Request routing through collaborative in‐network caching for bandwidth optimization: a methodology
yuemei xu,zihou wang,yang li,fu chen,tao lin,wenjia niu
DOI: https://doi.org/10.1002/ett.2947
IF: 3.6
2017-01-01
Transactions on Emerging Telecommunications Technologies
Abstract:To reduce data access latency, network traffic volume and server load, in-network caching was proposed and has become an intrinsic component of the content-centric network (CCN) architecture. The content-oriented characteristics of in-network caching, such as arbitrary topology and volatile content locations, make routers content-aware and supportive of fast content distribution. Meanwhile, they also raise new challenges in content placement and request routing, namely, how to optimally make content storage decisions and forward user requests towards a best' (e.g. closest) available content replicas, so as to minimize the bandwidth cost under storage and link capacity limit. To address this problem, we build a distributed in-network caching model to formulate the content placement and request routing in CCN, aiming at minimizing the bandwidth cost with strict storage and bandwidth constraints. Based on the proposed model, we design a scalable, adaptive and low-complexity in-network caching scheme for content placement and request routing and analyse the performance gains via simulations on a real Internet Service Provider (ISP) network topology and traffic traces. The experimental results show the proposed model and scheme are superior. Compared with the existing works, we also observe significant performance enhancements in terms of hit ratio of requests, reduction of server load and bandwidth cost. Copyright (c) 2015John Wiley & Sons, Ltd.