Fast convergence scheme for potential-based routing in wireless sensor networks

In the electrostatics-inspired methods, a potential function determines the routing mechanism for transportation of nodes´ traffic. In a potential-based routing, it is essential to calculate the potential function at the location of nodes. The best existing algorithm proposed for distributed potential calculation makes use of Jacobi iterations. These iterations are well-known for their slow convergence which makes them useless for practical applications. Our proposed method aims at improving the convergence rate of potential calculation. We propose a novel method inspired by wave propagation in physical media by making analogy between a simple path network and a single stub transmission line. We generalize our method to make it applicable to all possible network topologies. Analysis results show that the required number of iterations reduces significantly to O(N) for the path network of N nodes, compared to O(N²) for Jacobi iterations.