Edge Caching Strategy Based On User Mobility In Dense Small Cell Networks

Dense small cell network is a new network architecture targeted at the demand of the fifth generation mobile communication.The most prominent feature of small cell network is the intensive deployment of base stations,which provides supplemental data coverage and high-speed data services for hotspots.Dense small cell networks,as a further step,can promote wireless link performance through reducing the distance between base stations and users.In addition,the users' growing demand for mobile data traffic also greatly increasing pressure on the network capacity and backhaul link.The application of edge caching into dense heterogeneous small cell network,therefore,offers a solution to the challenge of massive mobile data flow.User mobility is a unique trait of Wireless Communication Networks(WCN).As is shown in recent studies,both system performance and user experience of WCN can be enhanced with the large amount of useful information contained in the user mobile context,making it an issue of great concern for small cell edge caching technique to make good use of user mobile data.For issues mentioned above,the main work of this paper is as follows:First,we proposes a Proactive Mobility-Aware Caching Strategy(PMA)grounded in user mobility prediction.Firstly,with user mobility prediction taken as a factor,a scene model for data caching and updating in small cell base stations aimed at maximizing the cache hit ratio is established.Then,we introduces the concept of cache content weight to simplify the problem and proposes the PMA strategy grounded in user mobility prediction.Last,We evaluate efficacy through simulation and compare our strategy against a commonly used cache strategy.Simulation results of our caching strategy show a considerable improvement on the users cache hit ratio.Second,we also proposes a Mobility-Aware Genetic-Annealing Cache Distribution(MGCD)algorithm.As the first step,we establishes a scene model for the cache device allocation in dense heterogeneous small cell.The model has taken long-term mobility trajectory of user groups and file popularity distribution into analysis and formulate it as an integer programming problem.Next,we proposes the MGCD algorithm,which combines the advantages of Genetic algorithm and Simulated annealing algorithm,redesigning the algorithm process according to specific scenes,with the aim of maximizing cache hit ratio.Last,simulation result shows that the algorithm is valid in increasing the cache hit ratio with limited cache devices.In addition,the MGCD algorithm can be combined with the PMA algorithm,which can make the system performance improved further.