ABSTRACT
This chapter employs a semi-Markov model for the prediction of contact opportunities. It introduces a semi-Markov model and detailed a number of equations suitable to predict different aspects of user mobility behavior, including detection of mobility patterns. The chapter describes a relevant scenario, a first architectural description of the studied platform along with the proposed semi-Markov stochastic model. It considers a heterogeneous network scenario consisting of both mobile and infrastructure nodes. The chapter investigates an software-defined network-inspired communication paradigm where infrastructure and opportunistic networks can efficiently interoperate. It argues that Opportunistic networks can bridge distant infrastructure networks using sophisticated routing protocols capable of detecting and exploiting user mobility patterns. The chapter also argues that Centralized infrastructure nodes can support opportunistic communication with mechanisms that: detect system-wide mobility patterns; and perform resource-expensive estimation calculations for the benefit of the mobile devices.
