Joint scheduling and association for α-fairness Network Utility Maximization in cellular networks

Enhancing system throughput and improving user fairness are two basic but contradictory objectives for resource allocation in wireless cellular networks. To obtain an efficient tradeoff between these two goals, Network Utility Maximization (NUM) framework has been adopted with log-utility to obtain proportional fairness among all the users in the network. However, such tradeoff can not control the bias towards throughput or fairness. In this paper, we focus on α-fairness NUM in Soft Frequency Reuse (SFR) based cellular networks, where SFR is an attractive frequency reuse technique to mitigate Inter-Cell-Interference (ICI) and α can be utilized to adjust the tradeoff. The difficulty of the considered issue comes from that it is a Mixed Integer Programming (MIP) problem taking into account both intra-cell user scheduling and inter-cell user association. To overcome this challenge, the α-fairness NUM problem is decomposed into two subproblems, which are dealt with one by one. First, maximize intra-cell utility by user scheduling and second, maximize network utility by distributed user association. Numerical results show that the proposed algorithm approaches the optimal solution of the α-fairness NUM problem. Also, we get a better tradeoff between throughput and fairness, where fairness is measured by Jain´s index. Particularly, we improve the maximum Jain´s index from about 0.3 to about 1.