Gradient-based Routing for Energy Consumption Balance in Multiple Sinks-based Wireless Sensor Networks

Multiple sinks routing is envisioned as a possible solution to the ‘bottleneck’ research problem in Wireless Sensor Networks (WSN). In addition to focusing on minimizing the energy consumption in a WSN, it is also equally important to design routing protocols that fairly and evenly distribute the network traffic; in order to prolong the network life time and improve its scalability. Gradient Based Routing (GBR) techniques such as the Generic GBR (GBR-G) and the Competing-GBR (GBR-C) have been previously proven to be energy efficient in single sink WSNs. These methods consider that each sensor node constructs a gradient with respect to a unique base station. The drawback of this approach is that, due to the position of sensor nodes next to the sink, their energy is usually overused compared to the one of the other sensor nodes in the network. To overcome this, this paper introduces enhanced GBR-G and GBR-C routing approaches (GB-GBR and CB-GBR) which consider the definition of a new gradient model to maximize network lifetime. In the proposed new approach, the GB-GBR and CB-GBR techniques not only consider the selection of the highest gradient link but also the link that avoids the most overloaded sensor nodes when forwarding packets. Using OMNET++ simulation and the MiXiM framework, it is shown that proposed GB-GBR and CB-GBR approaches achieve better performance in terms of network lifespan when compared to the single sink GBR-G and GBR-C approaches respectively.