Evolution Game Theoretic Optimization of Realistic Cooperative Networks Using Power Control with Imperfect Feedback

Distributed spatial diversity systems, such as multi-relay and multi-user networks, have drawn significant attention from the research community. However, their feasibility in a practical environment remains an open concern. Most notably, how to optimize cooperation between multiple users with individual interests and how practical systems issues can erode diversity gains. Whilst existing information theoretical analysis may yield insightful bounds, they provide an inadequate solution for optimal power allocation in realistic systems due to the mutual information saturation of non-Gaussian inputs. In our work, we use feasible modulation and error correction codes to implement a system and demonstrate how multiple users cannot only improve their performance through cooperation, but also optimize their performance through power allocation with imperfect feedback. We do so, by considering an evolution game theoretic (EGT) approach, whereby the status-quo between users change with each decision.