Research On Mobile Cache Deployment And Update Strategies

Recently,emerging wireless service forms have arisen sequentially fueled by the improving capabilities of intelligent terminals and the developing quality of service from users.To support such requirments,various new technologies have be proposed.Among them,small cell technology has be widely discussed due to its superior performance to tradeoff bewteeen explosive growth of terminals and limited spetrum resources.Small cell is also called micro Base Station,which is distinguished from Macro Base Station.Although the extensive deployment of small cell can relieve the contradiction problem effectively,it poses great pressure to backhaul links.In order to alleviate the adverse effects,wireless caching technology has been proposed as an effective solution.Traditional wireless caching technologies concentrate on the paradigm where the base station select caching files according to their popularity.The base stations can update their caches using FIFO,LRU,and LFU.Nevertheless,becase of the limited cache memory,their cached files are limited and hit rate are low,i.e.the cache updating algorithm cannot improve the hit rate either.In this paper,to optimize the hit rate,we discuss the cache deployment and updating strategies taking account of the capability that moving UEs and BSs can share their caches(base stations can share caching memory by forward link,and UEs can share their cache by D2D).The main contributions of this thesis are as follows:Firstly,we consider hit rate problems under differenct network frameworks.Simlation results confirmed that our scheme can reach the maximum hit rate with the same system caching memory.Secondly,to jointly consider the BS cache and mobile UE cache,the optimization of hit rate is a NP-hard problem that can be transformed to a submode problem,which can be solved by greedy algorithm.It is also verified that the proposed wireless cache deployment strategy is better than the random cache deployment strategy and the popular cache deployment strategy by simulation results.Thirdly,aiming to maxmize cache hit radio,a wireless cache updating algorithm to determine whether to perform cache replacement or not based on the marginal value is proposed.When cache missing occurs,one can decide whether to replace original file with the new file through calculating the marginal value of new file and existing files.The key of this algorithm is to renewal caching memory space without any decrease of hit rate.At last,to achieve the maximum UE service achievable rate,we discussed the base station selection strategy when UE download files from base station.Arithmetical results proved that the objective function is a monotone submode function which can get a suboptimal solution by greedy algorithm,when UE rate is e times of the number of active UE.We compared service rate achieved by the BS selection strategy aimed to maximum bit rate as well as the strategy based on the received SNR through numerical simulation,which shows that the proposed strategy in this thesis can provide a better service rate.