Mobility Pattern Prediction to Support Opportunistic Networking in Smart Cities

Abstract

The ever increasing number of mobile devices in Smart Cities and their heavy use, not only for personal communication but also as a distributed network of sensors, generate a data deluge that stresses the traditional wireless communication infrastructure. The opportunistic networking paradigm seems particularly well suited to the Smart City scenario because it exploits resources that temporarily fall into the connection range of mobile devices as communication proxies, thereby providing cheaper and more energy efficient alternatives to the use of the cellular city network and actively contributing to its offloading. However, its efficacy highly depends on the effectiveness of discovering and using those resources. To improve the effectiveness of opportunistic networking in Smart Cities, we propose a solution which exploits a prediction model tailored for the urban environment that, by detecting complex recurring patterns in nodes' contacts, can forecast the future availability of strategic communication resources. Experimental results obtained in a simulated environment show that our solution can improve the dissemination process and ease the access to the wired network infrastructure.