Coalition formation for uplink device to device coordination with cooperation costs

Cooperation in the uplink of wireless networks allows users to form coalitions and improve their throughput by jointly transmitting to the base station. As the users are geographically separated, they must first share the information to be transmitted, among themselves, before joint encoding can be performed. In this paper, we propose a framework to maximize the throughput of each coalition by optimizing the fraction of time and energy spent in sharing data between the coalition members in comparison to the time spent in subsequent joint transmission to the base station. Using the throughput obtained from this framework, we model the problem of determining which users should cooperate as a coalitional game with non-transferable utility and propose a coalition formation algorithm which determines an optimal coalition structure using a Pareto improvement criterion. Numerical simulations show that while users close to the base station obtain little benefit from cooperation, the throughput of cell edge users more than doubles, showing the benefits of cooperation even with significant cooperation costs.