Development of an energy efficient data aggregation technique based upon fish swarm optimization for underwater wireless sensor networks

Underwater Wireless Sensor Networks (UWSNs) have become an important enabler for various underwater and oceanographic applications such as environment monitoring, underwater surveillance, offshore exploration, disaster prevention, pollution monitoring and military applications. Underwater acoustic communication, however, causes several critical issues of high propagation delay, limited bandwidth, high attenuation, dynamic topology, multipath fading and high energy usage. The limitation significantly decreases the network lifetime, communication reliability and overall system-efficiency. Hence, the design of an energy efficient, scalable and reliable data clustering and aggregation scheme for UWSN still remains a challenging research problem.

In order to resolve these problems, a novel Fish Swarm Optimization-based Clustering and Cluster Head selection framework for energy-efficient underwater communication and data aggregation problem is proposed in this paper. The proposed method combines an adaptive clustering scheme, an intelligent cluster head (CH) selection scheme and an energy-efficient data aggregation scheme in a single optimization framework inspired by fish schools’ collective behavior. Unlike traditional clustering methods, the proposed FSOCCH algorithm adopts coordinated individual, instinctive and volitive movement behaviors to achieve a good balance in the optimization process of the global exploration and the local exploitation. This adaptive swarm intelligence mechanism allows the algorithm to adaptively change according to the complex communication conditions of the underwater acoustic environment.

The proposed model employs a multi-objective fitness function which takes into account residual node energy, intra-cluster communication distance, acoustic transmission cost, as well as factors related to cluster balancing in order to achieve the optimum selection of cluster heads. In addition, an energy conscious weighted mean based data aggregation mechanism is adopted at the cluster heads to decrease the redundant underwater acoustic transmissions and communication overhead. Unlike most current UWSN-based studies which have used simplified models for energy usage on the terrestrial solid earth, the proposed study implements a realistic propagation and energy consumption model for underwater acoustics using Thorp attenuation and geometric spreading loss. This makes the proposed FSOCCH framework for real-world underwater deployment more physically realistic and usable.

A large number of trials were performed for 5000 rounds of communication with realistic underwater acoustic communication conditions. The performance of the proposed FSOCCH algorithm was thoroughly evaluated and compared with the traditional underwater clustering protocols as well as the modern bio-inspired optimization techniques such as LEACH-UWSN, HEED-UWSN, Ant Colony Optimization (ACO), Grey Wolf Optimizer (GWO), Moth Flame Optimization (MFO) and Energy-Aware Fish Swarm-based Clustering and Cluster Information Protocol (EAFSCCIP). The simulation results verify that FSOCCH method is much more superior to the benchmark algorithms for various performance metrics such as network lifetime, residual energy, packet delivery ratio (PDR), communication delay, clustering stability. Specifically, the proposed method achieves the results of consuming less energy as compared to LEACH (38.8%), HEED (28.6%), ACO (28.6%), MFO (21.1%), GWO (11.8%), prolonging network lifetime, reducing communication delay and the average packet delivery ratio above 94% during the simulation time period.

In order to further confirm a robustness and reliability to the proposed framework, statistical analysis, convergence analysis, parameter sensitivity analysis and computational complexity analysis were also conducted. The results achieved validate the stable convergence properties of FSOCCH, its energy dissipation balance, its high scalability and its statistical significance in comparison to its performance improvements in the different underwater deployment scenarios. In conclusion, the proposed FSOCCH framework is efficient, scalable, physically realistic solution to the next generation of UWSN and shows the great potential of fish swarm intelligence in optimizing underwater energy utilization.