Design of Resilient Heterogeneous Wireless Networks

Abstract

With the growing use of new telecommunications technologies such as 4G and wireless hotspots, heterogeneous wireless networks (HetNets) are gaining more attention. The source of heterogeneity of a HetNet can either be the differences in nodes (such as transmission ranges, failure rates and energy levels) or the differences in services offered in the network (such as GSM and WiFi). Quality of service (QoS) is an issue for users while the cost of the infrastructure is an issue for the network provider. In telecommunications network design problems, survivability and reliability are well known QoS metrics. Most previous studies considered survivability and reliability as constraints (vertex connectivity or edge-disjoint paths), while other papers used traditional reliability metrics (such as two-terminal reliability or all-terminal reliability). In this dissertation, a new metric that combines network reliability with network resilience in capacitated networks is devised. Exact and approximate methods to evaluate this capacitated resilience metric are formulated and solved. Capacitated resilience is used to solve HetNets network designs ranging from 10 to 150 users. Results are compared to the popular reliability and survivability metrics in the literature. It is shown that networks designed by this new measure are significantly different than other network designs. This metric is the first to consider rerouting under capacity constraints in the instance of failure and thus reflects more realistic practice. It is also computationally tractable for use during network design.