An evaluation of low-power microphone array sound source localization for deforestation detection

• Microphone array chain-saw localization is evaluated, for deforestation monitoring.
• Delay-and-sum beam-forming is performed on FPGA.
• Localization accuracy does not increase significantly beyond 4 microphones per array.
• Using more than 1024-bin FFT does not increase localization accuracy significantly.
• The energy required for localization is at most 2 μJ for 16-microphone arrays.

Illegal deforestation is a worldwide problem which may be alleviated through technological means of deforestation monitoring, e.g. wireless sensor networks capable of identifying chain-saw noise, performing sound source localization (SSL), and alerting the authorities to the location of the illegal deforestation activity. In this paper we evaluate the feasibility of performing SSL on low-power, energy-constrained, microphone-array-equipped sensor nodes (SNs) with the Delay-and-Sum (DS) beam-forming algorithm. Our work is the first application of this technique for chain-saw noise. We evaluate array configurations of 4, 8, and 16 microphones, and a multitude of DS algorithm configurations, utilizing chain-saw recordings from the ESC dataset, which is available online. We implement the DS algorithm as a digital circuit in a Xilinx Spartan-6 FPGA and analyze its energy consumption. Our analysis indicates that accurate chain-saw localization can be achieved with much simpler microphone arrays and DS configurations compared to previous work. Furthermore, adding FPGA-based SSL capability to the SN increases the energy consumption by less than 10%, compared to a baseline SN capable only of chain-saw identification through spectral analysis executed in software on the SN microcontroller.

Figure optionsDownload as PowerPoint slide