Analysis of New Gradient Based Aggregation Algorithms for Data-Propagation in Mobile Networks

In large scale networks, agents must use partial knowledge obtained from local interactions to reason about their environment. They require efficient mechanisms to allow them to retrieve and aggregate information beyond their communication range. Even though proposals have been presented for gathering information in large scale wireless sensor networks, it is still a challenge to find an efficient and robust technique for gathering information in large scale mobile wireless networks. In this paper we propose gradients as a multi-path structure for routing and aggregating information across a network of computational mobile nodes. We use simulation to demonstrate that progressive aggregation done on top of a gradient improves the bandwidth usage and memory consumption. We also demonstrate self-* properties of our proposed algorithms including scalability, robustness and adaptability.