Agile neighbor discovery in MEMS-based free space optical networks: a computer vision approach

Abstract

Highly directional laser-based wireless optical radios are able to spatially reuse several THz of spectrum but suffer from link setup delays. In delay averse and high-throughput self-configuring networks, agile neighbor discovery is of critical importance. Given an optical wireless network with microelectromechanical systems based receivers, we explore the use of adaptive combinatorial group testing and contour tracing to achieve very low neighbor discovery latency. We propose a pattern-based algorithm that leverages group testing and contour tracing techniques to significantly reduce latency. Evaluating our algorithms, we compare hierarchical group testing algorithms to state-of-the-art raster and Lissajous pattern-based scanning and report 99.92% and 87% reduction in latency, respectively, for an array of one million micromirrors. The proposed pattern-based algorithm, when compared to hierarchical group testing and the Moore-neighbor contour tracing algorithms, achieves 63.4% and 4.91% reduction in latency, respectively.