Energy-efficient dual-sink algorithms for sink mobility in event-driven wireless sensor networks

Abstract

Improving energy-efficiency especially in routing mechanisms is one of the main goals in wireless sensor networks (WSNs). One of the issues of multi-hop routing is the phenomenon of fast energy depletion around the sink known as “sink neighborhood problem”. Recently, employing a dual-sink algorithm has become a popular trend to solve this problem. However, sink selection problem, optimizing the next destination for mobile sink, and finding the optimum next-hop in routing scheme are three other issues that need to be addressed properly in dual-sink approaches. This research firstly presents an energy-efficient dual-sink algorithm with role switching mechanism (EEDARS) to address the sink selection problem in scenarios with non simultaneous events. To this end, a role switching mechanism is applied to the dual-sink algorithm for sending the nearest sink to the event area, hence shorten the path. Secondly, an energy-efficient dual-sink algorithm with fuzzybased sink mobility (EDAFSM) is developed in which the mobile sink adaptively relocates to an optimum location among multiple events using fuzzy logic. Finally, a fuzzy logic scheme for routing optimization is proposed to improve further energyefficiency in EEDARS and EDAFSM. The aforementioned proposed algorithms are known as joint dual-sink and fuzzy-based geographic routing in single-event (JDFGR-S) and multi-event (JDFGR-M) WSNs. These algorithms are compared to seven recent and established techniques. Extensive simulation of these algorithms with different conditions through NS2 framework showed significant improvements on the network metrics especially lifetime, residual energy, number of nodes alive, delivery ratio and load distribution without negative effect on the end-to-end delay. The lifetime of JDFGR-S is 10% higher than EEDARS and the lifetime of JDFGR-M is 22% more than EDAFSM. The validation of simulation results show 96.53% and 98.98% reliability for lifetime and energy consumption metrics, respectively. As a conclusion, the proposed algorithms have improved the energy-efficiency in eventdriven based WSNs.