Reducing congestion in complex networks using targeted capacity improvement

Abstract

Regulating traffic flow in complex interconnected systems is practical when a system’s traffic volume is greater than its traffic capacity. Here we study congestion in a scale-free network with traffic volume greater than the network’s traffic capacity. We propose a method to reduce congestion by increasing the local capacity β of a minimum fraction f_c of nodes. We find that improving the capacity (β > 1) of nodes recovers the travel time delay distribution of a low volume traffic using random and targeted capacity improvement. Furthermore, we show the trade-off between the cost of capacity improvement and the congestion level in the system.