SearchLight: A Systematic Probing-based Asynchronous Neighbor Discovery Protocol

The rapid deployment of millions of smart phones has resulted in a demand for proximity-based social networking applications. However, the usefulness of these applications is limited by the lack of effective and energy efficient neighbor discovery protocols. While probabilistic approaches perform well for the average case, they exhibit long tails resulting in high upper bounds on neighbor discovery time. Recent deterministic protocols, including Disco and U-Connect, improve on the worst case bound, but do so by sacrificing average case performance. In response to these limitations, we present Searchlight, an asynchronous neighbor discovery protocol that combines both deterministic and probabilistic components. For symmetric nodes that all have the same duty cycle, this novel combination achieves an average case performance comparable to the probabilistic approaches and improves on the deterministic worst case bounds. Additionally, we show that for asymmetric cases, Searchlight performs comparably to the deterministic protocols when there is a high degree of asymmetry, but can improve on their performance if any of the nodes maintain the same duty cycle. We validate Searchlight through a series of analysis, simulation and real-world experiments on smartphones that show considerable improvement in discovery latency over existing approaches.