On the Use of Information Sharing in Wireless Networks

In this paper, we argue that network performance can be improved when nodes explicitly exchange protocol information with each other. We present a vision of network society for next generation wireless networks that entails such social networking. We begin by describing our cognitive social network (CosNet) system architecture that enables network nodes to effectively gather, analyze, compact, repositorize, and exchange network experience information. We also describe a host of technical issues that need to be addressed for successful realization of a network society, including the management of control overhead, protecting user privacy, vulnerability to selfish and malicious attacks, and scalability concerns. We demonstrate the utility of our approach by studying CosNet inspired modifications of the IEEE 802.11 MAC algorithm that we call CosMAC, respectively. Simulation results show that CosMAC outperforms the throughput of 802.11 MAC in single-hop wireless LAN scenarios. CosMAC also appears to result in a fairer allocation, recorded by 10-15% improvement in Jain´s fairness index.