FuzzyJam: Reducing traffic jams using a fusion of fuzzy logic and vehicular networks

Traffic congestion is a growing problem worldwide causing time/fuel waste, pollution, and even stress. Various approaches have been proposed to reduce traffic jams. Recently, researchers have started to employ connected vehicle (CV) technology. Most solutions, however, rely on a binary approach to determine a traffic jam, i.e., whether it exists or not. Accordingly, output given to a driver in the form of driving advisory also tends to be binary and static. However, a traffic jam is a dynamic phenomenon, the intensity of which changes over time depending on various factors including randomness of driving behavior and road conditions. In this paper, we propose to integrate a fuzzy inference system into a traffic-jam-control algorithm such that the dynamics of a traffic jam is effectively represented, thereby providing diversified driving advisory depending upon the intensity of a traffic jam. Through simulations, it is shown that the integrated approach reduces traffic delay by up to 6.5% compared with the state-of-the-art solution.