Opportunistic Device-To-Device Communication In Cellular Networks: From Theory To Practice

Abstract

Cellular service providers have been struggling with users’ demand since the emergence of mobile Internet. As a result, each generation of cellular network prevailed over its predecessors mainly in terms of connection speed. However, the fifth generation (5G) of cellular network promises to go beyond this trend by revolutionizing the network architecture. Device-to- Device (D2D) communication is one of the revolutionary changes that enables mobile users to communicate directly without traversing a base station. This feature is being actively studied in 3GPP with special focus on public safety as it allows mobiles to operate in adhoc mode. Although under the (partial) control of the network, D2D communications open the door to many other use-cases.

This dissertation studies different aspects of D2D communications and its impact on the key performance indicators of the network. We design an architecture for the collaboration of cellular users by means of timely exploited D2D opportunities. We begin by presenting the analytical study on opportunistic outband D2D communications. The study reveals the great potential of opportunistic outband D2D communications for enhancing energy efficiency, fairness, and capacity of cellular networks when groups of D2D users can be form and managed in the cellular network. Then we introduce a protocol that is compatible with the latest release of IEEE and 3GPP standards and allows for implementation of our proposal in a today’s cellular network. To validate our analytical findings, we use our experimental Software Defined Radio (SDR)-based testbed to further study our proposal in a real world scenario. The experimental results confirm the outstanding potential of opportunistic outband D2D communications. Finally, we investigate the performance merits and disadvantages of different D2D “modes”. Our investigation reveals, despite the common belief, that all D2D modes are complementary and their merits are scenario based.