation (multiple constrained minimum variance, the MCMV) detecting of the multiple limitation of tradition
Device may achieve good output signal and add jamming incoherent signal ratio (signal-to- in the case where no spreading codes mismatch situation
Interference plus noise ratio, SINR), but when signal mismatch occurs, especially when input signal noise
It, will be so that spreading codes be regarded with the increase of spreading codes mismatched degree when increasing than (signal-to-noise ratio, SNR)
The degree being eliminated for interference signal increases therewith, and output SINR has the situation of slump of disastrous proportions.Spreading codes mismatch
One of an important factor for influencing wireless communication system efficiency.
Summary of the invention
The application provides a kind of signal processing method based on sef-adapting filter, device, electronic equipment and computer can
Storage medium is read, can be improved the performance of wireless communication, reduces the interference in signal.
A kind of signal processing method based on sef-adapting filter, comprising:
Construct the weight vectors of sef-adapting filter, the weight vectors include Adaptive Weight vector sum fixed weight to
Amount；
The Adaptive Weight vector is split according to selected sub- weight segmentation is several, is obtained and the sub- weight point
Cut the sub- weight vectors of several corresponding numbers；
Each sub- weight vectors are calculated, and obtains the adaptability according to each sub- weight vectors being calculated and weighs
Weight vector；
It is carried out according to input signal of the Adaptive Weight vector sum fixed weight vector to the sef-adapting filter
Processing, obtains output signal.
A kind of signal processing apparatus based on sef-adapting filter, comprising:
Vector constructing module, for constructing the weight vectors of sef-adapting filter, the weight vectors include adaptability power
Weight vector sum fixed weight vector；
Divide module, for being split according to selected sub- weight segmentation is several to the Adaptive Weight vector, obtains
With the sub- weight vectors of the sub- weight segmentation number corresponding number；
Adaptive Weight vector calculation module, for calculating each sub- weight vectors, and it is each according to what is be calculated
A sub- weight vectors obtain the Adaptive Weight vector；
Processing module, for according to the Adaptive Weight vector sum fixed weight vector to the sef-adapting filter
Input signal is handled, and output signal is obtained.
A kind of electronic equipment, including memory and processor are stored with computer program, the calculating in the memory
When machine program is executed by the processor, so that the processor realizes method as described above.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
Method as described above is realized when row.
The above-mentioned signal processing method based on sef-adapting filter, device, electronic equipment and computer readable storage medium,
The weight vectors for constructing sef-adapting filter divide the Adaptive Weight vector according to selected sub- weight segmentation is several
It cuts, obtains the sub- weight vectors for dividing number corresponding number with sub- weight, calculate each sub- weight vectors, and according to being calculated
Each sub- weight vectors obtain Adaptive Weight vector, and according to Adaptive Weight vector sum fixed weight vector to adaptive
The input signal of filter is handled, and output signal is obtained, and can be improved the performance of wireless communication, is reduced dry in signal
It disturbs, and by being split to weight vectors, reduces computation complexity, improve calculating speed.
Detailed description of the invention
Fig. 1 is the system architecture diagram of the signal processing method based on sef-adapting filter in one embodiment；
Fig. 2 is the flow diagram of the signal processing method based on sef-adapting filter in one embodiment；
Fig. 3 is the flow diagram that load terms are added in one embodiment；
Fig. 4 is the block diagram of the signal processing apparatus based on sef-adapting filter in one embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only to explain the application, not
For limiting the application.
It is appreciated that term " first " used in this application, " second " etc. can be used to describe various elements herein,
But these elements should not be limited by these terms.These terms are only used to distinguish the first element from the other element.Citing comes
It says, in the case where not departing from scope of the present application, the first client can be known as the second client, and similarly, can incite somebody to action
Second client is known as the first client.The first client and the second client both client, but it is not same visitor
Family end.
Spreading codes (spreading code) have certain characteristic, can separate each other, receiver is according to different spreading codes
The information of needs is taken out to filter out other users signal.In without news communication system, spreading codes mismatch is to influence communication
One of an important factor for efficiency.Since in actual transmission channel environment, spreading codes are impossible right-on, receiving end
The spreading codes used needed for receiving, which will receive other multi-paths and channel attenuation, to be influenced, thus the production with receiving end signal
Spreading codes produced by raw circuit have non-matching phenomenon and then generate non-orthogonal situation.Once received required use
Non-orthogonal situation occurs for the spreading codes that person's spreading codes and receiving end generate, then between other users and required user
Correlation just will increase, and MAI (Multiple Access Interference, multi-access inference) problem is just opposite to become very tight
Weight.
The application provides a kind of signal processing method based on sef-adapting filter, device, electronic equipment and computer can
Storage medium is read, can be improved without in news communication system, in the unmatched situation of spreading codes, improve the performance of wireless communication,
The interference in signal is reduced, and reduces computation complexity, improves calculating speed.
Fig. 1 is the system architecture diagram of the signal processing method based on sef-adapting filter in one embodiment.Such as Fig. 1 institute
Show, the weight vectors of sef-adapting filter can be constructed, which includes Adaptive Weight vector waWith fixed weight vector
wq.Sef-adapting filter obtains input signal x (n), can be according to fixed weight vector wqInput signal x (n) is handled, is obtained
To desired output signal d (n).Matrix B can be obstructed according to signal to handle input signal x (n), obtain M signal z
(n), further according to Adaptive Weight vector waM signal z (n) is handled, processing result u (n) is obtained.It can be according to processing
As a result u (n) and desired output signal d (n) obtain real output signal y (n).Sef-adapting filter can pass through adaptive algorithm pair
Adaptive Weight vector waIt is updated.
As shown in figure 4, in one embodiment, a kind of signal processing method based on sef-adapting filter is provided, including
Following steps:
Step 210, the weight vectors of sef-adapting filter are constructed, weight vectors include the fixed power of Adaptive Weight vector sum
Weight vector.
Sef-adapting filter refers to the change according to environment, changes the parameter and structure of filter using adaptive algorithm
Filter.Sef-adapting filter is can be estimated as foundation with the statistical property of input signal and output signal, is taken specific
Algorithm automatically adjusts filter coefficient, reaches a kind of algorithm or device of optimum filtering characteristic.
The weight vectors of sef-adapting filter can be constructed, this can weight vectors include Adaptive Weight vector waIt is weighed with fixed
Weight vector wq, wherein fixed weight vector wqIt can remain unchanged, Adaptive Weight vector waUsing adaptive algorithm, with defeated
Enter the variation of signal and output signal and is changed update.In one embodiment, the weight vectors w=of sef-adapting filter
wq-Bwa, wherein B is that signal obstructs matrix, and signal barrier matrix B can be fixed matrix, can obstruct rectangular array by signal
B obstructs input x (n), to obtain z (n).
In one embodiment, Adaptive Weight vector can be as shown in formula (1):
Wherein, waFor Adaptive Weight vector, RzFor z (n) autocorrelation matrix of the M signal, Rz=BHRB, pzFor
The intercorrelation vector of desired output signal d (n) and M signal z (n), pz=BHRwq.The output signal of sef-adapting filter
Step 220, divide several pairs of Adaptive Weight vectors according to selected sub- weight to be split, obtain and sub- weight point
Cut the sub- weight vectors of several corresponding numbers.
Sef-adapting filter can divide several pairs of Adaptive Weight vectors according to selected sub- weight and be split, by suitable
Answering property weight vectors can provide fast convergence speed, and reduce computation complexity.Sub- weight segmentation number can be according to actual needs
It is set, in one embodiment, can also be set according to the length of spreading codes, it is optional if the length of spreading codes is longer
Biggish sub- weight segmentation number is selected, even partition is carried out to Adaptive Weight vector.
Input signal x (n) can be divided into z (n)=[z by the M signal z (n) that barrier obtains1(n), z2(n) ...,
zM(n)]T, wherein M is selected sub- weight segmentation number, ziIt (n) is a Li× 1 vector, andBy z
(n) it is formed by autocorrelation matrix RzSubmatrix R can be redeveloped intoij=E [zi(n)zj H(n)], wherein i, j=1,2 ..., M；Then
Intercorrelation vector also may be partitioned into pz=BHRwq=[p1, p2..., pM]T, so that each subvector is
Adaptive Weight vector waIt can be for wa=[w1, w2... ..., wM]T.Available sub- weight vectors can be such as formula
(2) shown in:
Wherein, wiIndicate i-th of sub- weight vectors.
Step 230, each sub- weight vectors are calculated, and obtain adaptability power according to each sub- weight vectors being calculated
Weight vector.
Divide several couples of Adaptive Weight vector w according to selected sub- weightaIt is split, it is several right to obtain dividing with sub- weight
Answer the sub- weight vectors w of quantityiAfterwards, each sub- weight vectors w can be calculated separately according to formula (2)iValue.
In one embodiment, it can be calculated according to the Adaptive Weight vector sum real output signal that this is obtained next
Secondary Adaptive Weight vector, to Adaptive Weight vector waIt is iterated update, until obtained real output signal y (n) is full
Sufficient preset condition.The preset condition can be set according to actual needs, for example, can be real output signal y (n) and expectation
The variance of output signal d (n) reaches minimum etc., but not limited to this.Optionally, RLS (Recursive Least can be passed through
Square, least square method of recursion) etc. adaptive algorithms Adaptive Weight vector is iterated, other algorithms can also be used
Calculating is iterated to Adaptive Weight vector.Herein in order to enable convenience of calculation, can enable
To Adaptive Weight vector waThe formula calculated can be as shown in formula (3)~formula (6):
wa=wa(n-1)+gz(n)d*(n) formula (3)；
Pz(n)=μ-1[Pz(n-1)-gz(n)zH(n)Pz(n-1)] formula (5)；
It can be to Adaptive Weight vector waAnd Pz(n) initializing set is carried out, for example, w can be enableda(0)=0, Pz
(n)=δ-1I, wherein δ can take lesser positive constant in high SNR, and biggish positive constant is taken in low SNR.
Divided after several pairs of Adaptive Weight vectors are split according to selected weight, every sub- weight vectors are counted
Calculate, this second son weight vectors can be calculated according to the sub- weight vectors that be calculated of last time and output signal, to sub- weight to
Amount is iterated, constantly the value of the sub- weight vectors of adjustment update.According to formula (3) to formula (6), sub- weight vectors w is calculatediIt can be according to
Formula (7) is to shown in formula (10):
WhereinAnd i
=1,2 ..., M.μ can be defined as forgetting factor, and 0 < μ < 1.μ can be acquired with formula (11):
Wherein,Adaptive Weight vector is split, and calculates separately every height
Weight vectors can be effectively reduced computation complexity and increase convergence rate, improve calculating speed.
Step 240, it is carried out according to input signal of the Adaptive Weight vector sum fixed weight vector to sef-adapting filter
Processing, obtains output signal.
It, can be according to w after the sub- weight vectors for the weight segmentation number corresponding number that sef-adapting filter is calculated and selectesa
=[w1, w2... ..., wM]TAdaptive Weight w is calculateda, and can be according to Adaptive Weight vector waWith fixed weight vector wq
The input signal of sef-adapting filter is handled, output signal is obtained.In one embodiment, cocoa is according to fixed weight
Vector wqInput signal x (n) is handled, desired output signal d (n) is obtained, matrix B can be obstructed according to signal and input is believed
Number x (n) is handled, and M signal z (n) is obtained, further according to Adaptive Weight vector waM signal z (n) is handled,
And real output signal y (n) is obtained according to desired output signal d (n).Wherein, real output signal y (n) can basisIt is calculated.
In the present embodiment, the weight vectors for constructing sef-adapting filter, it is several to described according to selected sub- weight segmentation
Adaptive Weight vector is split, and is obtained the sub- weight vectors for dividing number corresponding number with sub- weight, is calculated each sub- weight
Vector, and Adaptive Weight vector is obtained according to each sub- weight vectors being calculated, and according to Adaptive Weight vector sum
Fixed weight vector is handled to the input signal to sef-adapting filter, obtains output signal, can be improved wireless communication
Performance, reduce the interference in signal, and by being split to weight vectors, reduce computation complexity, improve calculating
Speed.
As shown in figure 3, in one embodiment, each sub- weight vectors are calculated in step 230, and according to being calculated
It is further comprising the steps of after each sub- weight vectors obtain Adaptive Weight vector:
Step 302, restraint-types are not waited according to the building of fixed weight vector is secondary, and is being adapted to according to secondary not equal restraint-types
Load terms are added in property weight vectors.
In one embodiment, secondary not equal restraint-types limitation adaptibility weight vectors w can be constructedaNorm, can be with
Robustness of the sef-adapting filter in the unmatched situation of spreading codes is improved, norm (norm) is that one of mathematics is substantially general
Read, it is typically defined in normed linear space, and be frequently used to measure each of some vector space (or matrix) to
The length or size of amount.The norm of Adaptive Weight vector is limited toWherein, two
The value of secondary not equal restraint-types is To=| | wq||2Or α2=| | wq||2。
Step 304, it is limited according to the norm that secondary not equal restraint-types obtain sub- weight vectors.
It can be limited according to the norm that the secondary not equal restraint-types of building obtain the sub- weight vectors after segmentation.Implement at one
In example, number can be divided according to selected sub- weight and the secondary norm for not waiting restraint-types to obtain sub- weight vectors limits, sub- weight
The norm limitation that the norm limitation of vector can be Adaptive Weight vector divides number divided by sub- weight.Adaptive Weight vector
Secondary not equal restraint-types areThe value of the secondary not equal restraint-types can be set as To=| | wq
||2Or α2=| | wq||2, then the norm limitation of sub- weight segmentation number can be set asWherein, M is
Sub- weight divides number.
Step 306, the expression formula of sub- weight vectors is obtained according to load terms, expression formula includes that sub- weight vectors are corresponding negative
Carry level value.
In one embodiment, formula (1) can be added in the secondary not equal restraint-types of building, then Adaptive Weight vector is available
Formula (12) indicates:
wa=(R+ γ I)-1C(CH(R+γI)-1C)-1F formula (12)；
Wherein, γ I may be defined as load terms, and γ then can be used for indicating load level, Adaptive Weight vector wa=(Rz+γ
I)-1pzIt is the R in formula (1)zγ I is added afterwards.As γ I >=0, the matrix R of load termsz+ γ I be positive number, then Adaptive Weight to
Measure waNorm can change therewith because of γ value；As γ I=0, Adaptive Weight vector waIt is just identical as formula (1), at this point, negative
Load value can satisfy secondary not grade restraint-types, therefore, can incite somebody to actionIt is defined as being not used in sef-adapting filter
The Adaptive Weight vector of secondary not equal restraint-types, and by wa=(Rz+γI)-1pzIt is defined as using after secondary not equal restraint-types
Adaptive Weight vector.It is inciting somebody to actionAfter proposition, waI.e. available formula (13) indicate:
In one embodiment, for γ,Two progress before Taylor series expansion can be used
It indicates, can enable respectively Due to g γ ≈ g0+ γ g ' 0, available (I+ γ Rz-1) -1 ≈ I- γ Rz-1.It can be by formula (13)
It is changed to formula (14):
Wherein,Work as waIt is unable to satisfy norm limitation ‖ wa‖2>α2When, formula (14) can be unfolded, be obtained
To formula (15):
Wherein,It may be defined as
The real part of x.According to the available load level of solution of formula (15)
In one embodiment, Ke YilingSecondary differ is not used in expression
The sub- weight of restraint-type is limited the quantity, and wa=(Rz+γI)-1pzIt, can to use the Adaptive Weight vector after secondary not equal restraint-types
The expression formula of the sub- weight vectors comprising load terms is obtained to construct.It can be by wa=(Rz+γI)-1pzIn diagonal matrix γ in
γ diagonal element γiInstead of, and according to wa=[w1, w2... ..., wM]T, then waIt can indicate are as follows:
It, can be as shown in formula (17) according to the expression formula of the available sub- weight vectors of formula (16):
It is available again via Taylor series expansionThen formula (17)
It can be revised as shown in formula (18):
Wherein,In one embodiment, sub- weight vectors wiNorm be limited to
The value of norm limitation can be set asGroup weight vectors wiIt, can when being unable to satisfy norm limitation
It will according to formula (18)It is shown as formula (19), and is solved according to formula (19) and obtain sub- weight vectors wiWater load
Level values γi。
It can be solved to obtain sub- weight vectors w according to formula (19)γWater load level values
Then ai=‖ vi‖2> 0, Wherein,It can
It can be positive number, zero or negative.WhenWhen, thenThen formula (19)
Can there can be two positive number real solutions and meet sub- weight vectors wiNorm limitation, can be selected from two positive number real solutions
Lesser value is used as sub- weight vectors wiWater load level values γi.WhenWhen, formula (19) obtained solution can
A pair of of conjugate complex number pair can be unable to get, then it can be rightIt is minimized, and according to formula (19) to γiDifferential is carried out, it can
To obtainThen available sub- weight vectors wiWater load level values γi=-bi/
2αi.Available sub- weight vectors wiWater load level values
Sub- weight vectors w is calculatediWater load level values γiIt afterwards, can be according to sub- weight vectors wiWater load level values
γiEach sub- weight vectors w is calculatedi, further according to each sub- weight vectors wiSeek the Adaptive Weight of adaptive filter
Vector wa, handled so as to the input signal to adaptive filter, obtain output signal.
In the present embodiment, load terms are added in autocorrelation matrix, can be improved sef-adapting filter spreading codes not
Robustness in matched situation can be improved the performance of wireless communication, reduce the interference in signal.And by weight vectors
It is first split and calculates again, reduce computation complexity, improve calculating speed.
In one embodiment, adaptive filter can be by adaptive algorithm to Adaptive Weight vector waIt is iterated
It updates, Adaptive Weight vector next time can be calculated according to the Adaptive Weight vector sum real output signal that this is obtained,
To Adaptive Weight vector waIt is iterated update, until obtained real output signal y (n) meets preset condition.It can basis
Formula (3)~formula (6) is to Adaptive Weight vector waIt is iterated update.
Load terms γ I is added in autocorrelation matrix, does not wait restraint-types limitation adaptibility weight vectors w using secondarya's
Norm can be iterated update to the load level γ in load terms γ I according to adaptive algorithm.It can enable
For the Adaptive Weight vector of secondary not equal restraint-types is not used in sef-adapting filter, and enable wa=(Rz+γI)-1pzBeing set to makes
With the Adaptive Weight vector after secondary not equal restraint-types.When being iterated calculating, formula (3) can be changed to formula (20):
It can determine whether this Adaptive Weight vector that secondary not equal restraint-types are not used being calculatedWhether model is met
Number limitationIf meetingThen this Adaptive Weight vector w being calculateda
As shouldNamelyIf being unsatisfactory for norm limitationThen can first it calculate
This Adaptive Weight vector wa(n) corresponding load level γ (n), then obtain this Adaptive Weight vector wa(n)。
Calculate this Adaptive Weight vector wa(n) corresponding load level γ (n) can be as shown in formula (21):
Seek after obtaining this load level γ (n), can be calculated according to formula (22) this Adaptive Weight to
Measure wa(n):
Diagonalization variable load item γ I, the load level of each load terms are updated according to the adaptive algorithms iteration such as RLS
It is to be obtained from the solution of quadratic polynomial (21), Shandong of the sef-adapting filter in the unmatched situation of spreading codes can be improved
Stick, and improve convergence rate.
In one embodiment, sef-adapting filter, may during being split to Adaptive Weight vector
Generate dynamic estimation error, for the influence for understanding few dynamic error, can when being iterated update to every sub- weight vectors,
Update is iterated to the corresponding water load level values of sub- weight vectors.It can enable
It indicates that the secondary sub- weight limitation for not waiting restraint-types is not used, it, can be by formula (7) more when being iterated update to sub- weight vectors
It is changed to formula (23):
It can determine whether this sub- weight vectors that secondary not equal restraint-types are not used being calculatedWhether norm limit is met
SystemIf meetingThen this sub- weight vectors w being calculatediAs should
NamelyIf being unsatisfactory for norm limitationThen can first calculate this sub- weight to
Measure wi(n) corresponding water load level values γi(n), then this sub- weight vectors w is obtainedi(n)。
Calculate this sub- weight vectors wi(n) corresponding water load level values γiIt (n) can be as shown in formula (24):
It seeks obtaining this second son weight vectors wi(n) corresponding water load level values γi(n) it after, can be calculated according to formula (25)
To this sub- weight vectors wi(n):
Calculate this each sub- weight vectors wi(n) after, then this Adaptive Weight vector w can be obtaineda(n), it adapts to
Property weight vectors wa(n)=[w1(n), w2(n) ..., wM(n)]。
In the present embodiment, diagonalization variable load item is updated by adaptive algorithm iteration, each load terms are born
Load level is obtained from the solution of quadratic polynomial, and sef-adapting filter can be improved in the unmatched situation of spreading codes
Robustness, and improve convergence rate.And after variable water load level values are added, it can reduce and be split in Adaptive Weight vector
During, it is influenced caused by the dynamic estimation error of generation.
In one embodiment, above-mentioned various modes can be calculated separately and calculate Adaptive Weight vector waCalculation amount, can
Respectively include: by adaptive algorithm to the Adaptive Weight vector w that secondary not equal restraint-types are not usedaIt is iterated update
Mode (i.e. formula (3)~formula (6))；By Adaptive Weight vector waBe divided into sub- weight segmentation number corresponding number sub- weight to
It measures, and (the i.e. formula in such a way that adaptive algorithm is iterated update to the sub- weight vectors that secondary not equal restraint-types are not used
(7)~formula (10))；By adaptive algorithm to the Adaptive Weight vector w for using secondary not equal restraint-typesaIt is iterated update
Mode (i.e. formula (20)~formula (22))；By Adaptive Weight vector waIt is divided into and is weighed with the son of sub- weight segmentation number corresponding number
Weight vector, and adaptive algorithm to by way of using the sub- weight vectors of secondary not equal restraint-types to be iterated update (i.e.
Formula (23)~formula (25)) etc., but not limited to this.
Since barrier matrix B is fixed, and obtained M signal z (n) can be with after barrier by input signal x (n)
It is calculated with the hardware of high speed, therefore the computation complexity of the M signal z (n) obtained after barrier can be ignored.It is false
If by Adaptive Weight vector waIt is divided into M sub- weight vectors, and i-th of sub- weight vectors wiSmall greatly Li, then formula (8)
In each gi(n) it about needsComplex multiplication, and P in formula (9)i(n) needsComplex multiplication, formula
(10) f (n) in then needs the complex multiplication of L=(L-1), and qi(n) and Δ hjIt (n) is to need L respectivelyiWith's
Complex multiplication.Each of formula (23)Need LiComplex multiplication, each v in formula (24)i(n) it needsComplex multiplication
Method, ai、bi、ciWith the w in formula (25)i(n) it is respectively necessary for identical LiComplex multiplication.
Above-mentioned various modes can be calculated and calculate Adaptive Weight vector waComputation complexity, wherein by suitable
Answering property algorithm is to the Adaptive Weight vector w that secondary not equal restraint-types are not usedaThe calculating for being iterated the mode of update is complicated
Degree is 2L2+6L；By Adaptive Weight vector waThe sub- weight vectors with sub- weight segmentation number corresponding number are divided into, and by suitable
Answering property algorithm is to the computation complexity for the mode that the sub- weight vectors that secondary not equal restraint-types are not used are iterated updateThe adaptability for using secondary not equal restraint-types is weighed by adaptive algorithm
Weight vector waThe computation complexity for being iterated the mode of update is 3L2+10L；By Adaptive Weight vector waIt is divided into and is weighed with son
Divide the sub- weight vectors of number corresponding number again, and passes through adaptive algorithm to the sub- weight vectors for using secondary not equal restraint-types
The computation complexity for being iterated the mode of update is
In the present embodiment, different modes can be calculated and calculate Adaptive Weight vector waComputation complexity, to weight to
Amount is split, and reduces computation complexity, improves calculating speed.
It should be understood that although each step in above-mentioned each flow diagram is successively shown according to the instruction of arrow
Show, but these steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly state otherwise herein, this
There is no stringent sequences to limit for the execution of a little steps, these steps can execute in other order.Moreover, above-mentioned each stream
At least part step in journey schematic diagram may include that perhaps these sub-steps of multiple stages or stage be simultaneously for multiple sub-steps
It is not necessarily and executes completion in synchronization, but can execute at different times, the execution in these sub-steps or stage
Sequence, which is also not necessarily, successively to be carried out, but can be at least the one of the sub-step or stage of other steps or other steps
Part executes in turn or alternately.
As shown in figure 4, in one embodiment, providing a kind of signal processing apparatus 400 based on sef-adapting filter, wrap
Include vector constructing module 402, segmentation module 404, Adaptive Weight vector calculation module 406 and processing module 408.
Vector constructing module 402, for constructing the weight vectors of sef-adapting filter, weight vectors include Adaptive Weight
Vector sum fixed weight vector,
Divide module 404, be split for dividing several pairs of Adaptive Weight vectors according to selected sub- weight, obtain with
The sub- weight vectors of sub- weight segmentation number corresponding number.
Adaptive Weight vector calculation module 406, it is each for calculating each sub- weight vectors, and according to what is be calculated
Sub- weight vectors obtain Adaptive Weight vector.
Processing module 408, for the input according to Adaptive Weight vector sum fixed weight vector to sef-adapting filter
Signal is handled, and output signal is obtained.
In one embodiment, processing module 408, including first processing units, blocker unit and the second processing unit.
First processing units, for according to fixed weight vector wqInput signal x (n) is handled, obtains it is expected defeated
Signal d (n) out.
Blocker unit handles input signal x (n) for obstructing matrix B according to signal, obtains M signal z
(n)。
The second processing unit, for according to Adaptive Weight vector waM signal z (n) is handled, is handled
As a result u (n), and real output signal y (n) is obtained according to processing result u (n) and desired output signal d (n).
In one embodiment, Adaptive Weight vectorWherein, RzFor M signal z (n) from phase
Close matrix, pzFor the intercorrelation vector for it is expected output signal d (n) and M signal z (n).
In the present embodiment, the weight vectors for constructing sef-adapting filter, it is several to described according to selected sub- weight segmentation
Adaptive Weight vector is split, and is obtained the sub- weight vectors for dividing number corresponding number with sub- weight, is calculated each sub- weight
Vector, and Adaptive Weight vector is obtained according to each sub- weight vectors being calculated, and according to Adaptive Weight vector sum
Fixed weight vector is handled to the input signal to sef-adapting filter, obtains output signal, can be improved wireless communication
Performance, reduce the interference in signal, and by being split to weight vectors, reduce computation complexity, improve calculating
Speed.
In one embodiment, the above-mentioned signal processing apparatus 400 based on sef-adapting filter, in addition to including that vector constructs
Module 402, segmentation module 404, Adaptive Weight vector calculation module 406 and processing module 408, further include iteration module.
Iteration module, the Adaptive Weight vector sum output signal for being obtained according to this calculate adaptability next time and weigh
Weight vector, is iterated update to Adaptive Weight vector, until obtained output signal meets preset condition.
In one embodiment, iteration module is also used to calculate every sub- weight vectors, can be according to the last time
The sub- weight vectors and output signal being calculated calculate this second son weight vectors, are iterated update to sub- weight vectors.
In the present embodiment, it can be iterated update to Adaptive Weight vector, improve the performance of wireless communication, reduce letter
Interference in number.
In one embodiment, the above-mentioned signal processing apparatus 400 based on sef-adapting filter, in addition to including that vector constructs
Module 402, segmentation module 404, Adaptive Weight vector calculation module 406, processing module 408 and iteration module, further include bearing
It carries adding module, sub- weight system limitation module and expression formula and obtains module.
Adding module is loaded, not equal is limited for according to the secondary not equal restraint-types of fixed weight vector building, and according to secondary
Standard adds load terms in the Adaptive Weight vector, wherein secondary not equal restraint-types are the model of Adaptive Weight vector
Number limitation.
Sub- weight system limits module, for being limited according to the secondary norm for not waiting restraint-types to obtain sub- weight vectors.
In one embodiment, number is divided according to the sub- weight and secondary not equal restraint-types obtains the sub- weight vectors
Norm limitation, wherein the norm of the Adaptive Weight vector is limited to It is secondary
The value of grade restraint-types isOrThe norm of the sub- weight segmentation number is limited toM is that sub- weight divides number.
Expression formula obtains module, and for obtaining the expression formula of sub- weight vectors according to load terms, expression formula includes sub- weight
The corresponding water load level values of vector.
In one embodiment, expression formula obtains module, including replaces unit and acquiring unit.
Instead of unit, for the diagonal matrix in load terms to be replaced with diagonal element.
Acquiring unit, for obtaining the expression formula of sub- weight vectors according to diagonal matrix element, and by diagonal matrix element
Water load level values as sub- weight vectors.
In one embodiment, Adaptive Weight vector calculation module 406 is also used to basis and every sub- weight vectors pair
The water load level values answered calculate corresponding sub- weight vectors, and obtain adaptability power according to each sub- weight vectors being calculated
Weight vector, wherein Adaptive Weight vector wa=[w1, w2... ..., wM]T, wherein M is that sub- weight divides number, wiIt indicates i-th
Sub- weight vectors, i are the integer greater than 0.
In the present embodiment, diagonalization variable load item is updated by adaptive algorithm iteration, each load terms are born
Load level is obtained from the solution of quadratic polynomial, and sef-adapting filter can be improved in the unmatched situation of spreading codes
Robustness, and improve convergence rate.And after variable water load level values are added, it can reduce and be split in Adaptive Weight vector
During, it is influenced caused by the dynamic estimation error of generation.
In one embodiment, the embodiment of the present application also provides a kind of electronic equipment, including memory and processor, this is deposited
Computer program is stored in reservoir, when computer program is executed by processor so that processor realize it is above-mentioned based on adaptive
Answer the signal processing method of filter.
In one embodiment, the embodiment of the present application also provides a kind of computer readable storage medium, is stored thereon with meter
Calculation machine program realizes the above-mentioned signal processing method based on sef-adapting filter when computer program is executed by processor.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a non-volatile computer and can be read
In storage medium, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage is situated between
Matter can be magnetic disk, CD, read-only memory (Read-OnlyMemory, ROM) etc..
It may include as used herein non-volatile to any reference of memory, storage, database or other media
And/or volatile memory.Suitable nonvolatile memory may include read-only memory (ROM), programming ROM (PROM),
Electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include arbitrary access
Memory (RAM), it is used as external cache.By way of illustration and not limitation, RAM is available in many forms, such as
It is static RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDR SDRAM), enhanced
SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM).
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (11)
1. a kind of signal processing method based on sef-adapting filter characterized by comprising
The weight vectors of sef-adapting filter are constructed, the weight vectors include Adaptive Weight vector sum fixed weight vector；
The Adaptive Weight vector is split according to selected sub- weight segmentation is several, is obtained and the sub- weight segmentation number
The sub- weight vectors of corresponding number；
Calculate each sub- weight vectors, and according to each sub- weight vectors being calculated obtain the Adaptive Weight to
Amount；
The input signal of the sef-adapting filter is handled according to the Adaptive Weight vector sum fixed weight vector,
Obtain output signal.
2. according to the method described in claim 1, it is described according to the Adaptive Weight and fixed weight vector to it is described from
The input signal of adaptive filter is handled, and output signal is obtained, comprising:
According to fixed weight vector wqInput signal x (n) is handled, desired output signal d (n) is obtained；
Matrix B is obstructed according to signal to handle the input signal x (n), obtains M signal z (n)；
According to Adaptive Weight vector waThe M signal z (n) is handled, obtains processing result u (n), and according to described
Processing result u (n) and desired output signal d (n) obtain real output signal y (n).
3. according to the method described in claim 2, it is characterized in that, the Adaptive Weight vectorWherein,
RzFor z (n) autocorrelation matrix of the M signal, pzFor the intercorrelation for it is expected output signal d (n) and M signal z (n)
Vector.
4. the method according to claim 1, wherein the method also includes:
Adaptive Weight vector next time is calculated according to the Adaptive Weight vector sum output signal that this is obtained, to the adaptation
Property weight vectors be iterated update, until obtained output signal meets preset condition.
5. the method according to claim 1, wherein being obtained and the sub- weight segmentation number corresponding number described
Sub- weight vectors after, the method also includes:
Restraint-types are not waited according to fixed weight vector building is secondary, and according to the secondary not equal restraint-types in the adaptation
Property weight vectors in add load terms, wherein the secondary not equal restraint-types are that the norm of the Adaptive Weight vector limits；
The norm limitation of the sub- weight vectors is obtained according to the secondary not equal restraint-types；
The expression formula of the sub- weight vectors is obtained according to the load terms, the expression formula includes that the sub- weight vectors are corresponding
Water load level values.
6. according to the method described in claim 5, it is characterized in that, described obtain the son according to the secondary not equal restraint-types
The norm of weight vectors limits, comprising:
According to the sub- weight segmentation number and the secondary norm limitation for not waiting restraint-types to obtain the sub- weight vectors, wherein institute
The norm for stating Adaptive Weight vector is limited toThe value of the secondary not equal restraint-types
For To=| | wq||2Or α2=| | wq||2；The norm of the sub- weight segmentation number is limited toM
Divide number for sub- weight.
7. according to the method described in claim 5, it is characterized in that, described obtain the sub- weight vectors according to the load terms
Expression formula, comprising:
Diagonal matrix in the load terms is replaced with diagonal element；
The expression formula of sub- weight vectors is obtained according to the diagonal matrix element, and using the diagonal matrix element as the son
The water load level values of weight vectors.
8. according to the method described in claim 5, it is characterized in that, described calculate each sub- weight vectors, and according to meter
Each obtained sub- weight vectors obtain the Adaptive Weight vector, comprising:
Corresponding sub- weight vectors are calculated according to water load level values corresponding with every sub- weight vectors；
The Adaptive Weight vector, the Adaptive Weight vector w are obtained according to each sub- weight vectors being calculateda=
[w1, w2... ..., wM]T, wherein M is that sub- weight divides number, wiIndicate that i-th of sub- weight vectors, i are the integer greater than 0.
9. a kind of signal processing apparatus based on sef-adapting filter characterized by comprising
Vector constructing module, for constructing the weight vectors of sef-adapting filter, the weight vectors include Adaptive Weight to
Amount and fixed weight vector；
Divide module, for being split according to selected sub- weight segmentation is several to the Adaptive Weight vector, obtains and institute
State the sub- weight vectors of sub- weight segmentation number corresponding number；
Adaptive Weight vector calculation module, for calculating each sub- weight vectors, and according to each height being calculated
Weight vectors obtain the Adaptive Weight vector；
Processing module, for the input according to the Adaptive Weight vector sum fixed weight vector to the sef-adapting filter
Signal is handled, and output signal is obtained.
10. a kind of electronic equipment, including memory and processor, computer program, the calculating are stored in the memory
When machine program is executed by the processor, so that the processor realizes method as described in any of the claims 1 to 8.
11. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
Method as described in any of the claims 1 to 8 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910034406.4A CN109889187B (en) | 2019-01-15 | 2019-01-15 | Signal processing method and device based on adaptive filter and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910034406.4A CN109889187B (en) | 2019-01-15 | 2019-01-15 | Signal processing method and device based on adaptive filter and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109889187A true CN109889187A (en) | 2019-06-14 |
| CN109889187B CN109889187B (en) | 2023-06-02 |
Family
ID=66925940
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910034406.4A Active CN109889187B (en) | 2019-01-15 | 2019-01-15 | Signal processing method and device based on adaptive filter and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109889187B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1189364A1 (en) * | 1999-06-23 | 2002-03-20 | Japan,represented by President of Hokkaido University | Radio device |
| US20060133689A1 (en) * | 2004-12-22 | 2006-06-22 | Kenneth Andersson | Adaptive filter |
| CN105654959A (en) * | 2016-01-22 | 2016-06-08 | 韶关学院 | Self-adaptive filtering coefficient updating method and device |
| CN106842237A (en) * | 2017-01-18 | 2017-06-13 | 南京理工大学 | The quick arbitrary shape conformal Adaptive beamformer method of the major lobe of directional diagram |
-
2019
- 2019-01-15 CN CN201910034406.4A patent/CN109889187B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1189364A1 (en) * | 1999-06-23 | 2002-03-20 | Japan,represented by President of Hokkaido University | Radio device |
| US20060133689A1 (en) * | 2004-12-22 | 2006-06-22 | Kenneth Andersson | Adaptive filter |
| CN105654959A (en) * | 2016-01-22 | 2016-06-08 | 韶关学院 | Self-adaptive filtering coefficient updating method and device |
| CN106842237A (en) * | 2017-01-18 | 2017-06-13 | 南京理工大学 | The quick arbitrary shape conformal Adaptive beamformer method of the major lobe of directional diagram |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109889187B (en) | 2023-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zoraster | Minimum peak range sidelobe filters for binary phase-coded waveforms | |
| Mehrotra et al. | Finding an interior point in the optimal face of linear programs | |
| Chang et al. | Analysis of conjugate gradient algorithms for adaptive filtering | |
| Shan et al. | Adaptive algorithms with an automatic gain control feature | |
| Osborne | Nonlinear least squares—the Levenberg algorithm revisited | |
| Fornasini et al. | Doubly-indexed dynamical systems: State-space models and structural properties | |
| Teunissen | On the integer normal distribution of the GPS ambiguities | |
| US7110722B2 (en) | Method for extracting a signal | |
| Reeves | An efficient implementation of the backward greedy algorithm for sparse signal reconstruction | |
| Muhammad et al. | A graph theoretic approach for synthesizing very low-complexity high-speed digital filters | |
| Chien et al. | On Bayesian learning and stochastic approximation | |
| Broersen et al. | Automatic identification of time-series models from long autoregressive models | |
| US20040248515A1 (en) | Multi-user detection | |
| CN109889187A (en) | Signal processing method, device and electronic equipment based on sef-adapting filter | |
| Dogancay et al. | Selective-partial-update NLMS and affine projection algorithms for acoustic echo cancellation | |
| Dunne et al. | Analysis of gradient algorithms for TLS-based adaptive IIR filters | |
| Paleologu et al. | A proportionate NLMS algorithm for the identification of sparse bilinear forms | |
| Hambly et al. | Erratum: stochastic evolution equations for large portfolios of stochastic volatility models | |
| van der Hofstad et al. | Performance of DS-CDMA systems with optimal hard-decision parallel interference cancellation | |
| CN110535452A (en) | System identifying device and method and record have the storage medium of System Discrimination program | |
| Stanciu et al. | FPGA implementation of an optimized NLMS algorithm | |
| Beck et al. | Minimax mean-squared error estimation of multichannel signals | |
| Nazih et al. | Computation of low-rank tensor approximation under existence constraint via a forward-backward algorithm | |
| Macdonald | Density of complex critical points of a real random SO (m+ 1) polynomial | |
| Fair | Computing median unbiased estimates in macroeconometric models |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230714 Address after: Room 1912, Zhuoyue Building, No. 2, Lane 642, Bachelor's Road, Shounan Street, Yinzhou District, Ningbo City, Zhejiang Province, 315100 Patentee after: Ningbo Qixin Semiconductor Technology Co.,Ltd. Address before: 315414 room 3-1-40, building 033, building 2, No. 15, Lane 587, Juxian Road, high tech Zone, Ningbo, Zhejiang Province Patentee before: NINGBO LIANHONG ELECTRONIC TECHNOLOGY Co.,Ltd. |
|
| TR01 | Transfer of patent right |