Proactive Cache Updating In Heterogeneous Cellular Networks

'Heterogeneous cellular network ' is a promising network architecture that can efficiently handle the rapidly growing volume of mobile data traffic,but its performance may be affected by the limited capacity of the backhaul links of the heterogeneous base stations in the network.As a result,researchers have proposed the idea of 'oactive Caching' in recent years.By proactively caching hot video files in heterogeneous small base stations'caches,the user access delay and the base stations'backhaul link traffic will be reduced significantly,which has become a hot topic of current research.Most of the existing research work focuses on how to optimize the cache placement strategy according to the popularity of video files,when the popularity of video files changes,a new placement strategy is generated based on the current popularity and the cache update is performed accordingly.However,this kind of cache update strategy ignores the original cache state on the small base station,which will bring a lot of communication overhead and energy costs.Regarding the above problems,this paper conducts thorough research on how to efficiently perform proactive cache update in heterogeneous cellular networks.Combined with the differences in access methods of backhaul links for small base stations,this paper proposes correspondingly efficient cache update strategies.The main contributions of the paper are as follows:1.Efficient cache updating based on unicast.In such a heterogeneous cellular network that eachheterogeneous base station can download files from the server through an independent backhaul link,we study how to update the heterogeneous base stations'caches to maxmize the overall hit ratio of the caches under the condition that the base station' energy is limited.First,this problem is modeled as a combinatorial optimization problem,and it is proved that the problem is NP-hard.Further,two efficient cache update strategies based on greedy selection and local search are proposed.Finally,we carry out simulation experiments according to the real data of Youku.com.The experimental results show that the proposed update strategy can achieve high hit ratio with lower costs.We also can seek the balance between caches'hit ratio and update cost by adjusting the update cycle.2.Efficient cache updating based on multicast.For a heterogeneous cellular network in which each heterogeneous base station connects to the macro base station through a wireless method,all base stations can only obtains files from the server through the macro base station.In order to effectively use energy,the macro base station multicasts files to the small base stations.The cache updating in this multicast scenario also faces the limitations of the available energy of the macro base station and the buffer capacity of the small base stations.In order to maximize the system cache hit ratio after updating,we model this problem as a combinatorial optimization problem,and prove it as NP-hard.In the situation that users can connect to at most one small base station,an update strategy based on the Multi-Choice Knapsack Problem(MCKP)is proposed.When the users can connect to multiple base stations,another efficient strategy is proposed.Finally,based on the real video access data of Youku.com,the proposed multicast-based update strategies are simulated respectively.The experimental results demonstrate the efficiency of the proposed update strategies using multicast.It is also showed that the multicast-based update stategy can significantly reduce energy consumption compared to the unicast-based update strategy.