Reliability analytical measurement to design Wireless Mesh Networks

Abstract

Wireless Mesh Networks (WMNs) are reliable. However, in dense WMNs, interferences cause a significant degradation of end-to-end throughput of multi-hop paths. If WMNs are designed without reliability in mind, multipath routing and non-overlapping routes (i.e., protection routes) to gateways are not sufficient to ensure a reliable WMN design. Indeed, when taking into account Quality of Service (QoS) requirements, alternative paths may fail to re-route traffic in case of failures. Thus, the key challenge in WMN design is the definition and the measurement of WMNs reliability. In this paper, we define a new Reliability Analytical Measurement, called RAM, that allows the measurement/evaluation of WMN reliability in the planning phase and thus the comparison between different designs in terms of reliability. RAM is based on a Multi-Commodity Network Flow (MCNF) model. It considers probabilities of node failures and analytically computes the impact of non-reparable failures.