Resource efficient connectivity restoration algorithm for mobile sensor/actor networks
1 Deanship of e-Transactions and Communication, King Saud University, Riyadh, Saudi Arabia
2 Department of Computer Science & Electrical Eng, University of Maryland, Baltimore, MD, USA
3 Department of Electrical and Electronics Eng, Universiti Teknologi PETRONAS, Seri Iskandar, Malaysia
4 College of Computer and Information Sciences, King Saud University, Riyadh, Saudi Arabia
EURASIP Journal on Wireless Communications and Networking 2012, 2012:347 doi:10.1186/1687-1499-2012-347Published: 21 November 2012
Maintaining inter-node connectivity is of a paramount concern in most applications of mobile sensor/actor networks because nodes have to report their data and coordinate their operations. Failure of a node may partition the inter-node network into disjoint segments, and may thus hinder data delivery and inter-node coordination. This article presents a novel resource efficient connectivity restoration algorithm (RECRA) that opts to repair severed connectivity while imposing minimal overhead on the nodes. To avoid overreacting to non-critical failure, RECRA identifies critical/non-critical nodes and only triggers the recovery when a critical node fails. The failure of a node is detected by its neighbors and a recovery procedure is executed based on their proximity, status (critical/non-critical), and transmission range. RECRA prefers to employ a non-critical node and moves it to the place of failed node in order to limit the impact on coverage, scope of the recovery, and strangle successive cascaded relocations. In case non-critical nodes in the neighborhood are not available, the neighbors restore connectivity by exploiting their partially utilized transmission power and repositioning closer to the failed node. RECRA is validated analytically and through extensive simulations. The simulation results confirm the effectiveness of RECRA for both dense and sparse network segments and its performance advantage over contemporary schemes found in the literature.