Abstract
The Ant Colony Optimization (ACO) routing technique is an adaptive and reliable approach to find the paths in routing for Mobile Ad hoc Networks (MANETs). In ACO algorithms, the ant as agents traverses across the network to find the shortest path from the source to its destination. During the traversal, the ants deposits pheromones on its path. The path with high probability of pheromone is chosen as the optimized path to transfer data packets between the source and the destination. The quantity of the pheromone laid down on a path depends on its quality metrics viz. minimum number of hops, minimum energy path cost and packet transfer delay. Due to the high dynamic nature of MANETs, the path with high probability of the pheromone deposition may rapidly become unavailable due to link failures. Therefore, it is a challenging task in MANETs to construct a reliable path that is less likely to be disconnected for a long period of time. In this paper, a novel robust energy efficient ACO routing algorithm named AntHocMMP has been proposed. This algorithm is meant to achieve robustness of paths for reliable communication with adaptive re-transmission delays in MANETs. This proposed algorithm enhances the performance of Max-Min-Path (MMP) approach by using ant as agents to find the optimal path in the network. The selection of the optimal path is based on the residual energy of the nodes in that path. A robust route with minimum energy path cost with a short hop-count is chosen for pheromone deposition. Therefore, to keep the paths connected in a dynamic topology, the proposed AntHocMMP redistributes the pheromone based on the energy path cost and it constructs an alternative path for reliable communication between the source and the destination nodes. In this paper, a study has also been done on the effects of re-transmission delays spent in reliably delivering the packets to the destination nodes. The simulation result shows that AntHocMMP achieves a good packet delivery ratio in a wider communication range than the existing algorithms viz. R-ACO1, LAR, MMP and AntHocNet routing algorithms. The statistical analysis viz. t test and Anova-test have been performed that proves which in turn the proposed AntHocMMP dominates the other existing algorithms in a more significant way.