NSF Award
0232169
This Small Business Innovation Research (SBIR) Phase I project is focused on developing novel network routing techniques that provide power-conserving, latency-minimized, throughput-maximized data delivery in mobile ad-hoc network (MANET) environments. It is well known that existing shortest path algorithms used for optimal routing tend to yield routing topologies, where a small number of the available paths and the available network nodes are heavily used. This tendency induces both path contention (and associated increased latency and reduced throughput) and disproportionately heavy loads on a small number of critical nodes (which yields premature ad hoc network failure when the batteries on one of those critical nodes become drained). The approach introduces a retroactive accounting mechanism in the core dynamic programming-based shortest path algorithm to establish balanced, parallel pathways through a MANET in order to avoid these problems. Preliminary results using a pair of crossing flows have shown a factor of two improvements in throughput and latency, and a factor of three improvements in network lifetime. The key objectives for this effort are exploring the performance bounds of this approach on a larger scale, where the investigative team expects the differences between traditional techniques and the proposed approach to emerge as even greater than they are now.<br/><br/>The commercial goal for this effort is developing a highly responsive, critically needed, enabling technology to provide situational awareness for the first responders and emergency management teams in large-scale emergencies and crisis situations. The proposed effort directly supports the ongoing thrust into this commercial area, which provides a natural conduit for technology transition. Standardization of this technology, as needed for interoperability, will also lead to licensing opportunities with other suppliers.
