Energy efficient location-based cooperative communications in 4G networks

Abstract

This thesis is concerned with the development of energy efficient cooperative communication protocols for 4G networks. A feature that permeates throughout our work is the use of location information in the cooperative environment in order to minimize the power consumed in the network. Importantly, we consider the effect of location errors in our analysis. We make six contributions in the thesis. In the first part of this thesis we consider multi-hop relay networks and make the following contributions. First, we study the energies required for a message to reach its destination with a given latency bound for two well-known location-based routing protocols. We highlight the different scenarios under which each protocol offers better energy efficiency. Second, we show how energy efficiency is decreased dramatically in most location-based routing protocols once location error is accounted for. We propose a new Location-Error Aware Routing (LEAR) protocol which minimizes the negative impact of location error on the energy efficiency. Third, we develop a new MAC-layer Location-based Cooperative Relay Forwarding (LCRF) protocol, which exploits both location information and cooperation among relays in order to minimize the power consumed in the network. We show how LCRF can result in factor-of-two energy savings, relative to current MAC protocols which use only location information. In the second part of this thesis, we consider an M-user, two-hop, single relay system (an M-1-1 system) and make the following additional contributions. Fourth, we develop a power allocation scheme at the relay that obtains a near-optimal throughput for an M-1-1 system where instantaneous channel state information (CSI) is available. We show that the complexity of our power allocation scheme is of order O(M log M ) and is free of logarithmic operations. Fifth, we develop a new power allocation scheme at the relay that obtains a near-optimal throughput for an M-1-1 system where CSI is not available. This scheme is more difficult to develop due to instantaneous CSI being unavailable. We show that the complexity of this new scheme remains of order O(M log M ). Sixth, we show how our work results in a trade-off between the lifetime and throughput of an M-1-1 system. We examine how the system throughput is influenced by the introduction of location errors, and by the introduction of user mobility.