Topology-aware transmission scheduling for highway wireless sensor networks

This study considers a highway wireless sensor network that is deployed along major highways as an infrastructure for supporting a wide range of traffic monitoring operations. Contrary to other low data rate wireless sensor networks, the highway sensor network is expected to carry rich media (e.g., video) with high throughput and low latency requirements. While line power is potentially available at selected locations of selected highways, it is expected most sensors will be individually solar-powered to enable practical implementation at large scales and flexible locations. The paper presents a set of centralized and distributed slotted transmission scheduling methods for IEEE 802.11 based sensor platforms. The methods divide sensors into groups based on their deployed road segment and optimize the bidirectional relay throughput, delay, and energy efficiency for each segment. The approach conforms well to the prevalent practice of placing traffic controllers at regular intersections. Based on local network topology, the methods utilize controllers to coordinate sensor transmission time slots with a pipelined, rotational, and load weighted approach. Simulations in ns-2 showed 20-40% higher throughput, 65-90% less energy, and similar delays compared to unscheduled IEEE 802.11 mesh networks. With multi-intersection coordination, we show that the distributed method was able to achieve 96% of the centralized method's throughput at much lesser communication overheads.