An Effective Method for the Direction of Arrival Estimation with Impact of Error Rectification in Acoustic Sensor Network

Spreadness of elephant and estimating the way of Arrival is one of the needed research focuses in our Acoustic Sensor Network.  In our work, a comparison is made between and estimate the real time data and offline data collected by passive acoustic sensor at 3 different forests for estimating Direction of Arrival and localization using enhanced hyperbolic circular array. Using MATLAB WAVE tool the performance of data pre-processing is done by us (A/D conversion, Filtering and Compression) and the various external factors affecting the sound propagation and localization accuracy is analyzed. The day to day activities and real time experimental results shows that difference in temperature and wind speed shows the major impact on localization and error accuracy. After the examine the results shows that unexpected changes occurred in metrological values strongly affects the DOA localization, RootMeanSquareError (RMSE) and error constant factor accuracy. Using the Nearest Neighbour Search Algorithm the Average estimated and expected location for real time and tool data and RMSE error are corrected and obtained the nearest error margin within Metres.