Title

Towards Building High Performance, Reliable, and Secure Underwater Sensor Networks

Date of Completion

January 2010

Keywords

Computer Science

Degree

Ph.D.

Abstract

Wireless sensor networks have been widely used in various applications. Recently, deploying wireless sensor nodes in aquatic environments, such as lakes and oceans, has received increasing interests from researchers. It has become an emerging network technology, called Underwater Wireless Sensor Networks (UWSNs), to facilitate the exploration of aquatic environments. We identified three fundamental issues for building high performance, reliable, and secure UWSNs: 1) Wireless communication devices with high data rate; 2) Reliable and efficient routing protocols for packet delivery; 3) Security services on underwater sensor nodes. In this thesis, we proposed possible solutions to address these issues. Firstly, we designed and implemented OFDM acoustic modems based on DSP development boards. Our SISO- and MIMO-OFDM acoustic modem implementations achieve the data rates of 3.3 kbps and 6.6 kbps, respectively. The floating-point implementations of the SISO- and MIMO-OFDM acoustic modems can decode one OFDM block with duration of 210 ms in 38.06 ms and 76.65 ms, respectively. Hence, the real-time processing is achieved with impressive margins. Secondly, we proposed a novel routing protocol, depth-based routing (DBR), for UWSNs. DBR can handle highly dynamic network topologies and requires only depth information of sensor nodes for packet delivery. Our simulation results show that DBR achieves reliable packet delivery (at least 95% delivery ratio) for dense networks with good energy efficiency. With the recovery algorithm, the adaptive DBR protocol significantly improves the packet delivery ratio in sparse networks to about 90% with reasonable energy consumption. Finally, we investigated the efficient implementations and optimizations of widely used cryptographic algorithms, AES, SHA-256, and Elliptic Curve Cryptography (ECC), on DSP and low-end 8-bit micro-controllers. The efficient implementations of these cryptographic primitives provide the essential building blocks for realizing security services for underwater wireless sensor networks. ^

COinS