| Mobile communication technology not only brings convenience to people,but also produces a large amount of data traffic,which brings great pressure to the transmission link.To alleviate this pressure,caching technology has been proposed and widely used.Based on the high-density cellular network and Device-to-Device(D2D)communication technology,data can be cached in the edge base stations or users in advance to save the burden of the core link and reduce the file transmission delay.The traditional popularity caching strategy can improve the cache hit rate of users,but it is not suitable for all scenarios.Therefore,in order to meet the needs of different scenarios,it is necessary to formulate a more reasonable caching strategy.This thesis adopts game theory,greedy algorithm and other tools to study the caching strategy based on transmission cost in cellular network.The specific contents and main contributions are as follows:(1)For Scalable Video Coding(SVC)files,a file caching strategy for D2 D users is proposed.First,the user distribution is modeled as Poisson point processes(PPP),and based on the hierarchical characteristics of SVC files,the activation probability and successful transmission probability of data packets at each layer of the file are derived.Then an incentive mechanism is introduced,that is,Content Provider(CP)gives incentives to encourage users to cache corresponding files to save transmission costs.Based on the derived probabilities and incentives,the utility functions of users and CPs are defined as their respective benefits.The interaction between users and CP are modeled as a Stackelberg game,CP is the leader,and the incentive price is optimized.Users are followers,and the caching strategy is optimized through the Lagrangian multiplier method.The simulation results show that compared with the existing literature,the proposed caching strategy can effectively reduce the average transmission cost of user requested files when the file hit rate is not much lower than the optimal value.(2)For energy harvesting scenarios,a caching strategy for users and Small Base Station(SBS)based on user mobility and social relations is proposed.Firstly,according to the contact model between users,the estimated energy consumption of file transmission between users and SBS is deduced.Combined with the social relationship strength between users,the cost function of users is defined as the sum of the product of the social relationship strength and the transmission cost,and the cost function of SBS is the sum of the transmission costs of all users.Then,under the constraints of cache space and residual energy,the cache optimization problems of users and SBS are constructed to minimize their cost functions.Finally,it is proved that both user and SBS cache optimization problems have monotonic sub-model properties,which can be solved by greedy algorithm.The simulation results show that when the user mobility is strong,the proposed cache strategy can save about 40% of the transmission cost compared with the traditional algorithm.(3)Aiming at the problem that there may be malicious users,a transmission user selection strategy based on trust mechanism is proposed.Considering that there are malicious users among users,a trust index is introduced to measure the credibility of users.The trust index of users in each time slot is composed of the historical data and the evaluations received from other users in this time slot.With the goal of maximizing the file hit rate,the user’s caching problem can be formulated as multiple independent 0-1 knapsack problems,where the value is the user’s request probability and the weight is the file size,and the optimal caching strategy can be obtained by a dynamic programming algorithm.Then,under the transmission interference constraint,the selection problem of transmission users is constructed to maximize the transmission cost that can be saved through D2 D transmission,and a greedy algorithm is proposed to optimize the selection of transmission users.Simulation results show that the proposed algorithm has higher secure transmission probability and can save more transmission cost. |
