Research Of Base Station Social Network Based Caching Strategy In Wireless Cellular Network

Wireless communication has widely effected people's normal life and economic development.Besides,The 5G(fifth generation)of wireless networks is supposed to support 1000-fold higher data rates and support a much larger number of user and more diverse applications.Thus,how to improve the network throughput and reduce the network average delay become the key problem in 5G networks.Among all the technologies,the dense deployment of small cells and caching at the edge are two promising approaches to increase the area spectral efficiency.Therefore,the effective deployment of caching at the base stations for ultra-dense network becomes a popular study for 5G technology research.In this paper,we studied the relationship between the characteristic of users requiring files and the variation of base station traffic,and utilized the base station social network and users' social behavior to study the base station social network based caching strategy in wireless cellular networks.The main contributions of this paper as summarized as follows:1.Research of base station social network based caching strategy:We studied the relationship between base stations by utilizing the theory of social network model and selected the VIBS(Very Important Base Station),of which is more typical and representative,and divided the rest base stations into NBS(Normal Base Station)and UBS(Unique Base Station)depending on the degree of similarity with VIBS's traffic variation.In our caching strategy,the VIBSs and UBSs are cache enabled,the total caching capacity and backhaul capacity are limited.To prove the superiority of the proposed social aware caching strategy,performance of network throughput,power consumption,network latency and EE(Energy Efficiency)are analyzed with the variation of STF(Social Tie Factor)and the proportion of VIBS.We can found that the throughput decreased when the STF increased,and the throughput increased at first,then decreased and no longer changed at last when the proportion of VIBS increased from the numerical results.Therefore,by optimizing the STF and the proportion,the strategy will achieve the largest throughput,and will gain 83%increase comparing with traditional caching strategy which all the base stations cached evenly.2.Research of base station and user combined social network based caching strategy:In this part,we utilized the India Buffet Process to model users' social interests and the Gamma distribution to model users' social behavior on encounter duration.Besides,the traffic of base station will be offloaded and the QoS(Quality of Service)will be improved by taking the D2D(Device to Device)communication as an underlay of cellular networks.Based on the combined social network,we proposed an effective caching strategy.First we divided the files into old files and new files depends on whether the file has a history of requirement.When the user requires an old file,he will find a D2D partner and take D2D communication to get the needed file,if the D2D communication failed,he will get the file by accessing the base stations through cellular networks.If the user requests a new file,he will get it from the cellular network directly.The base stations are divided into VIBS,NBS and UBS.The performance of network throughput,power consumption,network latency,EE and traffic amount offloaded by D2D layer are analyzed with the variation of STF,VIBS cover radius and D2D communication radius.The theoretical and numerical results show that the proposed caching strategy can achieve the 60%improvement of network throughput by optimizing the three variables.3.Research of combined social network based smart accessing and caching strategy:In this part,we optimized our caching strategy by introducing the smart accessing strategy comparing with the second part of our paper.Although the users take D2D communication can reduces the network power consumption,the probability of users successfully get the files are much higher by accessing a base station through cellular network.Therefore,we define two impact factors for users taking D2D communication and accessing cellular network respectively to indicate the probability of users get the needed files successfully.The impact factor of D2D layer contains three elements:the social interest,the physical location and the encounter duration.The impact factor of cellular network contains two elements:the caching capability and the physical location of base stations.Users will choose the accessing strategy according to the comparison of the two impact factors.The performance of network throughput,power consumption,network latency and EE are analyzed with the variation of VIBS cover radius and STF.By utilizing the smart accessing strategy,our proposed caching strategy for the combined social network achieves 16.7%gain of network throughput comparing with the caching strategy proposed in the second part by optimizing the VIBS cover radius and STF value.