Optimizing the throughput‐lifetime tradeoff in wireless sensor networks with link scheduling, rate adaptation, and power control

• Published in: Wireless communications and mobile computing Vol. 16; no. 12; pp. 1510 - 1525
• Main Author: Kaddour, Mejdi
• Format: Journal Article
• Language: English
• Published: Oxford John Wiley & Sons, Inc 25.08.2016
• Subjects: Adaptation; column generation; Frames; Heuristic methods; link scheduling; Links; Mathematical models; Networks; Power control; Remote sensors; SINR; TDMA; WSNs
• ISSN: 1530-8669; 1530-8677
• DOI: 10.1002/wcm.2613

Throughput and lifetime are usually conflicting objectives in designing wireless sensor networks; hence, the right balance needs to be found. With this aim in view, we address in this paper the problem of minimizing the frame length defined within a time division multiple access scheme and the problem of maximizing network lifetime subject to a maximum frame length. The pursued solution in either case leverages a wide range of parameters related to coverage, routing, transmission power, and data rate. Furthermore, it is consistent with the physical interference model. To this end, we rely on column generation technique to derive near‐optimal solutions even when the integrality constraints on coverage and flow variables are enforced. Moreover, we propose a polynomial‐time heuristic algorithm to solve efficiently the underlying NP‐hard problem of concurrent link selection with discrete power control and rate adaptation. Simulation results show that our heuristic algorithm leads to solutions within 3% of optimality while saving around 99% of computation time. Besides, the results illustrate the significant impact of power control and rate adaptation on throughput and lifetime improvement. Interestingly, we found that network lifetime can be significantly prolonged when traffic demands are sufficiently low at the affordable cost of small decrease in throughput. Copyright © 2015 John Wiley & Sons, Ltd. Throughput and lifetime are usually conflicting objectives in designing wireless sensor networks; hence, the right balance needs to be found. We address in this paper the problem of minimizing the frame length defined within a time division multiple access scheme and the problem of maximizing network lifetime subject to a maximum frame length. Moreover, we propose a polynomial‐time heuristic algorithm to solve efficiently the underlying NP‐hard problem of concurrent link selection with discrete power control and rate adaptation. Experimental results illustrate the significant impact of power control and rate adaptation on throughput and lifetime improvement. Interestingly, we found that network lifetime can be significantly prolonged when traffic demands are sufficiently low at the affordable cost of small decrease in throughput.