A modeling for hole problem in wireless sensor networks

Abstract

Void areas (holes) are hardly avoided in sensor networks either because of various actual geographical environments, e.g., puddles, buildings or obstacles, or uneven energy consumption. To bypass holes, most existing geographic routing protocols tend to route data packets along the boundaries of holes. Generally, a data packet will be either forwarded along a hole boundary by the right hand rule or pushed back to find another route to its destination when the data packet encounters the hole boundary. The right hand rule, on one hand, consumes more of the nodes' energy on the boundaries of holes, thus possibly enlarging the holes; on the other hand, it may incur data collisions if multiple communication sessions share the same boundaries of holes simultaneously. In this paper, we will propose a hole geometric modeling to solve hole problem in wireless sensor networks. Our hole geometric modeling has two goals: one is to prevent data packets from traveling along the boundaries of holes; the other is to avoid the problem of local minimum phenomenon. By achieving the first goal we can not only reduce the energy consumption of the nodes on the boundaries of holes, thus preventing holes' diffusion, but also reduce the data collisions in the nodes on the boundaries of holes. The second goal can reduce the packets rerouting overhead.