APCN: A Scalable Architecture for Balancing Accountability and Privacy in Large-scale Content-based Networks

Abstract

Balancing accountability and privacy has become extremely important in cyberspace, and the Internet has evolved to be dominated by content transmission. Several research efforts have been devoted to contributing to either accountability or privacy protection, but none of them has managed to consider both factors in content-based networks. An efficient solution is therefore urgently demanded by service and content providers. However, proposing such a solution is very challenging, because the following questions need to be considered simultaneously: (1) How can the conflict between privacy and accountability be avoided? (2) How is content identified and accountability performed based on packets belonging to that content? (3) How can the scalability issue be alleviated on massive content accountability in large-scale networks? To address these questions, we propose the first scalable architecture for balancing Accountability and Privacy in large-scale Content-based Networks (APCN). In particular, an innovative method for identifying content is proposed to effectively distinguish the content issued by different senders and from different flows, enabling the accountability of a content based on any of its packets. Furthermore, a new idea with double-delegate (i.e., source and local delegates) is proposed to improve the performance and alleviate the scalability issue on content accountability in large-scale networks. Extensive NS-3 experiments with real trace are conducted to validate the efficiency of the proposed APCN. The results demonstrate that APCN outperforms existing related solutions in terms of lower round-trip time and higher cache hit rate under different network configurations.