A Pricing Power Control Scheme with Statistical Delay QoS Provisioning in Uplink of Two-tier OFDMA Femtocell Networks

Abstract

Femtocell is a promising technique to enhance indoor coverage and improve network capacity. Nevertheless, because of the random and co-channel deployment of femtocells, the macrocell will suffer serious cross-tier interference from femtocells in two-tier femtocell networks. Thus, interference mitigation in femtocell networks has been an indispensable task. Meanwhile, with the explosive popularity of smart terminals, especially smart phones and tablets, the wireless networks have loaded a mount of data services with diverse delay quality of service (QoS) requirements. However, due to the stochastically varying nature of wireless physical channel, it is extremely difficult to offer a deterministic delay guarantee in wireless networks. Therefore, the effective capacity of femtocell users (FU) has been introduced to provide a statistical delay QoS provisioning. For that reason, in this paper, we will study the interference mitigation with statistical delay QoS guarantee in uplink two-tier orthogonal frequency division multiple access (OFDMA) femtocell networks. In order to mitigate the cross-tier interference at macrocell base station (MBS), we adopt a price-based power control strategy, in which the MBS protects itself by pricing the interference from FU. Additionally, to guarantee the statistical delay QoS for each FU, effective capacity is introduced into their utility functions. Then, a Stackelberg game is formulated to study the joint utility maximization of the MBS and the FUs subject to a maximum tolerable interference power constraint at the MBS. Subsequently, based on the mathematical analysis of the equilibrium of the formulated Stalkeberg game, a particle swarm optimization (PSO) aided power allocation (PSOPA) algorithm is proposed to solve this optimization problem. At last, simulation results show that our proposed PSOPA algorithm can not only improve significantly the average effective capacity of each FU and guarantee their statistical delay QoS, but also converge successfully.