A transmission method for device-to-device encoding cache problem based on random requester transmission

Abstract

The present invention discloses a transmission method for a device-to-device encoding cache problem based on random requester transmission. The implementation process of the method includes the establishment of a problem model, the selection of a cache scheme, the determination of a lead requester, the set splitting of requested files, and the encoding of requested sub-files. Based on real-world scenarios, we propose random requester transmission and establish a system model of the device-to-device encoding cache problem under this transmission. We then select a cache scheme based on the model characteristics and formulate a transmission scheme during the transmission phase based on the number of users, the number of files, the cache size, and the number of requesters. This method is applicable to the transmission phase of a device-to-device encoding cache problem based on random requester transmission, and the transmission rate of this scheme is, in most cases, lower than the transmission rate of a general device-to-device encoding cache problem with non-random requester transmission.

Description

Technical Field

[0001] The present invention belongs to the technical field of information theory, and in particular relates to a transmission method for device-to-device encoding cache problem based on random requester transmission. Background Art

[0002] Device-to-device caching is an effective method for reducing transmission rates between servers and users, enabling users to communicate directly with each other and thus alleviating network congestion on the server. The original device-to-device code caching problem significantly improved communication efficiency by applying multicast technology to device-to-device networks. It consists of two phases: placement and transmission. Most current research on device-to-device code caching assumes that all users will request content. However, in reality, not all users will request content during the transmission phase, and users who do not request content can still help transmit information to further enhance communication efficiency, making it meaningful to study transmission with random requesters. Summary of the Invention

[0003] The present invention aims to provide a transmission method for a device-to-device code cache problem based on random requester transmission, and to provide a transmission scheme for a device-to-device code cache problem based on random requester transmission with a requester number r, for any number of users K, any number of files N, and a cache size M within a certain range, M∈[1,N]. By establishing a system model for the device-to-device code cache problem based on random requester transmission, analyzing the model characteristics, and selecting an appropriate cache scheme, a transmission scheme for a general device-to-device code cache problem is obtained that is superior in transmission rate to that of non-random requester transmission.

[0004] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0005] A method for transmitting device-to-device code cache based on random requester transmission includes the following steps:

[0006] Step 1: First, define the random requester transmission, that is, for the number of users K, the user set At the beginning of the transmission phase, r random users request a single file with equal probability, r∈[1,K]. These users are called requesters and are represented by the set To express, The remaining users do not request files; then the device-to-device encoding cache problem is systematically modeled for any number of users K, any number of files N, and a cache size M within a certain range, M∈[1,N];

[0007] For file collections The nth file uses W nTo represent, n∈[1,N]. The system consists of a placement phase and a transmission phase; in the placement phase, all N files are connected The central server fills the caches of all K users; each file is F bits long; the cache of user k is In the transfer phase, r users act as requesters to request a single file The request vector is All K users transmit the encoded signal Make all r requesters able to decode the requested file; the goal is to find the cache and transmission solution with the minimum feasible rate R in the worst case;

[0008] Step 2: According to the system model, in the placement phase, the server does not know the characteristics of the number of future requesters r, and selects the non-coding symmetric placement scheme as the placement scheme;

[0009] Step 3: Design a transmission plan based on the placement scheme. First, select a lead requester from all requesters to satisfy subsequent transmissions. Then, classify each requested subfile into three categories. Finally, users who cache the requested subfiles encode them in different forms based on their type and send them to all requesters via a broadcast channel, allowing all requesters to correctly decode the requested files. The sum of the transmission rates for the three subfile categories is the feasible transmission rate for the entire transmission plan.

[0010] Furthermore, the non-coding symmetrical placement scheme specifically selected in step 2 is as follows:

[0011] Define H(X) as the information entropy of variable X; define the cache feature integer t as t=KM / N, t∈[0:K] to represent the cache step; each file W n Divided into disjoint subfiles, each subfile is represented by W n,T , where n∈[N], for The user set with length t in and For each user k, That is all Cache sub-files All bits of , n∈[N]; Since each file contains Therefore, each user satisfies the following storage constraints:

[0012]

[0013] Furthermore, the specific steps of step 3 are as follows, including determining the lead requester, splitting the set of request files, and encoding the requested sub-files:

[0014] The steps to determine the lead requester are as follows:

[0015] For each user k who does not request the file, Any choice requesters; these requesters request Different files, represented as is called the leader requester of user k.

[0016] Furthermore, the steps for splitting the collection of request files are as follows:

[0017] Each subfile is divided into three categories; for file n, the first category of subfiles is only cached by users who do not request the file, denoted as in for The user set with length t in The second type of sub-files are cached only by the requester and the user who does not request the file, which is represented as in Collection for users With user collection The union of i is the cache partition integer, i∈[1,t-1], B i is a user set of length i in R, that is, for The user set with length ti in the middle is The third type of sub-file is cached only in the requester, represented by in for The user set with length t in When t∈[K-r+1,K] or r=K, the first type of subfile does not exist; when t=1 or r=K, the second type of subfile does not exist, and when r=1 or t=K, the second type of subfile will not be needed by the requester; when t∈[r+1,K], the third type of subfile does not exist, and when r=1 or t=r, the third type of subfile will not be needed by the requester.

[0018] Furthermore, depending on the sub-file type, the steps for encoding the requested sub-file are as follows:

[0019] Step a: For requester k, The first type of sub-files required are first divided into t equal-length, non-overlapping sub-slices; each sub-slice is of length Bit, represented by Where a is The users in because It is needed by multiple requesters, so user a only needs to broadcast the following codeword:

[0020]

[0021] For all leader requesters of user a, the set is The codeword broadcast by user a is expressed as:

[0022]

[0023] These codewords are directly what all requesters need, and no decoding is required; because There are t users in There requesters, with a total of Different subsets So the feasible rate R that satisfies all requesters to obtain the first type of sub-files they need is 1st for:

[0024]

[0025] Step b: For requester k, The second type of subfile required is to split the set B into two disjoint subsets by integer i. and When i ≥ r, the second-category subfiles are all cached in the requester and do not need to be transmitted, so we only focus on i∈[1,min{t-1,r-1}]; first, fix the split integer i and perform the file splitting step: each second-category subfile is first divided into ti equal-length, non-overlapping sub-pieces; each sub-piece is of length Bit, represented by Where b is The users in Since given an arbitrary subset of i+1 requesters Each requester All need sub-pieces These sub-slices have been cached in Among the other requesters and user b in, user b only needs to broadcast the following codeword:

[0026]

[0027] Give satisfaction All subsets of All requesters in can be; due to the sub-slices on the right side of the above codeword, except for the sub-slices required by requester k The rest have been cached by requester k, so requester k successfully decodes the required sub-slice; the codeword broadcast by user b is expressed as:

[0028]

[0029] These codewords are useful to at least one leader requester of user b; since each user Need to broadcast A piece of film, There are ti users, total Different subsets And the partition integer i∈[1,min{t-1,r-1}], so the feasible rate R of the second type of sub-files required by all requesters is 2nd for:

[0030]

[0031] Step c: For requester k, The required third-category sub-files are first split into t equal-length, non-overlapping sub-pieces. Bit, expressed as Where e is The users in Requester Collection Each user j in the set Choose any group containing A subset of users, denoted as Users in the subset request different files; for all subsets containing t users Each user j broadcasts a codeword:

[0032]

[0033]

[0034] Since each user Both need to decode the required sub-slice from user j's transmission Using the idea of ​​leader requester, each user j only needs to broadcast satisfy of That's it, For each user A total of sub-slices, so the feasible rate R that satisfies all requesters to obtain the required third-category sub-files is 3rd for:

[0035]

[0036] The total feasible rate of the transmission scheme that enables all requesters to decode the requested single file is:

[0037] R=R 1st +R 2nd +R 3rd .

[0038] The transmission method for the device-to-device code cache problem based on random requester transmission of the present invention has the following advantages: due to the design of steps 1 and 2, the present invention can be applied to any number of users K, any number of files N, and any number of requesters r less than K; and due to the design of step 3, the present invention can achieve a lower transmission rate and is significantly superior to the transmission scheme for the general device-to-device code cache problem based on non-random requester transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A system model of a device-to-device encoding cache with random requester transmission according to the present invention. DETAILED DESCRIPTION

[0040] In order to better understand the purpose, structure and function of the present invention, the following is a further detailed description of a device-to-device code cache transmission method based on random requester transmission of the present invention in conjunction with the accompanying drawings.

[0041] An embodiment is given below:

[0042] Figure 1 System model for device-to-device encoding caching with random requester transfers. In this example, there are three users, and user 2 does not request content. The solid and dashed lines represent the placement and transfer phases, respectively.

[0043] Step 1: Consider a device-to-device encoding cache problem with K = 6 users, N = 2 files, and M = 2 / 3 cache size. In this problem, the parameter t = KM / N = 2.

[0044] Step 2: Select the non-encoded symmetric cache solution based on the system model. In the placement phase, each file is divided into sub-files, and the index of each sub-file is in the set Γ, and Γ is the set of all sub-files that satisfy A collection of conditions User k∈[6] caches the following sub-files for each file n∈{1,2}:

[0045]

[0046] Step 3: Implement the transmission plan according to the system model and the selected cache plan. In the transmission phase, only consider users 1, 2, 3, and 4 as requesters requesting a single file and the request vector is D {1,2,3,4} =(1,2,1,1). Note that r = 4 and N e (D {1,2,3,4} )=2.

[0047] Without loss of generality, assume that user 5 selects users 1 and 2 and user 6 selects users 2 and 3 as leader requesters, i.e. The missing sub-files required by requesters 1, 2, 3, and 4 are as follows:

[0048] W1\Z1={W 1,{2,3} ,W 1,{2,4} ,W 1,{2,5} ,W 1,{2,6} ,W 1,{3,4} ,W 1,{3,5} ,W 1,{3,6} ,W 1,{4,5} ,W 1,{4,6} ,W 1,{5,6}},

[0049] W2\Z2={W 2,{1,3} ,W 2,{1,4} ,W 2,{1,5} ,W 2,{1,6} ,W 2,{3,4} ,W 2,{3,5} ,W 2,{3,6} ,W 2,{4,5} ,W 2,{4,6} ,W 2,{5,6}},

[0050] W1\Z3={W 1,{1,2} ,W 1,{1,4} ,W 1,{1,5} ,W 1,{1,6} ,W 1,{2,4} ,W 1,{2,5} ,W 1,{2,6} ,W 1,{4,5} ,W 1,{4,6} ,W 1,{5,6}},

[0051] W1\Z4={W 1,{1,2} ,W 1,{1,3} ,W 1,{1,5} ,W 1,{1,6} ,W 1,{2,3} ,W 1,{2,5} ,W 1,{2,6} ,W 1,{3,5} ,W 1,{3,6} ,W 1,{5,6}}

[0052] Based on the definition of sub-files, the sub-files required above can be quickly divided into three categories.

[0053] 1) For the first type of sub-files, they are only cached in users 5 and 6. These sub-files need to be split into two sub-segments of equal length. Then, users 5 and 6 directly transmit the following codewords required by all requesters:

[0054]

[0055]

[0056] For the transmission of the first type of sub-files, the transmission rate is R 1st =1 / 30×2×2=2 / 15.

[0057] 2) For the second type of sub-files, they are only cached in the requester and users 5 and 6. Due to the partitioning integer i∈[1,min{t-1,r-1}], in this example, only i=1 needs to be considered. At the same time, since ti=1, no sub-files need to be further partitioned. According to and , both user 5 and user 6 can find Subset Make in Therefore, users 5 and 6 only need to transmit the following codewords:

[0058]

[0059]

[0060]

[0061]

[0062] Without transmitting codewords and This is due to and Regardless of how user 5 and user 6 choose the leader requester, through this scheme, they can transmit one less codeword, which is expressed as Therefore, for the transmission of the second type of sub-files, the transmission rate is R 2nd =1 / 15×5×2=2 / 3.

[0063] 3) For the third type of sub-files, they are only cached in the requester. and For i∈{1,3,4}, without loss of generality, consider user 1 selecting users 2 and 4, user 2 selecting user 3, user 3 selecting users 1 and 2, and user 4 selecting users 2 and 3 as leader requesters, i.e.

[0064] After dividing the third type of sub-file into two equal-length sub-slices, since there is no subset for users 1, 3, and 4 satisfy So these users can directly transmit the following codewords:

[0065]

[0066]

[0067]

[0068] However, according to u 2 choice, existence Subset Make in Therefore, user 2 only needs to transmit the following codeword:

[0069]

[0070] Without transmitting codewords This is due to No matter how user 2 chooses the leader requester, through this scheme, he can transmit one less codeword, which is expressed as Therefore, for the transmission of the third type of sub-file, the transmission rate is R 3rd =1 / 30×(3×3+2)=11 / 30.

[0071] By combining all codewords, all requesters can obtain the required sub-files. The total transmission rate in this example is R r-r =R 1st +R 2nd +R 3rd =7 / 6.

[0072] It will be understood that the present invention is described by way of some embodiments, and it will be appreciated by those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are intended to be protected by the present invention.

Claims

1. A method for transmitting device-to-device code cache based on random requester transmission, characterized in that: The steps include: Step 1: First, define the random requester transmission, that is, for the number of users K, the user set At the beginning of the transmission phase, r random users request a single file with equal probability, r∈[1,K]. These users are called requesters and are represented by the set To express, The remaining users do not request files; then the device-to-device encoding cache problem is systematically modeled for any number of users K, any number of files N, and a cache size M within a certain range, M∈[1,N]; For file collections The nth file uses W n To represent, n∈[1,N]; the system consists of a placement phase and a transmission phase; in the placement phase, all N files are connected The central server fills the caches of all K users; each file is F bits long; user k’s cache is Z k , In the transfer phase, r users act as requesters to request a single file d k ,d k ∈[N], The request vector is All K users transmit the encoded signal Make all r requesters able to decode the requested file; The goal is to find a cache and transmission scheme with the minimum feasible rate R in the worst case; Step 2: According to the system model, in the placement phase, the server does not know the characteristics of the number of future requesters r, and selects the non-coding symmetric placement scheme as the placement scheme; Step 3: Design a transmission plan based on the placement plan. First, select a leader requester from all requesters to satisfy subsequent transmissions. Then, classify each requested subfile into three categories. Finally, based on the type of the requested subfile, the user who caches these subfiles encodes them in different forms and sends them to all requesters via a broadcast channel, so that all requesters can correctly decode the requested file. The sum of the transmission rates of the three types of sub-files is the feasible transmission rate of the entire transmission solution.

2. The method for transmitting device-to-device code cache based on random requester transmission according to claim 1, characterized in that: The specific non-coding symmetrical placement scheme selected in step 2 is as follows: Define H(X) as the information entropy of variable X; define the cache feature integer t as t=KM / N, t∈[1:K] to represent the cache step; each file W n Divided into disjoint subfiles, each of which is represented by where n∈[N], for The user set with length t in and For each user k, That is all Cache sub-files All bits of , n∈[N]; Since each file contains indivual Therefore, each user satisfies the following storage constraints:

3. The method for transmitting device-to-device code cache based on random requester transmission according to claim 2, characterized in that: Step 3 includes the following steps: determining the lead requester, splitting the set of requested files, and encoding the requested sub-files: The steps to determine the lead requester are as follows: For each user k who does not request the file, Any choice requesters; these requesters request Different files, represented as is called the leader requester of user k.

4. The method for transmitting device-to-device code cache based on random requester transmission according to claim 3, characterized in that: The steps to split a collection of request files are as follows: Each subfile is divided into three categories; for file n, the first category of subfiles is only cached by users who do not request the file, denoted as W n,A ,in for The user set with length t in The second type of sub-files are cached only by the requester and the user who does not request the file, which is represented as in Collection for users With user collection The union of i is the cache partition integer, i∈[1,t-1], for The user set with length i in for The user set with length ti in the middle is The third type of sub-file is only cached in the requester, denoted as W n,ε , where ε is The user set with length t in ε=t; When t∈[K-r+1,K] or r=K, the first type of subfile does not exist; when t=1 or r=K, the second type of subfile does not exist, and when r=1 or t=K, the second type of subfile will not be needed by the requester; when t∈[r+1,K], the third type of subfile does not exist, and when r=1 or t=r, the third type of subfile will not be needed by the requester.

5. The method for transmitting device-to-device code cache based on random requester transmission according to claim 4, characterized in that: Depending on the sub-file type, the steps for encoding the requested sub-file are as follows: Step a: For requester k, The first type of sub-files required are first divided into t equal-length, non-overlapping sub-slices; each sub-slice is of length Bit, expressed as Where a is The users in because It is needed by multiple requesters, so user a only needs to broadcast the following codeword: For all leader requesters of user a, the set is The codeword broadcast by user a is expressed as: These codewords are directly what all requesters need, and no decoding is required; because There are t users in There are requesters, with a total of Different subsets So the feasible rate R that satisfies all requesters to obtain the first type of sub-files they need is 1st for: Step b: For requester k, The second type of subfile required is to split the set B into two disjoint subsets by integer i. and When i ≥ r, the second-category subfiles are all cached in the requester and do not need to be transmitted, so we only focus on i∈[1,min{t-1,r-1}]; first, fix the split integer i and perform the file splitting step: each second-category subfile is first divided into ti equal-length, non-overlapping sub-pieces; each sub-piece is of length Bit, expressed as Where b is The users in Since given an arbitrary subset of i+1 requesters Each requester All need sub-pieces These sub-slices have been cached in Among the other requesters and user b in, user b only needs to broadcast the following codeword: Give satisfaction All subsets of All requesters in can be; due to the sub-slices on the right side of the above codeword, except for the sub-slices required by requester k The rest have been cached by requester k, so requester k successfully decodes the required sub-slice; the codeword broadcast by user b is expressed as: These codewords are useful to at least one leader requester of user b; since each user Need to broadcast A piece of film, There are ti users, total Different subsets And the partition integer i∈[1,min{t-1,r-1}], so the feasible rate R of the second type of sub-files required by all requesters is 2nd for: Step c: For requester k, The required third-category sub-files are first split into t equal-length, non-overlapping sub-pieces. Bit, denoted as W n,ε,e , where e is the user in ε, that is, e∈ε; the requester set Each user j in the set Choose any group containing A subset of users, denoted as Users in the subset request different files; for all subsets containing t users Each user j broadcasts a codeword: Since each user Both need to decode the required sub-slice from user j's transmission Using the idea of ​​leader requester, each user j only needs to broadcast satisfy of That's it, For each user A total of sub-slices, so the feasible rate R that satisfies all requesters to obtain the required third-category sub-files is 3rd for: The total feasible rate of the transmission scheme that enables all requesters to decode the requested single file is: R=R 1st +R 2nd +R 3rd 。

Images