UACE: Target detection in an underwater acoustic network: Lake experiment

    Back


    • Day: June 19, Thursday
        Location / Time: D. CHLOE at 18:00-18:20
    • Last minutes changes: -
    • Session: 21. Underwater Communications and Networking
      Organiser(s): Charalampos Tsimenidis, Paul Mitchell, Konstantinos Pelekanakis
      Chairperson(s): Charalampos Tsimenidis, Paul Mitchell, Konstantinos Pelekanakis
    • Lecture: Target detection in an underwater acoustic network: Lake experiment [Invited]
      Paper ID: 2127
      Author(s): Lu Shen, Yuriy Zakharov, Benjamin Henson, Nils Morozs, Benjamin Sherlock, Paul Mitchell
      Presenter: Lu Shen
      Abstract: The aim of this work is to investigate the target detection performance of a distributed underwater acoustic (UWA) communication network with lake experimental data. In recent work, a target detection algorithm was proposed based on measurement of the time-varying channel impulse response (CIR) between a communication transmitter and receiver. The detector computes the normalized mean squared deviation (MSD) between a CIR of interest and a set of reference CIRs (estimated when the target is absent) and uses the minimum normalized MSD (nMSD) as the statistic for target detection. Before computing the nMSD, a filter is applied to the CIR to compensate for the change in delay and amplitude caused by environmental factors, such as node drift or moving surface waves. The performance of the detector was investigated in a small-scale lake experiment and showed clear indication of the target crossing the communication link. In this work, we improve the detector performance in fast-varying channels by updating the reference CIRs in time. The detection threshold is also time-varying and is computed using the current set of reference CIRs and a predefined false alarm rate. Instead of using the minimum nMSD, we exploit information from all nMSDs and make the detection decision by computing the probability of potential target crossing (percentage of nMSDs higher than the detection threshold). The performance of the new detector is investigated in an UWA network with nine communication links. Results show clear indication of target detection and a significant reduction of false alarms, compared to the previous work.
        Download the full paper
      This paper is a candidate for the "Prof. John Papadakis award for the best paper presented by a young acoustician(under 40)"
    • Corresponding author: Dr Lu Shen
      Affiliation: University of York
      Country: United Kingdom

    Back