Research On Resource Allocation In Content Delivery Networks

With the development of mobile communication networks and the rapid proliferation of intelligent devices,global mobile data traffic has experienced an explosive growth throughout the world,which leads to a series of problems such as network congestion and user experience decline and brings huge challenges for mobile operators.Content Delivery Network(CDN)has emerged to address the increasing demands of network by deploying caches on the edge of the network closer to users.This can effectively improve the system performance and enhance the quality of service(QoS)of users.Meanwhile,the constraint of radio resource and caching strategy have great impacts on system performance.On this basis,this paper focuses on resource allocation problem in CDN where targeted optimization algorithms are proposed to solve those problems.Simulation results show the effectiveness of our algorithms in different scenarios.The contributions of this work are summarized as follows:1.We study the concept of CDN and investigate the current research situation at home and abroad.We jointly consider the radio resource and cache resource and construct the optimization model of CDN from the perspectives of resource allocation and clustering collaboration.2.We analyze radio resource allocation problem in CDN.The formulated optimization task are proposed to minimize the average latency of all users with consideration of radio resource constraint and cache placement.An effective algorithm is proposed to solve this problem.Firstly,we try to find maximum clique sets of base stations(BSs)by using maximum clique strategy in graph theory.Then we put different files at these BSs in the same clique to improve local cache hit probability.Secondly,we try to associate as many users as possible to the BS with consideration of limited radio resource and users'rate requirements.Simulation results show that our algorithm can effectively reduce the average delay of all users as compared with other algorithms in different experimental parameter settings scenarios.3.We analyze the cluster collaboration problem in CDN.Consider in a mobile adhoc network,each node has a certain amount of space for caching.We try to maximize the nodes which can find their requested files locally by optimizing cache configuration and clustering collaboration.Our proposed algorithm is divided into two steps.Firstly,we cluster nodes into different clusters by using the Iterative Self-Organizing Data Analysis Techniques Algorithm(ISODATA),then optimize the cache placement for each cluster.Secondly,an effective cache collaboration mechanism is proposed to maximize the number of locally hit cache nodes by making use of cache diversity gain through intra-cluster and inter-cluster collaboration.Simulation results verify the effectiveness of our algorithm.In summary,this paper studies the allocation of radio resource and cache resource in CDN.Optimization problems are generated from two different perspectives where feasible algorithms are proposed to solve these problems.Simulation results show that reasonable resource allocation and cache collaboration can effectively improve network performance and provide a feasible idea for the research of CDN.