Robust $k$-Coverage Algorithms for Sensor Networks

Abstract

Robustness, fault tolerance, and long lifetime are key requirements of sensor networks used in real-world applications. Dense sensor networks with high sensor redundancy offer the possibility of redundant sensing and low duty-cycle operation at the same time, thus the required robust sensing services can be provided along with elongated lifetime. In this paper the Controlled Greedy Sleep Algorithm is analyzed. With low local communication overhead the proposed algorithm is able to solve the k-coverage sensing problem while it effectively preserves energy in the network. In addition, it can adapt to dynamic changes in the network such as node failures. The quality of service (network-wide k-coverage) is guaranteed independently of communication errors in the network (as long as it is physically possible); message losses affect only the network lifetime. Node failures may cause temporary decrease in the coverage service. The robustness of the algorithm is proven and its behavior is illustrated by simulation examples.

Performance Analysis of CGS, a k-Coverage Algorithm Based on One-Hop Neighboring Knowledge

“…These results show that CGS has been £rst designed to ensure full k-coverage (Q = 1) as in [7]. These results also show CGS robustness because when messages are lost, CGS manages to maintain Q = 1.…”

Robustness and Performance Analysis of a Dynamic Sensor Network Scheduling Algorithm

“…To make a good trade-off between k-coverage guarantee and network resource utilization, we proposed in [15] a new coordinated algorithm. This proposition take into account both the power status and the sensing assignment of the SNs.…”

“…To keep tabs on the power level of SNs and on the kcoverage criteria in parallel, a new metric describing the state of SN was proposed recently in [15].…”

“…This paper evaluates performance and robustness of the Controlled Greedy Sleep (CGS, [15]) algorithm which aims to provide a trade-off between network lifetime and measurement accuracy so that k-coverage is guaranteed as long as the network topology enables it. CGS is a novel solution considering both the remaining energy of sensors and the importance of their measurement results.…”

“…The algorithm tolerates the loss of messages (cf. [15]), which may cause only over-coverage, i.e., sensors may stay unnecessary awake.…”