Generalization of Relay Node Deployment Problem for Wireless Sensor Network on jMetal

In Wireless Sensor Networks (WSNs), the relay nodes deployment problem is essential, as the relay nodes serve as cluster heads aggregating the data packages from sensors. Further, the relay nodes transmit collected packets to the Sink node (central processing point) via wired or wireless paths, thus forming a connected network. Various sensor networks tools, such as NetTopo, PiccSIM integrated with NS2, and Atarraya, improve the performance of simulation and protocols in WSNs. Those tools are not easy to start with for researchers whose research field is uncorrelated to the sensor network. Moreover, the studies based on heuristic algorithms on this problem focus more on a two-dimensional plane or an entire three-dimensional space with homogeneous nodes and omnidirectional sensors. Few pieces of research concentrate on relay nodes on three-dimensional terrain contour deployment, which simulates more natural circumstances to a real-life situation. Besides, due to the diverse optimization objectives, arrangement strategies, and sensor node requirements for distinct networks, conducting experiments to compare with existing deployed methods become more challenging. A generalization tool integrating the problem, network environment, and optimization objectives into a framework is essential, where a user can plug in blocks easily with framework interfaces, which makes the comparison more convenient. This design concentrates on users unfamiliar with network settings/models, such as NS2, adopting heuristic algorithms to solve WSN-related problems. In this study, we fill this gap by implementing a generalization framework based on the jMetal toolbox, which makes the WSN problem a benchmark, and benefits its solving by metaheuristic algorithms.