Cache-assisted high-performance coding transmission method

Abstract

The invention relates to a coding transmission method in an MIMO (Multiple Input Multiple Output) communication network. The invention discloses a cache-assisted high-performance coding transmission method, which comprises the following steps of: based on a cache-assisted MIMO (Multiple Input Multiple Output) interference network system and a cache-assisted MIMO interference network model, introducing multiple antennas and multiple carriers, and reconstructing a network model with a multi-antenna transceiver and a file library; for the KT * KR cache assisted MIMO interference network, analyzing the degree of freedom constraint and the system total gain decomposition of the network communication model; on this basis, file segmentation and cache processing are performed in a cache placement stage, so that the network load is reduced; in the content transmission stage, on the basis of a diagonal block matrix, a high-degree-of-freedom coding transmission method is designed by combining coding caching and interference alignment around the achievable system transmission degree of freedom, a multicast signal processing and transceiving mechanism is established, and interference network expansion and improvement of system transmission efficiency and degree of freedom are achieved.

Description

Technical Field

[0001] The present invention relates to the field of communication, and in particular to a high-freedom coding transmission method in a cache-assisted MIMO communication network. Background Art

[0002] Traditional wireless caching schemes focus on considering the two phases separately and carefully designing the transmission strategy. This independent phase design makes the coding strategy simpler and reduces the coding redundancy of the caching scheme, but it still has certain limitations. It was not until 2014 that Maddah-Ali and Niesen first combined the cache placement phase and the content transmission phase for coding design and proposed the coding cache technology for the first time. From the perspective of information theory, the coding cache technology reveals that the receiver's cache can not only obtain the local cache gain of the traditional non-coding cache scheme, but also when the user requests different files, the server can obtain the global cache gain through file encoding and multicast transmission, so that the efficiency of information transmission can be improved significantly. The coding cache technology has broad application prospects in both academia and industry. At present, the promotion and application of coding cache technology to various network models has become a hot topic in the field of wireless communication research.

[0003] At present, with the gradual maturity of 5G technology and its commercial use and popularization around the world, it has promoted social changes to a certain extent and put forward new requirements for the field of wireless communications, such as application scenarios of new technologies, large-scale signal coverage, and integration of "air, land, and sea". The academic community has carried out research on 6G around these fields. The 6G communication system aims to achieve global interconnection of all things. The research on the physical layer is based on the latest proposed de-cellular architecture. In terms of network model, scholars have adopted the large-scale multi-input multi-output technology that is most advantageous in 5G. The coding cache technology greatly reduces the network load and guarantees a certain network transmission gain. Therefore, how to promote the application of coding cache technology in multi-input multi-output networks has become a future development trend of mobile communications. Many scholars have studied the extension of coding cache technology to wireless cellular communication networks. However, facing the actual complex multi-user interference network, coding cache still needs to solve many theoretical and technical problems.

[0004] Among them, the main issues include the following two aspects:

[0005] (1) The problem of generalized multi-user MIMO interference network expansion. With the development of mobile communication systems, the network topology has become more heterogeneous and complex. Base stations and user terminals with caching functions will be equipped with more antennas, that is, massive MIMO technology. Therefore, how to expand the transmission strategy of coding caching to generalized MIMO multi-user interference networks is one of the main problems.

[0006] (2) Cache-assisted coding transmission problem to maximize network performance. At present, existing research has characterized the network performance limit achieved by using cache from the perspective of information theory, but cache-assisted coding transmission strategies that achieve or approach this network performance limit need further research. At the same time, how to eliminate the interference in the parallel coding transmission process is also a problem that needs to be considered. Summary of the invention

[0007] In view of the deficiencies in the prior art, the present invention proposes a cache-assisted high-performance encoding transmission method.

[0008] Based on the cache-assisted MIMO interference network, on the basis of the traditional cache-assisted wireless interference network model, by introducing multiple antennas and multiple carriers, the network communication model consisting of multiple receivers, multiple senders and a file library with multiple files is rebuilt. In this scenario, combined with the coding cache and interference alignment technology, in the cache placement stage, file segmentation and cache setting are used to pre-process data and reduce network load; in the content transmission stage, efficient multicast transmission and interference elimination are achieved, so as to improve the system transmission efficiency and freedom, eliminate interference in the transmission process, and optimize and improve the system transmission performance.

[0009] The technical solution adopted by the present invention is:

[0010] A cache-assisted high-performance coding transmission method builds a communication model of multi-antenna transceivers and file libraries for the cache-assisted MIMO interference network model. On this basis, a coding cache optimization method is designed, and interference alignment technology is used to eliminate inter-user interference to achieve system performance improvement; the coding transmission method process includes the following steps:

[0011] S1, for the cache-assisted MIMO interference network model, a system model with multi-antenna transceivers and a file library is constructed;

[0012] S2, for K T ×K R Cache-assisted MIMO interference network, analyze the degree of freedom constraints and system total gain decomposition of the network model. On this basis, the file segmentation and cache processing in the cache placement stage, as well as the multicast signal processing and transceiver mechanism in the content transmission stage are explained. Finally, based on the diagonal block matrix, a high-degree-of-freedom coding transmission method is designed by combining coding cache and interference alignment technology;

[0013] S3, based on the overall framework of step S2, introduces a 2×K R The algorithm for solving the cache-assisted MIMO interference network example is designed. Focusing on the interference problem in content transmission, the interference alignment principle and channel reciprocity are used to design the precoding matrix and filter matrix to achieve interference elimination, further optimizing the overall performance of the system;

[0014] Beneficial effects of the invention:

[0015] 1. In order to solve the problem that the traditional cache-assisted single-antenna interference network method is not suitable for the future mobile communication network environment, the present invention expands the application of coding cache technology from single-antenna interference network to MIMO interference network. In the new system model, multi-carrier technology is added, and the channel matrix under the model is defined as a new diagonal block channel matrix, and the communication model of the system is reconstructed. Research is carried out around the defined diagonal block channel matrix, and based on the characteristics of the matrix, a model of coding cache and interference alignment related problems is established, realizing the optimal configuration of system resources.

[0016] 2. The present invention focuses on the feasibility of high degrees of freedom and designs an interference-free parallel coding transmission method for diagonal block channel matrices for cache-assisted MIMO interference networks. This method combines coding cache technology and interference alignment technology, reduces the transmission load with the help of fine file segmentation and coding strategies, and jointly designs precoding matrices and filter matrices at the transceiver end to eliminate interference, thereby achieving a higher degree of freedom in system transmission.

[0017] 3. The efficient coding transmission method in the cache-assisted MIMO interference network of the present invention can obtain a higher degree of freedom of system transmission. First, the degree of freedom constraints and the decomposition of the total system gain under the network model are explained. Then, based on the achievable system transmission degrees of freedom, a coding transmission method for high-freedom interference-free parallel transmission through a MIMO interference network is designed by combining coding cache and interference alignment. In the coding cache design, the present invention adopts a sophisticated secondary file segmentation and coding transmission strategy to obtain higher local and global gains of the coding cache while reducing the network load.

[0018] 4. The present invention aims at the interference problem between transmissions, considers the duality of the precoding matrix and the filter matrix under the symmetric channel, and designs a detailed algorithm to solve it. Finally, the quantitative analysis method is used to verify the performance improvement of the method proposed in the present invention on the cache-assisted MIMO interference network. By changing the system parameter settings, the impact on the interference alignment gain, the local gain of the coding cache and the global gain is observed, the availability of each gain is analyzed, and the degree of freedom of the method proposed in the present invention is further verified by comparing the degree of freedom with the traditional cache-assisted single-antenna interference network. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 K is shown T ×K R Example diagram of cache-assisted MIMO interference network model deduction;

[0020] Figure 2 for Cache-assisted interference network model diagram;

[0021] Figure 3 The sender cache size M T Impact analysis diagram for three types of gains;

[0022] Figure 4 The receiver buffer size M R Impact analysis diagram for three types of gains;

[0023] Figure 5 is the number of receivers K R Impact analysis diagram for three types of gains;

[0024] Figure 6 This is a comparison diagram of the degrees of freedom of the two systems;

[0025] Figure 7 This is a comparison chart of degrees of freedom under SA extended channel;

[0026] Figure 8 The following is a comparative analysis diagram of the gains of the two systems. DETAILED DESCRIPTION

[0027] In order to make the technical concept and advantages of the present invention more clearly understood, the technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the following embodiments are only used to explain and illustrate the preferred implementation methods of the present invention, and should not be regarded as and do not constitute a limitation on the scope of patent protection claimed by the present invention.

[0028] Example 1

[0029] The present invention proposes a cache-assisted high-performance coding transmission method, based on a cache-assisted MIMO interference network system, to achieve interference network expansion and improve system transmission efficiency and degree of freedom through the following steps:

[0030] Step S1, based on the cache-assisted MIMO interference network model, multiple antennas and multiple carriers are introduced to reconstruct the network model with a multi-antenna transceiver and a file library;

[0031] Step S2, for K T ×K R Cache-assisted MIMO interference network, analyze the degree of freedom constraints and system total gain decomposition of the network communication model; and on this basis, perform file segmentation and cache processing in the cache placement stage to reduce network load;

[0032] Step S3, in the content transmission stage, based on the diagonal block matrix, a high-freedom coding transmission method is designed around the achievable system transmission freedom, combined with coding cache and interference alignment, to establish a multicast signal processing and transceiver mechanism, which includes:

[0033] First, based on the network communication model framework in step S2, the information is encoded and transmitted based on the encoding cache technology according to the information cache content of the transmitting and receiving ends:

[0034] Adopting sophisticated secondary file segmentation and encoding transmission strategy, we can obtain higher local and global gains of encoding cache while reducing network load.

[0035] Second, to address the interference problem in content transmission, interference alignment technology is used to eliminate the interference in content transmission:

[0036] Introducing 2×K R The cache-assisted MIMO interference network solution algorithm uses the interference alignment principle and channel reciprocity to design the precoding matrix and filter matrix to achieve interference elimination and further optimize the overall performance of the system.

[0037] Example 2

[0038] A cache-assisted high-performance coding transmission method builds a communication model of multi-antenna transceivers and file libraries for the cache-assisted MIMO interference network model. On this basis, a coding cache optimization method is designed, and interference alignment technology is used to eliminate inter-user interference to achieve system performance improvement; the coding transmission method process includes the following steps:

[0039] S1, for the cache-assisted MIMO interference network model, a system model with multi-antenna transceivers and a file library is constructed;

[0040] S2, for K T ×K R Cache-assisted MIMO interference network, analyze the degree of freedom constraints and system total gain decomposition of the network model. On this basis, the file segmentation and cache processing in the cache placement stage, as well as the multicast signal processing and transceiver mechanism in the content transmission stage are explained. Finally, based on the diagonal block matrix, a high-degree-of-freedom coding transmission method is designed by combining coding cache and interference alignment technology;

[0041] S3, based on the overall framework of step S2, introduces a 2×K R The cache-assisted MIMO interference network example is designed to solve the problem. Focusing on the interference problem in content transmission, the interference alignment principle and channel reciprocity are used to design the precoding matrix and filter matrix to achieve interference elimination and further optimize the overall performance of the system.

[0042] Due to the reciprocity of uplink and downlink communication systems, this article only considers the downlink transmission situation, but the present invention is not limited to the downlink transmission network.

[0043] Example 3

[0044] The efficient coding transmission method in the cache-assisted MIMO interference network of this embodiment is different from that of Embodiment 1 and Embodiment 2 in that: further, the specific scheme of each step is disclosed as follows:

[0045] In step S1, the cache-assisted MIMO interference network model building process is as follows:

[0046] like Figure 1 As shown, the present invention adds multiple antennas and S orthogonal subcarriers to the traditional cache-assisted wireless interference network to establish a T senders and K R The network model has a file library containing N files, each of which can be denoted by W. n ,n∈{1,2,…,N}, the size is F packets. The size of the file that each sender and receiver can cache is M T and M R , the number of antennas configured are A T and A R .

[0047] The above interference network model can be expressed as:

[0048]

[0049] Among them, x i (t) represents the signal sent by the i-th sender at the t-th time, y j (t) represents the received signal of the jth receiver at the tth time, n j (t) represents the Gaussian white noise at the receiver j. ji (t) represents the SA from the i-th sender to the j-th receiver at the t-th time R ×SA T dimensional multi-carrier MIMO channel matrix.

[0050] The channel matrix under this interference network model is no longer a diagonal matrix, and it has the following characteristics:

[0051]

[0052] Among them, each matrix sub-block It is an S-order diagonal matrix, representing the subcarrier extension channel from the m-th transmitting antenna of sender i to the n-th receiving antenna of receiver j.

[0053] In the traditional cache-assisted interference network, according to Shannon's theorem, without loss of generality, C ρ represents the total capacity of the system, and the degrees of freedom are expressed as:

[0054]

[0055] When the difference between signal and noise power is large, log 2 (ρ) is a first-order function of the independent variable that approximates the system capacity. In this case, the degree of freedom can be used as the slope of the first-order function. Equation (3) can be rewritten as:

[0056] C ρ =DoF·log 2 ρ+o(log 2 ρ), (4)

[0057] Where ρ is the signal-to-noise ratio, and olog(ρ) represents the higher-order infinitesimal of log(ρ).

[0058] The present invention designs a transmission method from two stages: cache placement and content transmission. In the content transmission stage, for any receiver, it needs to request the uncached data packet of the required file from the sender and send it through K T ×K R MIMO interference network is used for transmission. Considering the channel matrix as a diagonal block matrix, the degrees of freedom are reset and the specific definitions are as follows.

[0059] Under a given signal-to-noise ratio condition, let L represent the T ×K R The total link transmission load after optimization based on coding cache design in MIMO interference network, δ represents the sum of degrees of freedom after the symmetric multicast channel eliminates interference by interference alignment in the content transmission stage, then the network transmission degrees of freedom of the joint cache placement and content transmission strategy can be expressed as,

[0060] d=δ / L (5)

[0061] The network transmission freedom defined in the present invention can also be understood as the number of information flows that can be transmitted without interference by the system.

[0062] The efficient coding transmission method in the cache-assisted MIMO interference network, step S2, the coding transmission method of the joint coding cache and interference alignment design is as follows:

[0063] First, according to the network model constructed in step S1, the concept of degree of freedom is introduced to measure the interference-free parallel transmission capability in the wireless network. T ×K R Cache-assisted MIMO interference network, where the file library contains N files W n , each file has F packets. The size of the file that each sender and receiver can cache is M T and M R , the number of antennas configured are A T and AR , using S orthogonal subcarriers for information transmission. Let d be K T ×K R The network freedom of the cache-assisted MIMO interference network satisfies,

[0064]

[0065] Where κ = K R M R / N is the normalized cumulative buffer function of the receiving end, and d in the present invention can also be understood as K T ×K R The number of independent data streams that can be transmitted without interference in a cache-assisted MIMO interference network. In order to more clearly demonstrate the improvement of the network transmission freedom brought about by the transmission strategy proposed in the present invention, the degree of freedom d is multiplied by the number of receivers K. R represents the approximate total system gain and splits it into three parts,

[0066]

[0067] Among them, g IA is the interference alignment gain, which means that the interference is eliminated by using interference alignment on the extended channel, thereby achieving the transmission gain of the system; g LC is the local cache gain, reflecting that each user has cached some sub-files of the requested file in its local cache. Therefore, g LC Can be considered as M R / N, that is, the proportion of each file stored in the local cache of receiver j; g GC The additional global cache gain for the code cache strategy comes from the encoding opportunities created by the code cache for storing different files for different receivers, and the multicast links created to provide useful encoding information to multiple receivers at the same time. This gain depends on the total amount of memory on the receiving end. Therefore, g GC Affected by the cumulative cache function κ = K R M R / N influence.

[0068] Based on the degree of freedom constraints in the above network model, the present invention designs a sophisticated coding transmission method based on diagonal block matrix, joint coding cache and interference alignment. The specific strategy is as follows.

[0069] For the cache placement phase, select the sender set and a subset of receivers Among them, the receiver subset The number of This selection does not mean that every sender can cache all files in the repository. In fact, any file W nOnly a portion can be cached at each sender, which makes it easier to design a caching strategy on the sender. The cache placement phase strategy is as follows.

[0070] First, each file in the file library W n All split into K T sub-files of equal size, each containing F / K T packets, represented by These subfiles are cached on the sender Secondly, the sub-file W n,i Continue to split into sub-files of equal size. After the second split, the size of each sub-file is Expressed as These subfiles are subset by the recipient cache. Assuming F is large enough, F / K T and is an integer.

[0071] For any sender i, it caches N subfiles W n,i , each sub-file size is F / K T packets, then the cached packets at sender i satisfy its cache constraints.

[0072] NF / K T =NF / K T ≤M T F.(8)

[0073] Similarly, consider any receiver The cached data packets also meet the receiver cache constraints.

[0074]

[0075] In the content transmission phase, considering the worst case, all receivers request different files, which requires that the number of files in the file library is greater than the number of receivers, that is, N ≥ K R In this case, the total transmission load of the network reaches the maximum value L≤(K R -κ) / (κ+1), according to the network transmission freedom formula defined in the present invention, d=δ / L, at this time, the network transmission freedom is the minimum.

[0076] Receiver Request File The request vector for all recipients can be f j ∈{1,2,…,N} represents the subfile requested by receiver j. There is no cache at receiver j, so it needs to be collected by the sender Provided by transmission via a communication network.

[0077] like Figure 2 As shown, according to the encoding cache method, the receiver subset is constructed The network model at this time can be regarded as a Cache-assisted interference network model. The sender sends the sub-file that the receiver does not cache. XOR processing as multicast signal Afterwards, through The cache-assisted MIMO interference network sends to a subset of receivers Receiver Subset The number of The information multicast size of this network model is σ=κ+1.

[0078] Consider any sender definition It means that at time t, sender i sends a message to a subset of receivers. The multicast signal sent; Represents a multicast signal The purpose of the designed precoding matrix is ​​to compress the interference subspace. Then the signal sent by sender i can be expressed as:

[0079]

[0080] According to the communication model formula (1), for any receiver The received signal is:

[0081]

[0082] Separate the desired signal from the interference signal in equation (11),

[0083]

[0084] in, is the expected signal sent by sender i to receiver j at the tth moment, It is represented as the interference received by receiver j, n j is Gaussian white noise. In order to eliminate interference and effectively restore the expected information Design the filter matrix at receiver j

[0085]

[0086] Designing precoding matrix using interference alignment technique and filter matrix Should satisfy,

[0087]

[0088] Where d′ is the number of interference-free information transmission data streams from sender i to receiver j.

[0089] In the method for efficient coding transmission in a cache-assisted MIMO interference network, in step S3, the algorithm for solving the precoding matrix and the filter matrix is ​​as follows:

[0090] Under the conditional constraints of formula (14), in order to successfully solve the precoding matrix and the filter matrix, the present invention introduces a 2×K R A cache-assisted MIMO interference network example design solution algorithm.

[0091] The interference network is set up as follows. The file library contains N files Each file has F packets. The size of the file that each sender and receiver can cache is M T and M R , the number of antennas configured is A, and S orthogonal subcarriers are used. In the cache placement phase, the sender caches the split subfiles W n,i , the size of each sub-file is F / 2; the receiving end buffer size is Secondary split sub-file In this case, the data packets cached by both the sender and the receiver satisfy the local cache constraint. In the content transmission phase, suppose that the receiver j requests file W j , sender i sends a multicast signal to a subset of receivers Interference alignment technology is used to eliminate interference.

[0092] The present invention takes into account 2×K R Symmetric channel model of cache-assisted MIMO interference network, namely K R ×2 cache-assisted MIMO interference network model. In this interference network, the content delivery strategy is that the sender subset To the receiver Send a multicast signal, the sender subset size is σ. For any sender subset All senders Sending the same signal To the receiver Design the precoding matrix at each transmitter j To compress the interference subspace.

[0093] For receiver 1, the received signal can be described as,

[0094]

[0095] in, 2×K RCache-Assisted MIMO Interference Network Channel Matrix The dual matrix of represents the channel matrix from sender j to receiver 1. Subset of senders The expected signal sent to receiver 1, Denotes the interference at receiver 1. Define the interference signal matrix P 1 ,

[0096]

[0097] in, m∈{1,2,…,M}, is the precoding matrix, is the set of all sender subsets, the matrix P 1 have At receiver 1, the interference from other senders is perfectly aligned with the interference from sender 1. Therefore, the precoding matrix is ​​selected In order to and Aligned to the same subspace,

[0098]

[0099] Select from independent continuous distributions Then the precoding matrix It can be expressed as,

[0100]

[0101] in, At this time, by designing the precoding matrix at receiver 1, the interference is aligned to the same subspace. The interference subspace dimension is given by Compressed to The expected signal space dimension is At this point, the interference alignment strategy has achieved the separation of the desired signal subspace and the interference signal subspace. The dimension of a global vector space can be described as

[0102]

[0103] Similarly, at receiver 2, the interference from sender 1 is Interference with other senders Aligned to the same subspace,

[0104]

[0105] Repeat the above interference alignment strategy and select Then at receiver 2, the precoding matrix It can be expressed as

[0106]

[0107] in,

[0108] Through the interference alignment strategy described above, dimensional subspace. R ×2 cache-assisted MIMO interference network precoding matrix and filter matrix as the original 2×K R The filter matrix and precoding matrix of the cache-assisted MIMO interference network are:

[0109] For K shown in formula (14) T ×K R The cache-assisted MIMO interference network precoding and filter matrix condition constraints are designed through interference subspace compression similar to formula (17), and the receivers j, j∈{1,2,...,K R By selecting an independent and continuously distributed precoding matrix at}, the precoding matrix and the filter matrix can be solved by iterative calculation. At the same time, it is easy to know that under the interference alignment strategy designed by the present invention, the expected signal vector and the interference signal vector are linearly independent.

[0110] Figure 3 The present invention has been verified to be effective for M T Analysis of the system parameters.

[0111] The present invention designs a sophisticated file segmentation strategy and cache constraints in the problem setting. The constraint condition of the local cache at the sender is K T M T ≥ N, that is, the sender can cache all files in the file library. Therefore, the system's transmission freedom is related to the sender's cache size M T This feature provides flexibility in system design and enables more efficient use of cache resources for transmission optimization.

[0112] like Figure 4 As shown, when the system parameter N is fixed, R =A T =4, K T =A R =2, the receiver's local cache size M R Set it as a variable and observe its effect on the interference alignment gain g IA , encoding cache local gain g LC and the encoding cache global gain g GC The horizontal axis in the figure is M R / N, represents the number of sub-files of each file in the file library that the receiver can cache. The present invention uses M R / N is called the cache function, and the ordinate is the total system gain K R d is the logarithm of e.

[0113] When M R =0 and M R =N=4, the system is considered distorted. R = 0, the receiving end has no local cache memory, that is, in the cache placement phase, the receiving end does not cache any content. In the content placement phase, the sending end will send the union of the requested files to the receiving end. This content transmission process has no encoding processing and can be regarded as unicast transmission. R =N=4, in the cache placement phase, each receiver can cache all files in the file library, so there is no need to send a request vector to the sender in the content placement phase, that is, there is no encoding processing and information transmission in the content placement phase. R =0 and M R =N=4, the system does not have a coding cache gain g LC and g GC .

[0114] With M R The growth of local cache gain g LC Then it increases slowly. When M R When / N≥0.5, each receiver can cache many packets, g LC Therefore, it is relevant. Compared with the local cache gain g LC , global cache gain g GC It grows faster when the receiver's cumulative buffer size is larger, i.e., K R M R / N≥1, and it also has correlation. However, the interference alignment gain g IA is constant, because the present invention performs interference alignment design around the diagonal block matrix, i.e. g IA It is closely related to the number of antennas and orthogonal subcarriers at the transmitting and receiving ends.

[0115] like Figure 5 As shown, assuming the number of receivers K R is a variable, and the fixed receiver buffer size is M R =2, the total system gain K in the figure R d is decomposed. It can be observed that g LC Size is not affected by K R impact.

[0116] Consider K R When the size of K approaches infinity, R →∞,gIA Converges to K T SA, while g GC It can be approximated as

[0117]

[0118] It is worth noting that g GC Unlike the other two gains, g GC Will not follow K R The increase of converges to a limit and is linearly related to the number of receivers. Therefore, it can be considered that for a system with a large number of receivers, g GC Dominates in transmission gain.

[0119] From the above analysis, we can conclude that as the number of receivers K increases R Or receiver cache memory M R As the number of senders K increases, the total degree of freedom of the system will become arbitrarily large. T The change does not hold true when K T →∞,g IA Approximately K R SA, while g GC Converges to 1, and the total system gain is approximately

[0120]

[0121] Based on this result analysis, for the cache-assisted MIMO interference network, when the number of system senders reaches a certain level, channel expansion and interference alignment strategies can effectively establish K from each sender to the receiver. R SA orthogonal links, each link has a transmission degree of freedom close to 1. At the same time, since these orthogonal links do not have multicast signals, the global cache gain g GC disappears, leaving only the local cache gain g LC .

[0122] like Figure 6 As shown, the system degrees of freedom achieved by the traditional cache-assisted single-antenna interference network and the cache-assisted MIMO interference network are compared, and analysis is performed around the comparison results.

[0123] The present invention sets the inverse of the degree of freedom 1 / d as the vertical coordinate, which is consistent with the previous research. Compared with the cache-assisted single-antenna interference network, the transmission degree of freedom of the cache-assisted MIMO interference network is significantly improved, which is attributed to the expansion of the channel by antennas and orthogonal subcarriers.

[0124] exist Figure 4 The present invention has been described in M R =0 and M R=4 distortion. When M R →0, that is, the local cache memory of the receiving end is very small, the huge difference in the degrees of freedom of the two methods further verifies that Figure 3 Part of the present invention relates to g LC and g GC Follow M R Analysis of the changes. At this time, the interference alignment gain g IA It occupies a dominant position in the system gain, and the difference in degrees of freedom reflects the improvement of the network model by introducing multi-antenna and multi-carrier factors in the present invention. R →4, both systems have a higher degree of freedom. This result complements the previous simulation analysis: due to the interference alignment gain g IA is constant, and its size depends on the system settings. R The growth of the number of files N, g LC and g GC The proportion of the total transmission gain of the system gradually increases, that is, for a file number N and a user local cache M R For larger systems, the interference alignment gain g IA This result verifies the potential of combining coding buffer and interference alignment technology in MIMO interference networks, providing strong support for improving system performance.

[0125] Likewise, the increase in d can be explained: as M R With the increase of , the receiver can cache more sub-files in the cache placement phase. In the content transmission phase, the receiver only needs to request a smaller number of data packets for a specific file, or directly obtain the complete file from the local cache. In this network transmission scenario, the total network load L is relatively low, thereby obtaining a higher total network transmission freedom. This improvement in transmission freedom can also be understood by the definition of freedom d = δ / L.

[0126] like Figure 7 As shown, based on the above degree of freedom analysis, let M R =2, and study the impact of extended channels on the system's freedom. The horizontal axis SA in the figure represents the number of orthogonal subcarriers S and the minimum value of the transmitting and receiving antennas A = min{A T ,A R}, indicating the expansion of the system channel by multiple antennas and multiple carriers.

[0127] from Figure 7It can be clearly observed that when the extended channel SA = 1, the cache-assisted MIMO interference network is regarded as a single-antenna interference network, so the two systems have the same degree of freedom in transmission. With the increase in the number of antennas and subcarriers, the coding transmission method based on the cache-assisted MIMO interference network proposed in the present invention has achieved considerable improvement in degrees of freedom. However, the improvement of the degrees of freedom by the extended channel is not unlimited. With the increase of the extended channel, the problem of interference between transmissions will become more serious, the joint coding cache and interference alignment design will be more difficult, and the system calculation complexity will also increase accordingly.

[0128] like Figure 8 As shown in the figure, by comparing all the gains of the two systems, it can be intuitively observed that the extended channel significantly improves the system performance. R The increase of g in both systems LC and g GC The increase trend of g is the same, and the increase range is also equal. This is because both systems use similar coding cache technology and implement similar coding cache strategies in design. IA , the cache-assisted MIMO interference network shows a significant improvement. This improvement is mainly due to the clever design of the high-degree-of-freedom coding transmission strategy and the interference alignment strategy. The application of coding cache technology in the system aims to improve transmission efficiency and create multicast opportunities, thus laying the foundation for the overall improvement of network performance.

[0129] The present invention aims at the interference problem between transmissions in a cache-assisted MIMO interference network, considers the duality of the precoding matrix and the filter matrix under a symmetric channel, and designs a coding transmission method for high-freedom interference-free parallel transmission through a MIMO interference network by combining coding cache and interference alignment based on the achievable system transmission degrees of freedom. By comparing the degrees of freedom with the traditional cache-assisted single-antenna interference network and verifying it by quantitative analysis, the present invention can obtain a higher system transmission degree of freedom and achieve performance improvement on the cache-assisted MIMO interference network.

[0130] The above description is only a preferred embodiment of the present invention and does not constitute a limitation of the present invention. Under the guidance of the prior art, those skilled in the art can make other modifications to the implementation of the present invention without creative work. Any modification made within the spirit and principle of the present invention or simple replacement or equivalent substitution using conventional technical means in the field should be included in the protection scope of the present invention.

Claims

1. A cache-assisted high-performance encoding transmission method, characterized in that: Based on the cache-assisted MIMO interference network system, the interference network expansion and the improvement of system transmission efficiency and degree of freedom are achieved through the following steps: Step S1, based on the cache-assisted MIMO interference network model, multiple antennas and multiple carriers are introduced to reconstruct the network model with a multi-antenna transceiver and a file library; Step S2, for K T ×K R Cache-assisted MIMO interference network, analyze the degree of freedom constraints and system total gain decomposition of the network communication model; and on this basis, perform file segmentation and cache processing in the cache placement stage to reduce network load; Step S3, in the content transmission stage, based on the diagonal block matrix, around the achievable system transmission freedom, a high-freedom coding transmission method is designed in conjunction with coding cache and interference alignment, and a multicast signal processing and transceiver mechanism is established.

2. The cache-assisted high-performance encoding transmission method according to claim 1, characterized in that: In step S3, the process of designing a high-degree-of-freedom coding transmission method by combining coding buffering and interference alignment includes: First, based on the network communication model framework in step S2, the information is encoded and transmitted based on the encoding cache technology according to the information cache content of the transmitting and receiving ends: Adopting sophisticated secondary file segmentation and encoding transmission strategy, we can obtain higher local and global gains of encoding cache while reducing network load. Second, to address the interference problem in content transmission, interference alignment technology is used to eliminate the interference in content transmission: Introducing 2×K R The cache-assisted MIMO interference network solution algorithm uses the interference alignment principle and channel reciprocity to design the precoding matrix and filter matrix to achieve interference elimination and further optimize the overall performance of the system.

3. The cache-assisted high-performance encoding transmission method according to claim 1 or 2, characterized in that: In step S1, the multi-antenna transceiver and file library cache assisted MIMO interference network model construction process is as follows: In the traditional cache-assisted wireless interference network, multiple antennas and S orthogonal subcarriers are added to establish a K T senders and K R The cache-assisted MIMO interference network model of the receiver is expressed as: Among them, x i (t) represents the signal sent by the i-th sender at the t-th time, y j (t) represents the received signal of the jth receiver at the tth time, n j (t) represents the Gaussian white noise at the receiver j; H ji (t) represents the SA from the i-th sender to the j-th receiver at the t-th time R ×SA T dimensional multi-carrier MIMO channel matrix; The network model has a file library containing N files, each of which can be denoted by W. n ,n∈{1,2,…,N}, the size is F packets; the size of the file that each sender and receiver can cache is M T and M R , the number of antennas configured are A T and A R ; The channel matrix under this interference network model is no longer a diagonal matrix, and it has the following characteristics: Each matrix sub-block They are all S-order diagonal matrices, representing the subcarrier extension channel from the m-th transmitting antenna of sender i to the n-th receiving antenna of receiver j; In the traditional cache-assisted interference network, according to Shannon's theorem, without loss of generality, C ρ represents the total capacity of the system, and the degrees of freedom are expressed as: When the difference between the signal and noise power is large, the system capacity is approximately represented by a first-order function with log2(ρ) as the independent variable. In this case, the degree of freedom can be used as the slope of the first-order function, and equation (3) can be rewritten as: C ρ =DoF·log2ρ+o(log2ρ), (4) Where ρ is the signal-to-noise ratio, olog(ρ) represents the higher-order infinitesimal of log(ρ); The transmission method is designed from two stages: cache placement and content transmission. In the content transmission stage, for any receiver, it needs to request the uncached data packet of the required file from the sender and send it through K T ×K R MIMO interference network for transmission; Consider the channel matrix as a diagonal block matrix, reset the degrees of freedom, and let L represent the K T ×K R The total link transmission load after optimization based on coding cache design in MIMO interference network, δ represents the sum of degrees of freedom after interference is eliminated by interference alignment in the content transmission stage of symmetric multicast channel, then the network transmission degrees of freedom of the joint cache placement and content transmission strategy can be expressed as: d=δ / L (5) Among them, the network transmission freedom can also be understood as the number of information flows that the system can transmit without interference.

4. The cache-assisted high-performance encoding transmission method according to claim 1 or 2, characterized in that: The specific implementation process of step S2 includes: Step S2.1, based on the network model constructed in step S1, the concept of degree of freedom is introduced to measure the interference-free parallel transmission capability in the wireless network. The number of antennas configured is A T and A R , using S orthogonal subcarriers for information transmission; Let d be K T ×K R The network freedom of the cache-assisted MIMO interference network satisfies, Where κ = K R M R / N is the normalized cumulative buffer function at the receiving end; Step S2.2: To increase the degree of freedom of network transmission, multiply the degree of freedom d by the number of receivers K R represents the approximate total system gain and splits it into three parts, Among them, g IA is the interference alignment gain, which means that the interference is eliminated by using interference alignment on the extended channel, thereby achieving the transmission gain of the system; g LC is the local cache gain, reflecting that each user has cached some sub-files of the requested file in its local cache. Therefore, g LC Can be considered as M R / N, that is, the proportion of each file stored in the local cache of receiver j; g GC The additional global cache gain for the code cache strategy comes from the encoding opportunities created by the code cache for storing different files for different receivers, and the multicast links created to provide useful encoding information to multiple receivers at the same time. This gain depends on the total amount of memory on the receiving end. Therefore, g GC Affected by the cumulative cache function κ = K R M R / N influence.

5. The cache-assisted high-performance encoding transmission method according to claim 3, characterized in that: Step S3, based on the degree of freedom constraints in the network model, a sophisticated coding transmission strategy is designed by combining coding cache and interference alignment: For the cache placement phase, select the sender set and a subset of receivers Among them, the receiver subset The number of First, each file in the file library W n All split into K T sub-files of equal size, each containing F / K T data packets, represented by These subfiles are cached in the sender i, middle; Secondly, the subfile W n,i Continue to split into sub-files of equal size; after the second split, the size of each sub-file is Expressed as These subfiles are subset by the recipient cache; assuming F is large enough, F / K T and is an integer; For any sender i, it caches N subfiles W n,i , each sub-file size is F / K T packets, then the cached packets at sender i satisfy its cache constraints. NF / K T =NF / K T ≤M T F. (8) Similarly, consider any receiver j, The cached data packets also meet the receiver cache constraints. In the content transmission phase, considering the worst case, all receivers request different files, which requires that the number of files in the file library is greater than the number of receivers, that is, N ≥ K R ; In this case, the total transmission load of the network reaches the maximum value L≤(K R -κ) / (κ+1), according to the network transmission freedom formula d=δ / L, the network transmission freedom is the smallest at this time; Receiver j, Request File The request vector for all recipients can be f j ∈{1,2,…,N} represents the subfile requested by receiver j. There is no cache at receiver j, so it needs to be collected by the sender Provided by transmission via a communication network; Construct a subset of receivers based on the encoding cache method The network model at this time can be regarded as a Cache-assisted interference network model; the sender sends the receiver a subfile that is not cached XOR processing as multicast signal Afterwards, through The cache-assisted MIMO interference network sends to a subset of receivers Receiver Subset The number of The information multicast size of this network model is σ=κ+1; Consider any sender i, definition It means that at time t, sender i sends a message to the receiver subset The multicast signal sent; Represents a multicast signal The purpose of the designed precoding matrix is ​​to compress the interference subspace; then the signal sent by sender i can be expressed as, According to the communication model formula (1), for any receiver j, The received signal is: Separate the desired signal from the interference signal in equation (11), in, is the expected signal sent by sender i to receiver j at the tth moment, It is represented as the interference received by receiver j, n j is Gaussian white noise; In order to eliminate interference and effectively restore the desired information Design the filter matrix at receiver j Designing precoding matrix using interference alignment technique and filter matrix Should satisfy, Where d′ is the number of interference-free information transmission data streams from sender i to receiver j.

6. The cache-assisted high-performance encoding transmission method according to claim 5, characterized in that: In step S3, the algorithm for solving the precoding matrix and the filter matrix is ​​as follows: Under the conditional constraints of formula (14), a 2×K R Cache-assisted MIMO interference network example design solution algorithm: The interference network is set up as follows. The file library contains N files Each file has F packets; each sender and receiver can cache files of M sizes respectively. T and M R , the number of antennas configured is A, and S orthogonal subcarriers are used; In the cache placement phase, the sender caches the split sub-files W once. n,i , the size of each sub-file is F / 2; the receiving end buffer size is Secondary split sub-file In the content transmission phase, suppose receiver j requests file W j , sender i sends a multicast signal to a subset of receivers Use interference alignment technology to eliminate interference: In K R ×2 cache-assisted MIMO interference network model, content delivery strategy is a sender subset To the receiver i, Send a multicast signal, the sender subset size is σ; for any sender subset Among all senders j, Sending the same signal To receiver i, Design the precoding matrix at each transmitter j To compress the interference subspace; For receiver 1, the received signal can be described as, in, 2×K R Cache-assisted MIMO interference network channel matrix H j1 The dual matrix of (t) represents the channel matrix from sender j to receiver 1. Subset of senders The expected signal sent to receiver 1, represents the interference at receiver 1; Define the interference signal matrix P1: in, is the precoding matrix, is the set of all sender subsets, and the matrix P1 has dimension; At receiver 1, the interference from other senders is perfectly aligned with the interference from sender 1, and the precoding matrix is ​​selected In order to and Aligned to the same subspace, Select from independent continuous distributions Then the precoding matrix It can be expressed as, in, At this time, by designing the precoding matrix at receiver 1, the interference is aligned to the same subspace; the interference subspace dimension is given by Compressed to The expected signal space dimension is At this point, the interference alignment strategy has achieved the separation of the desired signal subspace and the interference signal subspace. The dimension of a global vector space can be described as Similarly, at receiver 2, the interference from sender 1 is Interference with other senders Aligned to the same subspace, Repeat the above interference alignment strategy and select Then at receiver 2, the precoding matrix It can be expressed as in, Through the interference alignment strategy described above, The desired signal can be efficiently recovered on the subspace of dimensionality.

7. The cache-assisted high-performance encoding transmission method according to claim 6, characterized in that: Select a symmetric K R ×2 cache-assisted MIMO interference network precoding matrix and filter matrix as the original 2×K R The filter matrix and precoding matrix of the cache-assisted MIMO interference network are: For K shown in formula (14) T ×K R The cache-assisted MIMO interference network precoding and filter matrix condition constraints are designed through interference subspace compression similar to formula (17), and the receivers j, j∈{1,2,...,K R By selecting an independent and continuously distributed precoding matrix at}, the precoding matrix and the filter matrix can be solved by iterative calculation. At the same time, it is easy to know that under the interference alignment strategy designed by the present invention, the expected signal vector and the interference signal vector are linearly independent.

Images