Search

Abstract: We propose a novel vision for roadway safety warning based on sensor networks, aiming at providing user-friendly zero-delay safety warnings to motorists. Our idea leverages the advanced sensing, networking and storage technologies. Roadway sensors detect events and store event records at multiple designated locations along the against traffic direction, such that the passing-by drivers can be alerted to potential dangers or traffic delays through the wireless communication between roadway sensors and the vehicle. We design a location-centric storage (LCS) protocol, which manages the propagation and storage of event records based on the time needed to clear the road. In LCS, the density of the sensors storing an event record decreases logarithmically with respect to the distance to the event location. Thus, the closer to the event position, the more number of warnings a driver may obtain. LCS is further tailored for the case of “highway” sensor networks when all sensors are deployed along a straight line mimicking a highway, and the more complex case when two roads intersect at some place. We conduct both theoretic analysis and simulation study to verify the performance of LCS when applied to roadway sensor networks for safety warning. The results indicate that LCS is fair to all sensors. We conclude that roadway safety warning based on sensor networks is a promising idea for realizing ITS''s “Zero Fatality, Zero Delay” roadway safety philosophy. [Copyright &y& Elsevier]

Abstract: In this paper, we present time-based positioning scheme (iTPS), a purely localized location detection scheme for sensor networks with long-range beacons. iTPS relies on time difference of arrival (TDoA) of radio frequency (RF) signals measured locally at each sensor to detect range differences from the sensor to four base stations. These range differences are combined to estimate the sensor location through trilateration. iTPS is an improvement over TPS (Cheng et al., IEEE INFOCOM, 2004), which produces two ambiguous position estimates when sensors are close to any base station. iTPS substantially reduces the number of ambiguous estimates and can improve accuracy. Features of iTPS include low communication overhead for sensors, no requirements for time synchronization, low computational overhead due to simple algebraic operations, and high scalability. We conduct extensive simulation to test iTPS and compare it with TPS. The obtained results show that iTPS is an efficient and effective scheme for location discovery in sensor networks with long-range beacon stations. [Copyright &y& Elsevier]