Network protocols and time synchronization for underwater acoustic networks

NITTHITA CHIRDCHOO

Publication

Network protocols and time synchronization for underwater acoustic networks

NITTHITA CHIRDCHOO

Abstract

Unlike terrestrial networks that mainly rely on radio waves for communications, underwater acoustic (UWA) networks utilize acoustic waves, which poses a new research challenge in the networking area. In this dissertation, we address three important networking problems which are: (1) how to improve the normalized throughput via MAC (Medium Access Control) design; (2) how to enhance the packet delivery ratio (PDR) when packets are routed in mobile UWA networks; and (3) how to reduce the time synchronization error in mobile UWA networks. To address the first problem, we propose two simple random access MAC protocols which are suitable for small UWA networks. The protocols attempt to avoid collisions in a distributed manner, by making use of the information that it overhears, as well as the knowledge of inter-nodal propagation delays. The normalized throughput performance has been evaluated in both simulations and theoretical framework against the terrestrial-designed random access MAC protocols. For larger networks, we propose a handshaking-based MAC protocol for multi-hop UWA networks. The design addresses the channel?s long propagation delay characteristic by utilizing receiver-initiated reservations, as well as by coordinating packets from multiple neighboring nodes to arrive in a packet train manner at the receiver. As for the second problem, we introduce a location-based routing protocol that only assumes rough knowledge of the node?s own location and the destination?s location (e.g. sink) which enables the node to select the next best forwarder efficiently when it is combined with our sector-based routing mechanism. Furthermore, the need for precise knowledge in locating the destination node can be relaxed when the protocol is coupled with the location prediction mechanism. In addressing the last problem, we develop a cluster-based synchronization algorithm for mobile UWA networks. Our design is the first to take both the long and time-varying propagation delays into account in the skew and offset estimation process. As a result, it can achieve a significant improvement in reducing the synchronization errors of mobile UWA networks.