A Research Of Collaborative Cache Management In Information-Centric Network

With the expansion of the scale of the computer network and the growing of the users' demand for content,the amount of content being delivered in the network is increasing.The traditional network system,which is called the host-centric network,has more and more problems,such as safety,mobility and content delivery efficiency.And it has been unable to adapt to the growing demand for the access to information.Information-centric Network(ICN)architecture then comes into being.The embedded cache of the network is an important advantage of Information-centric Network and managing the cache efficiently is of great significance to the improvement of network performance.The cache management approaches can be classified into collaborative management approaches and non-collaborative management approaches,and the collaborative cache management approaches require a high degree of coordination between nodes in the network.Software Defined Network(SDN),known as the important network architecture in next generation,separates the control plane from the data plane and realizes the centralization of network control.Bringing Software Defined Networking into the design of Information-centric Network can realize the collaborative cache management in a global way.This thesis introduces the realization of Software Defined Information-centric Network,and on this basis,designs the collaborative cache management architecture.In general,cache management takes the information of the content popularity of network as one of the key inputs,so it is important to design a high performance popularity statistical architecture for the cache management.And this thesis proposes PopFilter popularity statistical approach,which realizes the line-speed processing of content request and reduces the complexity of it.Besides,this approach allocates the computing power of switch and of controller reasonably by delivering the decoding work of the statistical result to the controller who has great computing power.Compared with the traditional statistical approach of content popularity,this approach has better performance in accuracy.In the research of this thesis,the collaborative cache management approache based on the whole network aims at minimizing the cost of content request in the network.Under the condition of constant content routing strategy,this thesis introduces the concept of content local profit,which takes the request cost and popularity of the content into account.And the simulation results show that the scheme can reduce the cost of content request in the network to a large extent.On this basis,this thesis proposes a more flexible cache management approache under the condition of variable content routing strategy and solves the problem of joint optimization of content routing and cache placement by using genetic algorithms.The simulation results show that the joint optimization of content routing and cache placement can achieve better result in reducing the cost of content request in the network.