FLSS: A Fault-tolerant Topology Control Algorithm For Wireless Networks
Abstract
Topology control algorithms usually reduce the number of links in a wireless network, which in turn decreases the degree of connectivity. The resulting network topology is more susceptible to system faults such as node failures and departures. In this paper, we consider k-vertex connectivity of a wireless network. We first present a centralized algorithm, <i>Fault-tolerant Global Spanning Subgraph</i> (FGSS<sub>k</sub>), which preserves k-vertex connectivity. FGSS<sub>k</sub> is min-max optimal, i.e., FGSS<sub>k</sub> minimizes the maximum transmission power used in the network, among all algorithms that preserve k-vertex connectivity. Based on FGSS<sub>k</sub>, we propose a localized algorithm, <i>Fault-tolerant Local Spanning Subgraph</i> (FLSS<sub>k</sub>). It is proved that FLSS<sub>k</sub> preserves k-vertex connectivity while maintaining bi-directionality of the network, and FLSS<sub>k</sub> is min-max optimal among all strictly localized algorithms. We then relax several widely used assumptions for topology control to enhance the practicality of FGSS<sub>k</sub> and FLSS<sub>k</sub>. Simulation results show that FLSS<sub>k</sub> is more power-efficient than other existing distributed/localized topology control algorithms.