CN112597388A - Cache-enabled D2D communication joint recommendation and caching method - Google Patents
Cache-enabled D2D communication joint recommendation and caching method Download PDFInfo
- Publication number
- CN112597388A CN112597388A CN202011508666.XA CN202011508666A CN112597388A CN 112597388 A CN112597388 A CN 112597388A CN 202011508666 A CN202011508666 A CN 202011508666A CN 112597388 A CN112597388 A CN 112597388A
- Authority
- CN
- China
- Prior art keywords
- user
- content
- recommendation
- cache
- scheme
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Abstract
The invention discloses a cache-enabled D2D communication joint recommendation and cache method, which comprises the steps of constructing an optimization function by taking the minimum file acquisition cost of a user as an optimization target under the conditions of limited cache capacity and limited recommended quantity of the user, obtaining a recommendation scheme for each user by adopting an annealing algorithm, obtaining the cache scheme of each user by using a knapsack algorithm, performing iterative optimization on the cache scheme and the recommendation scheme, obtaining the optimal solution of the recommendation scheme and the cache scheme, and realizing the minimum average cost of the user. The invention combines recommendation and caching, reduces the influence of different preferences of users on caching, can reduce the average cost of the users by the proposed recommendation and caching scheme, stimulates the users to use D2D for communication, and reduces the network pressure.
Description
Technical Field
The invention relates to a recommendation and caching method, in particular to a cache-enabled D2D communication joint recommendation and caching method.
Background
The explosive growth of intelligent mobile devices and intelligent applications has led to a dramatic increase in mobile data traffic, and traditional cellular networks have been greatly challenged. Device-to-Device (D2D) technology, one of the key technologies for fifth generation mobile communication, allows direct communication between nearby devices in a communication network without the help of infrastructure such as core devices or central devices, thereby greatly reducing the data pressure of the core network of the communication system. The caching technology is to store file content items (including videos, web pages and the like) required by a user in a caching entity in advance, so that the requirement of the user on the content can be effectively realized, the transmission delay is greatly reduced, and the energy loss caused by repeated transmission is reduced. Combining D2D with caching is an effective way to handle the large amount of traffic on the network.
However, there are some problems in cache-enabled D2D communication, such as cellular networks typically have smaller cache capacities compared to wired networks, and users 'preferences for different content vary, and caching a portion of content ahead of time in a limited cache capacity gives the network the same degree of gain depending on the user's preference. If the user's preferences for content are very different, it is difficult to get significant gains even if the caching strategy is optimized. The recommendation system is introduced to reshape the request probability of the user for different contents to a certain extent, so that the user preference distribution tends to be uniform, but if the recommended contents are low in user preference degree, the cache contents cannot be fully utilized, and the problems of resource shortage and serious base station load are caused. The existing combined recommendation and caching method considers that the recommendation is utilized to remodel the request probability of the user for the content, and then files are cached at a base station end, so that the number of times of using a backhaul link is reduced. However, even if the request probability is reformed through recommendation, the content requested by the user still has a certain difference, only caching part of the file at the base station end cannot meet the content requests of all users, and when the user is far away from the base station, the user still needs more cost to obtain the required content.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the defects of the prior art, the invention provides a cache-enabled D2D communication joint recommendation and caching method, which designs a recommendation and caching strategy aiming at the condition of small cache capacity, reduces the average cost of users, stimulates the users to use D2D for communication and relieves the pressure on the network.
The technical scheme is as follows: cache memory of the inventionThe energy D2D communication combined recommendation and cache method is characterized in that the average cost of a user is established as a target, the problem is decomposed into two sub-problems of recommendation and cache by taking the cache capacity and the system recommendation number of the user as an optimization function of a constraint condition, the recommendation sub-problem is solved by using an annealing algorithm, the cache sub-problem is solved by using a knapsack algorithm, finally the two sub-problems are iterated circularly to obtain an optimal recommendation and cache scheme, R content items are recommended to each user according to the preference degrees of different users to different content items in a system consisting of K users and I content items, and the size of each content item is LiEach user can cache the total size not exceeding CkCharacterized in that said method comprises the steps of:
(1) calculating the request probability of each user for the content;
(2) constructing an optimization function taking the minimum average cost of the files acquired by the system as an optimization target;
(3) a buffer scheme is randomly initialized for each user, namely the total size of the random buffer does not exceed CkA set of contents of (a);
(4) based on the caching scheme in the step (3), caching conditions of other users in the communication range of the user k can be obtained, the average cost of the current user k is calculated through an objective function according to the content request probability of the user k and the caching files of the surrounding users, and an optimization function is solved through an annealing algorithm to obtain a recommendation scheme with the minimum average cost of the user k;
(5) calculating the new request probability of each user for the content based on the recommendation scheme in the step (4), fixing the cache schemes of other users, and solving an optimization function for each user by using a knapsack algorithm in sequence to obtain the cache scheme with the minimum total average cost of the system;
(6) and (5) replacing the caching scheme obtained in the step (5) into the step (3), circularly performing the steps (4) and (5), and stopping iteration when the recommended scheme and the caching scheme are not updated any more.
The step (1) comprises the following specific steps:
calculating the request probability of the user to the recommended R items of content as follows:
the probability of a request for the remaining content that is not recommended is:
wherein the content of the first and second substances,indicating how well user k accepts the recommendation,to the extent of the preference of user k for content i,an increase in the probability that the R items of content are requested for recommendation;
the probability that end user k requests content i is:
wherein xk,iE {0,1} is an indicator variable of the recommendation policy, if xk,iA value of 1 indicates that content i is recommended to user k, and otherwise, no recommendation is made.
The optimization function in the step (2) is as follows:
xk,i∈{0,1}, (4e)
yk,i∈{0,1}, (4f)
the optimization function of the recommended scheme obtained based on the optimization function is as follows:
the optimization function of the caching scheme is as follows:
wherein d iskjiIndicating the cost of user k to obtain content i to user j, when i equals b the user obtains content to the base station, yk,iE {0,1} is an indication variable of the cache policy, if yk,iIf the number is 1, caching the content i in the own cache space by the user k, otherwise, not caching; l isiIs the size of content i, CkRepresenting the buffer capacity of user k; when the user's preference degree for the content is larger thanWhen so, accepting the recommendation of the content; (4a) the formula is an objective function of the optimization problem, (4b) is the cache capacity limit of the user, (4c) indicates that the number of contents recommended to the user by the system is R, (4d) indicates that the contents recommended to the user by the system are the contents in which the user is interested, and (4e) and (4f) indicate that the indicating variable is 0 or 1.
The step (4) comprises the following steps:
(41) based on the caching scheme in the step (3), randomly recommending R items of content to the user k, and calculating the average cost C (k) of the user k;
(42) randomly selecting one of the R items of content recommended to the user k, exchanging the selected one with the unrecommended content, and calculating new average cost C' (k);
(43) judging whether C' (k) is smaller than C (k), if so, taking the recommended scheme of the step 42 as a new recommended scheme; if not, then the probability is usedUpdate the recommendation according to the step of step 42, wherein BcIs the boltzmann constant, and is,is the minimum average cost of user k in the cycle;
(44) repeating the step (42) when the cycle number reaches tmaxStopping the iteration of the recommendation strategy, and setting the cycle number as tmaxThe R items of content are recommended to the user.
The step (5) specifically comprises the following steps:
(51) fixing the caching schemes of other users, and updating the optimal caching scheme for the users 1 to K by using a knapsack algorithm in sequence; the dynamic programming is used for solving the knapsack problem of the user k, and the method specifically comprises the following steps: si,cRepresenting a subset, 0, of files 1,2, …, i<i<I, the total size of this subset does not exceed c (0)<c<Ck) And the total cost is minimal, f (i, c) denotes the use of the subset Si,cTotal cost of the system, LiDenotes the size of the file i, denoted by f' (i +1, c)The total system value of f (i, c) is given by the following equation:
all f (I, C), f (I, C) are obtained by the formula (7)k) Corresponding toNamely the caching scheme of the user k;
(52) calculating the total average cost of all users before and after the update of the user k, if the total average cost after the update is reduced, keeping the update, otherwise, taking a cache scheme before the update;
(53) after the step (52) is performed on the user K, the next round of circulation is performed by returning to the step (51) until the total average cost in the whole round of circulation is not changed.
Has the advantages that: compared with the prior art, the invention has the following remarkable advantages: the optimal recommendation and cache strategy is obtained by jointly optimizing the recommendation and cache strategies, the influence of different preferences of users on the cache is reduced, the cache strategy is designed at a user side, the distance from the user side to a request user is short, the number of nodes is large, different files can be cached according to different request probabilities, the method is high in operation efficiency, and the pressure of the system is reduced.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 shows D2D communication radius RD2DWhen the number is 80m, the method convergence graph under different cache capacities;
FIG. 3 shows the communication radius R of D2DD2DWhen the number is 150m, the method convergence graph under different cache capacities;
FIG. 4 shows the average user cost as a function of buffer capacity at D2D communication radius RD2DChange plot at 80 m;
FIG. 5 shows the average user cost as a function of buffer capacity at D2D communication radius RD2D150 m.
Detailed Description
The technical scheme of the invention is further explained by combining the attached drawings.
The invention designs a recommendation and cache method in a D2D communication system applying a cache technology, considers the conditions of limited user cache capacity and limited recommendation quantity, and optimizes the recommendation and cache strategy by taking the minimum average cost of the user for acquiring files as an optimization target. Dividing the problem into two sub-problems of a caching strategy and a recommendation strategy, and obtaining recommended contents for each user by adopting an annealing algorithm under the condition of fixing the caching strategy; then, a recommendation strategy is fixed, and a cache scheme of each user is obtained by using a knapsack algorithm. The two algorithms are iterated through recycling to minimize the average cost of the user.
First, the probability of each user's request for different content is calculated: the system consists of K users and I content items, with different users having different degrees of preference for different content. The system recommends R items of content to each user, the size of the content items being LiEach user can cache the total size not exceeding CkThe content of (1). Finally, calculating the request probability of the user to the recommended R items of content as follows:
the probability of a request for the remaining content that is not recommended is:
wherein the content of the first and second substances,indicating how well user k accepts the recommendation,to the extent of the preference of user k for content i,to recommend an increase in the probability that the R items of content are requested. The probability that end user k requests content i is:
wherein xk,iE {0,1} is an indicator variable of the recommendation policy, if xk,iA value of 1 indicates that content i is recommended to user k, and otherwise, no recommendation is made.
Constructing an optimization function taking the minimum average cost of the files acquired by the user as an optimization target:
xk,i∈{0,1}, (4e)
yk,i∈{0,1}, (4f)
wherein d iskjiIndicating the cost of user k to obtain content i to user j, when i equals b the user obtains content to the base station, yk,iE {0,1} is an indication variable of the cache policy, if yk,iIf the number is 1, the user k caches the content i in the own cache space, otherwise, the content i is not cached. L isiIs the size of content i, CkIndicating the buffer capacity of user k. The user will not accept all recommended content, only if he prefers it more thanThen, the recommendation of the content is accepted. (4a) The formula is an objective function of the optimization problem, (4b) is the cache capacity limit of the user, (4c) indicates that the number of contents recommended to the user by the system is R, (4d) indicates that the contents recommended to the user by the system are the contents in which the user is interested, and (4e) and (4f) indicate that the indicating variable is 0 or 1.
The method comprises the following steps of decomposing a question into two sub-questions of a recommendation strategy and a cache strategy, wherein the recommendation sub-questions are as follows:
the cache sub-problem is:
a caching strategy is randomly initialized.
Designing a recommendation strategy of a user:
(41) the recommendation strategies of different users are independent from each other, so that the recommendation strategies can be formulated for each user respectively. Randomly recommending R items of content to a user k, and calculating the average cost C (k) of the user k according to an objective function:
(42) randomly selecting one of the R items of content recommended to the user k, exchanging the R item of content with the unrecommended content, and then obtaining a new average cost C' (k);
(43) judging whether C' (k) is smaller than C (k), if so, updating the recommendation strategy according to the content in the step 2; if not, then the probability P is usedkUpdating the recommendation strategy, probability P, according to the content in step 2kComprises the following steps:
(44) returning to the step 2, when the cycle number reaches tmaxAnd stopping circulation to obtain the recommendation strategy of the user k.
(5) Designing a caching strategy of a user:
(51) fixing the caching schemes of other users, and updating the optimal caching scheme for the users 1 to K by using a knapsack algorithm in sequence; the dynamic programming is used for solving the knapsack problem of the user k, and the method specifically comprises the following steps: si,cRepresenting a subset, 0, of files 1,2, …, i<i<I, the total size of this subset does not exceed c (0)<c<Ck) And the total cost is minimal, f (i, c) denotes the use of the subset Si,cTotal cost of the system, LiDenotes the size of the file i, denoted by f' (i +1, c)The total system value of f (i, c) is given by the following equation:
all f (I, C), f (I, C) are obtained by the formula (7)k) Corresponding toNamely the caching scheme of the user k;
(52) calculating the total average cost of all users before and after the update of the user k, if the total average cost after the update is reduced, keeping the update, otherwise, taking a cache scheme before the update;
(53) after the step (52) is performed on the user K, the next round of circulation is performed by returning to the step (51) until the total average cost in the whole round of circulation is not changed.
(6) Returning to the step (5), when the recommendation strategy and the cache strategy are not updated any more, stopping iteration.
In summary, the invention decomposes the problem into two sub-problems of recommendation and cache by establishing an optimization function which takes the lowest average cost of the user as an objective function and the cache capacity and the system recommendation number of the user as constraint conditions, solves the sub-problems of recommendation by using an annealing algorithm, solves the sub-problems of cache by using a knapsack algorithm, and finally circularly iterates the two sub-problems to obtain the optimal recommendation and cache scheme.
As can be seen from FIGS. 2 and 3, the communication radius R is shown at D2DD2DWhen the buffer capacity is 5, the communication radius R is 80m, the communication can be converged after two iterations, when the buffer capacity is 7, 9 and 11, the communication radius R is three times of iterations, and the communication radius R is D2DD2DWhen the distance is 80m, convergence can be realized through two iterations, the method has high convergence speed, and the joint optimization method is proved to have higher calculation efficiency and reduce the calculation burden of the system. Comparing fig. 2 to fig. 3, it can be seen that fig. 3 converges faster and has a smaller average cost, which shows that sharing content through D2D communication by more users can further reduce the average cost and bring more gain to the system.
As can be seen from fig. 4 and 5, when the buffer capacity is smaller than the system recommended number 8, the average cost decreases rapidly, and as the buffer capacity increases, the average cost decreases; and when the cache capacity exceeds the system recommendation number, the average cost reduction speed tends to be flat, because when the cache capacity is large enough, the system can easily recommend the content which is interested in the user and is cached already. It can be seen that when the cache capacity is small, the average cost generated by the user adopting the cache scheme of the present invention is obviously lower than that of the cache scheme without recommendation, when the user preferences are different, the gain brought by optimizing the cache strategy alone is limited, and the combined recommended cache method of the present invention effectively improves the network system performance.
Claims (5)
1. A D2D communication joint recommendation and caching method enabling caching recommends R content items for each user according to preference degrees of different users on different contents in a system consisting of K users and I content items, wherein the size of each content item is LiEach user can cache the total size not exceeding CkCharacterized in that said method comprises the steps of:
(1) calculating the request probability of each user for the content;
(2) constructing an optimization function taking the minimum average cost of the files acquired by the system as an optimization target;
(3) a buffer scheme is randomly initialized for each user, namely the total size of the random buffer does not exceed CkA set of contents of (a);
(4) obtaining the caching conditions of other users in the communication range of the user k based on the caching scheme in the step (3), calculating the average cost of the current user k through an objective function according to the content request probability of the user k and the caching files of the surrounding users, and solving an optimization function by using an annealing algorithm to obtain a recommendation scheme with the minimum average cost of the user k;
(5) calculating the new request probability of each user for the content based on the recommendation scheme in the step (4), fixing the cache schemes of other users, and solving an optimization function for each user by using a knapsack algorithm in sequence to obtain the cache scheme with the minimum total average cost of the system;
(6) and (5) replacing the caching scheme obtained in the step (5) into the step (3), circularly performing the step (4) and the step (5), and stopping iteration when the recommended scheme and the caching scheme are not updated any more.
2. The cache-enabled D2D joint recommendation and caching method according to claim 1, wherein the step (1) comprises the following steps:
calculating the request probability of the user to the recommended R items of content as follows:
the probability of a request for the remaining content that is not recommended is:
wherein the content of the first and second substances,indicating how well user k accepts the recommendation,to the extent of the preference of user k for content i,an increase in the probability that the R items of content are requested for recommendation;
the probability that end user k requests content i is:
wherein xk,iE {0,1} is an indicator variable of the recommendation policy, if xk,iA value of 1 indicates that content i is recommended to user k, and otherwise, no recommendation is made.
3. The cache-enabled D2D joint recommendation and caching method according to claim 1, wherein said optimization function is:
xk,i∈{0,1}, (4e)
yk,i∈{0,1}, (4f)
the optimization function of the recommended scheme obtained based on the optimization function is as follows:
s.t.(4c),(4d),(4e);
the optimization function of the caching scheme is as follows:
s.t.(4b),(4f)
wherein d iskjiIndicating the cost of user k to obtain content i to user j, when i equals b the user obtains content to the base station, yk,iE {0,1} is an indication variable of the cache policy, if yk,iIf the number is 1, caching the content i in the own cache space by the user k, otherwise, not caching; l isiIs the size of content i, CkRepresenting the buffer capacity of user k; when the user's preference degree for the content is larger thanWhen so, accepting the recommendation of the content; (4a) the formula is an objective function of the optimization problem, (4b) is the cache capacity limit of the user, (4c) indicates that the number of contents recommended to the user by the system is R, (4d) indicates that the contents recommended to the user by the system are the contents in which the user is interested, and (4e) and (4f) indicate that the indicating variable is 0 or 1.
4. The cache-enabled D2D joint recommendation and caching method according to claim 1, wherein (4) comprises the steps of:
(41) based on the caching scheme in the step (3), randomly recommending R items of content to the user k, and calculating the average cost C (k) of the user k;
(42) randomly selecting one of the R items of content recommended to the user k, exchanging the selected one with the unrecommended content, and calculating new average cost C' (k);
(43) judging whether C' (k) is smaller than C (k), if so, executing the step(42) The recommended scheme of (2) is a new recommended scheme; if not, then the probability is usedUpdating the recommendation according to the steps of step (42), wherein BcIs the boltzmann constant, and is,is the minimum average cost of user k in the cycle;
(44): repeating the step (42) when the cycle number reaches tmaxStopping the iteration of the recommendation strategy, and setting the cycle number as tmaxThe R items of content are recommended to user k.
5. The cache-enabled D2D joint recommendation and caching method according to claim 1, wherein the step (5) comprises the steps of:
(51) fixing the caching schemes of other users, and updating the optimal caching scheme for the users 1 to K by using a knapsack algorithm in sequence; the dynamic programming is used for solving the knapsack problem of the user k, and the method specifically comprises the following steps: si,cRepresenting a subset, 0, of files 1,2, …, i<i<I, the total size of this subset does not exceed c (0)<c<Ck) And the total cost is minimal, f (i, c) denotes the use of the subset Si,cTotal cost of the system, LiDenotes the size of the file i, denoted by f' (i +1, c)The total system value of f (i, c) is given by the following equation:
all f (I, C), f (I, C) are obtained by the formula (7)k) Corresponding toNamely the caching scheme of the user k;
(52) calculating the total average cost of all users before and after the update of the user k, if the total average cost after the update is reduced, keeping the update, otherwise, taking a cache scheme before the update;
(53) after the step (52) is performed on the user K, the next round of circulation is performed by returning to the step (51) until the total average cost in the whole round of circulation is not changed.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011508666.XA CN112597388B (en) | 2020-12-18 | 2020-12-18 | Cache-enabled D2D communication joint recommendation and caching method |
PCT/CN2021/074169 WO2022126836A1 (en) | 2020-12-18 | 2021-01-28 | Cache-enabled d2d communication joint recommendation and caching method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011508666.XA CN112597388B (en) | 2020-12-18 | 2020-12-18 | Cache-enabled D2D communication joint recommendation and caching method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112597388A true CN112597388A (en) | 2021-04-02 |
CN112597388B CN112597388B (en) | 2022-10-14 |
Family
ID=75199559
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011508666.XA Active CN112597388B (en) | 2020-12-18 | 2020-12-18 | Cache-enabled D2D communication joint recommendation and caching method |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN112597388B (en) |
WO (1) | WO2022126836A1 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113064907A (en) * | 2021-04-26 | 2021-07-02 | 陕西悟空云信息技术有限公司 | Content updating method based on deep reinforcement learning |
CN114827270A (en) * | 2022-03-25 | 2022-07-29 | 南京邮电大学 | Recommendation and cache combined optimization method based on multi-base-station cooperation |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116261211B (en) * | 2023-02-02 | 2024-02-09 | 北方工业大学 | Low-energy-consumption dynamic caching method for unmanned aerial vehicle auxiliary data transmission |
CN116761152B (en) * | 2023-08-14 | 2023-11-03 | 合肥工业大学 | Roadside unit edge cache placement and content delivery method |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106708923A (en) * | 2016-11-09 | 2017-05-24 | 武汉大学 | Shared method of native cache file based on mobile intelligence network |
US20180240030A1 (en) * | 2016-03-18 | 2018-08-23 | Youku Internet Technology (Beijing) Co., Ltd. | Content recommendation method, apparatus and system |
CN109587776A (en) * | 2018-12-07 | 2019-04-05 | 东南大学 | The combined optimization method of base station dormancy and cooperation caching in the super-intensive network of D2D auxiliary |
CN111314862A (en) * | 2020-02-19 | 2020-06-19 | 东南大学 | Caching method with recommendation under deep reinforcement learning in fog wireless access network |
CN111860595A (en) * | 2020-06-17 | 2020-10-30 | 南京邮电大学 | Heterogeneous network cache decision method based on user preference prediction |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8019777B2 (en) * | 2006-03-16 | 2011-09-13 | Nexify, Inc. | Digital content personalization method and system |
US7559072B2 (en) * | 2006-08-01 | 2009-07-07 | Sony Corporation | System and method for neighborhood optimization for content recommendation |
CN111432380B (en) * | 2020-03-25 | 2022-06-21 | 哈尔滨工程大学 | D2D-oriented auxiliary data unloading cache optimization method |
CN111488528A (en) * | 2020-04-28 | 2020-08-04 | 西安邮电大学 | Content cache management method and device and electronic equipment |
CN111935784B (en) * | 2020-08-12 | 2022-04-22 | 重庆邮电大学 | Content caching method based on federal learning in fog computing network |
-
2020
- 2020-12-18 CN CN202011508666.XA patent/CN112597388B/en active Active
-
2021
- 2021-01-28 WO PCT/CN2021/074169 patent/WO2022126836A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180240030A1 (en) * | 2016-03-18 | 2018-08-23 | Youku Internet Technology (Beijing) Co., Ltd. | Content recommendation method, apparatus and system |
CN106708923A (en) * | 2016-11-09 | 2017-05-24 | 武汉大学 | Shared method of native cache file based on mobile intelligence network |
CN109587776A (en) * | 2018-12-07 | 2019-04-05 | 东南大学 | The combined optimization method of base station dormancy and cooperation caching in the super-intensive network of D2D auxiliary |
CN111314862A (en) * | 2020-02-19 | 2020-06-19 | 东南大学 | Caching method with recommendation under deep reinforcement learning in fog wireless access network |
CN111860595A (en) * | 2020-06-17 | 2020-10-30 | 南京邮电大学 | Heterogeneous network cache decision method based on user preference prediction |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113064907A (en) * | 2021-04-26 | 2021-07-02 | 陕西悟空云信息技术有限公司 | Content updating method based on deep reinforcement learning |
CN113064907B (en) * | 2021-04-26 | 2023-02-21 | 陕西悟空云信息技术有限公司 | Content updating method based on deep reinforcement learning |
CN114827270A (en) * | 2022-03-25 | 2022-07-29 | 南京邮电大学 | Recommendation and cache combined optimization method based on multi-base-station cooperation |
CN114827270B (en) * | 2022-03-25 | 2023-07-25 | 南京邮电大学 | Recommendation and cache combined optimization method based on multi-base station cooperation |
Also Published As
Publication number | Publication date |
---|---|
WO2022126836A1 (en) | 2022-06-23 |
CN112597388B (en) | 2022-10-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112597388B (en) | Cache-enabled D2D communication joint recommendation and caching method | |
CN107295619B (en) | Base station dormancy method based on user connection matrix in edge cache network | |
CN108777853B (en) | Network edge caching method and system based on D2D | |
CN109673018B (en) | Novel content cache distribution optimization method in wireless heterogeneous network | |
CN108600998B (en) | Cache optimization decision method for ultra-density cellular and D2D heterogeneous converged network | |
Ma et al. | Low latency caching placement policy for cloud-based VANET with both vehicle caches and RSU caches | |
Zhang et al. | Collaborative hierarchical caching over 5G edge computing mobile wireless networks | |
CN108521640B (en) | Content distribution method in cellular network | |
Gu et al. | DR-MDS: An energy-efficient coding scheme in D2D distributed storage network for the internet of things | |
CN110062356B (en) | Cache copy layout method in D2D network | |
CN111277666B (en) | Online collaborative caching method based on freshness | |
Jiang et al. | A reinforcement learning-based computing offloading and resource allocation scheme in F-RAN | |
CN110035415A (en) | A kind of D2D network-caching method for down loading of latency model | |
CN112004265B (en) | Social network resource allocation method based on SRM algorithm | |
CN116546053A (en) | Edge computing service placement system in resource-constrained Internet of things scene | |
CN109547979B (en) | Content distribution method combining energy consumption and user fairness under D2D cache network | |
CN116347522A (en) | Task unloading method and device based on approximate computation multiplexing under cloud edge cooperation | |
CN112954026B (en) | Multi-constraint content cooperative cache optimization method based on edge calculation | |
Hua et al. | On cost minimization for cache-enabled D2D networks with recommendation | |
Javadi et al. | Decentralized hierarchical coded caching over heterogeneous wireless networks with multi-level popularity content | |
CN108738048B (en) | Active storage method of maximized fairness base station based on genetic algorithm | |
Wu et al. | Video placement and delivery in edge caching networks: Analytical model and optimization scheme | |
CN115314948A (en) | Content recommendation and path selection method in D2D auxiliary wireless cache network | |
CN105743975B (en) | Caching laying method and system based on data access distribution | |
Jiang et al. | Learning-Based Content Caching with Update Strategy for Fog Radio Access Networks |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |