WO2023045253A1 - 分布式移动网络视频缓存放置方法、系统及相关设备 - Google Patents
分布式移动网络视频缓存放置方法、系统及相关设备 Download PDFInfo
- Publication number
- WO2023045253A1 WO2023045253A1 PCT/CN2022/078360 CN2022078360W WO2023045253A1 WO 2023045253 A1 WO2023045253 A1 WO 2023045253A1 CN 2022078360 W CN2022078360 W CN 2022078360W WO 2023045253 A1 WO2023045253 A1 WO 2023045253A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- base station
- video data
- micro
- sub
- micro base
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 230000001413 cellular effect Effects 0.000 claims abstract description 65
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 27
- 238000004364 calculation method Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 14
- 238000012217 deletion Methods 0.000 claims description 12
- 230000037430 deletion Effects 0.000 claims description 12
- 238000005457 optimization Methods 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 3
- 230000005540 biological transmission Effects 0.000 abstract description 5
- 238000004891 communication Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 16
- 238000005070 sampling Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/234—Processing of video elementary streams, e.g. splicing of video streams or manipulating encoded video stream scene graphs
- H04N21/23406—Processing of video elementary streams, e.g. splicing of video streams or manipulating encoded video stream scene graphs involving management of server-side video buffer
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching
- H04L67/5682—Policies or rules for updating, deleting or replacing the stored data
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/231—Content storage operation, e.g. caching movies for short term storage, replicating data over plural servers, prioritizing data for deletion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/10—Flow control between communication endpoints
- H04W28/14—Flow control between communication endpoints using intermediate storage
Definitions
- the invention is applicable to the technical field of network communication, and in particular relates to a distributed mobile network video cache placement method, system and related equipment.
- a mobile terminal in the cellular network can only obtain the requested content from the macro base station, which is equivalent to forming a traditional cellular network, and the devices in the cellular network form a star network centered on the macro base station structure, which has high requirements on the carrying capacity of the macro base station, and will reach a performance bottleneck as the number of mobile terminal devices in the cellular network increases.
- micro base stations with caching capabilities can be used for local caching to reduce the backhaul link load of cellular base stations and improve the utilization of cached content.
- the carrying capacity of the cellular network structure is increased.
- the small base station cache of the mobile video distribution network needs to face the challenge of storing massive multimedia content in the limited node cache space.
- users have higher requirements for the picture quality and transmission time of mobile video. It is impossible to solve the delay requirement of mobile video only by introducing micro base stations to improve the load capacity of the cellular network.
- Embodiments of the present invention provide a distributed mobile network video cache placement method, system and related equipment, aiming to solve the problem that traditional cellular networks cannot reduce mobile video data transmission delay in a macro-micro base station cellular network structure.
- the embodiment of the present invention provides a distributed mobile network video cache placement method, comprising the following steps:
- a distributed cache model of a cellular network is established, the cellular network includes a macro base station and a plurality of micro base stations, and the cellular network provides video services to micro users within coverage;
- the micro base station obtains a sub-video data from the macro base station through a regret minimum algorithm and places it in a cache;
- the micro base station calculates the popularity of all the sub-video data cached by itself through a probability algorithm, and deletes the sub-video data in the micro base station according to the popularity;
- the micro base station calculates a cache probability according to the popularity, and caches the sub-video data that does not exist in the micro base station in the micro base station.
- the macro base station in the cellular network includes a total video data, and the total video data also includes a plurality of sub-video data, and the micro base station requests the sub-video data from the macro base station and performs Cache placement, the micro base station is an access node facing the micro user.
- the distributed cache model is about joint clustering and cache optimization problems, where:
- any of the micro base stations in B is denoted by b;
- any of the micro-users in M is represented by m;
- the total video data in the macro base station is defined as V, and there are i sub-video data, and any sub-video data in V is represented by v;
- micro base station Define a set of micro users served by the micro base station as U, there are S micro users in U, and any micro user in U is represented by u;
- v b One of the sub-video data cached locally by the micro base station from the macro base station is denoted as v b ;
- the total service delay of the micro-user requesting the sub-video data satisfies the following conditions:
- Q(R,m b ) represents the minimum delay value of the cache optimization problem
- m b represents the relationship between any micro-user m in M and the micro-base station b
- m b,i represents any The micro user m and the micro base station b request the relationship between the ith sub-video data
- c b represents the buffer space capacity of the micro base station b.
- the step of the cellular network providing video services to the micro-users within the coverage specifically includes the following sub-steps:
- the micro user requests the micro base station to obtain the first sub-video data at time t;
- the micro base station requests the macro base station to obtain the first sub-video data corresponding to the request in the total video data, and places it in a cache;
- the micro base station sends the first sub-video data cached by itself to the micro user.
- the step of the micro base station obtaining a sub-video data from the macro base station through the regret minimum algorithm and placing it in the buffer specifically includes the following sub-steps:
- the micro base station calculates the regret minimum value according to the first sub-video data requested by the micro user at the moment t;
- the micro base station selects a piece of second sub-video data from the total video data of the macro base station according to the regret minimum value, and places it in a cache.
- step of calculating the minimum value of regret by the micro base station according to the first sub-video data requested by the micro user at the time t is specifically:
- m b, zb (Nb) (t) is the probability that the micro base station b performs the action Z b (Nb) at the time t;
- the probability calculation formula for the micro base station b to perform the action Z b (Nb) satisfies the following conditions:
- ⁇ b is the Boltzmann regret temperature coefficient controlling the probability calculation formula
- r b + (t) represents a positive regret vector
- each parameter in the above-mentioned conditions also satisfies:
- ⁇ b (t), ⁇ b (t), ⁇ b (t) are learning parameters for calculation, and:
- ⁇ b (t), ⁇ b (t), and ⁇ b (t) satisfy the above constraints.
- the micro base station calculates the popularity of all the sub-video data cached by itself through a probability algorithm, and deletes the sub-video data in the micro base station according to the popularity. Include the following sub-steps:
- the micro base station calculates the popularity for the micro user for all the sub-video data cached by itself;
- the micro base station sorts all the sub-video data according to the popularity.
- step of calculating the popularity for the micro user for all the sub-video data cached by the micro base station is specifically:
- p m [p m, 1 , ..., p m, k , ..., p m, C ]
- p m,k represents the request frequency of the micro user m in the cellular network to the buffer space k of the micro base station
- p m,c represents the request frequency of the micro user m in the cellular network to the micro base station The request frequency of the cache space c content of the base station
- the calculation process of obtaining the popularity with respect to the cache deletion probability of the request frequency satisfies the following conditions:
- ⁇ remove is a cache update coefficient
- "-" is for giving higher probability to the sub-video data that is less frequently requested.
- the micro base station calculates the popularity of all the sub-video data cached by itself through a probability algorithm, and deletes the sub-video data in the micro base station according to the popularity. Include the following sub-steps:
- the micro base station does not perform any operation
- the micro base station deletes the sub-video data with a lower usage frequency according to the size of the cache space.
- the micro base station calculates the caching probability according to the popularity, and the step of caching the sub-video data that does not exist in the micro base station in the micro base station specifically includes the following sub-steps:
- the micro base station calculates a caching probability according to the popularity of all the sub-video data cached by itself at a previous time at the time t;
- the micro base station selects the sub-video data that does not exist in the buffer space from the macro base station according to the buffer probability for buffer placement.
- the process of calculating the caching probability by the micro base station according to the popularity of all the sub-video data cached by the micro base station at the previous time of the time t is specific for:
- m b (t) The result of m b (t) is to calculate the cache probability distribution of the sub-video data requested by the micro user from the micro base station b at the time t according to the regret vector.
- the embodiment of the present invention also provides a distributed mobile network video cache placement system, including the following modules:
- a distributed cache model establishment module configured to establish a distributed cache model of a cellular network, the cellular network includes a macro base station and a plurality of micro base stations, and the cellular network provides video services to micro users within coverage;
- the desired cache placement module is used to control the micro base station to obtain a sub-video data from the macro base station through the regret minimum algorithm and perform cache placement;
- a cache deletion module configured to control the micro base station to calculate the popularity of all the sub-video data cached by itself through a probability algorithm, and delete the sub-video data in the micro base station according to the popularity ;
- a cache update module configured to control the micro base station to calculate a cache probability according to the popularity, and cache the sub-video data that does not exist in the micro base station in the micro base station.
- the embodiment of the present invention also provides a computer device, including: a memory, a processor, and a computer program stored on the memory and operable on the processor, and the processor executes the computer program
- the program implements the steps in the distributed mobile network video cache placing method described in any one of the above-mentioned embodiments.
- an embodiment of the present invention also provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, any one of the above-mentioned embodiments can be implemented. Steps in the method for placing the distributed mobile network video cache.
- the beneficial effects achieved by the present invention are due to the design of a video data cache placement method based on the regret minimum value in the structure of the macro-micro base station, and combined with the probability algorithm to clean up and pre-store the video data in the buffer space of the micro base station, Solve the problem of high transmission delay of mobile video data in cellular network.
- FIG. 1 is a schematic diagram of a scene of a distributed mobile network video cache placement method provided by an embodiment of the present invention
- Fig. 2 is the block flow diagram of the distributed mobile network video cache placement method provided by the embodiment of the present invention.
- Fig. 3 is a sub-flow diagram of step S101 of the distributed mobile network video cache placement method provided by the embodiment of the present invention.
- FIG. 4 is a sub-flow diagram of step S102 of the distributed mobile network video cache placement method provided by an embodiment of the present invention.
- Fig. 5 is a subflow diagram of step S103 of the distributed mobile network video cache placement method provided by the embodiment of the present invention.
- FIG. 6 is a sub-flow diagram of step S104 of the distributed mobile network video cache placement method provided by an embodiment of the present invention.
- Fig. 7 is a structural block diagram of a distributed mobile network video cache placement system provided by an embodiment of the present invention.
- Fig. 8 is a schematic diagram of computer equipment provided by an embodiment of the present invention.
- FIG. 1 is a schematic diagram of a scenario of a distributed mobile network video cache placement method provided by an embodiment of the present invention.
- a cell where a cellular network is located is deployed with a macro base station and multiple The micro base station has a plurality of micro users in the cell, and the micro users obtain other network communication terminals through intelligent terminal equipment to connect to the cellular network composed of the macro base station and the micro base station to connect to the network And obtain the service provided by the cellular network.
- the cellular network provides the video service for the micro user.
- FIG. 2 is a block flow diagram of a distributed mobile network video cache placement method provided by an embodiment of the present invention, which specifically includes the following steps:
- the cellular network includes a macro base station and multiple micro base stations, and the cellular network provides video services to micro users within a coverage area.
- FIG. 3 is a sub-flow diagram of step S101 of the distributed mobile network video cache placement method provided by the embodiment of the present invention, which specifically includes the following sub-steps:
- the micro user requests the micro base station to obtain first sub-video data at time t.
- the cellular network provides video services to the micro users.
- the video data that may be required is defined as total video data, that is, the macro base station has all the video content in the cellular network, and the total video data is composed of multiple sub-video data;
- the micro base station is A structure for assisting the macro base station in realizing data distribution, that is, the micro base station does not have the total video data in the macro base station at the beginning, but the density of the micro base station is high, and the coverage area is comparable to the The distance of the micro users is smaller, so the micro base station is designed to directly provide services to the micro users.
- the micro user When the micro user needs to obtain a certain video data, assuming that it sends a request to the micro base station at the time t, and no video data has been cached in the micro base station, then at the time t, the micro user
- the requested video data is defined as first sub video data.
- the micro base station requests the macro base station to acquire the first sub-video data corresponding to the request in the total video data, and places it in a cache.
- the micro base station acquires the sub-video data corresponding to the first sub-video data from the total video data of the macro base station according to the request content of the first sub-video data, and stores it in its own
- the cache is placed in the cache space.
- the micro base station sends the first sub-video data cached by itself to the micro user.
- the micro base station After the micro base station acquires the first sub-video data, it sends the first sub-video data existing in its buffer space to the micro user, so as to complete the micro user request.
- the distributed cache model of the cellular network is established according to the operation logic of the above steps, wherein the set of the micro base stations in the cell is defined as B, and any micro base station in B is represented by b, Define the set of all the micro-users in the cellular network as M, any of the micro-users in M is represented by m, and the position of the micro-user m in the cell is uniformly randomly distributed, according to the above steps
- the micro base station has a buffer space and has a buffer capability
- the cache object of the micro base station is the sub-video data included in the total video data in the macro base station
- the sub-video data is defined
- the quantity is N
- the total video data is set V
- the total video data can be expressed as:
- V ⁇ v 1 , v 2 , . . . , v N ⁇
- v represents any sub-video data in the total video data V.
- a subset of the micro users served by the micro base station is defined as a micro user set, denoted as U, wherein the number of the micro users is S, then the micro user set U can be expressed as:
- u represents any one of the micro-users in the micro-user set.
- the capacity of the buffer space of all described micro base stations as C
- the capacity C of the buffer space of all described micro base stations can be expressed as:
- the sub-video data v b obtained by the micro base station b from the macro base station and cached locally is from V, that is:
- the probability that the content of one of the sub-video data v i requested by the micro-user comes from V follows a Poisson distribution ⁇ u,v with a mean value.
- the main goal of the micro base station in the embodiment of the present invention is to find the best caching strategy that minimizes the total service delay of the micro user requesting the sub-video data, wherein the total service delay can be expressed as follows Condition means:
- the micro-user sends a request to the micro-base station closest to it, and each micro-base station provides services for a plurality of micro-users that are closer to the distance. Therefore, in order to minimize the total service delay, the micro-base station, According to the content of the sub-video data requested by the micro-user at the time t, the content of the sub-video data that may be requested next by the micro-user needs to be calculated and cached.
- R as the correlation vector between the micro-user and the micro-base station, so that the following expression holds true:
- r u represents the micro base station b that provides services for the micro user u i .
- the distributed cache model of the cellular network on the joint clustering and cache optimization problem is as follows:
- the micro base station obtains a sub-video data from the macro base station through a regret minimum algorithm and places it in a cache.
- FIG. 4 is a sub-flow diagram of step S102 of the distributed mobile network video cache placement method provided by the embodiment of the present invention, which specifically includes the following sub-steps:
- the micro base station calculates the minimum regret value according to the first sub-video data requested by the micro user at a moment preceding the moment t.
- the micro base station b will select an action Z b (Nb) from the operation space Z b , where the operation space Z b is:
- the action Z b (nb) is a binary value indicating whether to cache the sub-video data v i .
- N b is the total number of operations, which is equal to the total number of the sub-video data in the total video data V.
- the micro base station selects an action Z b (Nb) according to an action probability distribution m b (t ) , and the action probability distribution m b (t) is:
- m b,zb (Nb) (t) is the probability that the micro base station b performs the action Z b (Nb) at the time t.
- Each of the micro base stations will update its action according to the vector m b , wherein each of the micro base stations will choose a probability distribution to minimize the regret of whether the micro base station caches a certain sub-video data vector
- the regret vector Specifically:
- the small base station b will also estimate the regret vector Simultaneously estimate the utility vector for The utility vector for:
- the Femto base station selects a piece of second sub-video data from the total video data of the Macro base station according to the regret minimum value, and places it in a cache.
- a probability distribution based on Gibbs-Sampling (Gibbs sampling) is used to capture the action Z b (Nb) , then, the micro base station b executes the action Z b (Nb ) probability calculation formula can be expressed as:
- ⁇ b is the Boltzmann regret temperature coefficient controlling the probability calculation formula
- r b + (t) represents a positive regret vector
- the positive regret vector r b + ( t) is:
- the micro base station b estimates the regret vector using a calculation process based on the regret minimum value
- the utility vector and the specific process of the action probability distribution m b (t) satisfies the following conditions:
- ⁇ b (t), ⁇ b (t), ⁇ b (t) are learning parameters for calculation, ⁇ b (t), ⁇ b ( t), , ⁇ b (t) satisfy the following constraints:
- the probability calculation formula can be converged to an ⁇ rough correlation balance.
- the micro base station caches one sub-video data from the macro base station as the cached second sub-video data at the time t.
- the micro base station calculates popularity for all the sub-video data cached by itself through a probability algorithm, and deletes the sub-video data in the micro base station according to the popularity.
- FIG. 5 is a sub-flow diagram of step S103 of the distributed mobile network video cache placement method provided by the embodiment of the present invention, which specifically includes the following sub-steps:
- the micro base station calculates the popularity for the micro user for all the sub-video data cached by the micro base station.
- each of the micro base stations will obtain the popularity of all the sub-video data in the current service domain, wherein each of the micro base stations is based on the popularity of each of the sub-video data in its service domain.
- the request frequency P m of video data is used to construct popularity, and the request frequency is different for each micro base station, and the request frequency P m is defined as:
- p m [p m, 1 , ..., p m, k , ..., p m, C ]
- p m,k represents the request frequency of the micro user m in the cellular network to the buffer space k of the micro base station
- p m,c represents the request frequency of the micro user m in the cellular network to the micro base station The request frequency of the cache space c content of the base station
- ⁇ remove is a cache update coefficient, and "-" is to give higher probability to the sub-video data that is less frequently requested.
- the use of the Gibbs-Sampling probability distribution allows updating the cache removal strategy using the ⁇ remove parameter.
- ⁇ remove equal to 0
- the execution removes all the sub-video data with the same probability, while a higher ⁇ remove value means that it will be deleted at a higher
- the sub-video data whose request frequency is low is probabilistically deleted.
- the Femtocell sorts all the sub-video data according to the popularity.
- Each micro base station sorts all the sub-video data in its cache space according to the order of the cache deletion probability according to the calculated popularity and the calculation result of the cache deletion probability, for the next step.
- the micro base station After the micro base station calculates the cache deletion probability corresponding to the sub-video data in each cache, it checks the size of its own cache space.
- the total video data of the macro base station is composed of N pieces of sub-video data, and the buffer spaces occupied by the N pieces of sub-video data are equal in size. If the cache space of the micro base station is still sufficient to store a sub-video data that has not been cached, the micro base station will not perform the delete operation.
- the micro base station deletes the sub-video data that is used less frequently according to the size of the cache space.
- the micro base station If the result obtained by the micro base station inspection shows that the buffer space of the micro base station is not enough to cache a new sub-video data, then the micro base station, according to the sorting result of the cache deletion probability, from the Selecting the sub-video data with the highest cache deletion probability in the cache space to delete.
- the micro base station calculates a caching probability according to the popularity, and caches the sub-video data that does not exist in the micro base station in the micro base station.
- FIG. 6 is a sub-flow diagram of step S104 of the distributed mobile network video cache placement method provided by the embodiment of the present invention, which specifically includes the following sub-steps:
- the micro base station calculates a caching probability according to the popularity of all the sub-video data cached by the micro base station at a time immediately preceding the time t.
- the micro base station b will execute the action Z b ( Nb) , and calculate the micro user's Acquire the actual delay of the sub-video data, so as to obtain the actual utility e(t-1) of the micro base station.
- the micro base station b uses iterative update to calculate the utility vector at the time t and the regret vector And calculate the cache probability distribution m b (t) of the sub-video data requested by the micro-user at the moment t according to the regret vector, and the cache probability distribution m b (t) satisfies the following conditions:
- the micro base station selects the sub-video data that does not exist in the buffer space from the macro base station according to the buffer probability, and places it in buffer.
- the micro base station b selects, from the total video data of the macro base station according to the calculated buffer probability distribution m b (t), a video data that does not exist in the buffer space of the micro base station b.
- the sub-video data is cached.
- a video data cache placement strategy based on the regret minimum value is designed in the macro-micro base station structure, and a probabilistic algorithm is combined with the video data in the buffer space of the micro base station.
- the data is cleaned and pre-stored, which solves the problem of high transmission delay of mobile video data in the cellular network.
- the embodiment of the present invention also provides a distributed mobile network video cache placement system.
- FIG. 7 is a structural block diagram of the distributed mobile network video cache placement system provided by the embodiment of the present invention.
- the cache placement system 200 includes a distributed cache model building module 201, a desired cache placement module 202, a cache deletion module 203, and a cache update module 204, wherein:
- the distributed cache model establishment module 201 is used to establish a distributed cache model of a cellular network, the cellular network includes a macro base station and a plurality of micro base stations, and the cellular network provides video services to micro users within the coverage area ;
- the desired cache placement module 202 is configured to control the micro base station to obtain a sub-video data from the macro base station through the regret minimum algorithm and perform cache placement;
- the cache deletion module 203 is configured to control the micro base station to calculate the popularity of all the sub-video data cached by itself through a probability algorithm, and to calculate the popularity of the sub-video data in the micro base station according to the popularity. data to be deleted;
- the cache update module 204 is configured to control the micro base station to calculate a cache probability according to the popularity, and cache the sub-video data that does not exist in the micro base station in the micro base station.
- the distributed mobile network video cache placement system 700 provided in the embodiment of the present invention can implement the steps in the distributed mobile network video cache placement method in the above-mentioned embodiment, and can achieve the same technical effect, refer to the description in the above-mentioned embodiment , which will not be repeated here.
- the embodiment of the present invention also provides a computer device, please refer to FIG. 8, which is a schematic diagram of the computer device provided by the embodiment of the present invention.
- the computer device 300 includes: a memory 302, a processor 301, and and a computer program that can run on the processor 301.
- the processor 301 invokes the computer program stored in the memory 302 to execute the steps in the distributed mobile network video cache placement method provided by the embodiment of the present invention. Please refer to FIG. 1, specifically including:
- the cellular network includes a macro base station and multiple micro base stations, and the cellular network provides video services to micro users within coverage;
- the macro base station in the cellular network includes a total video data, and the total video data also includes a plurality of sub-video data, and the micro base station requests the sub-video data from the macro base station and performs Cache placement, the micro base station is an access node facing the micro user.
- the step of the cellular network providing video services to the micro-users within the coverage specifically includes the following sub-steps:
- the micro user requests the micro base station to obtain the first sub-video data at time t;
- the micro base station requests the macro base station to obtain the first sub-video data corresponding to the request in the total video data, and places it in a cache;
- the micro base station sends the first sub-video data cached by itself to the micro user.
- the micro base station obtains a sub-video data from the macro base station through a regret minimum algorithm and places it in a cache;
- the step of the micro base station obtaining a sub-video data from the macro base station through the regret minimum algorithm and placing it in the buffer specifically includes the following sub-steps:
- the micro base station calculates the regret minimum value according to the first sub-video data requested by the micro user at the moment t;
- the micro base station selects a piece of second sub-video data from the total video data of the macro base station according to the regret minimum value, and places it in a cache.
- the micro base station calculates popularity for all the sub-video data cached by itself through a probability algorithm, and deletes the sub-video data in the micro base station according to the popularity;
- the micro base station calculates the popularity of all the sub-video data cached by itself through a probability algorithm, and deletes the sub-video data in the micro base station according to the popularity. Include the following sub-steps:
- the micro base station calculates the popularity for the micro user for all the sub-video data cached by itself;
- the micro base station sorts all the sub-video data according to the popularity.
- the micro base station calculates the popularity of all the sub-video data cached by itself through a probability algorithm, and deletes the sub-video data in the micro base station according to the popularity. Include the following sub-steps:
- the micro base station does not perform any operation
- the micro base station deletes the sub-video data with a lower usage frequency according to the size of the cache space.
- the micro base station calculates a caching probability according to the popularity, and caches the sub-video data that does not exist in the micro base station in the micro base station.
- the micro base station calculates the caching probability according to the popularity, and the step of caching the sub-video data that does not exist in the micro base station in the micro base station specifically includes the following sub-steps:
- the micro base station calculates a caching probability according to the popularity of all the sub-video data cached by itself at a previous time at the time t;
- the micro base station selects the sub-video data that does not exist in the buffer space from the macro base station according to the buffer probability for buffer placement.
- the computer equipment provided by the embodiment of the present invention can realize the steps in the distributed mobile network video cache placement method in the above-mentioned embodiment, and can achieve the same technical effect, refer to the description in the above-mentioned embodiment, and will not repeat them here.
- the embodiment of the present invention also provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the distributed mobile network video cache placement method provided by the embodiment of the present invention is implemented.
- the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM for short).
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Multimedia (AREA)
- Mobile Radio Communication Systems (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
本发明适用于网络通信技术领域，提供了一种分布式移动网络视频缓存放置方法、系统及相关设备，所述方法包括以下步骤：建立一个蜂窝网络的分布式缓存模型，蜂窝网络包括一个宏基站与多个微基站，蜂窝网络对覆盖范围内的微用户提供视频服务；微基站通过后悔最小值算法从宏基站中获取一个子视频数据并进行缓存放置；微基站通过概率算法对自己缓存的所有子视频数据计算受欢迎度，并根据受欢迎度对所述微基站中的子视频数据进行删除；所述微基站根据受欢迎度计算缓存概率，并将一个微基站中不存在的子视频数据在微基站中进行缓存。本发明解决了传统蜂窝网络无法在宏-微基站的蜂窝网络结构中降低移动视频数据传输延时的问题。
Description
本发明适用于网络通信技术领域，尤其涉及一种分布式移动网络视频缓存放置方法、系统及相关设备。
在传统蜂窝网络中，蜂窝网络中的一个移动终端只能从宏基站获取请求的内容，相当于在构成传统的蜂窝网络时，蜂窝网络中的设备就形成了以宏基站为中心的星型网络结构，这对宏基站的承载能力有很高的要求，并随着蜂窝网络内移动终端设备数量的增长而达到性能瓶颈。
而在无线异构蜂窝网络中，通过引入具备内容缓存能力的小区基站，可以利用具有缓存能力的微基站进行本地缓存来减小蜂窝基站的回程链路负载，提高缓存内容利用率，这在一定程度上增加了蜂窝网络结构的承载能力。然而随着互联网的发展，数据的爆发式增长，使得移动视频分发网络的微基站缓存中，需要面对有限的节点缓存空间存储海量的多媒体内容的挑战，这主要是因为移动视频已经成为互联网中的主要数据类型，同时，用户对于移动视频的画面质量和传输时间的要求变得更高，若仅仅通过引入微基站来提高蜂窝网络的负载能力，是无法解决移动视频的延迟要求的。
发明内容
本发明实施例提供一种分布式移动网络视频缓存放置方法、系统及相关设备，旨在解决传统蜂窝网络无法在宏-微基站的蜂窝网络结构中降低移动视频数据传输延时的问题。
本发明实施例提供了一种分布式移动网络视频缓存放置方法，包括以下步骤：
建立一个蜂窝网络的分布式缓存模型，所述蜂窝网络包括一个宏基站与多个微基站，所述蜂窝网络对覆盖范围内的微用户提供视频服务；
所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置；
所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除；
所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存。
更进一步地，所述蜂窝网络中的所述宏基站包括一个总视频数据，所述总视频数据还包括多个子视频数据，所述微基站从所述宏基站中请求所述子视频数据并进行缓存放置，所述微基站为面向所述微用户的访问节点。
更进一步地，所述分布式缓存模型关于联合集群和缓存优化问题，其中：
定义所述蜂窝网络中的所述微基站的集合为B，B中任一所述微基站用b表示；
定义所述蜂窝网络中所有所述微用户的集合为M，M中任一所述微用户用m表示；
定义所述宏基站中的所述总视频数据为V，所述子视频数据有i个，V中任一所述子视频数据用v表示；
定义一个所述微基站服务的微用户集合为U，U中所述微用户有S个，U中任一所述微用户用u表示；
定义所有的所述微基站的缓存空间的容量总和为C，C中任一所述微基站的缓存空间为c；
所述微基站从所述宏基站缓存到本地的一个所述子视频数据记为v
b；
根据以上，所述微用户请求所述子视频数据的总服务延迟满足如下条件：
其中，
代表关于U中所述微用户u从所述微基站b请求所述子视频数据v的延迟，u
i表示所述微用户u所请求的第i个所述子视频数据。
定义所述微用户与所诉微基站之间的关联向量为R，r
u∈R，r
u代表为所述微用户u
i提供服务的所述微基站b，根据以上，关于联合集群和缓存优化问题的所述分布式缓存模型满足如下表达式：
其中，Q(R,m
b)代表缓存优化问题的最小延迟值，m
b代表M中任一所述微用户m与所述微基站b之间的关系，m
b,i代表M中任一所述微用户m与所述微基站b之间请求第i个所述子视频数据之间的关系，c
b表示所述微基站b的缓存空间容量。
更进一步地，所述蜂窝网络对覆盖范围内的微用户提供视频服务的步骤具体包括以下子步骤：
所述微用户在t时刻时向所述微基站请求获取第一子视频数据；
所述微基站向所述宏基站请求获取所述总视频数据中对应所述请求的所述第一子视频数据，并将其缓存放置；
所述微基站将自己缓存放置的所述第一子视频数据发送给所述微用户。
更进一步地，所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置的步骤具体包括以下子步骤：
在所述t时刻的下一时刻时，所述微基站根据所述t时刻时所述微用户请求的所述第一子视频数据计算所述后悔最小值；
所述微基站从所述宏基站的所述总视频数据中根据所述后悔最小值选择一个第二子视频数据进行缓存放置。
更进一步地，所述微基站根据所述t时刻时所述微用户请求的所述第一子视频数据计算所述后悔最小值的步骤具体为：
定义后悔向量为
以及与所述后悔向量对应的效用向量
定义所述微基站在所述t时刻从所述总视频数据中获取一个所述子视频数据的动作为Z
b
(nb)，其中，N
b是操作总数，N
b等于所述总视频数据V中的所述子视频数据的总数；
定义m
b,zb
(Nb)(t)是所述微基站b在所述t时刻执行所述动作Z
b
(Nb)的概率；
根据以上，所述微基站b执行所述动作Z
b
(Nb)的概率计算公式满足如下条件：
其中，
为所述概率计算公式计算出的概率结果，β
b是控制所述概率计算公式的玻尔兹曼后悔温度系数，r
b
+(t)表示正后悔向量，上述条件中的各个参数还满足：
其中，
是指在所述t时刻的前一时刻的瞬时观测效用函数，δ
b(t)、ξ
b(t)、η
b(t)是计算用的学习参数，且：
δ
b(t)、ξ
b(t)、η
b(t)满足如上约束条件。
更进一步地，所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删 除的步骤具体包括以下子步骤：
在每一个所述t时刻时，所述微基站对自己缓存的所有所述子视频数据计算对于所述微用户来说的所述受欢迎度；
所述微基站将所有所述子视频数据按照所述受欢迎度进行排序。
更进一步地，所述微基站对自己缓存的所有所述子视频数据计算对于所述微用户来说的所述受欢迎度的步骤具体为：
定义每一个所述微基站基于其服务域中每一个所述子视频数据的请求频率为P
m，其中任一所述微基站中的所述子视频数据的请求频率满足：
p
m＝[p
m，1，...，p
m，k，...，p
m，C]
其中，p
m,k表示所述蜂窝网络中所述微用户m向所述微基站的缓存空间k内容的请求频率，p
m,c表示所述蜂窝网络中所述微用户m向所述微基站的缓存空间c内容的请求频率；
根据以上，关于所述请求频率的缓存删除概率以得到所述受欢迎度的计算过程满足如下条件：
其中，β
remove是缓存更新系数，“-”是为了赋予被请求频率较低的所述子视频数据更高的概率。
更进一步地，所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除的步骤还包括以下子步骤：
若所述微基站的缓存空间还足够存放另一个所述子视频数据，那么所述微基站不进行任何操作；
若所述微基站的所述缓存空间不足够存放另一个所述子视频数据，所述微基站根据所述缓存空间的大小，将所述使用频率较低的所述子视频数据删除。
更进一步地，所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存的步骤具体包括以 下子步骤：
在每一个所述t时刻时，所述微基站根据所述t时刻的上一时刻中自己缓存的所有所述子视频数据的所述受欢迎度计算缓存概率；
所述微基站根据所述缓存概率从所述宏基站中选择所述缓存空间中不存在的所述子视频数据进行缓存放置。
更进一步地，所述在每一个所述t时刻时，所述微基站根据所述t时刻的上一时刻中自己缓存的所有所述子视频数据的所述受欢迎度计算缓存概率的过程具体为：
定义所述缓存概率分布为m
b(t)，且满足如下条件：
m
b(t)的结果为根据所述后悔向量计算所述t时刻时所述微用户向所述微基站b请求的所述子视频数据的缓存概率分布。
第二方面，本发明实施例还提供了一种分布式移动网络视频缓存放置系统，包括以下模块：
分布式缓存模型建立模块，用于建立一个蜂窝网络的分布式缓存模型，所述蜂窝网络包括一个宏基站与多个微基站，所述蜂窝网络对覆盖范围内的微用户提供视频服务；
期望缓存放置模块，用于控制所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置；
缓存删除模块，用于控制所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除；
缓存更新模块，用于控制所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存。
第三方面，本发明实施例还提供了一种计算机设备，包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，所述处理器 执行所述计算机程序时实现如上述实施例中任一项所述的分布式移动网络视频缓存放置方法中的步骤。
第四方面，本发明实施例还提供了一种计算机可读存储介质，所述计算机可读存储介质上存储有计算机程序，所述计算机程序被处理器执行时实现如上述实施例中任一项所述的分布式移动网络视频缓存放置方法中的步骤。
本发明所达到的有益效果，由于在宏-微基站的结构中设计了基于后悔最小值的视频数据缓存放置方法，并结合概率算法来对微基站的缓存空间中的视频数据进行清理和预存，解决了在蜂窝网络中移动视频数据的传输延时高的问题。
图1是本发明实施例提供的分布式移动网络视频缓存放置方法的场景示意图；
图2是本发明实施例提供的分布式移动网络视频缓存放置方法的流程框图；
图3是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S101的子流程框图；
图4是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S102的子流程框图；
图5是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S103的子流程框图；
图6是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S104的子流程框图；
图7是本发明实施例提供的分布式移动网络视频缓存放置系统的结构框图；
图8是本发明实施例提供的计算机设备示意图。
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
请参照图1，图1是本发明实施例提供的分布式移动网络视频缓存放置方法的场景示意图，在本发明实施例中，一个蜂窝网络所在的一个小区的内部部署有一个宏基站和多个微基站，所述小区中拥有对多个微用户，所述微用户通过智能终设备获取其他网络通信终端连接由所述宏基站和所述微基站组成的所述蜂窝网络中，以连接至网络并获取所述蜂窝网络提供的服务，在本发明实施例中，所述蜂窝网络提供给所述微用户视频服务。
请参照图2，图2是本发明实施例提供的分布式移动网络视频缓存放置方法的流程框图，具体包括以下步骤：
S101、建立一个蜂窝网络的分布式缓存模型，所述蜂窝网络包括一个宏基站与多个微基站，所述蜂窝网络对覆盖范围内的微用户提供视频服务。
请参照图3，图3是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S101的子流程框图，具体包括以下子步骤：
S1011、所述微用户在t时刻时向所述微基站请求获取第一子视频数据。
在本发明实施例中，所述蜂窝网络对所述微用户提供视频服务，具体的，考虑到所述蜂窝网络中的视频内容的多样性，假设所述宏基站中包括了所有所述微用户可能需要的视频数据，将其定义为总视频数据，即所述宏基站中拥有所述蜂窝网络中的所有视频内容，并且，所述总视频数据由多个子视频数据组成；所述微基站是用于辅助所述宏基站实现数据分发的结构，即所述微基站一开始并不具有所述宏基站中的所述总视频数据，但是所述微基站的密度大，覆盖范围内与所述微用户的距离更小，因此所述微基站被设计为用于直接给所述微用户提供服务。所述微用户需要获取某一个视频数据时，假设其在所述t时刻向所述微基站发出请求，且所述微基站中尚未缓存任何视频数据，那么在所 述t时刻时所述微用户请求的视频数据定义为第一子视频数据。
S1012、所述微基站向所述宏基站请求获取所述总视频数据中对应所述请求的所述第一子视频数据，并将其缓存放置。
所述微基站根据所述第一子视频数据的请求内容，从所述宏基站的所述总视频数据中获取与所述第一子视频数据对应的所述子视频数据，并将其在自己的缓存空间中缓存放置。
S1013、所述微基站将自己缓存放置的所述第一子视频数据发送给所述微用户。
所述微基站获取到所述第一子视频数据后，将存在于自己的缓存空间内的所述第一子视频数据发送给所述微用户，以完成本次在所述t时刻的所述微用户的请求。
本发明实施例根据以上步骤的运行逻辑建立所述蜂窝网络的分布式缓存模型，其中，定义所述小区中的所述微基站的集合为B，B中任一所述微基站用b表示，定义所述蜂窝网络中所有所述微用户的集合为M，M中任一所述微用户用m表示，所述微用户m在所述小区中的位置为均匀的随机分布，根据以上步骤的运行逻辑，所述微基站拥有缓存空间，并具有缓存能力，所述微基站的缓存对象为所述宏基站中的所述总视频数据所包含的所述子视频数据，定义所述子视频数据的数量为N，所述总视频数据为集合V，那么所述总视频数据可以表示为：
V＝{v
1，v
2，...，v
N}
其中，v表示所述总视频数据V中的任一所述子视频数据。
定义一个所述微基站服务的所述微用户的一个子集为微用户集合，记为U，其中所述微用户的数量为S，那么所述微用户集合U可以表示为：
U＝{u
1，u
2，...，u
S}
其中，u表示所述微用户集合中的任一所述微用户。
定义所有的所述微基站的缓存空间的容量为C，那么所有的所述微基站的 缓存空间的容量C可以表示为：
C＝{c
1,c
2,...,c
M}
所述微基站b从所述宏基站获取并缓存到本地的所述子视频数据v
b来自V，即：
v
b∈V
所述微用户请求的某一个所述子视频数据v
i的内容来自V的概率遵循具有平均值的柏松分布λ
u，v。
综合以上，在本发明实施例中所述微基站的主要目标是寻找使所述微用户请求所述子视频数据的总服务延迟最小的最佳缓存策略，其中，所述总服务延迟可以用以下条件表示：
所述微用户向其距离最近的所述微基站发出请求，而每一个所述微基站会为距离较近的多个微用户提供服务，因此所述微基站为了使所述总服务延迟最小，根据所述t时刻时所述微用户请求的所述子视频数据的内容，需要计算出所述微用户可能的下一个请求的所述子视频数据的内容并进行缓存。定义R为所述微用户与所述微基站之间的关联向量，使得如下表达式成立：
r
u∈R
其中，r
u代表为所述微用户u
i提供服务的所述微基站b。
根据上述定义，关于联合集群和缓存优化问题的所述蜂窝网络的分布式缓存模型如下：
S102、所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置。
请参照图4，图4是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S102的子流程框图，具体包括以下子步骤：
S1021、在所述t时刻，所述微基站根据所述t时刻的上一时刻时所述微用户请求的所述第一子视频数据计算所述后悔最小值。
在所述t时刻时，所述微基站b会从操作空间Z
b中选择一个动作Z
b
(Nb)，其中，所述操作空间Z
b为：
所述动作Z
b
(nb)为一个是否缓存对所述子视频数据v
i进行缓存操作的二进制值。其中，N
b是操作总数，其等于所述总视频数据V中的所述子视频数据的总数。
所述微基站根据一个动作概率分布m
b(t)选择一个所述动作Z
b
(Nb)，所述动作概率分布m
b(t)为：
其中，m
b,zb
(Nb)(t)是所述微基站b在所述t时刻执行所述动作Z
b
(Nb)的概率。
每一个所述微基站都会根据向量m
b来更新自己的动作，其中，每一个所述微基站都会选择一种概率分布，以最小化所述微基站是否缓存某一个所述子视频数据的后悔向量
所述后悔向量
具体为：
其中，
表示所述微基站b在所述t时刻之前的所有时刻{1，2，…，t-1}中都执行了相同的所述动作Z
b
(Nb)。
另外，所述微基站b还会在估计所述后悔向量
的同时估计效用向量
所述效用向量
为：
S1022、所述微基站从所述宏基站的所述总视频数据中根据所述后悔最小值选择一个第二子视频数据进行缓存放置。
根据上述步骤S1021中的定义，采用基于Gibbs-Sampling(吉布斯采样) 的概率分布来进行所述动作Z
b
(Nb)的捕获，那么，所述微基站b执行所述动作Z
b
(Nb)的概率计算公式可以表示为：
其中，
为所述概率计算公式计算出的概率结果，β
b是控制所述概率计算公式的玻尔兹曼后悔温度系数，r
b
+(t)表示正后悔向量，所述正后悔向量r
b
+(t)为：
在所述t时刻，所述微基站b使用基于后悔最小值的计算过程估算所述后悔向量
所述效用向量
和所述动作概率分布m
b(t)的具体过程满足如下条件：
其中，
是指在所述t时刻的前一时刻的瞬时观测效用函数，δ
b(t)、ξ
b(t)、η
b(t)是计算用的学习参数，δ
b(t)、ξ
b(t)、、η
b(t)满足如下约束条件：
根据以上约束条件，能够使所述概率计算公式收敛到一个ε粗相关平衡。
根据所述概率计算公式的计算结果，所述微基站从所述宏基站中缓存一个所述子视频数据作为所述t时刻的所缓存的所述第二子视频数据。
S103、所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除。
请参照图5，图5是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S103的子流程框图，具体包括以下子步骤：
S1031、在所述t时刻时，所述微基站对自己缓存的所有所述子视频数据计算对于所述微用户来说的所述受欢迎度。
在所述t时刻时，每一个所述微基站都会获取当前服务域中所有所述子视频数据的所述受欢迎度，其中，每一个所述微基站基于其服务域中每一个所述子视频数据的请求频率P
m来构建受欢迎度，所述请求频率对于每一个所述微基站来说都是不同的，定义所述请求频率P
m为：
p
m＝[p
m，1，...，p
m，k，...，p
m，C]
其中，p
m,k表示所述蜂窝网络中所述微用户m向所述微基站的缓存空间k内容的请求频率，p
m,c表示所述蜂窝网络中所述微用户m向所述微基站的缓存空间c内容的请求频率；
那么根据所述请求频率的值基于Gibbs-Sampling的概率分布计算缓存删除概率的条件如下：
其中，β
remove是缓存更新系数，“-”是为了赋予被请求频率较低的所述子视频数据更高的概率。Gibbs-Sampling概率分布的使用允许使用β
remove参数更新缓存删除策略，当使用β
remove等于0时，执行删除所有所述子视频数据的概率相同，而较高的β
remove值表示将以较高的概率删除请求频率较低的所述子视频数据。
S1032、所述微基站将所有所述子视频数据按照所述受欢迎度进行排序。
每一个所述微基站根据计算得到的所述受欢迎度，结合所述缓存删除概率的计算结果，对其缓存空间内的所有所述子视频数据按照所述缓存删除概率的大小顺序进行排序，以便进行下一步骤。
S1033a、若所述微基站的缓存空间还足够存放另一个所述子视频数据，那么所述微基站不进行任何操作。
所述微基站在计算出每一个缓存里的所述子视频数据对应的所述缓存删除概率之后，会检查自己的缓存空间的大小。在本发明实施例中，所述宏基站的所述总视频数据由N个所述子视频数据组成，且N个所述子视频数据占用的缓存空间的大小相等。若所述微基站的所述缓存空间还足够存放一个未缓存过的所述子视频数据，那么所述微基站不会执行删除操作。
S1033b、若所述微基站的所述缓存空间不足够存放另一个所述子视频数据，所述微基站根据所述缓存空间的大小，将所述使用频率较低的所述子视频数据删除。
若所述微基站检查得到的结果表明所述微基站的所述缓存空间并不足以缓存一个新的所述子视频数据，那么所述微基站根据所述缓存删除概率的排序结果，从所述缓存空间中选择所述缓存删除概率最大的一个所述子视频数据进行删除。
S104、所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存。
请参照图6，图6是本发明实施例提供的分布式移动网络视频缓存放置方法的步骤S104的子流程框图，具体包括以下子步骤：
S1041、在所述t时刻时，所述微基站根据所述t时刻的上一时刻中自己缓存的所有所述子视频数据的所述受欢迎度计算缓存概率。
在所述t时刻时，所述微基站b会执行一个所述动作Z
b
(Nb)，并根据所述t时刻的上一时刻实际发生的所述动作Z
b
(Nb)计算所述微用户获取所述子视频数据的实际延迟，从而得到所述微基站的实际效用e(t-1)。
根据所述实际效用，所述微基站b使用迭代更新计算所述t时刻的所述效用向量
和所述后悔向量
并根据所述后悔向量计算所述t时刻时所述微用户请求的所述子视频数据的缓存概率分布m
b(t)，所述缓存概率分布m
b(t)满足如下条件：
S1042、所述微基站根据所述缓存概率从所述宏基站中选择所述缓存空间中不存在的所述子视频数据进行缓存放置。
所述微基站b根据计算得到的所述缓存概率分布m
b(t)从所述宏基站的所述总视频数据中，选择一个所述微基站b的所述缓存空间中不存在的一个所述子视频数据进行缓存。
本发明实施例所述的分布式移动网络视频缓存放置方法在宏-微基站的结构中设计了基于后悔最小值的视频数据缓存放置策略，并结合概率算法来对微基站的缓存空间中的视频数据进行清理和预存，解决了在蜂窝网络中移动视频数据的传输延时高的问题。
本发明实施例还提供一种分布式移动网络视频缓存放置系统，请参照图7，图7是本发明实施例提供的分布式移动网络视频缓存放置系统的结构框图，所述分布式移动网络视频缓存放置系统200包括分布式缓存模型建立模块201、期望缓存放置模块202、缓存删除模块203、缓存更新模块204，其中：
所述分布式缓存模型建立模块201，用于建立一个蜂窝网络的分布式缓存模型，所述蜂窝网络包括一个宏基站与多个微基站，所述蜂窝网络对覆盖范围内的微用户提供视频服务；
所述期望缓存放置模块202，用于控制所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置；
所述缓存删除模块203，用于控制所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除；
所述缓存更新模块204，用于控制所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存。
本发明实施例提供的分布式移动网络视频缓存放置系统700能够实现如上 述实施例中的分布式移动网络视频缓存放置方法中的步骤，且能实现同样的技术效果，参上述实施例中的描述，此处不再赘述。
本发明实施例还提供一种计算机设备，请参见图8，图8是本发明实施例提供的计算机设备示意图，所述计算机设备300包括：存储器302、处理器301及存储在所述存储器302上并可在所述处理器301上运行的计算机程序。
处理器301调用存储器302存储的计算机程序，执行本发明实施例提供的分布式移动网络视频缓存放置方法中的步骤，请结合图1，具体包括：
S101、建立一个蜂窝网络的分布式缓存模型，所述蜂窝网络包括一个宏基站与多个微基站，所述蜂窝网络对覆盖范围内的微用户提供视频服务；
更进一步地，所述蜂窝网络中的所述宏基站包括一个总视频数据，所述总视频数据还包括多个子视频数据，所述微基站从所述宏基站中请求所述子视频数据并进行缓存放置，所述微基站为面向所述微用户的访问节点。
更进一步地，所述蜂窝网络对覆盖范围内的微用户提供视频服务的步骤具体包括以下子步骤：
所述微用户在t时刻时向所述微基站请求获取第一子视频数据；
所述微基站向所述宏基站请求获取所述总视频数据中对应所述请求的所述第一子视频数据，并将其缓存放置；
所述微基站将自己缓存放置的所述第一子视频数据发送给所述微用户。
S102、所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置；
更进一步地，所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置的步骤具体包括以下子步骤：
在所述t时刻的下一时刻时，所述微基站根据所述t时刻时所述微用户请求的所述第一子视频数据计算所述后悔最小值；
所述微基站从所述宏基站的所述总视频数据中根据所述后悔最小值选择一个第二子视频数据进行缓存放置。
S103、所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除；
更进一步地，所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除的步骤具体包括以下子步骤：
在每一个所述t时刻时，所述微基站对自己缓存的所有所述子视频数据计算对于所述微用户来说的所述受欢迎度；
所述微基站将所有所述子视频数据按照所述受欢迎度进行排序。
更进一步地，所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除的步骤还包括以下子步骤：
若所述微基站的缓存空间还足够存放另一个所述子视频数据，那么所述微基站不进行任何操作；
若所述微基站的所述缓存空间不足够存放另一个所述子视频数据，所述微基站根据所述缓存空间的大小，将所述使用频率较低的所述子视频数据删除。
S104、所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存。
更进一步地，所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存的步骤具体包括以下子步骤：
在每一个所述t时刻时，所述微基站根据所述t时刻的上一时刻中自己缓存的所有所述子视频数据的所述受欢迎度计算缓存概率；
所述微基站根据所述缓存概率从所述宏基站中选择所述缓存空间中不存在的所述子视频数据进行缓存放置。
本发明实施例提供的计算机设备能够实现如上述实施例中的分布式移动网络视频缓存放置方法中的步骤，且能实现同样的技术效果，参上述实施例中的 描述，此处不再赘述。
本发明实施例还提供一种计算机可读存储介质，所述计算机可读存储介质上存储有计算机程序，该计算机程序被处理器执行时实现本发明实施例提供的分布式移动网络视频缓存放置方法中的各个过程及步骤，且能实现相同的技术效果，为避免重复，这里不再赘述。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的程序可存储于一计算机可读取存储介质中，该程序在执行时，可包括如上述各方法的实施例的流程。其中，所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory，ROM)或随机存取存储器(Random Access Memory，简称RAM)等。
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者装置不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者装置所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、方法、物品或者装置中还存在另外的相同要素。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端(可以是手机，计算机，服务器，空调器，或者网络设备等)执行本发明各个实施例所述的方法。
上面结合附图对本发明的实施例进行了描述，所揭露的仅为本发明较佳实施例而已，但是本发明并不局限于上述的具体实施方式，上述的具体实施方式仅仅是示意性的，而不是限制性的，本领域的普通技术人员在本发明的启示下， 在不脱离本发明宗旨和权利要求所保护的范围情况下，还可做出很多形式用等同变化，均属于本发明的保护之内。
Claims (14)
- 一种分布式移动网络视频缓存放置方法，其特征在于，包括以下步骤：建立一个蜂窝网络的分布式缓存模型，所述蜂窝网络包括一个宏基站与多个微基站，所述蜂窝网络用于对其覆盖范围内的微用户提供视频服务；所述微基站基于后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置；所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据按预设规则进行删除；所述微基站根据所述受欢迎度计算缓存概率分布，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存。
- 如权利要求1所述的分布式移动网络视频缓存放置方法，其特征在于，所述蜂窝网络中的所述宏基站包括一个总视频数据，所述总视频数据还包括多个子视频数据，所述微基站从所述宏基站中请求所述子视频数据并进行缓存放置，所述微基站为面向所述微用户的访问节点。
- 如权利要求2所述的分布式移动网络视频缓存放置方法，其特征在于，所述分布式缓存模型关于联合集群和缓存优化问题，其中：定义所述蜂窝网络中的所述微基站的集合为B，B中任一所述微基站用b表示；定义所述蜂窝网络中所有所述微用户的集合为M，M中任一所述微用户用m表示；定义所述宏基站中的所述总视频数据为V，所述子视频数据有i个，V中任一所述子视频数据用v表示；定义一个所述微基站服务的微用户集合为U，U中所述微用户有S个，U中任一所述微用户用u表示；定义所有的所述微基站的缓存空间的容量总和为C，C中任一所述微基站 的缓存空间为c；所述微基站从所述宏基站缓存到本地的一个所述子视频数据记为v b；根据以上，所述微用户请求所述子视频数据的总服务延迟满足如下条件：
其中，代表关于U中所述微用户u从所述微基站b请求所述子视频数据v的延迟，u i表示所述微用户u所请求的第i个所述子视频数据；
定义所述微用户与所述微基站之间的关联向量为R，r u∈R，r u代表为所述微用户u i提供服务的所述微基站b，根据以上，关于联合集群和缓存优化问题的所述分布式缓存模型满足如下表达式：
其中，Q(R,m b)代表缓存优化问题的最小延迟值，m b代表M中任一所述微用户m与所述微基站b之间的关系，m b,i代表M中任一所述微用户m与所述微基站b之间请求第i个所述子视频数据之间的关系，c b表示所述微基站b的缓存空间容量。 - 如权利要求3所述的分布式移动网络视频缓存放置方法，其特征在于，所述蜂窝网络用于对其覆盖范围内的微用户提供视频服务的步骤具体包括以下子步骤：所述微基站获取所述微用户在t时刻时向其发出的获取第一子视频数据的请求；所述微基站向所述宏基站请求获取所述总视频数据中对应所述请求的所述第一子视频数据，并将其缓存放置；所述微基站将自己缓存放置的所述第一子视频数据发送给所述微用户。
- 如权利要求4所述的分布式移动网络视频缓存放置方法，其特征在于， 所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置的步骤具体包括以下子步骤：在所述t时刻的下一时刻时，所述微基站根据所述t时刻时所述微用户请求的所述第一子视频数据计算所述后悔最小值；所述微基站从所述宏基站的所述总视频数据中根据所述后悔最小值选择一个第二子视频数据进行缓存放置。
- 如权利要求5所述的分布式移动网络视频缓存放置方法，其特征在于，所述微基站根据所述t时刻时所述微用户请求的所述第一子视频数据计算所述后悔最小值的步骤具体为：定义后悔向量为以及与所述后悔向量对应的效用向量
定义所述微基站在所述t时刻从所述总视频数据中获取一个所述子视频数据的动作为Z b (nb)，其中，N b是操作总数，N b等于所述总视频数据V中的所述子视频数据的总数；定义m b,zb (Nb)(t)是所述微基站b在所述t时刻执行所述动作Z b (Nb)的概率；根据以上，所述微基站b执行所述动作Z b (Nb)的概率计算公式满足如下条件：
其中，为所述概率计算公式计算出的概率结果，β b是控制所述概率计算公式的玻尔兹曼后悔温度系数，r b +(t)表示正后悔向量，上述条件中的各个参数还满足：
其中，是指在所述t时刻的前一时刻的瞬时观测效用函数，δ b(t)、ξ b(t)、η b(t)是计算用的学习参数，且：
δ b(t)、ξ b(t)、η b(t)满足如上约束条件。 - 如权利要求6所述的分布式移动网络视频缓存放置方法，其特征在于，所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除的步骤具体包括以下子步骤：在每一个所述t时刻时，所述微基站对自己缓存的所有所述子视频数据计算对于所述微用户来说的所述受欢迎度；所述微基站将所有所述子视频数据按照所述受欢迎度进行排序。
- 如权利要求7所述的分布式移动网络视频缓存放置方法，其特征在于，所述微基站对自己缓存的所有所述子视频数据计算对于所述微用户来说的所述受欢迎度的步骤具体为：定义每一个所述微基站基于其服务域中每一个所述子视频数据的请求频率为P m，其中任一所述微基站中的所述子视频数据的请求频率满足：p m＝[p m，1，...，p m，k，...，p m，c]其中，p m,k表示所述蜂窝网络中所述微用户m向所述微基站的缓存空间k内容的请求频率，p m,c表示所述蜂窝网络中所述微用户m向所述微基站的缓存空间c内容的请求频率；根据以上，关于所述请求频率的缓存删除概率以得到所述受欢迎度的计算过程满足如下条件：
其中，β remove是缓存更新系数，“-”是为了赋予被请求频率较低的所述子视频数据更高的概率。 - 如权利要求8所述的分布式移动网络视频缓存放置方法，其特征在于，所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除的步骤还包括以下子步骤：若所述微基站的缓存空间还足够存放另一个所述子视频数据，那么所述微基站不进行任何操作；若所述微基站的所述缓存空间不足够存放另一个所述子视频数据，所述微基站根据所述缓存空间的大小，将所述使用频率较低的所述子视频数据删除。
- 如权利要求9所述的分布式移动网络视频缓存放置方法，其特征在于，所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存的步骤具体包括以下子步骤：在每一个所述t时刻时，所述微基站根据所述t时刻的上一时刻中自己缓存的所有所述子视频数据的所述受欢迎度计算缓存概率；所述微基站根据所述缓存概率从所述宏基站中选择所述缓存空间中不存在的所述子视频数据进行缓存放置。
- 如权利要求10所述的分布式移动网络视频缓存放置方法，其特征在于，所述在每一个所述t时刻时，所述微基站根据所述t时刻的上一时刻中自己缓存的所有所述子视频数据的所述受欢迎度计算缓存概率的过程具体为：定义所述缓存概率分布为m b(t)，且满足如下条件：
m b(t)的结果为根据所述后悔向量计算所述t时刻时所述微用户向所述微基站b请求的所述子视频数据的缓存概率分布。 - 一种分布式移动网络视频缓存放置系统，其特征在于，包括以下模块：分布式缓存模型建立模块，用于建立一个蜂窝网络的分布式缓存模型，所 述蜂窝网络包括一个宏基站与多个微基站，所述蜂窝网络对覆盖范围内的微用户提供视频服务；期望缓存放置模块，用于控制所述微基站通过后悔最小值算法从所述宏基站中获取一个子视频数据并进行缓存放置；缓存删除模块，用于控制所述微基站通过概率算法对自己缓存的所有所述子视频数据计算受欢迎度，并根据所述受欢迎度对所述微基站中的所述子视频数据进行删除；缓存更新模块，用于控制所述微基站根据所述受欢迎度计算缓存概率，并将一个所述微基站中不存在的所述子视频数据在所述微基站中进行缓存。
- 一种计算机设备，其特征在于，包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现如权利要求1至11中任一项所述的分布式移动网络视频缓存放置方法中的步骤。
- 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有计算机程序，所述计算机程序被处理器执行时实现如权利要求1至11中任一项所述的分布式移动网络视频缓存放置方法中的步骤。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111125908.1A CN113573103B (zh) | 2021-09-26 | 2021-09-26 | 分布式移动网络视频缓存放置方法、系统及相关设备 |
| CN202111125908.1 | 2021-09-26 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2023045253A1 true WO2023045253A1 (zh) | 2023-03-30 |
Family
ID=78174478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2022/078360 WO2023045253A1 (zh) | 2021-09-26 | 2022-02-28 | 分布式移动网络视频缓存放置方法、系统及相关设备 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113573103B (zh) |
| WO (1) | WO2023045253A1 (zh) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113573103B (zh) * | 2021-09-26 | 2022-01-28 | 深圳飞骧科技股份有限公司 | 分布式移动网络视频缓存放置方法、系统及相关设备 |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017108070A1 (en) * | 2015-12-21 | 2017-06-29 | Huawei Technologies Co., Ltd. | A micro base station comprising a cache-memory to provide files to a user-equipment |
| CN109327851A (zh) * | 2018-12-04 | 2019-02-12 | 吉林大学 | 覆盖与数据平面分离的超密集异构蜂窝网络用户接入方法 |
| CN109803338A (zh) * | 2019-02-12 | 2019-05-24 | 南京邮电大学 | 一种基于后悔度的双连接基站选择方法 |
| CN111488528A (zh) * | 2020-04-28 | 2020-08-04 | 西安邮电大学 | 内容缓存管理方法及装置、电子设备 |
| CN112671880A (zh) * | 2020-12-18 | 2021-04-16 | 中国科学院上海高等研究院 | 分布式内容缓存和寻址方法、系统、介质、宏基站及微基站 |
| CN113573103A (zh) * | 2021-09-26 | 2021-10-29 | 深圳飞骧科技股份有限公司 | 分布式移动网络视频缓存放置方法、系统及相关设备 |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017113373A1 (zh) * | 2015-12-31 | 2017-07-06 | 华为技术有限公司 | 一种缓存的方法及分组数据网关 |
| CN110035539B (zh) * | 2019-03-14 | 2021-07-13 | 北京邮电大学 | 一种基于相关均衡后悔值匹配的资源优化分配方法及装置 |
-
2021
- 2021-09-26 CN CN202111125908.1A patent/CN113573103B/zh active Active
-
2022
- 2022-02-28 WO PCT/CN2022/078360 patent/WO2023045253A1/zh unknown
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017108070A1 (en) * | 2015-12-21 | 2017-06-29 | Huawei Technologies Co., Ltd. | A micro base station comprising a cache-memory to provide files to a user-equipment |
| CN109327851A (zh) * | 2018-12-04 | 2019-02-12 | 吉林大学 | 覆盖与数据平面分离的超密集异构蜂窝网络用户接入方法 |
| CN109803338A (zh) * | 2019-02-12 | 2019-05-24 | 南京邮电大学 | 一种基于后悔度的双连接基站选择方法 |
| CN111488528A (zh) * | 2020-04-28 | 2020-08-04 | 西安邮电大学 | 内容缓存管理方法及装置、电子设备 |
| CN112671880A (zh) * | 2020-12-18 | 2021-04-16 | 中国科学院上海高等研究院 | 分布式内容缓存和寻址方法、系统、介质、宏基站及微基站 |
| CN113573103A (zh) * | 2021-09-26 | 2021-10-29 | 深圳飞骧科技股份有限公司 | 分布式移动网络视频缓存放置方法、系统及相关设备 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113573103A (zh) | 2021-10-29 |
| CN113573103B (zh) | 2022-01-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Elbamby et al. | Proactive edge computing in latency-constrained fog networks | |
| Zhong et al. | A deep reinforcement learning-based framework for content caching | |
| Jiang et al. | Multi-agent reinforcement learning based cooperative content caching for mobile edge networks | |
| Lee et al. | An online secretary framework for fog network formation with minimal latency | |
| CN112020103B (zh) | 一种移动边缘云中的内容缓存部署方法 | |
| CN110098969B (zh) | 一种面向物联网的雾计算任务卸载方法 | |
| CN103827858B (zh) | 移动网络中的缓存 | |
| CN108259628B (zh) | 异构蜂窝网络中内容缓存与用户关联联合优化方法 | |
| WO2023168824A1 (zh) | 一种基于联邦学习的移动边缘缓存优化方法 | |
| CN106331083B (zh) | 一种考虑内容分发能耗的异构网络选择方法 | |
| Jiang et al. | Learning-based cooperative content caching policy for mobile edge computing | |
| CN108156596B (zh) | 支持d2d-蜂窝异构网络联合用户关联及内容缓存方法 | |
| KR101785221B1 (ko) | 무선 통신에서 파일의 전송 성공 확률을 고려하여 캐싱 확률을 결정하는 방법 및 장치 | |
| US20190075501A1 (en) | Control method for network communication system and multi-access edge computing ecosystem device | |
| CN106851741B (zh) | 蜂窝网络中基于社交关系的分布式移动节点文件缓存方法 | |
| WO2023045253A1 (zh) | 分布式移动网络视频缓存放置方法、系统及相关设备 | |
| CN114390057A (zh) | Mec环境下基于强化学习的多接口自适应数据卸载方法 | |
| Yen et al. | Cooperative online caching in small cell networks with limited cache size and unknown content popularity | |
| CN108566636A (zh) | 面向不同用户偏好的d2d随机缓存布设方法 | |
| CN115633380A (zh) | 一种考虑动态拓扑的多边缘服务缓存调度方法和系统 | |
| Li et al. | Mobility-aware content caching and user association for ultra-dense mobile edge computing networks | |
| CN112911614B (zh) | 基于动态请求d2d网络中的协作编码缓存方法 | |
| CN111465057B (zh) | 一种基于强化学习的边缘缓存方法、装置及电子设备 | |
| CN108174395B (zh) | 基于转移动作评估学习框架的基站缓存管理方法及系统 | |
| Yao et al. | Joint caching in fronthaul and backhaul constrained C-RAN |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22871319 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |