Urban Mobility and Networking

Abstract

Understanding urban human mobility is crucial for various mobile and network applications. We address two challenges presented in these mobility-based mobile applications, namely urban mobility modelling and its applications in Delay Tolerant Networks (DTNs). First, We build a human mobility model based on two real-life GPS datasets containing approximately 20 and 10 million GPS samples with transportation mode information. Previous research has suggested that the trajectories in human mobility have statistically similar features as Levy Walks. We propose to explain the Levy walk behaviour by decomposing them into different classes according to the different transportation modes, such as Walk/Run, Bike, Train/ Subway or Car/Taxi/Bus. Second, we develop a solution framework, namely Ameba, for timely delivery in DTNs. We find that human mobility exhibits strong special and temporal patterns. We leverage these human mobility patterns to derive an optimal routing hop count of each content to maximize the number of needed nodes. Illustrative results verify that Ameba achieves comparable delivery ratio as a flooding algorithm but with much lower overhead.