CoLBA: A Collaborative Load Balancing Algorithm to Avoid Queue Overflow in WSNs

Most wireless sensor networks (WSNs) are used for data collection applications. WSNs are known for their ease of deployment which make them very popular. In most data collection applications traffic is destined towards one node, usually known as the sink. With high data rate, this many-to-one traffic scenario causes queue overflow in the nodes located near the sink. Sensor nodes are known for their limited storage capacities which makes the packet queue overflow even more critical. This issue is often avoided by the routing protocol. Several efforts have recently led to the proposal of several traffic load balancing routing protocols to avoid queue overflow in WSNs. These protocols use various metrics in order to guide the routing mechanism such as the next hop neighbor queue occupancy rate or the next hop current number of children. In this paper, we propose a new routing protocol based on a Collaborative Load Balancing Algorithm (CoLBA). CoLBA uses the time spent by data packets in the queues as a routing metric. In addition to this metric, each node is aware of its queue occupancy rate. In case the receiving node queue is close to be full, all transmitting neighbors will be alerted and asked to choose another next hop. Simulation results on Contiki OS Cooja simulator showed that CoLBA achieves a better packet delivery rate and much less queue overflow compared to a standard routing protocol based on minimal hop counts.