A Probabilistic Approach for Maximizing Travel Journey WiFi Coverage Using Mobile Crowdsourced Services

A public transport journey planning service often yields multiple alternative journeys plans to get from a source to a destination. In addition to journey preferences, such as connecting time and walking distance, passengers can select the optimal plan based on mobile crowdsourced WiFi coverage available along the journey. This requires discovering mobile crowdsourced WiFi services available along the journey path. However, this task is challenging due to the uncertain availability of discovered services. To enhance the availability of WiFi coverage, we propose a probabilistic approach to discover groups of available crowdsourced WiFi services along with the journey segments. We first analyze the log of their trajectories and use a density estimation technique to discover reference spots representing the frequently visited locations. Then, a joint discrete Fourier transform and autocorrelation analysis are applied to mine the periods of the presence of moving crowdsourced services with respect to each reference spot. A low-complexity cluster analysis based on Jensen-Shannon divergence is then used to mine the periodic movement behaviors of services during the identified periods. Finally, mobile crowdsourced WiFi services that are simultaneously available at intersecting reference spots are grouped. The QoS of discovered groups is computed in terms of availability confidence, failover capacity, aggregated bandwidth capacity, and coverage. Additionally, we propose an algorithm to determine the best public transport journey plan offering based on the QoS of available WiFi service groups along the journey path. We conduct a comprehensive comparative study to validate the effectiveness of the proposed framework.