Performance Evaluation of AODV and DSDV for Implementing Robot Swarm

Abstract

In Mobile Ad-Hoc Networks, routing the data towards a destination node is the most fundamental task. Different routing protocols have been proposed considering specific appli cation and scenario characteristics. The constant change in the positions of the nodes make this task challenging. In the proposed work, a set of agents, here referred to robots, are made to converge towards each other and an external stimulus, fire, using software based sensors that detect heat without knowing prior positions of other robots and starting off at random positions. Agents use input from their peers to make and modify individual decisions which allows for better synergy, more informed and efficient path finding, collision avoidance, etc. For accomplishing this objective, Particle Swarm Optimization(PSO), a population based approach in solving optimization problems, requires an efficient routing protocol which is fast and reliable making the best use of the feeble hardware that comprise the robots. The robots require frequent communication with each other and they have performed well when paired with the routing protocol AODV as compared to DSDV and the simulations performed have given promising results.