CN109698741B - A kind of novel full-duplex communication transmitter and receiver design method - Google Patents
A kind of novel full-duplex communication transmitter and receiver design method Download PDFInfo
- Publication number
- CN109698741B CN109698741B CN201811540407.8A CN201811540407A CN109698741B CN 109698741 B CN109698741 B CN 109698741B CN 201811540407 A CN201811540407 A CN 201811540407A CN 109698741 B CN109698741 B CN 109698741B
- Authority
- CN
- China
- Prior art keywords
- signal
- cluster
- symbol
- label
- label symbol
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/14—Two-way operation using the same type of signal, i.e. duplex
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/10—Monitoring; Testing of transmitters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/20—Monitoring; Testing of receivers
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Detection And Prevention Of Errors In Transmission (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention belongs to wireless communication technology fields, are related to the novel full-duplex communication transmitter of one kind and receiver design method.The present invention proposes a kind of signal detecting method for directly receiving to full duplex system receiving antenna and directly being clustered after N number of signaling point.Specifically, receiving signal will form M1M2A cluster.Signal cluster is indicated by label symbol and sends the corresponding relationship between symbol combination, to recover transmission symbol.The invention has the benefit that method of the invention directly clusters the signal that receiving antenna receives, and using the label symbol sent, to mark the signal cluster being polymerized to.Method proposed by the invention is greatly improved the detection performance of full duplex system, and the remaining self-interference for not needing complicated cumbersome channel estimation and numeric field inhibits step.
Description
Technical field
The invention belongs to wireless communication technology field, it is related to the novel full-duplex communication transmitter of one kind and receiver designs
Method.
Background technique
Although co-channel full duplex communication system can increase exponentially frequency spectrum resource utilization rate simultaneously, since signal is sent
It is carried out simultaneously with frequency with receiving, proximal end receiving antenna can also receive proximal end and be sent out while receiving distal end useful signal
The self-interference signal sent.In order to eliminate numeric field remaining self-interference signal influence, traditional digital self-interference technology for eliminating
It needs accurately to estimate self-interference channel, and rebuilds self-interference letter by feedback signal and self-interference channel estimated value
Number, then subtract from total reception signal the self-interference signal of reconstruction.Since channel estimation errors result in interference signal remnants
The phenomenon that, this severely limits the performances of full duplex system.
Summary of the invention
The present invention is in view of the above-mentioned problems, propose the novel full-duplex communication transmitter of one kind and receiver design method.
The technical solution adopted by the present invention are as follows:
A kind of novel full-duplex communication transmitter and receiver design method set full duplex communication system receiving antenna
Reception signal include useful signal and the self-interference signal transmitted by oneself that external transmitter is sent, total reception signal mode
Type are as follows:
Wherein,WithIt is proximal end respectively
Self-interference signal and distal end useful signal, order of modulation is respectively M1And M2,WithIt is corresponding modulation constellation point set；
P1It is the power of the self-interference signal of proximal end full duplex system, P2It is the available signal power that distal outer transmitter is sent, D is
The order of non-linear self-interference noise, c2d-1It is the power coefficient of the self-interference noise of (2d-1) rank, h0It is distal outer transmitter
Channel response between transmitting antenna and proximal end full duplex system receiving antenna, it is σ that w (n), which obeys power,2The multiple height of Cyclic Symmetry
This distribution, i.e.,W (n) and x (n), z (n) are independent；Wherein σ2Value can be learnt in offline middle test.
Traditional full duplex communication system needs know the information of all correlated channels in (1) formula do self-interference eliminate and
Signal detection.Different from traditional full duplex communication system signal detection scheme, the present invention proposes that a kind of novel full duplex is logical
Believe system design scheme, clustering directly is carried out to the signal that full duplex system receiving antenna receives, then by sending mark
Note symbol indicates the classification of clustering cluster and label symbol, to recover transmission symbol.
Since noise obeys CSCG distribution, so reception signal y (n) obedience mean value isVariance is σ2CSCG distribution, then y (n) probability distribution obey Gauss
Mixed distribution (Gaussian Mixture Model, GMM), it may be assumed that
Wherein K represents the number of the element of composition gaussian variable, is equal to x (n) and z (n) combination variety number M1M2, such as:
BPSK, K=22=4；QPSK, K=42=16, each classification indicates a combination of the two, μk, σ2It is that k-th of classification is corresponding
Reception signal mean value and variance.Therefore, if it is known that the combination of each cluster corresponding x (n) and z (n), then signal detection
Problem can be converted into a clustering problem.
The present invention proposes that a kind of directly receive to full duplex system receiving antenna directly clusters after N number of signaling point
Signal detecting method.Specifically, receiving signal will form M1M2A cluster.It indicates signal cluster by label symbol and sends and accord with
Number combination between corresponding relationship, to recover transmission symbol.
The present invention proposes a kind of direct label symbol design scheme for full duplex system.By in formula 1 it is recognised that connecing
It receives signaling point y (n) to be determined by x (n) and z (n), wherein x (n) and z (n) respectively have M1And M2Kind situation.So y (n) has M1M2Kind can
Energy property, therefore received signal points form M in receiver constellation point1M2Cluster.All M can be emitted1M2Different label symbol combinations
To all M1M2Cluster is marked.So, the combined situation of the x (n) and z (n) of the signaling point in respective cluster and transmitted label
Symbol combination situation is identical.It is specific as follows: the M that will be received1M2The signal following rectangular of different label symbol combinations
Formula:
Wherein, subscript (l) indicates the received signal points of label symbol,It is corresponding
Channel/power coefficient vector form,It is the square of all transmission label symbol combinations
Formation formula, wherein(0≤l≤M1-1).As a result it will be appreciated that in S
(M2L+k) group that column represent label symbol is combined into XlWith Zk-1.So withIn (M2L+k) the signaling point institute of the label symbol arranged
All reception data signal points in cluster are all XlWith Zk-1。
Further, the invention proposes a kind of method that the label symbol for reducing label symbol expense is rebuild, this method is simultaneously
M is not needed1M2A label symbol is marked and only needs that M can be identified with r label symbol1M2The corresponding transmission of a cluster
Symbol, wherein r is the order and r≤M of matrix S1M2.Concrete reason with steps are as follows: assuming that channel is constant in a period of time, i.e.,
All N number of channels for receiving signal experience are identical, so phase shift caused by channel and amplitude variation are identical, and transmitter
Constellation point amplitude-phase is all fixed and it is known that so M in receiver constellation point1M2The central point of a cluster is equally fixed.
Further, it is possible to know all signaling points and dependent, linear correlation can simultaneously be indicated mutually.Therefore the center of all signals
Point can be indicated by the maximal system of linear independence in matrix S.Using this characteristic, the present invention proposes the method that label symbol is rebuild:
Corresponding full duplex system is combined with r label symbol of transmission of far end transmitter, and (A is any one in S labeled as A
Maximal system of linear independence), then it is known that the combination of all label symbols can be write as following form S=AB.It is so corresponding
The received signal points of label symbol can be expressed as
B is multiplied to its right side, it is available
By (5) as can be seen that having rebuild M with r label symbol A1M2A label symbol S, and recover and connect accordingly
Receiving signaling point isAnd M is marked with this1M2A cluster thus can mark each cluster, cluster by the received signal points of recovery
Interior corresponding all data points and label symbol have same transmission symbol combination.Therefore it is right to can detecte out external transmitter institute
The transmission symbol answered.
Further, the selection strategy of optimal label symbol has also been devised in the present invention.Specifically, because clustering algorithm
Performance is very high to the selection sensitivity of initialization.Therefore, optimal label symbol design method, which can maximize, reduces noise
It influences to improve last detection performance.Specific practice is to minimize reconstructed error function:And it solves following
Optimization problem:
Once S is just corresponding fixed since modulation system is given.Then A can be found by offline poor searching method and be made
The smallest optimal label symbol combination of reconstructed error function.
Further, the present invention propose a kind of data preprocessing method with and meanwhile reduce algorithm complexity and improving performance.
Specifically due to oneself known transmitting signal, can by the N number of signaling point received according to oneself identical transmitting signal
Signaling point be grouped.Then it is clustered in each group.Such as defining and receiving data signal point is y(d)=[y (r+1), y
(r+2), y (r+3) ..., y (N)], and define received signal points collection and be combined intoIt therefore, can be with
It learnsWherein subsetIt is defined as follows:
For above-mentioned groupingSince the transmission signal of oneself is all Xl, then wherein of the potential cluster of signaling point
It counts from M1M2It is reduced to M2.Therefore the complexity and improving performance of clustering algorithm can be reduced in clustering.
The invention also provides the improvement clustering algorithm (Modulation-Constraint based on modulation limitation
Clustering), the algorithm consider communication system some inherent features (such as transmitting terminal equiprobability transmission send symbol and
Self-interference signal it is known that and receive constellation point internal relation) as prior information come accelerate cluster speed and improve cluster
Performance.Herein using GMM clustering algorithm as example, but other clustering algorithms such as be equally applicable to K-means.
The feature of communication system is considered first:
1) self-interference signal known to: it is grouped using known self-interference signalTherefore, j-th of signaling point is labeled as in the combinationIts probability density letter
Number is
2) transmitting terminal equiprobability transmission sends symbol:
3) constellation point internal relation: using the corresponding relationship S=AB of all label symbol group sums, r central point is used here
GA reconstructs the central point μ, i.e. μ=gAB=μ of K clusterAB, wherein μA=gA is rebuild similar to label symbol.
Then features described above, the likelihood function for the received signal points that following is distributed based on Gaussian Mixture are utilized.
WhereinFor (the M of B2L+K it) arranges.In order to maximize above-mentioned likelihood function, the random hidden variable of binary is introduced,Wherein φl,j,kZ is represented equal to 1j,jEqual to Zk-1, otherwise φl,j,kEqual to 0 and represent zj,jNo
Equal to Zk-1。zj,jForIn signal to be detected in j-th of received signal points.So, joint probability density is further obtained such as
Shown in lower:
Wherein the l row j column element of Φ is by φl,jVector composition.Further to maximize the input data in formula (9)
Likelihood function, in the hope of Gaussian Mixture distribution parameter Φ and μA.Specific method for solving is with expectation maximization (EM) algorithm
For calculate μAWith Φ until convergence.
E step: following equalities are acquired
WhereinFor zj,jEqual to Zk-1Posterior probability also,
M- step: it is updated with following equalities
It finally checks convergence, if do not restrained, E step is just returned to according to the value of estimation M step estimation, is continuing to change
Generation；If convergence stops iteration.
The invention has the benefit that method of the invention directly clusters the signal that receiving antenna receives, and
Using the label symbol (Label Symbol) of transmission, to mark the signal cluster being polymerized to (Cluster).Side proposed by the invention
Method is greatly improved the detection performance of full duplex system, and does not need the remaining from dry of complicated cumbersome channel estimation and numeric field
Disturb inhibition step.
Detailed description of the invention
Fig. 1 shows system model of the invention；
Fig. 2 shows inventive receivers to detect block diagram；
Fig. 3 shows the clustering algorithm flow chart after the present invention improves；
Fig. 4 shows the performance comparison of the bit error probability of 16QAM/64QAM；
Fig. 5 shows the influence of the long bit error probability to 16QAM of transmission signal packet.
Specific embodiment
The present invention will be described in detail with simulated example with reference to the accompanying drawing, so that those skilled in the art can be more
Understand the present invention well.
Fig. 1 shows system model of the invention, and for full duplex system, full duplex system receiving antenna can be received and be removed
Outside the useful signal that external transmitter is sent, the self-interference signal transmitted by oneself, total receipt signal model can be also received
It is expressed as
Wherein,WithIt is proximal end respectively
Interference signal and distal end useful signal, order of modulation is respectively M1And M2,WithIt is corresponding modulation constellation point set.P1
It is the power of the self-interference signal of proximal end full duplex system, P2It is the available signal power that distal outer transmitter is sent.h0It is remote
The channel response between external transmitter transmitting antenna and proximal end full duplex system receiving antenna is held, it is σ that w (n), which obeys power,2's
Cyclic Symmetry multiple Gauss (CSCG) distribution, i.e.,W (n) and x (n), z (n) are independent.Wherein σ2Value can from
Test is learnt in line.
Fig. 2 shows receivers of the invention to detect block diagram:
Carry out process of data preprocessing first to reduce algorithm complexity and improving performance simultaneously.Specifically due to known
The transmitting signal of oneself can will be grouped in the N number of signaling point received according to the signaling point of oneself identical transmitting signal.
Then it is clustered in each group.Such as defining and receiving data signal point is y(d)=[y (r+1), y (r+2), y (r+3) ..., y
(N)] it, and defines received signal points collection and is combined intoTherefore, it can be seen thatWhereinIt is defined as follows:
Due to for above-mentioned groupingItsIn signaling point potential cluster from M1M2It is reduced to M2。
Then full duplex and external transmitter send r label symbol simultaneously, and (public using the method that label symbol is rebuild
Formula 5) restore the reception constellation points of all label symbols, and to all M1M2One-to-one label is marked in a cluster.
Clustering (formula 11 and 12) is carried out to all constellation points again, and after convergence willSignaling point cluster arrives
(M2l+k*) in cluster, wherein
Finally the relationship according to corresponding to the constellation point and transmission and the label symbol restored that cluster obtains detects outside
Transmitter sends signaling point: such as in the manner described above S=AB, it can be appreciated that (M2l+k*) the corresponding external transmission of a cluster
The transmission symbol of machine isTo detect that external transmitter sends signaling point.
Fig. 3 illustrates the improvement clustering algorithm (Modulation-Constraint based on modulation limitation
Clustering), the algorithm consider communication system some inherent features (such as transmitting terminal equiprobability transmission send symbol and
Self-interference signal it is known that and receive constellation point internal relation) as prior information come accelerate cluster speed and improve cluster
Performance.Herein using GMM clustering algorithm as example, but other clustering algorithms such as be equally applicable to K-means.
The feature of full duplex communication system is considered first:
1) self-interference signal known to: it is grouped using known self-interference signalTherefore, the probability density function of j-th of signaling point is in the combination
2) transmitting terminal equiprobability transmission sends symbol:
3) constellation point internal relation: using the corresponding relationship S=AB of all label symbol group sums, we are only with r here
The central point gA of vertical cluster reconstructs M1M2The central point μ of a cluster, i.e. μ=gAB=μAB, wherein μAThe central point of r independent clusters of=gA
Value.The method principle is consistent with the idea that label symbol is rebuild.
Then features described above, the likelihood function for the received signal points that following is distributed based on Gaussian Mixture are utilized.
WhereinFor (the M of B2L+K it) arranges.In order to maximize above-mentioned likelihood function, we introduce binary hidden change at random
Amount,Wherein φl,j,kZ is represented equal to 1j,jEqual to Zk-1, otherwise φl,j,kEqual to 0 and represent zj,j
Not equal to Zk-1。zj,jForIn signal to be detected in j-th of received signal points.So, it is general further to obtain joint for we
Rate density is as follows:
Wherein the l row j column element of Φ is by φl,jVector composition.Further to maximize the input data in formula (16)
Likelihood function, in the hope of Gaussian Mixture distribution parameter Φ and μA.Specific method for solving is with expectation maximization (EM) algorithm
For calculate μAWith Φ until convergence.
E step: following equalities are acquired
WhereinFor zj,jEqual to Zk-1Posterior probability also,
M- step: it is updated with following equalities
It finally checks convergence, if do not restrained, E step is just returned to according to the value of estimation M step estimation and updates γl,j,kIn
Continue iteration；If convergence stops iteration.
Fig. 4 shows the performance comparison of the bit error probability based on 16QAM/64QAM.It can be seen that this programme proposed
Scheme performance approaches optimal performance very much.(P1=40dB and N=128)
Fig. 5 shows the performance comparison of the long bit error probability to 16QAM of transmission signal packet.It can be seen that the present invention mentions
Even if scheme out similarly plays the role of preferable detection when signal packet is long smaller, performance still approaches optimality very much
Energy.(P1=40dB.).
Claims (5)
1. a kind of novel full-duplex communication transmitter and receiver design method, setting full duplex communication system receiving antenna
Receiving signal includes the useful signal and the self-interference signal transmitted by oneself that external transmitter is sent, total receipt signal model
Are as follows:
Wherein,WithIt is that proximal end is certainly dry respectively
Signal and distal end useful signal are disturbed, order of modulation is respectively M1And M2,WithIt is corresponding modulation constellation point set；P1It is
The power of the self-interference signal of proximal end full duplex system, P2It is the available signal power that distal outer transmitter is sent, D is non-thread
The order of property self-interference noise, c2d-1It is the power coefficient of the self-interference noise of (2d-1) rank, h0It is the transmitting of distal outer transmitter
Channel response between antenna and proximal end full duplex system receiving antenna, it is σ that w (n), which obeys power,2Cyclic Symmetry multiple Gauss point
Cloth, i.e.,W (n) and x (n), z (n) are independent；It is characterized in that,
Signal y (n) will be received and be defined as that there are M1M2Kind possibility, i.e. received signal points form M in receiver constellation point1M2
Cluster, transmitter emission mark symbol is to all M1M2Cluster is marked；Receiver according to signal cluster and is sent by label symbol
Corresponding relationship between symbol combination recovers transmission symbol.
2. the novel full-duplex communication transmitter of one kind according to claim 1 and receiver design method, feature exist
In described couple of all M1M2The method that cluster is marked are as follows: use M1M2Different label symbol combinations are to all M1M2Cluster is marked
Note；Then, the M that will be received1M2Different label symbol combinations are as follows:
Wherein, subscriptIt indicates to receive the reception signal that signal is label symbol,It is phase
The vector form of channel response coefficient is answered,It is that all transmission label symbols combine
Matrix form, wherein0≤l≤M1(M in -1, i.e. S2L+k generation) is arranged
The group of list notation symbol is combined into XlWith Zk-1, while withIn (M2L+k) the institute where the signaling point of the label symbol arranged in cluster
Having reception data signal point is all XlWith Zk-1, w is the corresponding noise for sending symbol S.
3. the novel full-duplex communication transmitter of one kind according to claim 1 and receiver design method, feature exist
In described couple of all M1M2The method that cluster is marked are as follows: M is identified using r label symbol1M2The corresponding transmission symbol of a cluster,
Wherein r is the order and r≤M of matrix S1M2,It is all transmission label symbol combinations
Matrix form, wherein0≤l≤M1- 1, specifically:
The central point for setting all signals can be indicated that defining A is any one in S by the maximal system of linear independence in matrix S
Maximal system of linear independence, and the r label symbol combination sent for indicating transmitter, the combination of all label symbols is defined
It is the sequency spectrum matrix as corresponding to A in S for S=AB, B；Then, the received signal points of respective markers symbol are as follows:
Wherein, subscriptIt indicates to receive the reception signal that signal is label symbol,It is phase
Answer the vector form of channel response coefficient, wAIt is the corresponding noise for sending label symbol A；
The right side multiplies B, obtains:
M has been rebuild with r label symbol A1M2A label symbol S, and recover corresponding received signal points and beAnd it is marked with this
Remember M1M2A cluster marks each cluster by the received signal points of recovery, there is corresponding all data points with label symbol in cluster
It is same to send symbol combination.
4. the novel full-duplex communication transmitter of one kind according to claim 3 and receiver design method, feature exist
In the design method of the label symbol are as follows:
Minimize reconstructed error function:And solve following optimization problem:
S.t.S=AB,
Rank (A)=r.
S is set according to modulation system, and A is found by offline poor searching method makes the smallest optimal label of reconstructed error function
Symbol combination.
5. according to the novel full-duplex communication transmitter of one kind described in claim 2~4 any one and receiver design side
Method, which is characterized in that it further include data preprocessing method, specifically:
In full duplex system, due to oneself known transmitting signal, by the N number of signaling point received according to oneself identical hair
The signaling point for penetrating signal is grouped, and is then clustered in each group；Definition receives data signal point
y(d)=[y (r+1), y (r+2), y (r+3) ..., y (N)]
Received signal points collection is defined to be combined into
ThenWherein subsetIs defined as:
Above-mentioned groupingIn oneself transmission signal be all Xl, then by the number of the potential cluster of signaling point from M1M2It is reduced to
M2。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811540407.8A CN109698741B (en) | 2018-12-17 | 2018-12-17 | A kind of novel full-duplex communication transmitter and receiver design method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811540407.8A CN109698741B (en) | 2018-12-17 | 2018-12-17 | A kind of novel full-duplex communication transmitter and receiver design method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109698741A CN109698741A (en) | 2019-04-30 |
| CN109698741B true CN109698741B (en) | 2019-11-01 |
Family
ID=66231811
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811540407.8A Active CN109698741B (en) | 2018-12-17 | 2018-12-17 | A kind of novel full-duplex communication transmitter and receiver design method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109698741B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111585644B (en) * | 2020-05-27 | 2021-07-20 | 北京邮电大学 | Radar communication integrated system, signal processing method and apparatus, and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103974433A (en) * | 2014-05-15 | 2014-08-06 | 西安电子科技大学 | Resource distribution method oriented towards service quality guarantee and suitable for wireless full-duplex network |
| US9515809B2 (en) * | 2012-12-06 | 2016-12-06 | Intel Corporation | Testing using analog near end loop back circuit |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9036749B2 (en) * | 2013-08-09 | 2015-05-19 | Kumu Networks, Inc. | Systems and methods for frequency independent analog self-interference cancellation |
| CN104539341B (en) * | 2014-12-26 | 2018-02-13 | 西安电子科技大学 | The active echo self-interference suppressing method of wireless broadband multi-antenna MIMO full duplex systems |
| CN106059705B (en) * | 2016-05-17 | 2018-10-26 | 北京邮电大学 | A kind of relaying safe transmission method of physical layer |
| CN107124245B (en) * | 2017-03-05 | 2019-03-05 | 北京工业大学 | Full duplex one-way junction self-interference suppressing method based on combined transceiving beam forming |
| CN107395542B (en) * | 2017-04-28 | 2020-05-22 | 电子科技大学 | Signal transmitting and receiving method of bidirectional relay communication system |
| CN107070603B (en) * | 2017-04-28 | 2019-11-01 | 电子科技大学 | Space-time block code system signal method of sending and receiving |
| CN107862278A (en) * | 2017-11-02 | 2018-03-30 | 黄冈师范学院 | A kind of high-resolution remote sensing image road information extraction system |
-
2018
- 2018-12-17 CN CN201811540407.8A patent/CN109698741B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9515809B2 (en) * | 2012-12-06 | 2016-12-06 | Intel Corporation | Testing using analog near end loop back circuit |
| CN103974433A (en) * | 2014-05-15 | 2014-08-06 | 西安电子科技大学 | Resource distribution method oriented towards service quality guarantee and suitable for wireless full-duplex network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109698741A (en) | 2019-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Park et al. | Learning to demodulate from few pilots via offline and online meta-learning | |
| CN110071750B (en) | Two-dimensional space code index modulation method based on machine learning | |
| US8537922B2 (en) | Methods and systems for providing feedback for beamforming and power control | |
| CN109687897B (en) | Superposition CSI feedback method based on deep learning large-scale MIMO system | |
| CN101547182B (en) | Method for mapping and demapping QAM constellation diagrams | |
| CN110113288A (en) | A kind of design and demodulation method of the ofdm demodulator based on machine learning | |
| CN108540185B (en) | Differential space modulation method combined with space-time block code | |
| CN106453162A (en) | Channel estimation method for multiple-input-multiple-output orthogonal frequency division multiplexing system | |
| CN107332599A (en) | A kind of up non-orthogonal multiple cut-in method for combining domain based on power and code word | |
| CN109167649A (en) | A kind of GSM-MBM system low complex degree detection method | |
| Ji et al. | Modulation recognition in maritime multipath channels: A blind equalization-aided deep learning approach | |
| CN109547376A (en) | A kind of Modulation Identification method of the gauss hybrid models based on modulating characteristic | |
| CN116257750A (en) | Radio frequency fingerprint identification method based on sample enhancement and deep learning | |
| CN109698741B (en) | A kind of novel full-duplex communication transmitter and receiver design method | |
| Ke et al. | Application of adversarial examples in communication modulation classification | |
| Zhang et al. | Lightweight automatic modulation classification via progressive differentiable architecture search | |
| US11258649B2 (en) | Transmitter and receiver communication apparatus for non-coherent communication | |
| CN113114603B (en) | Information recovery method and device for MIMO-OFDM system | |
| Qiao et al. | Unsourced massive access-based digital over-the-air computation for efficient federated edge learning | |
| CN107070603B (en) | Space-time block code system signal method of sending and receiving | |
| CN106375065A (en) | Iterative MMSE soft detection method based on spatial modulation system having frequency offset | |
| CN107017929B (en) | MIMO system signal transmitting and receiving method | |
| CN114759997B (en) | MIMO system signal detection method based on data model double driving | |
| US20220405589A1 (en) | Methods for reliable over-the-air computation with pulses for distributed learning and with federated edge learning without channel state information | |
| CN107707493A (en) | A kind of channel estimation methods based on compressed sensing |
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 |