In this paper we describe the deployment of an offshore wireless sensor network and the lightweight intelligence that was integrated into the data acquisition and forwarding software. Although the conditions were harsh and the hardware readily available, the deployment managed to characterise an extended period due the some unique algorithms. In-situ adaptation to differences in data types, hardware condition and user requirements are demonstrated and analysis of performance carried out. Statistical tests, local feedback and global genetic style material exchange ensure limited resources such as battery and bandwidth are used efficiently by manipulating data at the source.
Chris M. Roadknight, Antonio González, Laur