Self-Organized Resource Allocation in LTE Systems with Weighted Proportional Fairness

We consider the problem of Long Term Evolution (LTE) network self-organization and optimization of resource allocation. One particular challenge for LTE networks is that, by applying OFDMA, a transmission may use multiple resource blocks scheduled over the frequency and time. We identify that there are three key components involved in the resource allocation and network optimization: resource block scheduling, transmit power control, and client association. We propose a distributed protocol that aims to achieve weighted proportional fairness among clients by jointly consider the three components. This cross-layer design and network optimization includes: (i) an optimal online scheduling policy for resource block scheduling, (ii) a heuristic for transmit power control, and (iii) a selfish strategy for client association. The proposed scheme only requires limited local information exchange, and thus can be easily implemented for large networks. Simulation results have shown its effectiveness in both the system throughput and user fairness.