Abstract
With the rapid growth of network traffic and the enhancement of the quality of experiences of users, Information-Centric Networking (ICN), which is a content-centric network architecture with named data caching and routing, is proposed to improve the multimedia content distribution efficiency. In arbitrary topology, cache nodes and users are randomly distributed and connected, hence it is challenging to achieve an optimal caching placement under this situation. In this paper, we propose a caching placement algorithm for arbitrary topology in ICN. We formulate an optimization problem of proactive caching placement for arbitrary topology combined with multi-hop forwarding, with an objective to optimize the user delay and the load balancing level of the nodes simultaneously. Since the original problem is NP-hard, we solve the formulated caching placement problem in two sub-problems, content replica allocation sub-problem and content replica placement sub-problem. First, in the content replica allocation sub-problem, the replica number of each content is obtained by utilizing the auction theory. Second, the replica number of each content is used as a constraint for the content replica placement sub-problem, which is solved by matching theory. The caching placement algorithm combined with multi-hop NRR forwarding maximizes the utilization of cache resources in order to achieve better caching performance. The numerical results show that significant hop count savings and load balancing level improvement are attainable via the proposed algorithm.
Highlights
Recent traffic measurements have clearly shown that IP video traffic will account for 82 percent of the overall Internet traffic by 2022, as illustrated in the Cisco VNI forecasts [1]
We proposed a content distribution architecture for arbitrary topology in Information-Centric Networking (ICN), where a caching placement optimization problem combined with the multi-hop Nearest Replica Routing (NRR) forwarding is formulated, with an objective to optimize user delay and the load balancing level of the nodes simultaneously
We define the preference value of a cache node over a content. Since it is the initial preference of the cache node, it is assumed that other cache nodes do not exist, but changing the matching content i will affect the value of the cache miss part in the formula of the optimization objective, so the hop count required for acquiring all the contents is considered in the preference value of a cache node, which can be expressed as dj (i) = pkidkj +
Summary
Recent traffic measurements have clearly shown that IP video traffic will account for 82 percent of the overall Internet traffic by 2022, as illustrated in the Cisco VNI forecasts [1]. We proposed a content distribution architecture for arbitrary topology in ICN, where a caching placement optimization problem combined with the multi-hop NRR forwarding is formulated, with an objective to optimize user delay and the load balancing level of the nodes simultaneously. Reference [11] proposes a cache-aware network planning strategy, including location selection of cache nodes and content placement, as well as determination of request forwarding path, and establishes an optimization objective with multiple costs as the main body. The network topology, the working process of the proposed algorithm and the content request distribution in our scenario is described, and the formulation of a caching placement problem combined with the multi-hop NRR forwarding is given.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
