QoS aware path selection in content centric networks

With the availability of higher network bandwidths, greater computing power, and a shift to more content centric use of the internet, researchers are revisiting the Internet architecture. One such effort is the Content Centric Networking (CCN) project where content driven data, rather than nodes, is primarily used for enabling communications. CCN include state maintenance at the intermediate network nodes and allow multiple paths to the content; two aspects absent in the current Internet architecture. In this paper, we focus on the problem of path selection in CCN. Our contribution is threefold - we propose: (i) a technique for downloading content from multiple sources simultaneously, maximizing aggregate bandwidth, (ii) a method for determining the highest bandwidth path to a single source for high bandwidth applications, and (iii) an approach for minimizing per packet delay for real time traffic. Compared to our approach, random forwarding gives only 63% of attainable bandwidth and twice the average per packet delay.