Performance analysis of blind routing algorithms over content centric networking architecture

摘要

Content Centric Networking (CCN) Architecture has been suggested to overcome weaknesses of current Internet architecture. This architecture is based on the concept of decoupling data from location and retrieving contents by name instead of location. Transport layer protocol i.e. CCNx has been proposed for use with CCN. CCNx performs longest prefix match between requested content name and entries stored in Forwarding Information Base (FIB) table of CCN router. A request is discarded if no match is found between them. CCN routers will be required to maintain a large number of entries in FIB table due to shifting of address space from one billion IPs to at least one trillion content names. Currently, CCN routers cannot store mapping for all content names in FIB table. This will cause large number of FIB's misses and can lead to deadlock situation despite the fact that requested content might be held by a nearby router. In this paper, we have accessed suitability of using different alternatives of blind routing algorithms as a fall back scheme for CCNx when FIB miss occurs. We have implemented different blind routing algorithms i.e. flooding, expanding ring and random walk over CCNx in an OPNET simulator using C language. Matrices of routing overhead, end-to-end delay and success rate have been used for making comparisons. Our results suggest that random walk with queries sent on half of the total links can be used as fall back scheme to avoid deadlock problem. To the best of our knowledge, this is the first time that a simulation study of blind routing algorithms over CCNx architecture has been carried out.