High-throughput and energy-efficient data gathering in heterogeneous multi-channel wireless sensor networks using genetic algorithm

The need for high throughput in Wireless Sensor Networks (WSNs) necessitates the use of the Multi-Channel (MC) scheme. Existing multi-channel approaches used single or multi-radio sensors to exploit channel resources. The former scheme yields low throughput as well as considerable overhead for channel switching. On the other hand, the latter is costly due to installing multiple radios per node. This study applies the MC scheme to heterogeneous WSNs to bridge this gap. It assumes some super nodes that are equipped with multiple radios. These nodes are used as Cluster Heads (CHs) and are responsible for delivering the gathered data from normal sensors to the Base Station (BS). A high-throughput and energy-efficient algorithm is proposed to perform clustering, routing, and channel assignment in the heterogeneous setting. The proposed method decouples the original problem into two phases and solves them using genetic algorithm. The first phase is devoted to constructing a spanning tree over super nodes and assigning proper channels to their radios. A novel multi-objective cost function is proposed which prolongs network lifetime considerably compared to existing tree construction approaches. It also increases throughput by balancing the observed interference throughout the network. In the second phase of the proposed algorithm, the proper CH and channel per normal node are determined. The extensive simulations demonstrate that the proposed algorithm obtains high throughput using multiple channels. Additionally, it improves total energy consumption and energy exhaustion per delivered bit to the BS by 21.6% and 48.3% compared to the previous schemes.