Subcarrier And Power Allocation In OFDMA Systems Based On Noncooperative Game Theory

Nowadays, with the rapid development of the information industry, the use of all kinds ofmobile intelligent digital terminals is more and more frequent, so wireless communicationtechnology is facing a new challenge. As the shortage of resources, how to use the limitedbandwidth and power plays a crucial role for wireless network. Being a branch of mathematics,Game theory aims to solve the competition for resources between users by providing an appropriateanalysis framework for the interaction between users. The problem of noncooperative resourceallocation in single-cell and multi-cell orthogonal frequency division multiple access systems isconsidered in this paper.First of all, a subcarrier allocation algorithm in uplink single-cell multiuser OFDMA systemsbased on potential games is proposed, aiming at minimizing the amout of interference generated.The whole process is that the user updates the subcarrier and power allocation in turn until theequilibrium is reached where resource allocation is modeled as a potential game. This game has adesirable property that it will always converge to a Nash Equilibrium (NE). The simulation resultsshow that the proposed algorithms can achieve better spectral efficiency compared to the casewhere only a single subcarrier is allowed to transmit for each user.Secondly, this paper considers the scenario where the co-channel interference exists betweenadjacent cells. Besides, each base station only needs to know users’ information of its jurisdiction. Adistributed subcarrier allocation algorithm in the downlink multi-cell OFDMA systems is proposed.We assume that transmit powers are equal and fixed, and then subcarriers are allocated to all usersindependtly based on the criteria of interference minimization for each cell. The simulation resultindicates that fairness is increased compared to the subcarrier allocation algorithm based on the userscheduling.Finally, given the result of subcarrier allocatioin, a noncooperative game for power control isconsidered based on potential games, aiming at maximizing the users’SINRs. Then the user selectsthe best response strategy in turn until the stability is reached. The simulation results show that theproposed algorithm is superior to the noncooperative resource allocation algorithm based on theuser scheduling, and is easy to implement as its low computational complexity.