A transport-aware cooperative caching method

Abstract

This invention relates to a transmission-aware cooperative caching method, belonging to the field of wireless communication. The method first caches primary content by analyzing the location of primary users (PUs) and the popularity of their requested content. Then, through horizontal cooperation among secondary base stations (SBSs), the caching layout of primary content is further optimized. Using established PU service rules, the method aims to minimize PU request latency in both transmission and storage dimensions, providing faster service to PUs. Finally, in the remaining time, secondary content is cached through cache replacement to maximize the throughput of secondary users (SUs). This invention can significantly reduce the request latency for PUs to retrieve content in a CRN, while simultaneously bringing greater throughput benefits to SUs, thus improving the overall performance of the radio access network.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication and relates to a transmission-aware cooperative caching method. Background Technology

[0002] The surge in mobile devices (including smartphones, wearables, and tablets) and IoT devices, coupled with the enormous demand for emerging services, has led to an explosive growth in mobile data traffic. Multimedia transmission, in particular, has become the primary burden on Radio Access Networks (RAN) and backhaul networks. However, this content follows a Zipf distribution, meaning that 20% of the most popular content accounts for 80% of the total traffic. This results in users repeatedly downloading the same content. In traditional RANs, users frequently need to retrieve popular content from content servers in the core network via backhaul, which not only causes significant latency for users but also severely wastes limited backhaul resources, further burdening the backhaul network.

[0003] Cached content at base stations is considered an effective way to support the massive growth of mobile data traffic. To address this issue without causing latency or congestion for users, edge caching has been proposed. This involves proactively caching popular content at base stations closer to users during off-peak hours. When users request content, they don't need to retrieve it from the content server via backhaul, thus reducing backhaul load and improving user experience. However, this approach rarely considers base station collaboration, resulting in significant waste of base station caching resources.

[0004] Another proposed solution to support the increasing mobile network traffic is the dense deployment of base stations with small coverage areas. However, this dense deployment presents another challenge: spectrum scarcity. Radio Networks (CRNs) offer a promising paradigm for improving spectrum efficiency and alleviating spectrum scarcity. CRs feature adaptive spectrum access technology, allowing wireless devices to adaptively select appropriate spectrum holes for communication based on currently available spectrum resources. This allows them to operate without interfering with existing communication systems, effectively improving spectrum utilization and meeting the needs of different application scenarios. Furthermore, unlicensed users can utilize spectrum resources not used by licensed users.

[0005] Based on the above technical advantages, consider a CRN consisting of Primary Base Stations (PBSs) and Service Base Stations (SBSs), and User Providers (PUs) requesting primary content and Service Providers (SUs) requesting secondary content. In the CRN, unlicensed secondary networks are allowed opportunistic access to spectrum licensed to primary networks. When a PU on a channel is inactive, SUs are allowed opportunistic access to that channel for transmission. SBSs can assist the PBS in transmitting PUs' data, thus reducing the total duration of channel occupancy by PUs. As a reward, SUs gain access to licensed spectrum, increasing their transmission opportunities, thus benefiting both PUs and SUs and achieving a mutually beneficial CRN. In the considered CRN, PUs can be served by both SBSs and PBSs. If the primary content requested by a PU is simultaneously cached in multiple densely deployed SBSs, the SBSs can transmit the content to the PU through Joint Transmission Technology (JT), a technique called CoMP-JT. This technique helps reduce inter-cell interference, improves cell edge throughput, and maximizes transmission opportunities for SUs by serving PUs faster. Summary of the Invention

[0006] In view of this, the purpose of this invention is to provide a traffic sign detection method based on dense connection attention. In cognitive radio scenarios, through cooperative caching between base stations, under service rules aimed at minimizing PU request latency, the system maximizes throughput by providing faster service to PUs, thus giving SUs more opportunities to occupy channels. This cooperative caching includes two aspects: firstly, unlicensed SBSs can cache certain key content to serve PUs in exchange for SUs having the opportunity to access licensed spectrum; secondly, cooperation between SBSs is considered to fully utilize potential JT and ST opportunities.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A transport-aware cooperative caching method, comprising the following steps:

[0009] S1: System cold start: Initialize and determine network topology, relevant content library information, primary user (PU) and secondary user (SU) locations, channel conditions, and time cost of obtaining cached files at different locations;

[0010] S2: Establishing the primary content cache decision: Without considering the collaborative caching between secondary base stations (SBS), to obtain the primary content cache placement in constant time, a joint optimization of content cache based on a genetic algorithm is used to solve for the suboptimal cache layout.

[0011] S3: Horizontal Collaboration: In Suboptimal Cache Layout Based on this, by horizontal collaboration between adjacent SBSs, the latency of PU request content is reduced, and the optimal caching strategy for each SBS in the scenario of large-scale content is obtained.

[0012] S4: Content Routing: Under the final content placement strategy X, the optimal transmission strategy is determined by the defined PU service rules when users in different locations request different content, so as to achieve the minimum content delivery delay.

[0013] S5: Cache Replacement: Replace some of the less effective main content with less effective secondary content. This increases the throughput of SU by sacrificing some PU service time to cache the secondary content.

[0014] Optionally, the relevant content library information includes content size and content popularity.

[0015] Optionally, S1 specifically includes the following steps:

[0016] S11: Determine network topology: using δ ir To represent SBSb i With SBSb r Are they neighbors? If b i With b r If they are neighbors, then δ ir =1, otherwise δ ir =0, if the network topology is given, δ ir The value of is known, then by The adjacent SBSs are represented by the following formula:

[0017]

[0018] in, Let b0 represent the set of SBSs, and b0 represent PBS. Represents the set of all base stations;

[0019] S12: Retrieve the main contents of the file library The probability that secondary content N = {1, 2, ..., N} is requested. and Specifically, it is expressed as follows:

[0020]

[0021]

[0022] in, and The average request reach rate for primary and secondary content; and This indicates the popularity of the main and secondary content.

[0023] S13:SBSb i According to long-term average To determine if PU is located in SBSb i Core or periphery of the community:

[0024]

[0025] In the formula, T is the update Time length; ρ i For SBSb i The transmit power; v ij Channel power gain; From SBSb i To PUp j Path loss; σ 2 Noise power; I j For other interference;

[0026] Determined with SBSb i The associated set of PUs at the cell edge:

[0027]

[0028] In the formula, τ refers to the SINR threshold at which qualified ST transmission can be performed; the user set of the cell core is represented as:

[0029]

[0030] SBS, according to the report Regular updates and

[0031] S14: For each main content m∈M This indicates that the main content m is in SBSb i The cache state on, where This indicates that the main content is cached; similarly, for each secondary content n∈N, Indicates secondary content n in SBSb i The cache state on, where This indicates that secondary content is cached; therefore, x = (x p ,x s ) indicates that the content is in SBSb i The cache state on, where Furthermore, the cache status meets the SBS capacity limit:

[0032]

[0033] S15: SBSb i The total latency for a PU within the core of a residential community to access the main content m using ST is calculated as follows:

[0034]

[0035] In the formula, the latency of the user using ST transmission is:

[0036]

[0037] SBSb i The total latency for a PU within the cell edge to access the main content m using JT is calculated as follows:

[0038]

[0039] In the formula, a JT decision variable is introduced. To indicate the selected transmission method, if This indicates that JT transmission is used; the latency of JT transmission is:

[0040]

[0041] The latency calculation for PU to obtain the main content m through cooperation with neighboring base stations is as follows:

[0042]

[0043] To find the content containing the requested main content m with the lowest transmission time. Optimization issues:

[0044]

[0045] In the formula, V is a constant to ensure that the main content m is not included. Ignored; if one of its directly connected SBSs stores the main content m, then SBSb i The content will be downloaded from one of these nodes, rather than from the core network or another SBS in the network via SBS.

[0046] The PU will obtain the main content m from the core network, and the latency is calculated as follows:

[0047]

[0048] In the formula, To download the main content m from PBSb0 to PBSb i The delay.

[0049] Optionally, S2 specifically includes the following steps:

[0050] S21: For a network with B BSs and M contents, the chromosome of an individual is a B×M binary matrix, where the b-th row of the matrix represents the contents that may be cached in SBSb, expressed as:

[0051]

[0052] Assume the initial population Q contains N pop A matrix of individuals; the top N individuals in the population pop -2 individuals are initialized to X n =0 M×B ,n∈[1,N pop A zero matrix of [-2], then elements in the matrix are processed row by row. Randomly set to 0 or 1, This represents the cached content m in SBSb, and it satisfies the following conditions:

[0053]

[0054] Finally, the two matrix individuals after processing the MPC and LCD schemes and It is added to the initial population Q to improve the convergence speed of the entire algorithm;

[0055] S22: Perform the fitness function to evaluate N as follows: pop Fitness values ​​of individual individuals:

[0056]

[0057] in, In the formula,

[0058] In each selection process, R individuals are randomly selected from the population to enter the tournament group, and the individual with the best fitness value is selected to enter the next generation as the parent individual; this continues until a new parent population Q composed of elite individuals is formed. new Sufficient; to reduce algorithm complexity, select the top N from the population in descending fitness order. ele Each individual is used as a parent and directly copied into the next generation population E.

[0059] S23: To improve the crossover efficiency of the algorithm, a multi-row parallel multi-point crossover strategy is adopted. Each time, two adjacent matrix individuals are selected, and then multiple crossover points are selected in each row of each individual to perform crossover with probability, generating two new offspring individuals. Secondly, to avoid getting trapped in local optima, a mutation operation is also required. During the mutation process, the offspring individuals are randomly positioned with probability p. mutPerform flipping; check and repair individuals after crossover and mutation. During the repair process, sort the cached content in base stations that violate capacity constraints in ascending order of popularity and flip it sequentially. That is, for individuals smaller than the base station capacity, "0" needs to be flipped to "1", and for individuals larger than the base station capacity, "1" needs to be flipped to "0"; until the capacity constraint is met to ensure that the size of the cached content is equal to the storage space of each SBS.

[0060] S24: Repeat S22 and S23 until the fitness difference between the two populations reaches the preset threshold, then the iteration ends and a suboptimal cache layout of the main content is obtained.

[0061] Optionally, S3 specifically includes the following steps:

[0062] S31: Initial state for obtaining content It then checks the size of the cached content at the current base station, and then obtains the delivery delay under the current caching conditions using the following formula, denoted as .

[0063]

[0064] stC1:T P ≤T

[0065] C2:

[0066] C3: In the formula, θ is the cache segmentation ratio, i.e., the SBS uses... We will use space to cache the main content, let θ = 1;

[0067] S32: Check SBSb i Does the neighboring base station cache the same content?

[0068]

[0069] If the same content m is cached, select content m other than content m that is not cached by its neighboring base stations. * Cache the cache while satisfying the SBS capacity constraint. If the capacity constraint is satisfied, then... Then perform a cache update. Calculate the size of the current SBS cached content. and SBSb i Delivery delays for all users within the coverage area

[0070] S33: After checking all SBSs, calculate the total delivery delay. If the total latency decreases at this point, then update the content and the new delivery delay to obtain the optimal caching strategy for each SBS in a large-scale content scenario.

[0071] Optionally, S4 specifically includes the following steps:

[0072] S41: SBS receives a request from PU and checks the content cache status, and determines the PUp. j Location within the residential area;

[0073] S42: If SBSb i Cached PUP j The requested file m, i.e. And PUp j Located in the heart of the community, Then ST is executed to provide services to the user; if PUp j Located on the edge of the residential area, Then, through the buffer state of neighboring base stations To determine the steps that need to be taken;

[0074] S43: If The request will then be routed to the secondary base station SBS b. r And using JT, where Otherwise, SBSb i The request is satisfied using ST;

[0075] S45: If SBSb i No cached PUP j The requested file m, i.e. Then it is necessary to determine the buffer status of neighboring base stations. like Then obtain it from the nearest neighboring base station; if It is obtained from PBS.

[0076] Optionally, S5 specifically includes the following steps:

[0077] S51: Calculate SBSb i The total size of the cached content is denoted as . SBSb i The main content already cached is sorted in ascending order by the number of PU requests, denoted as All secondary content to be cached should be configured according to SBSb. i Sort the number of SU requests in descending order, denoted as

[0078] S52: Check SBSb i All cached content, while meeting SBS capacity requirements, replaces some less useful primary files with more popular secondary files; if the requirements are met... but Calculate and replace After that, the time T that caused the service duration to increase. 1 The difference between the time it takes for the PU to retrieve the content from the PBS and the time originally required to request the content m;

[0079] S53: Until in The pre-set threshold for stopping replacement; the time T is the latency benefit after replacing the main content. 1 Replacement is complete when it is no longer sufficient to bring greater benefits to the system.

[0080] The beneficial effects of this invention are as follows:

[0081] This invention caches primary content by analyzing the location of primary users (PUs) and the popularity of their requested content. Through horizontal collaboration among secondary base stations (SBSs), the caching layout of primary content is further optimized. By establishing PU service rules, the invention aims to minimize PU request latency in both transmission and storage dimensions, providing faster service to PUs. Finally, during the remaining time, secondary content is cached to maximize the throughput of secondary users (SUs) through cache replacement. This invention can significantly reduce the request latency for PUs to retrieve content in a CRN, while simultaneously bringing greater throughput benefits to SUs, thus improving the overall performance of the radio access network.

[0082] This invention, within a cognitive radio network, considers cooperative caching between primary and secondary networks. Specifically, the secondary network serves the primary user (PU) by caching primary content. Simultaneously, joint optimization of transmission and storage dimensions is considered during the service process, enabling faster PU service. This allows the primary user (SU) to occupy the channel for a longer period, resulting in greater throughput and effectively addressing the drawbacks of high backhaul pressure and low spectrum utilization.

[0083] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0084] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0085] Figure 1 This is a network architecture diagram for collaborative caching and transmission in a cognitive radio scenario according to the present invention;

[0086] Figure 2 This is the PU service rule diagram of the present invention;

[0087] Figure 3 This is a flowchart of the transmission-aware collaborative caching method of the present invention. Detailed Implementation

[0088] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0089] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0090] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0091] Please see Figures 1-3This invention provides a transmission-aware collaborative caching method. Addressing the significant increase in video services' share of mobile traffic in current wireless access networks, pre-caching content at base stations has become an effective solution. Simultaneously, to alleviate resource scarcity, cognitive radio technology is employed to improve spectrum utilization. Within this framework, secondary networks (SUs) can cache some primary content to serve public users (PUs). While serving PUs, a combination of transmission and storage dimensions is considered, along with defined PU service rules, to provide PUs with lower content request latency. This allows SUs more opportunities to occupy channels, increasing their throughput. Based on this, with the goal of minimizing PU request latency, a primary content caching strategy is developed under available time and capacity constraints. Then, considering collaboration between service base stations (SBSs), the caching layout of primary content is optimized to further reduce PU request latency. Finally, to provide services to SUs and maximize their throughput, cache replacement is used to replace inefficient primary content in SBSs with secondary content, improving the overall system performance.

[0092] This method specifically includes the following steps:

[0093] Step 1: System Cold Start: Initialize network topology, relevant content library information (such as content size, content popularity, cache status, etc.), PU and SU locations, channel conditions, and the time cost of retrieving cached files from different locations. This includes the following steps:

[0094] Step 1.1: Determine the network topology: using δ ir To represent SBSb i With SBSb r Are they neighbors? If b i With b r If they are neighbors, then δ ir =1, otherwise δ ir =0, and once the network topology is given, δ ir The value is known, by The adjacent SBS can be represented by the following formula:

[0095]

[0096] in, Let b0 represent the set of SBSs, and b0 represent PBS. This represents the set of all base stations.

[0097] Step 1.2: Obtain the main content from the file library The probability that secondary content N = {1, 2, ..., N} is requested. and Specifically, it is expressed as follows:

[0098]

[0099]

[0100] in, and The average request reach rate for primary and secondary content. and This indicates the popularity of the main and secondary content.

[0101] Step 1.3: SBSb i According to long-term average To determine if PU is located in SBSb i Core or periphery of the community:

[0102]

[0103] In the formula, T is the update Time length; ρ i For SBSb i The transmit power; v ij Channel power gain; From SBSb i To PUp j Path loss; σ 2 Noise power; I j Other interferences.

[0104] Then, determine with SBSb i The associated set of PUs at the cell edge:

[0105]

[0106] In the formula, τ refers to the SINR threshold at which a qualified ST transmission can be performed. Similarly, the user set of the cell core can be represented as:

[0107]

[0108] SBS, according to the report Regular updates and

[0109] Step 1.4: For each main content m∈M, This indicates that the main content m is in SBSb i The cache state on, where This indicates that the primary content is cached. Similarly, for each secondary content n∈N, Indicates secondary content n in SBSb iThe cache state on, where This indicates that secondary content is cached. Therefore, x = (x p ,x s ) indicates that the content is in SBSb i The cache state on, where Furthermore, the cache status meets the SBS capacity limit:

[0110]

[0111] Step 1.5: SBSb i The total latency for a PU within the core of a residential community to access the main content m using ST is calculated as follows:

[0112]

[0113] In the formula, the latency of the user using ST transmission is:

[0114]

[0115] SBSb i The total latency for a PU within the cell edge to access the main content m using JT is calculated as follows:

[0116]

[0117] In the formula, a JT decision variable is introduced. To indicate the selected transmission method, if This indicates that JT transmission is being used. The latency for JT transmission is:

[0118]

[0119] The latency calculation for PU to obtain the main content m through cooperation with neighboring base stations is as follows:

[0120]

[0121] To find the content containing the requested main content m with the lowest transmission time. Optimization issues:

[0122]

[0123] In the formula, V is a very large constant, ensuring that it does not contain the main content m. Ignored. If one of its directly connected SBSs stores the main content m, then SBSb i The content will be downloaded from one of these nodes, rather than attempting to download it from the core network or from another SBS in the network via SBS.

[0124] The PU will obtain the main content m from the core network, and the latency is calculated as follows:

[0125]

[0126] In the formula, To download the main content m from PBSb0 to PBSb i The delay.

[0127] Step 2: Main Content Caching Decision: Without considering collaborative caching between SBSs, in order to obtain the main content cache placement in constant time, a suboptimal cache layout is obtained by solving the problem through joint optimization of content caching based on a genetic algorithm. Specifically, the following steps are included:

[0128] Step 2.1: For a network with B BSs and M contents, the chromosome of an individual can be a B×M binary matrix, where the b-th row of the matrix represents the contents that may be cached in SBSb, expressed as:

[0129]

[0130] Assume the initial population Q contains N pop A matrix of individuals. The top N individuals in the population. pop -2 individuals are initialized to X n =0 M×B ,n∈[1,N pop A zero matrix of [-2], then elements in the matrix are processed row by row. Randomly set to 0 or 1, This indicates that content m is cached in SBSb and meets the following conditions:

[0131]

[0132] Finally, the two matrix individuals after processing the MPC and LCD schemes and It is added to the initial population Q to improve the convergence speed of the entire algorithm.

[0133] Step 2.2: Evaluate N by performing the fitness function as follows. pop Fitness values ​​of individual individuals:

[0134]

[0135] in, In the formula,

[0136] In each selection process, R individuals are randomly chosen from the population to enter the tournament group, and the individual with the best fitness value is selected to enter the next generation as the parent individual. This continues until a new parent population Q composed of elite individuals is formed. new This is sufficient. To reduce algorithm complexity, the top N cells in the population are selected in descending order of fitness. ele Each individual is used as a parent and directly copied into the next generation population E.

[0137] Step 2.3: To improve the crossover efficiency of the algorithm, a multi-row parallel multi-point crossover strategy is adopted. Each time, two adjacent matrix individuals are selected, and then multiple crossover points are selected in each row of each individual to perform crossover with probability, generating two new offspring individuals. Secondly, to avoid getting trapped in local optima, a mutation operation is also needed. During the mutation process, the offspring individuals are randomly positioned with probability p. mut Perform a flipping process. Considering that crossover and mutation may violate capacity constraints, individuals after crossover and mutation need to be checked and repaired. During the repair process, the cached content in the base stations that violate capacity constraints is sorted in ascending order of popularity and flipped sequentially (i.e., for individuals smaller than the base station capacity, "0" needs to be flipped to "1", and for individuals larger than the base station capacity, "1" needs to be flipped to "0"), until the capacity constraint is met, to ensure that the size of the cached content is equal to the storage space of each SBS.

[0138] Step 2.4: Repeat steps 2.2 and 2.3 until the fitness difference between the two populations reaches the preset threshold, the iteration ends, and the suboptimal cache layout of the main content is obtained.

[0139] Step 3: Horizontal Collaboration: In Suboptimal Cache Layout Building upon this foundation, by leveraging horizontal collaboration between adjacent SBSs, the latency of PU request content is further reduced, resulting in the optimal caching strategy for each SBS in scenarios with large-scale content. Specifically, this includes the following steps:

[0140] Step 3.1: Obtain the initial state of the content It then checks the size of the cached content at the current base station, and then obtains the delivery delay under the current caching conditions using the following formula, denoted as .

[0141]

[0142] stC1:T P ≤T

[0143] C2:

[0144] C3: In the formula, θ is the cache segmentation ratio, that is, SBS should use We will use space to cache the main content, assuming θ = 1.

[0145] Step 3.2: Check SBSb i Does the neighboring base station cache the same content?

[0146]

[0147] If the same content m is cached, select content m other than content m that is not cached by its neighboring base stations. * Cache the cache while satisfying the SBS capacity constraint. If the capacity constraint is satisfied, then... Then perform a cache update. Calculate the size of the current SBS cached content. and SBSb i Delivery delays for all users within the coverage area

[0148] Step 3.3: After checking all SBSs, calculate the total delivery delay. If the total latency decreases at this point, then the content update and the new delivery delay will be reduced. Based on the optimization steps above, the optimal caching strategy for each SBS can be obtained in scenarios with large-scale content.

[0149] Step 4: Content Routing: Under the final content placement strategy X, the optimal transmission strategy is determined based on the defined PU service rules when users in different locations request different content, in order to achieve minimal content delivery latency. This includes the following steps:

[0150] Step 4.1: SBS receives the request from PU and checks the content cache status, and determines the PUp. j Location within the residential area.

[0151] Step 4.2: If SBSb i Cached PUP j The requested file m (i.e. ), and PUp j Located in the core of the community Then ST is executed to provide services to the user. If PUp j Located on the edge of the community Then, through the buffer state of neighboring base stations To determine the steps that need to be taken.

[0152] Step 4.3: If The request will then be routed to the secondary base station SBS b. r And using JT, where Otherwise, SBSb i The request is satisfied using ST.

[0153] Step 4.4: If SBSb i No cached PUP j The requested file m, i.e. Then it is necessary to determine the buffer status of neighboring base stations. like Then obtain it from the nearest neighboring base station; if It is obtained from PBS.

[0154] Step 5: Cache Replacement: Replace some of the less efficient primary content with less efficient secondary content. This is done by sacrificing some service time per unit (PU) to cache the secondary content, thereby increasing the throughput of the task manager (SU). Specifically, this includes the following steps:

[0155] Step 5.1: Calculate SBSb i The total size of the cached content is denoted as . SBSb i The main content already cached is sorted in ascending order by the number of PU requests, denoted as All secondary content to be cached should be configured according to SBSb. i Sort the number of SU requests in descending order, denoted as

[0156] Step 5.2: Check SBSb i All cached content, while meeting SBS capacity requirements, replaces some less useful primary files with more popular secondary files. If the requirements are met... but At the same time, calculate to replace After that, the time T that caused the service duration to increase. 1 (The difference between the time it takes for PU to retrieve this content from PBS and the time originally required to request content m).

[0157] Step 5.3: Until in The pre-set threshold for stopping the replacement. The latency gain after replacing the main content is time T. 1 Replacement is complete when it is no longer sufficient to bring greater benefits to the system.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A transmission-aware cooperative caching method, characterized in that: The method includes the following steps: S1: System cold start: Initialize and determine network topology, relevant content library information, primary user (PU) and secondary user (SU) locations, channel conditions, and time cost of obtaining cached files at different locations; S2: Establishing the primary content cache decision: Without considering the collaborative caching between secondary base stations (SBS), to obtain the primary content cache placement in constant time, a joint optimization of content cache based on a genetic algorithm is used to solve for the suboptimal cache layout. S3: Horizontal Collaboration: In Suboptimal Cache Layout Based on this, by horizontal collaboration between adjacent SBSs, the latency of PU request content is reduced, and the optimal caching strategy for each SBS in the scenario of large-scale content is obtained. S4: Content Routing: Under the final content placement strategy X, the optimal transmission strategy is determined by the defined PU service rules when users in different locations request different content, so as to achieve the minimum content delivery delay. S5: Cache replacement: Replace some of the less effective main content with less effective secondary content. This increases the throughput of SU by sacrificing some PU service time to cache the secondary content. The relevant content library information includes content size and content popularity; S1 specifically includes the following steps: S11: Determine network topology: using δ ir To represent SBS b i With SBS b r Are they neighbors? If b i With b r If they are neighbors, then δ ir =1, otherwise δ ir =0, if the network topology is given, δ ir The value of is known, then by The adjacent SBSs are represented by the following formula: in, Let b0 represent the set of SBSs, and b0 represent PBS. Represents the set of all base stations; S12: Retrieve the main contents of the file library and secondary content Probability of being requested and Specifically, it is expressed as follows: in, and The average request reach rate for primary and secondary content; and This indicates the popularity of the main and secondary content. S13: SBS b i According to long-term average To determine if PU is located in SBS b i Core or periphery of the community: In the formula, T is the update Time length; ρ i For SBS b i The transmit power; v ij Channel power gain; From SBS b i to PU p j Path loss; σ 2 Noise power; I j For other interference; Determined to be in contact with SBS b i The associated set of PUs at the cell edge: In the formula, τ refers to the SINR threshold at which qualified ST transmission can be performed; the user set of the cell core is represented as: SBS, according to the report Regular updates and S14: For each main content This indicates that the main content m is in SBS b i The cache state on, where This indicates that the main content is cached; similarly, for each secondary content... Indicates secondary content n in SBSb i The cache state on, where This indicates that secondary content is cached; therefore, x = (x p ,x s ) indicates that the content is in SBS b i The cache state on, where Furthermore, the cache status meets the SBS capacity limit: S15: SBS b i The total latency for a PU within the core of a residential community to access the main content m using ST is calculated as follows: In the formula, the latency of the user using ST transmission is: SBS b i The total latency for a PU within the cell edge to access the main content m using JT is calculated as follows: In the formula, a JT decision variable is introduced. To indicate the selected transmission method, if This indicates that JT transmission is used; the latency of JT transmission is: The latency calculation for PU to obtain the main content m through cooperation with neighboring base stations is as follows: To find the content containing the requested main content m with the lowest transmission time. Optimization issues: In the formula, V is a constant to ensure that the main content m is not included. Ignored; if one of its directly connected SBSs stores the main content m, then SBS b i The content will be downloaded from one of these nodes, rather than from the core network or another SBS in the network via SBS. The PU will obtain the main content m from the core network, and the latency is calculated as follows: In the formula, To download the main content m from PBS b0 to SBS b i The delay; S2 specifically includes the following steps: S21: For a network considering B BSs and M contents, the chromosome of an individual is a B×M binary matrix, where the b-th row of the matrix represents the contents that may be cached in SBS b, expressed as: Assume the initial population Q contains N pop A matrix of individuals; the top N individuals in the population pop -2 individuals are initialized to X n =0 M×B ,n∈[1,N pop A zero matrix of [-2], then elements in the matrix are processed row by row. Randomly set to 0 or 1, This indicates that the cached content m in SBS b satisfies the following conditions: Finally, the two matrix individuals after processing the MPC and LCD schemes and It is added to the initial population Q to improve the convergence speed of the entire algorithm; S22: Perform the fitness function to evaluate N as follows: pop Fitness values ​​of individual individuals: in, In the formula, In each selection process, R individuals are randomly selected from the population to enter the tournament group, and the individual with the best fitness value is selected to enter the next generation as the parent individual; this continues until a new parent population Q composed of elite individuals is formed. new Sufficient; to reduce algorithm complexity, select the top N from the population in descending fitness order. ele Each individual is used as a parent and directly copied into the next generation population E. S23: To improve the crossover efficiency of the algorithm, a multi-row parallel multi-point crossover strategy is adopted. Each time, two adjacent matrix individuals are selected, and then multiple crossover points are selected in each row of each individual to perform crossover with probability, generating two new offspring individuals. Secondly, to avoid getting trapped in local optima, a mutation operation is also required. During the mutation process, the offspring individuals are randomly positioned with probability p. mut Perform flipping; check and repair individuals after crossover and mutation. During the repair process, sort the cached content in base stations that violate capacity constraints in ascending order of popularity and flip it sequentially. That is, for individuals smaller than the base station capacity, 0 needs to be flipped to 1, and for individuals larger than the base station capacity, 1 needs to be flipped to 0. Until the capacity constraint is met, to ensure that the size of the cached content is equal to the storage space of each SBS. S24: Repeat S22 and S23 until the fitness difference between the two populations reaches the preset threshold, the iteration ends, and the suboptimal cache layout of the main content is obtained. S3 specifically includes the following steps: S31: Initial state for obtaining content It then checks the size of the cached content at the current base station, and then obtains the delivery delay under the current caching conditions using the following formula, denoted as . s.t.C1:T P ≤T In the formula, θ is the cache segmentation ratio, i.e., the SBS uses... We will use space to cache the main content, let θ = 1; S32: Check SBS b i Does the neighboring base station cache the same content? If the same content m is cached, select content m other than content m that is not cached by its neighboring base stations. * Cache the cache while satisfying the SBS capacity constraint. If the capacity constraint is satisfied, then... Then perform a cache update. Calculate the size of the current SBS cached content. And SBS b i Delivery delays for all users within the coverage area S33: After checking all SBSs, calculate the total delivery delay. If the total latency decreases at this point, then update the content and the new delivery delay to obtain the optimal caching strategy for each SBS in a large-scale content scenario. S4 specifically includes the following steps: S41: SBS receives a request from PU and checks the content cache status, and determines the PU p j Location within the residential area; S42: If SBS b i Cached PU p j The requested file m, i.e. And PU p j Located in the heart of the community, Then ST is executed to provide services to the user; if PU p j Located on the edge of the residential area, Then, through the buffer state of neighboring base stations To determine the steps that need to be taken; S43: If The request will then be routed to the secondary base station SBS b. r And using JT, where Otherwise, SBS b i Use ST to fulfill this request; S45: If SBS b i No cached PU p j The requested file m, i.e. Then it is necessary to determine the buffer status of neighboring base stations. like Then obtain it from the nearest neighboring base station; if Then obtain it from PBS; S5 specifically includes the following steps: S51: Calculate SBS b i The total size of the cached content is denoted as . SBS b i The main content already cached is sorted in ascending order by the number of PU requests, denoted as All secondary content to be cached will be processed according to SBS b i Sort the number of SU requests in descending order, denoted as S52: Check SBS b i All cached content, while meeting SBS capacity requirements, replaces some less useful primary files with more popular secondary files; if the requirements are met... but Calculate and replace After that, the time T that caused the service duration to increase. 1 The difference between the time it takes for the PU to retrieve the content from the PBS and the time originally required to request the content m; S53: Until in The pre-set threshold for stopping replacement; the time T is the latency benefit incurred after replacing the main content. 1 Replacement is complete when it is no longer sufficient to bring greater benefits to the system.

Images