Dynamic Rate and Power Allocation in Wireless Ad Hoc Networks with Elastic and Inelastic Traffic

Abstract

In this paper, we focus on the problem of dynamic rate and power allocation in wireless ad hoc networks with slow-fading channels, where a mixture of elastic and inelastic traffic is supported. A stochastic optimization problem incorporating the quality of service requirements of the two types of traffic is formulated, which aims to maximize the network performance by allocating the power for each link and controlling the service rates of all flows in the network. Since the utility functions of inelastic flows may be non-concave, which are difficult to be readily transformed to be concave, the proposed original problem is non-convex. However, despite the existing difficulty, a dynamic rate and power allocation algorithm (named DRPAA), is proposed to solve the original optimization problem. In DRPAA, both the stochastic duality theory and the particle swarm optimization approach are used, and this dynamic algorithm provides a good approximation to the optimal solution when the variation of the channel condition of each link gets larger. By using DRPAA, flow rates and link powers are dynamically allocated in each network state without the need for the distribution of the network states. Simulation results show that DRPAA has a good convergence speed and can efficiently utilize network resources to improve the network performance.