Network Coding

Network Coding in Underwater Sensor Networks

Claude Manville, Abdulaziz Miyajan, Ayman Alharbi, Haining Mo, Michael Zuba and Jun-Hong Cui

Computer Science & Engineering Department, University of Connecticut, Storrs, CT, USA 06269 {claude.manville, abdulaziz.miyajan, ayman.alharbi, haining.mo, michael.zuba, jcui}@engr.uconn.edu

Abstract—In this paper, we expand upon previous work in the application of network coding to Underwater Sensor Networks (UWSN). Network coding allows information from multiple packets to be encoded into a single packet and decoded upon receiving sufficient linearly independent encoded packets, improving throughput and providing redundancy for error recovery. Encoding and decoding algorithms for network coding are reviewed, and an algorithm capable of implementing network coding on an UWSN with arbitrary block size, packet length, and symbol size is presented. The network coding algorithms are implemented on real UWSN nodes utilizing our hardware and software platforms and tested using our UWSN testbed. Initial results indicate improved throughput with less overhead than other error-recovery techniques.

I. I NTRODUCTION AND M OTIVATION In recent years, underwater sensor networking has become an active field of research with applications ranging from environmental monitoring to underwater surveillance. Acoustic networking of aqueous sensors provides a broad range of possibilities for large-scale automated data collection. However, before such potential can be fully realized, the implementation of fast and efficient Underwater Sensor Networks (UWSN) presents many challenges [1]–[4]. The naturally lossy nature of underwater acoustic channels introduces significant barriers to the effective transmission of network traffic. High error rates and long propagation delays suffered by any acoustic underwater network require intelligent and efficient means of error-recovery. A number of approaches exist to mitigate the effects of high error rates, though...