CN109787924A - A kind of compressed sensing based LoRa channel estimation methods - Google Patents
A kind of compressed sensing based LoRa channel estimation methods Download PDFInfo
- Publication number
- CN109787924A CN109787924A CN201910189366.0A CN201910189366A CN109787924A CN 109787924 A CN109787924 A CN 109787924A CN 201910189366 A CN201910189366 A CN 201910189366A CN 109787924 A CN109787924 A CN 109787924A
- Authority
- CN
- China
- Prior art keywords
- channel
- signal
- pilot
- lora
- follows
- 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 present invention relates to a kind of compressed sensing based LoRa channel estimation methods, belong to wireless sensor network field.This method comprises: firstly, building meets the channel model of LoRa by LoRa linear frequency modulation spread spectrum technology and transmission characteristic with signal because the influence that multipath fading, Doppler frequency shift are decayed combines；Then, in channel estimation, propose that a kind of algorithm based on circulation replacement pilot search designs pilot frequency design, so that the cross correlation of calculation matrix in LoRa channel model is minimum；Finally, transmitter is by sending the subcarrier Jing Guo pilot design, after receiver receives subcarrier, using the degree of rarefication Adaptive matching tracing algorithm based on adaptive step, the status information that channel is reconstructed by pilot sub-carrier recovers the time and frequency domain characteristics value of LoRa channel.The present invention can reduce the complexity of restructuring procedure, improve the overall performance of channel estimation.
Description
Technical field
The invention belongs to wireless sensor network technology fields, are related to a kind of compressed sensing based channel estimation side LoRa
Method.
Background technique
In wireless sensor network, due to the complexity of space environment, the signal of transmitting terminal during transmission can
Phenomena such as being blocked, reflect by various barriers, reflecting, the signal for causing receiving end to receive are the superpositions of multiple signals,
Synchronous signal during transmission also constantly decaying by energy.This has resulted in the multipath fading of channel, so that entire logical
The performance of letter system is greatly affected, in order to realize the compensation to multipath fading in wireless channel transmission, it is necessary to nothing
The channel characteristics of line channel are accurately estimated.
With the development of communication and network technology, Internet of Things has obtained extensive concern, and low-power consumption wide area network (LPWAN) is answered
It transports and gives birth to.The LoRa technology wherein represented the most is even more to have obtained extensive utilization, therefore estimate to the channel of LoRa communication system
Meter, the feature for accurately obtaining its channel have certain realistic meaning.
In current channel estimation, first by sending training sequence in time domain, frequency domain or space, further according to receiving end
The signal received obtains the characteristic parameter of channel by the method for linear reconstruction.But these data are all that can occupy centainly
Channel resource, reduce the utilization rate of frequency band.With the proposition of compressive sensing theory, using to the greatest extent in the estimation of wireless channel
The impulse response that few training sequence carrys out high-precision estimation channel is measured, to improve the availability of frequency spectrum of system.
Summary of the invention
In view of this, reducing the purpose of the present invention is to provide a kind of compressed sensing based LoRa channel estimation methods
The computation complexity of LoRa channel estimation restructuring procedure based on pilot training sequence, to improve the performance of channel estimation.
In order to achieve the above objectives, the invention provides the following technical scheme:
A kind of compressed sensing based LoRa channel estimation methods, specifically includes the following steps:
S1: LoRa channel model is established: special according to the decaying of LoRa characteristic and modulation system and signal in transmission process
Property, establish LoRa channel model；
S2: design pilot schemes: according to the LoRa channel model of compressed sensing principle and foundation, meeting model measurement square
On the basis of battle array cross-correlation minimizes criterion, the scheme of circulation replacement pilot search algorithm is obtained；
S3: channel reconstruction: the restructing algorithm of the degree of rarefication Adaptive matching tracking based on adaptive step connects in receiver
It receives after the pilot signal of compression sampling, channel reconstruction is carried out using the algorithm, with current reconstituted state (signal energy)
Carry out adaptive adjusting step, so that reconstruction error fast convergence, reduces the computation complexity of restructuring procedure.
Further, the step S1 specifically includes the following steps:
S11:LoRa signal is after linear modulation spread spectrum and matched filter, and at a time t, the time domain of output are believed
Number indicate are as follows:
Wherein, A' indicates the amplitude of transmitting signal, f0For centre carrier frequency, k indicates frequency slope, S=B2/ k indicates letter
Number time-bandwidth product, B indicate signal bandwidth；
S12: carrier frequency f, spreading factor SF, channel width Bw and encoding rate are needed to configure in the transmission of LoRa signal
CR, according to different parameter configurations, the spreading rate r ' after obtaining band spectrum modulation is indicated are as follows:
S13: according to the spreading rate of step S12, it is assumed that the preamble length of a LoRa data packet is npreamble, effectively
Data length is PL, then the propagation time T of data packetpacketIt indicates are as follows:
Wherein, TpreambleIndicate preamble transmissions time, TpayloadIndicate the preamble transmissions time；H=0 is indicated using report
Literary head, H=1 indicate not using heading；DE is indicated when LowDataRateOptimize, register settings in sx1278 chip
When being 1, DE=1, otherwise DE=0；
S14: assuming that the transmission range of transmitter to receiver is D, then LoRa signal per unit road in transmission process
Diameter d consumption time beThe influence for considering multipath effect and Doppler effect, when obtaining the propagation under unit length
Between indicate are as follows:
T'=td+tj+αltp
Wherein, tjIndicate multidiameter delay, tpIndicate the time delay under Doppler effect, αlIndicate doppler frequency rate；Then LoRa
The impulse response of channel indicates are as follows:Wherein δ () indicates jump function, hl(t) table
Show the amplitude in path；
S15: assuming that having L paths in communication, doppler frequency rate is constant, the set H of up channelup=[h1,h2,…,
hK1], K1 indicates up channel number, and the reception signal obtained in uplink indicates are as follows:
Assuming that having L paths in communication, doppler frequency rate is constant, the set H of down channeldown=[h1, h1...,
hK2], K2 indicates down channel number, and the reception signal obtained in downlink transfer indicates are as follows:
Wherein, n (t) indicates white Gaussian noise；Consider all subcarriers, defines received signal vector Y, transmission signal vector
For X, noise vector n, then the input/output relation of system are as follows: Y=XH+n.
Further, according to compressed sensing principle in the channel estimation based on pilot frequency sequence, the defeated of pilot signal is being obtained
After entering output relation, need to obtain the function for solving channel state parameter, its step are as follows:
S161: assuming that the carrier number of signal is N, the pilot tone number needed is P, it assumes that pilot set is Λ={ k1,
k2..., kp, it is all t easet ofasubcarriers Ω={ 1,2 ..., N } subset, and the signal that pilot tone is sent indicates are as follows: [X (k1)X
(k2)…X(kP)]T, pilot reception signal expression are as follows: [Y (k1)Y(k2)…Y(kP)]T, noise signal expression are as follows: [n (1) n
(2)…n(P)]T, then send and receive signal between be expressed in matrix as:
Wherein, FP×LFor discrete fourier matrix, indicate are as follows:
Wherein, ω=e-j2π/N；Then YP=XPFP×Lh+nP=Ah+nP, A=XPFP×LIndicate calculation matrix；
S162: by nonlinear IEM model, work as mean square deviationWhen minimum, channel status is obtained
The expression formula of information matrix h are as follows: h=(AHA)-1AHy。
Further, the step S2 is specifically included: calculation matrix A will guarantee that sparse channel state matrix h can be from observation
It is recovered in value y, needs calculation matrix to meet cross-correlation and minimize criterion, obtain the optimal pilot schemes algorithm of myopia,
Steps are as follows:
S21: according to different pilot frequency designs, by the cross correlation expression formula of calculation matrix are as follows:
The pilot frequency mode p then finally exportedoutIt indicates are as follows:
pout=argming (p)
S22: initialization: cycle-index T, number of pilots P, it is random to generate M kind pilot frequency designIt is then corresponding
Calculation matrix cross correlation is
S23: parallel replacement, in s (s≤kp) secondary parallel replacement when, respectively by the s-1 times replacement of every kind of pilot frequency design
Pattern afterwardsIn s-th of element be replaced, substitution is for initial alternate node for the first time；I.e.
It is for pilot frequency sequenceAlternative patternFixed pilot frequency locationsTo own
The set omega of subcarrier withDifference set in element be placed in ps-1, pilot frequency design is then calculated according to cross correlation expression formula
Cross correlation amounts to and calculates N-kpIt is secondary；From the combination being calculated, the smallest pilot frequency design conduct of M cross correlation of reselection
The alternative pattern of the s times replacement, is replaced for the s+1 times, is denoted as
S24: circulation select it is excellent, by kpAfter secondary parallel replacement, M alternate pilot scheme is obtainedM=1,2 ... M, if
Cycle-index is less than T at this time, then continues step S22, otherwise end loop, obtain the smallest pilot schemes conduct of a correlation
Final result pout。
Further, the step S3 is specifically included: being the signal under near-optimization pilot schemes when receiving signal, is passed through
When pilot signal reconstructs channel status, channel is carried out using the restructing algorithm that the degree of rarefication Adaptive matching of adaptive step is tracked
Reconstruct, specifically includes the following steps: input: initial reception pilot signal is arranged, wherein the subcarrier of signal is YP, measurement
Matrix A, threshold residual value ε, additional threshold Γ, initial step length sI, store the set of calculation matrix A and residual error r inner product value
S31: initialization h=[0,0, K, 0]T, hold=[0,0, K, 0]T, rtemp=[0,0, K, 0] T, indexed set
Candidate SetResidual error r0=Yp, the final size L for supporting collectionF=s=sI, final to support collectionIteration index i=1；
S32: in signal parameter after (i-1)-th reconstruct, compare the practical residual error and setting residual error threshold of channel parameter
Whether value approaches, i.e., | | ri-1| | >=ε obtains the channel parameter h to be estimated, is unsatisfactory for then if it is satisfied, then directly terminating
Continue to execute following steps；
S33: calculating calculation matrix and respectively arrange and is directly worth with residual error, there are in SP, SP=| | AHri-1||；
S34: in the SP being calculated in step S33, preceding L is selectedFA maximum value, and by these value corresponding As
Column serial number be deposited into index combine BiIn；
S35: index obtained in step S34 is gathered and final support is merged into candidate support and concentrates, obtains Ci
=Bi∪Fi-1:
S36: it calculatesValue, and select preceding L whereinFThe index of a greatest member obtains finally supporting collection
Fi；
S37: estimated channel parameters obtain
S38: calculating current residual error,
S39: compare current residue | | rtemp| | with threshold residual value ε；
S310: the required channel state matrix h obtained is reconstructed.
Further, comparison current residue described in step S39 | | rtemp| | with threshold residual value ε, specifically include:
(1) if | | rtemp| |≤ε, ri=ri-1, stop iteration and enter S310；
(2) if | | rtemp| | > ε, and have | | rtemp||≥||ri-1| |, compare i-th and (i-1)-th signal reconstruction
Energy between difference and additional threshold Γ；
A. if | | h (Fi)||-||hold| | < Γ, undated parameter information step-lengthThe final size L for supporting collectionF
=LF+ s, the channel parameter h of reconstructold=h (Fi), residual error ri=ri-1, i=i+1, if i≤P, return S32 continue iteration,
Otherwise enter S310 and stop iteration；
B. if | | h (Fi)||-||hold| | >=Γ, then undated parameter finally supports the size L of collectionF=LF+ s, reconstruct
Channel parameter hold=h (Fi), residual error ri=ri-1, i=i+1, if i≤P, return S32 continue iteration, otherwise enter S310
Stop iteration；
(3) if | | rtemp| | > ε, and | | rtemp| | < | | ri-1| |, then undated parameter finally supports collection Fi-1=Fi,
Residual error ri=rtemp, the number of iterations i=i+1 returns to S32 and continues iteration if i≤P, otherwise into S310 stopping iteration.
The beneficial effects of the present invention are:
(1) the LoRa channel model that the present invention constructs, according to the modulation system binding signal of LoRa in transmission process
Multipath fading and Doppler effect, building meet the channel model of LoRa feature；According to the model of channel estimation, one kind is devised
The pilot schemes of near-optimization improve the overall performance of channel estimation so that the mutual phasic property of calculation matrix minimizes；
(2) the invention proposes based on adaptive step degree of rarefication Adaptive matching tracking restructing algorithm, utilize
The algorithm is reconstructed, according to current reconstituted state (signal energy) come adaptive adjusting step s, so that reconstruction signal is quick
Close to echo signal, reduce the complexity of restructuring procedure.
Other advantages, target and feature of the invention will be illustrated in the following description to a certain extent, and
And to a certain extent, based on will be apparent to those skilled in the art to investigating hereafter, Huo Zheke
To be instructed from the practice of the present invention.Target of the invention and other advantages can be realized by following specification and
It obtains.
Detailed description of the invention
To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention is made below in conjunction with attached drawing excellent
The detailed description of choosing, in which:
Fig. 1 is LoRa channel estimation methods flow chart of the present invention；
Fig. 2 is the flow chart of circulation replacement pilot search algorithm；
Fig. 3 is the flow chart of the restructing algorithm of the degree of rarefication Adaptive matching tracking based on adaptive step.
Specific embodiment
Illustrate embodiments of the present invention below by way of specific specific example, those skilled in the art can be by this specification
Other advantages and efficacy of the present invention can be easily understood for disclosed content.The present invention can also pass through in addition different specific realities
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints and application, without departing from
Various modifications or alterations are carried out under spirit of the invention.It should be noted that diagram provided in following embodiment is only to show
Meaning mode illustrates basic conception of the invention, and in the absence of conflict, the feature in following embodiment and embodiment can phase
Mutually combination.
As shown in Figure 1, a kind of compressed sensing based LoRa channel estimation methods of the present invention, specifically include following
Step:
S1: LoRa channel model is established: special according to the decaying of LoRa characteristic and modulation system and signal in transmission process
Property, establish LoRa channel model；
S2: design pilot schemes: according to the LoRa channel model of compressed sensing principle and foundation, meeting model measurement square
On the basis of battle array cross-correlation minimizes criterion, the scheme of circulation replacement pilot search algorithm is obtained；
S3: channel reconstruction: the restructing algorithm of the degree of rarefication Adaptive matching tracking based on adaptive step connects in receiver
It receives after the pilot signal of compression sampling, channel reconstruction is carried out using the algorithm, with current reconstituted state (signal energy)
Carry out adaptive adjusting step, so that reconstruction error fast convergence, reduces the computation complexity of restructuring procedure.
It is established in the Doppler's efficiency propagated in view of signal in conjunction with the peculiar linear frequency modulation spread spectrum modulation system of LoRa
LoRa channel model, its step are as follows:
S11:LoRa signal is after linear modulation spread spectrum and matched filter, and at a time t, the time domain of output are believed
Number indicate are as follows:
Wherein, A' indicates the amplitude of transmitting signal, f0For centre carrier frequency, k indicates frequency slope, S=B2/ k indicates letter
Number time-bandwidth product, B indicate signal bandwidth；
S12: carrier frequency f, spreading factor SF, channel width Bw and encoding rate are needed to configure in the transmission of LoRa signal
CR, according to different parameter configurations, the spreading rate r ' after obtaining band spectrum modulation is indicated are as follows:
S13: according to the spreading rate of step S12, it is assumed that the preamble length of a LoRa data packet is npreamble, effectively
Data length is PL, then the propagation time T of data packetpacketIt indicates are as follows:
Wherein, TpreambleIndicate preamble transmissions time, TpayloadIndicate the data payload transmission time；H=0 indicates to use
Heading, H=1 indicate not using heading；DE is indicated when LowDataRateOptimize, register set in sx1278 chip
When being set to 1, DE=1, otherwise DE=0；
S14: assuming that the transmission range of transmitter to receiver is D, then LoRa signal per unit road in transmission process
Diameter d consumption time beThe influence for considering multipath effect and Doppler effect, when obtaining the propagation under unit length
Between indicate are as follows:
T'=td+tj+αltp
Wherein, tjIndicate multidiameter delay, tpIndicate the time delay under Doppler effect, αlIndicate doppler frequency rate；Then LoRa
The impulse response of channel indicates are as follows:Wherein δ () indicates jump function, hl(t) table
Show the amplitude in path；
S15: actual LoRa multichannel communication, according to LoRa standard, by taking 470~510MHz frequency range as an example.It is passed in uplink
There are 96 channels in defeated, since 470.3MHz, 489.3MHz is incremented to each 200KHZ.Wherein the bandwidth of channel is
125KHz, encoding rate 4/5, the value range of adaptation rate is from DR0~DR5.Assuming that having L paths, Doppler in communication
Frequency modulation rate is constant, the set H of up channelup=[h1,h2,…,hK1], K1 indicates up channel number, obtains in uplink
Receiving signal indicates are as follows:
Similarly, the reception signal in downlink transfer is obtained.48 down channels are shared, since 500.3MHz, with each
200KHZ is incremented to 509.3MHz.Wherein the bandwidth of channel is 125KHz, encoding rate 4/5, the value range of adaptation rate
From DR0~DR5.Assuming that having L paths in communication, doppler frequency rate is constant, the set H of down channeldown=[h1,h1,…,
hK2], K2 indicates down channel number, and the reception signal obtained in downlink transfer indicates are as follows:
Wherein, n (t) indicates white Gaussian noise；Consider all subcarriers, defines received signal vector Y, transmission signal vector
For X, noise vector n, then the input/output relation of system are as follows: Y=XH+n.
Then, according to compressed sensing principle in the channel estimation based on pilot frequency sequence, in the input for obtaining pilot signal
After output relation, need to obtain the function for solving channel state parameter, its step are as follows:
S161: assuming that the carrier number of signal is N, the pilot tone number needed is P, it assumes that pilot set is Λ={ k1,
k2,…,kP, it is all t easet ofasubcarriers Ω={ 1,2 ..., N } subset, and the signal that pilot tone is sent indicates are as follows: [X (k1)X
(k2)…X(kP)]T, pilot reception signal expression are as follows: [Y (k1)Y(k2)…Y(kP)]T, noise signal expression are as follows: [n (1) n
(2)…n(P)]T, then send and receive signal between be expressed in matrix as:
Wherein, FP×LFor discrete fourier matrix, indicate are as follows:
Wherein, ω=e-j2π/N；Then YP=XPFP×Lh+nP=Ah+nP, A=XPFP×LIndicate calculation matrix；
S162: by nonlinear IEM model, work as mean square deviationWhen minimum, channel status is obtained
The expression formula of information matrix h are as follows: h=(AHA)-1AHy。
Fig. 2 is the flow chart of circulation replacement pilot search algorithm, as shown in Fig. 2, pilot schemes described in step S2 are set
Meter specifically includes: calculation matrix A will guarantee that sparse channel state matrix h can be recovered from observation y, need to measure square
Battle array meets cross-correlation and minimizes criterion, obtains the optimal pilot schemes algorithm of myopia, its step are as follows:
S21: according to different pilot frequency designs, by the cross correlation expression formula of calculation matrix are as follows:
The pilot frequency mode p then finally exportedoutIt indicates are as follows:
pout=argming (p)
S22: initialization: cycle-index T, number of pilots P, it is random to generate M kind pilot frequency designIt is then corresponding
Calculation matrix cross correlation is
S23: parallel replacement, in s (s≤kp) secondary parallel replacement when, respectively by the s-1 times replacement of every kind of pilot frequency design
Pattern afterwardsIn s-th of element be replaced, substitution is for initial alternate node for the first time；I.e.
It is for pilot frequency sequenceAlternative patternFixed pilot frequency locationsTo own
The set omega of subcarrier withDifference set in element be placed in ps-1, pilot frequency design is then calculated according to cross correlation expression formula
Cross correlation amounts to and calculates N-kpIt is secondary；From the combination being calculated, the smallest pilot frequency design conduct of M cross correlation of reselection
The alternative pattern of the s times replacement, is replaced for the s+1 times, is denoted as
S24: circulation select it is excellent, by kpAfter secondary parallel replacement, M alternate pilot scheme is obtainedM=1,2 ... M, such as
Cycle-index is less than T to fruit at this time, then continues step S22, otherwise end loop, obtains the smallest pilot schemes of a correlation and makees
For final result pout。
Fig. 3 is the flow chart of the restructing algorithm of the degree of rarefication Adaptive matching tracking based on adaptive step, such as Fig. 3 institute
Show, the channel reconstruction of step S3 specifically includes: being the signal under near-optimization pilot schemes when receiving signal, believed by pilot tone
Number reconstruct channel status when, using adaptive step degree of rarefication Adaptive matching track restructing algorithm carry out channel reconstruction,
It is specifically includes the following steps: input: initial reception pilot signal is arranged, wherein the subcarrier of signal is YP, calculation matrix
A, threshold residual value ε, additional threshold Γ, initial step length sI, store the set of calculation matrix A and residual error r inner product value
S31: initialization h=[0,0, K, 0]T, hold=[0,0, K, 0]T, rtemp=[0,0, K, 0]T, indexed set
Candidate SetResidual error r0=Yp, the final size L for supporting collectionF=s=sI, final to support collectionIteration index i=
1；
S32: in signal parameter after (i-1)-th reconstruct, compare the practical residual error and setting residual error threshold of channel parameter
Whether value approaches, i.e., | | ri-1| | >=ε obtains the channel parameter h to be estimated, is unsatisfactory for then if it is satisfied, then directly terminating
Continue to execute following steps；
S33: calculating calculation matrix and respectively arrange and is directly worth with residual error, there are in SP, SP=| | AHri-1||；
S34: in the SP being calculated in step S33, preceding L is selectedFA maximum value, and by these value corresponding As
Column serial number be deposited into index combine BiIn；
S35: index obtained in step S34 is gathered and final support is merged into candidate support and concentrates, obtains Ci
=Bi∪Fi-1；
S36: it calculatesValue, and select preceding L whereinFThe index of a greatest member obtains finally supporting collection
Fi；
S37: estimated channel parameters obtain
S38: calculating current residual error,
S39: compare current residue | | rtemp| | with threshold residual value ε:
(1) if | | rtemp| |≤ε, ri=ri-1, stop iteration and enter S310；
(2) if | | rtemp| | > ε, and have | | rtemp||≥||ri-1| |, compare i-th and (i-1)-th signal reconstruction
Energy between difference and additional threshold Γ；
A. if | | h (Fi)||-||hold| | < Γ, undated parameter information step-lengthThe final size L for supporting collectionF
=LF+ s, the channel parameter h of reconstructold=h (Fi), residual error ri=ri-1, i=i+1, if i≤P, return S32 continue iteration,
Otherwise enter S310 and stop iteration；
B. if | | h (Fi)||-||hold| | >=Γ, then undated parameter finally supports the size L of collectionF=LF+ s, reconstruct
Channel parameter hold=h (Fi), residual error ri=ri-1, i=i+1, if i≤P, return S32 continue iteration, otherwise enter S310
Stop iteration；
(3) if | | rtemp| | > ε, and | | rtemp| | < | | ri-1| |, then undated parameter finally supports collection Fi-1=Fi,
Residual error ri=rtemp, the number of iterations i=i+1 returns to S32 and continues iteration if i≤P, otherwise into S310 stopping iteration.
S310: the required channel state matrix h obtained is reconstructed.
Finally, it is stated that the above examples are only used to illustrate the technical scheme of the present invention and are not limiting, although referring to compared with
Good embodiment describes the invention in detail, those skilled in the art should understand that, it can be to skill of the invention
Art scheme is modified or replaced equivalently, and without departing from the objective and range of the technical program, should all be covered in the present invention
Scope of the claims in.
Claims (6)
1. a kind of compressed sensing based LoRa channel estimation methods, which is characterized in that this method specifically includes the following steps:
S1: it establishes LoRa channel model: according to the attenuation characteristic of LoRa characteristic and modulation system and signal in transmission process, building
Vertical LoRa channel model；
S2: design pilot schemes: according to the LoRa channel model of compressed sensing principle and foundation, meeting, model measurement matrix is mutual
On the basis of correlation minimizes criterion, the scheme of circulation replacement pilot search algorithm is obtained；
S3: channel reconstruction: the restructing algorithm of the degree of rarefication Adaptive matching tracking based on adaptive step is received in receiver
After the pilot signal of compression sampling, channel reconstruction is carried out using the algorithm, adaptive adjustment is carried out with current reconstituted state
Step-length, so that reconstruction error fast convergence, reduces the computation complexity of restructuring procedure.
2. a kind of compressed sensing based LoRa channel estimation methods according to claim 1, which is characterized in that the step
Rapid S1 specifically includes the following steps:
S11:LoRa signal is after linear modulation spread spectrum and matched filter, at a time t, the time-domain signal table of output
It is shown as:
Wherein, A' indicates the amplitude of transmitting signal, f0For centre carrier frequency, k indicates frequency slope, S=B2/ k indicates signal
Time-bandwidth product, B indicate signal bandwidth；
S12: carrier frequency f, spreading factor SF, channel width Bw and encoding rate CR, root are needed to configure in the transmission of LoRa signal
According to different parameter configurations, the spreading rate r ' after obtaining band spectrum modulation is indicated are as follows:
S13: according to the spreading rate of step S12, it is assumed that the preamble length of a LoRa data packet is npreamble, valid data
Length is PL, then the propagation time T of data packetpacketIt indicates are as follows:
Wherein, TpreambleIndicate preamble transmissions time, TpayloadIndicate the data payload transmission time；H=0 indicates to use message
Head, H=1 indicate not using heading；DE indicate when in chip register LowDataRateOptimize be set as 1 when, DE
=1, otherwise DE=0；
S14: assuming that the transmission range of transmitter to receiver is D, then LoRa signal per unit path d in transmission process disappears
The time of consumption isThe influence for considering multipath effect and Doppler effect, the propagation time obtained under unit length indicate
Are as follows:
T'=td+tj+αltp
Wherein, tjIndicate multidiameter delay, tpIndicate the time delay under Doppler effect, αlIndicate doppler frequency rate；Then LoRa channel
Impulse response indicate are as follows:Wherein δ () indicates jump function, hl(t) road is indicated
The amplitude of diameter；
S15: assuming that having L paths in communication, doppler frequency rate is constant, the set H of up channelup=[h1,h2,…,hK1],
K1 indicates up channel number, and the reception signal obtained in uplink indicates are as follows:
Assuming that having L paths in communication, doppler frequency rate is constant, the set H of down channeldown=[h1,h1,…,hK2], K2
Indicate down channel number, the reception signal obtained in downlink transfer indicates are as follows:
Wherein, n (t) indicates white Gaussian noise；Consider all subcarriers, define received signal vector Y, transmission signal vector X,
Noise vector is n, then the input/output relation of system are as follows: Y=XH+n.
3. a kind of compressed sensing based LoRa channel estimation methods according to claim 2, which is characterized in that according to pressure
Contracting perception principle after the input/output relation for obtaining pilot signal, needs to obtain in the channel estimation based on pilot frequency sequence
The function of channel state parameter is solved, its step are as follows:
S161: assuming that the carrier number of signal is N, the pilot tone number needed is P, it assumes that pilot set is Λ={ k1,k2,…,
kP, it is all t easet ofasubcarriers Ω={ 1,2 ..., N } subset, and the signal that pilot tone is sent indicates are as follows: [X (k1) X
(k2)…X(kP)]T, pilot reception signal expression are as follows: [Y (k1) Y(k2)…Y(kP)]T, noise signal expression are as follows: [n (1) n
(2)…n(P)]T, then send and receive signal between be expressed in matrix as:
Wherein, FP×LFor discrete fourier matrix, indicate are as follows:
Wherein, ω=e-j2π/N；Then YP=XPFP×Lh+nP=Ah+nP, A=XPFP×LIndicate calculation matrix；
S162: by nonlinear IEM model, work as mean square deviationWhen minimum, channel state information is obtained
The expression formula of matrix h are as follows: h=(AHA)-1AHy。
4. a kind of compressed sensing based LoRa channel estimation methods according to claim 3, which is characterized in that the step
Rapid S2 is specifically included: calculation matrix A will guarantee that sparse channel state matrix h can be recovered from observation y, need to measure
Matrix meets cross-correlation and minimizes criterion, obtains the optimal pilot schemes algorithm of myopia, its step are as follows:
S21: according to different pilot frequency designs, by the cross correlation expression formula of calculation matrix are as follows:
The pilot frequency mode p then finally exportedoutIt indicates are as follows:
pout=arg min g (p)
S22: initialization: cycle-index T, number of pilots P, it is random to generate M kind pilot frequency designThen corresponding measurement square
Battle array cross correlation be
S23: parallel replacement, in the s times parallel replacement, respectively by the s-1 times of every kind of pilot frequency design replaced pattern
In s-th of element be replaced, substitution is for initial alternate node for the first time；It is for pilot frequency sequenceAlternative patternFixed pilot frequency locationsBy the set omega of all subcarriers
WithDifference set in element be placed in ps-1, the cross correlation of pilot frequency design is then calculated according to cross correlation expression formula, is amounted to
Calculate N-kpIt is secondary；From the combination being calculated, the smallest pilot frequency design of M cross correlation of reselection is replaced standby as the s times
Pattern is selected, is replaced for the s+1 times, is denoted as
S24: circulation select it is excellent, by kpAfter secondary parallel replacement, M alternate pilot scheme is obtainedIf at this time
Cycle-index is less than T, then continues step S22, otherwise end loop, obtains the smallest pilot schemes of a correlation as final
As a result pout。
5. a kind of compressed sensing based LoRa channel estimation methods according to claim 4, which is characterized in that the step
Rapid S3 is specifically included: being the signal under near-optimization pilot schemes when receiving signal, is reconstructed channel status by pilot signal
When, channel reconstruction is carried out using the restructing algorithm that the degree of rarefication Adaptive matching of adaptive step is tracked, is specifically included following
Step: input: being arranged initial reception pilot signal, and wherein the subcarrier of signal is YP, calculation matrix A, threshold residual value ε are attached
Add threshold value Γ, initial step length sI, store the set of calculation matrix A and residual error r inner product value
S31: initialization h=[0,0, K, 0]T, hold=[0,0, K, 0]T, rtemp=[0,0, K, 0]T, indexed setIt is candidate
CollectionResidual error r0=Yp, the final size L for supporting collectionF=s=sI, final to support collectionIteration index i=1；
S32: in signal parameter after (i-1)-th reconstruct, the practical residual error and setting threshold residual value for comparing channel parameter are
It is no close, i.e., | | ri-1| | >=ε obtains the channel parameter h to be estimated, is unsatisfactory for, continues if it is satisfied, then directly terminating
Execute following steps；
S33: calculating calculation matrix and respectively arrange and is directly worth with residual error, there are in SP, SP=| | AHri-1||；
S34: in the SP being calculated in step S33, preceding L is selectedFA maximum value, and by the column sequence of these value corresponding As
It number is deposited into index and combines BiIn；
S35: index obtained in step S34 is gathered and final support is merged into candidate support and concentrates, obtains Ci=Bi∪
Fi-1；
S36: it calculatesValue, and select preceding L whereinFThe index of a greatest member obtains finally supporting collection Fi；
S37: estimated channel parameters obtain
S38: calculating current residual error,
S39: compare current residue | | rtemp| | with threshold residual value ε；
S310: the required channel state matrix h obtained is reconstructed.
6. a kind of compressed sensing based LoRa channel estimation methods according to claim 5, which is characterized in that step
Comparison current residue described in S39 | | rtemp| | with threshold residual value ε, specifically include:
(1) if | | rtemp| |≤ε, ri=ri-1, stop iteration and enter S310；
(2) if | | rtemp| | > ε, and have | | rtemp||≥||ri-1| |, compare the energy of i-th Yu (i-1)-th signal reconstruction
Difference and additional threshold Γ between amount；
A. if | | h (Fi)||-||hold| | < Γ, undated parameter information step-lengthThe final size L for supporting collectionF=LF
+ s, the channel parameter h of reconstructold=h (Fi), residual error ri=ri-1, i=i+1, if i≤P, return S32 continue iteration, otherwise
Stop iteration into S310；
B. if | | h (Fi)||-||hold| | >=Γ, then undated parameter finally supports the size L of collectionF=LF+ s, the channel of reconstruct
Parameter hold=h (Fi), residual error ri=ri-1, i=i+1 returns to S32 and continues iteration if i≤P, otherwise into S310 stopping
Iteration；
(3) if | | rtemp| | > ε, and | | rtemp| | < | | ri-1| |, then undated parameter finally supports collection Fi-1=Fi, residual error ri
=rtemp, the number of iterations i=i+1 returns to S32 and continues iteration if i≤P, otherwise into S310 stopping iteration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910189366.0A CN109787924B (en) | 2019-03-13 | 2019-03-13 | LoRa channel estimation method based on compressed sensing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910189366.0A CN109787924B (en) | 2019-03-13 | 2019-03-13 | LoRa channel estimation method based on compressed sensing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109787924A true CN109787924A (en) | 2019-05-21 |
| CN109787924B CN109787924B (en) | 2021-06-15 |
Family
ID=66487779
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910189366.0A Active CN109787924B (en) | 2019-03-13 | 2019-03-13 | LoRa channel estimation method based on compressed sensing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109787924B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112904386A (en) * | 2021-01-15 | 2021-06-04 | 武汉梦芯科技有限公司 | Method and system for compensating LoRa Doppler frequency offset based on GNSS |
| CN113037735A (en) * | 2021-03-02 | 2021-06-25 | 河北工程大学 | LoRaWAN bidirectional authentication method based on frame reconstruction |
| WO2022037074A1 (en) * | 2020-08-21 | 2022-02-24 | 华为技术有限公司 | Data transmission method and device, wireless communication system, and storage medium |
| CN114521866A (en) * | 2021-12-31 | 2022-05-24 | 西北大学 | Bioluminescence tomography reconstruction method based on self-adaptive Newton hard threshold tracking method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070087756A1 (en) * | 2005-10-04 | 2007-04-19 | Hoffberg Steven M | Multifactorial optimization system and method |
| CN101764652A (en) * | 2010-01-18 | 2010-06-30 | 哈尔滨工业大学 | Signal detection method having compression perception process based on orthogonal matching pursuit |
| CN104749573A (en) * | 2013-12-31 | 2015-07-01 | 中国科学院电子学研究所 | Sparse stepped-frequency SAR imaging method under spatial-frequency-domain two-dimensional condition |
| CN106357352A (en) * | 2016-08-27 | 2017-01-25 | 成都秦川科技发展有限公司 | Wireless radio frequency module performance test method and device |
| CN108306711A (en) * | 2018-01-02 | 2018-07-20 | 重庆邮电大学 | A kind of data frame based on LoRa quickly continuously transmits method |
| CN109347770A (en) * | 2018-12-10 | 2019-02-15 | 中南大学 | A kind of channel estimation methods and device based on compressive sensing theory |
-
2019
- 2019-03-13 CN CN201910189366.0A patent/CN109787924B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070087756A1 (en) * | 2005-10-04 | 2007-04-19 | Hoffberg Steven M | Multifactorial optimization system and method |
| CN101764652A (en) * | 2010-01-18 | 2010-06-30 | 哈尔滨工业大学 | Signal detection method having compression perception process based on orthogonal matching pursuit |
| CN104749573A (en) * | 2013-12-31 | 2015-07-01 | 中国科学院电子学研究所 | Sparse stepped-frequency SAR imaging method under spatial-frequency-domain two-dimensional condition |
| CN106357352A (en) * | 2016-08-27 | 2017-01-25 | 成都秦川科技发展有限公司 | Wireless radio frequency module performance test method and device |
| CN108306711A (en) * | 2018-01-02 | 2018-07-20 | 重庆邮电大学 | A kind of data frame based on LoRa quickly continuously transmits method |
| CN109347770A (en) * | 2018-12-10 | 2019-02-15 | 中南大学 | A kind of channel estimation methods and device based on compressive sensing theory |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022037074A1 (en) * | 2020-08-21 | 2022-02-24 | 华为技术有限公司 | Data transmission method and device, wireless communication system, and storage medium |
| CN112904386A (en) * | 2021-01-15 | 2021-06-04 | 武汉梦芯科技有限公司 | Method and system for compensating LoRa Doppler frequency offset based on GNSS |
| CN113037735A (en) * | 2021-03-02 | 2021-06-25 | 河北工程大学 | LoRaWAN bidirectional authentication method based on frame reconstruction |
| CN114521866A (en) * | 2021-12-31 | 2022-05-24 | 西北大学 | Bioluminescence tomography reconstruction method based on self-adaptive Newton hard threshold tracking method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109787924B (en) | 2021-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109787924A (en) | A kind of compressed sensing based LoRa channel estimation methods | |
| CN105637775B (en) | The method and device that reciprocity corrects between a kind of base station | |
| JP6539503B2 (en) | Transceiver and method for use in transmitting information in a large scale MIMO system | |
| CN103298124B (en) | Extensive mimo system pilot distribution method based on spatial orthogonality | |
| CN110401476B (en) | Codebook-based millimeter wave communication multi-user parallel beam training method | |
| CN102215186B (en) | Time varying TDD-MIMO (Time Division Duplex-Multiple Input Multiple Output) channel reciprocity compensating method based on LS-SVM (Least Square Support Vector Machine) | |
| RU2008117460A (en) | RADIO TRANSMITTER, RADIO RECEIVER, WIRELESS COMMUNICATION METHOD AND WIRELESS COMMUNICATION SYSTEM | |
| CN109039960A (en) | A kind of underwater sound condition of sparse channel estimation variable step degree of rarefication Adaptive matching method for tracing | |
| CN104320219B (en) | Multi-user's letter can simultaneous interpretation system low complex degree transceiver design method | |
| CN108289004A (en) | A kind of configuration method that channel state information measurement reports and relevant device | |
| KR20070061582A (en) | Method and apparatus for closed loop data transmission | |
| US8699611B2 (en) | Wireless communications system and apparatus | |
| JP2012165370A (en) | Method of estimating channel matrix for channel between transmitter and receiver in wireless multiple-input multiple-output (mimo) network | |
| CN101834814B (en) | Time-variant TDD-MIMO communication channel reciprocity compensation method based on channel prediction | |
| CN105978674A (en) | FDD large-scale MIMO channel estimation pilot frequency optimization method based on compressed sensing | |
| US9036745B2 (en) | Use of neural network based matched filter for fast response time in high-speed communications channels | |
| CN102594739B (en) | Channel estimation methods, pilot frequency information selection method, user equipment and base station | |
| CN105245310A (en) | Downlink pilot signal processing method and system | |
| CN105812083A (en) | Radio-frequency rapid self-interference cancellation method for use in simultaneous same-frequency full duplex system | |
| CN108259397A (en) | Extensive mimo system channel estimation based on adaptive regularization subspace tracking compressed sensing algorithm | |
| CN102160346A (en) | Methods for precoding signals for transmission in wireless mimo system | |
| CN103457647B (en) | A kind of dual-stream beamforming method and device | |
| CN104836605B (en) | A kind of novel dual-mode antenna Combination selection method based on spatial reuse | |
| Zhu et al. | DRL-based STAR-RIS-assisted ISAC secure communications | |
| CN103237356B (en) | Selective cognitive access method on basis of quality of channels |
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 |