A Novel Prediction Methodology For Detecting Failures & Instabilities In Directional Wireless Networks

Even though advances in wireless technology have yielded lower power consumption, higher data rates, and numerous other improvements, the ability to develop a proactive strategy towards handling degradations and failures in directional wireless networks has evaded the research community. In this paper, we introduce a methodology using an analogy to molecular systems in which a directional wireless network utilizing free space optical (FSO) or RF links is modeled as a molecule whose links can grow/retract similarly to bonds. A normal mode analysis (NMA) is performed to identify link instabilities (degradations and failures) and an N-dimensional potential energy surface (PES) is derived with respect to network and environmental parameters to aide in the detection of when a new topology is available ahead of the topology computation stage. Together, the NMA and PES form a basis for a proactive network methodology aimed at improving performance in directional wireless networks.