Probing Message Based Local Optimization of Rotational Sweep Paths

In localized geographic routing, the beaconless recovery problem with guaranteed delivery refers to a nodes' task of selecting the next hop on a local minimum recovery path without use of beaconing. The state-of-the-art solution, the Rotational Sweep algorithm RS, solves this problem using only three messages per hop. However, immediate data forwarding, as part of the next-hop selection process, causes problems concerning routing efficiency. Within transmission range of a single forwarding node the RS recovery path may consist of multiple nodes each of which transmits the data. In this paper, we suggest and evaluate a simple, yet promising extension of RS to avoid this. Instead of immediate data forwarding, a lightweight probing packet is used to explore the RS recovery path and to determine a locally optimal hop for data transmission w.r.t. some user-defined optimization function. Our simulations suggest that independent of the node density, our extension can lead to significant reductions of data packet transmissions, and reduced energy consumption, compared to RS, which helps, e.g., to save valuable energy resources of nodes.