Delay-bounded resource allocation for femtocells exploiting the statistical multiplexing gain

摘要

Femtocell is an efficient solution for mobile operators to expand indoor coverage and increase network capacity. In this paper, we study the downlink resource allocation problem of two-tier macrocell-femtocell networks. We first formulate the problem as a Mixed Integer Non-Linear Program (MINLP) which aims to maximize the capacity of clustered femtocell networks subject to hard delay constraints of flows with different priorities. Next, we build arrival model for the traffics and apply Stochastic Network Calculus (SNC) to transforming the delay constraints into alternative minimum transmission rate requirements, then we propose a resource allocation algorithm called S-SAPCS to solve the MINLP. Simulation results show that the proposed algorithm has near-optimal performance. We also design a scheme based on deterministic network calculus to show that S-SAPCS is able to exploit the statistical multiplexing gain among multiple flows, which improves the throughput significantly.