Measurements of a Real-time Transit Feed Service Architecture for Mobile Participatory Sensing

摘要

Abstract—We spend a substantial part of our time withudtraveling, in crowded cities usually taking public transportation.udIt is important, making travel planning easier, to have accurateudinformation about vehicle arrival times at the stops. Most ofudthe public transport operators make their timetables freelyudavailable either on the web or in some special format, likeudGTFS (General Transit Feed Specification). However, they containudstatic information only, not reflecting the actual traffic conditions.udMobile participatory sensing can help extend the basic serviceudwith real-time updates letting the crowd collect the required data.udWith this respect we believe that such participatory sensing basedudapplication must offer a day zero service following incrementaludservice extension. In this paper, we discuss how to realize real-timeudrefinements to static GTFS data based on mobile participatoryudsensing. We show how this service can be implemented byudan XMPP (Extensible Messaging and Presence Protocol) basedudmobile participatory sensing architecture and we evaluate itsudperformance.