Research On In-network Cache Resolution Scheme In Information-Centric Networking

With huge demand of user's content access requirement, the driving forces behind the exponential growth in network traffic have fundamentally shifted from host-centric end-to-end communication to receiver-driven content retrieval. Under this model, users are concerned about more scalable and efficient content retrievals instead of where the content 'comes from'. To cope with the shift, Information-centric Networking (ICN) has been conceived by means of name-based routing and on-path caching, the new content-centric network architecture involves a change of the underlying implement but a traditional network middleware,which leads to efficient content distribution and high resource utilization that user could easily captured in the whole network. Given that caching capacity of ICN is relatively small in comparison to the amount of forwarded content without optimization techniques and methods, also the new features like fine-grained characteristic, transparency and ubiquity appear in in-network caching results in unbalanced distribution of content assignment. In order to alleviate the problems above, optimization design in cache resolution scheme and cache replacement algorithm of ICN are carried out in this paper, and the balance of content assignment can also be guaranteed.This paper studies the cache resolution scheme, firstly we analyze the problem caused by universal caching represented by on-path caching,which is the default scheme of ICN, the default scheme generate lots of invalid replacement and out of balance, so we propose a new cache resolution scheme based on content popularity and node's centrality, called PCB. PCB scheme selectively cache passing content after a multiple parameter evaluation to nodes and contents, It reduces caching redundancy and in turn, make more efficient utilization of available cache resources along a delivery path through selective caching of content passing.This paper also investigates the cache replacement algorithm. We research the difference between ICN in-network caching and traditional web or streaming media caching system, then prove that the single parameter replacement algorithm like LRU, FIFO and others which widely used in web or streaming media caching system can't fit the ICN well, it may bring out some problem such as cache pollution. So the lifetime and size based caching replacement algorithm is proposed in this paper, called LSB. This algorithm combined the expiration time of content that is native support by ICN and size of content, constructing utility function for each content, then it makes selection for the most suitable replacement in current node, narrowing the gaps between popular objects and users that usually distributed in edge nodes of network, so the popular objects count get storage for a relative long time, also the overall caching hit ratio can be improved.In the experiment section, we make improvements to experimental model via joining the locality principle for exogenous requests. Using the model and a developed testbed based on CCNx, this paper finds that the proposed cache resolution scheme and cache replacement algorithm can realize a more balanced assignment for content, a reduction in cache redundancy and improvement in stability of cache system.