Distributed Sequential Coalition Formation Algorithm for Spectrum Allocation in Underlay Cognitive Radio Networks

In this paper, we study the resource allocation problem, including the joint consideration of user grouping and power allocation, of an, underlay cognitive radio (CR) network employing non-orthogonal multiple access (NOMA), which is referred to as the NOMA-CR network. Based on the non-transferable utility coalition formation game theory, the so-called distributed sequential coalition formation (DSCF) algorithm is leveraged to solve the user grouping problem, where multiple cognitive users within a group can access the primary user's spectrum band while satisfying a minimum rate requirement of the primary user. The simulation results show that a near-optimal performance can be achieved by implementing the proposed DSCF as well as the optimal power allocation algorithms. Furthermore, the results validate that NOMA-CR can significantly outperform the traditional orthogonal multiple access CR, in terms of throughput. Finally, it is shown that by applying a novel adaptive trellis turbo coded modulation technique to the underlay NOMA-CR network, a significant throughput gain can be achieved.