A Fast Algorithm and Testbed Evaluation for Sound Source Localization Using Sensor Networks
Sensor Networks Sound Source Localization GNOMES
The theoretical idea of a network composed of small, low-cost sensor nodes able to localize the source of acoustic events has the potential to make a multitude of exciting applications possible. However, this potential can only be realized if computationally tractable algorithms are employed. In this paper, we propose a novel algorithm that offers not only low computational complexity, but also high accuracy, making it an excellent choice for implementation on low-power systems. Moreover, we have developed a flexible sensor network testbed and utilized it to determine the performance of our algorithm relative to existing methods. Finally, we present the results of our experimentation using real data collected with our sensor nodes, demonstrating the effectiveness of our localization scheme, and showing that our hardware platform is well-suited to the needs of sensor network applications.