Optimal capacity distribution on complex networks

School of Engineering, the Australian National University - Canberra ACT 0200, Australia

Corresponding author: xiayx@ieee.org

Received: 16 September 2009
Accepted: 24 February 2010

Abstract

The capacity distribution plays a crucial role to determine the traffic performance. In this paper, we use a flow model to study the optimal capacity distribution, with the consideration of the uneven nature of traffic generation rates at different nodes. The problem is mathematically modelled by an optimization problem, which adjusts the flow rate and capacity distribution to maximize the system utility and the utilization ratio of capacity. An iteration algorithm is proposed to find the solution of this optimization problem. We show that using this algorithm, the node capacity can be optimally assigned. By comparison, we also indicate that considering the uneven nature of traffic generation rate, the network structural characteristics, such as the betweenness and degree, do not give the best estimation of real traffic load. This study is helpful for the planning of real traffic networks.