Equilibrium design for multi-channel random access networks with selfish users

Abstract: Recently, there has been increasing research in the inefficency in spectrum utilization, which is mainly caused by fixed spectrum allocation policies. There are some proposed approaches to solve this inefficiency, like Dynamic Spectrum Access (DSA), which allows users to share spectrum resources. Implementing DSA in a distributed way can avoid problems with system complexity that can arise in centralized DSA systems; however, it can create incentives for the users to behave selfishly. Selfish behavior reduces the sfficiency of the DSA system, and it causes the system to end up in one of many possible operating points, which makes the performance analysis difficult. In this work, we study a multichannel random access system with selfish users and we propose two mechanisms in which the access point charges users for transmissioon. We analyze the performance of these mechanisms using Game Theory. Results show that by charging users for transmission, we can reduce the possible operating points of the system to a single one. Of the two proposed mechanisms, the per-channel cost mechanism performs rather well, bringing the system sum utility close to that of scheduling systems.