Mitigation of rain impact on microwave backhaul networks

Microwave backhaul networks are a cost-efficient option to support increasing capacity demands of mobile networks. However, inherent vulnerability of wireless backhauling to random fluctuations of the wireless channel complicates the design of reliable backhaul links. Long-lasting channel fluctuations such as rain fading may bring significant network performance degradation, and therefore, need to be carefully treated. This paper proposes a novel rain detection algorithm utilizing both temporal and spatial correlation of link status, aiming at efficiently distinguishing between long-term and short-term channel fading. With this distinction, a central controller decides whether network-wide strategies, such as rerouting, are required to mitigate the effects of rain. The accuracy of the proposed detection method is evaluated by measuring false alarm and misdetection probabilities. Numerical results show high rain detection accuracy of the proposed algorithm. Consequently, the impact of imperfect rain detection on the network throughput performance and on the overhead imposed to the central controller becomes negligible.