University of Houston

This paper presents a microscopic traffic estimation algorithm for smartphones by employing their built-in probes (such as GPS and acceleration sensors) to increase the accuracy of estimating the real-time traffic condition without significantly increasing these smart phones' energy consumption. In this approach, real-time traffic data is collected through cooperating smartphone users traveling on urban roads. A new reporting algorithm is provided on smartphone users' side to minimize smartphone probes' connection with the server. Based on these probes' data, real- time traffic condition is estimated on the server side, by applying the Kalman-Filtering (KF) algorithm. Also the real-time traffic level of service (LOS) is estimated on urban roads. A smartphone app is also developed to work as one sample GPS probe together with different simulation-based traffic scenario. Simulation result shows that the proposed algorithm requires less energy usage that existing methods without sacrificing the accuracy of the real-time traffic estimation.