Abstract
This paper considers the problem of implementing an Active Noise Control (ANC) system over a network of distributed acoustic nodes. For this purpose we define a general acoustic node consisting on one or several microphones and one or several loudspeakers, together with a unique processor with communication capabilities. In the case that the centralized processing of all the involved information is not possible, several centralized systems could work at the same time in a non-collaborative mode, but this can impair the overall performance of the ANC system. To avoid this drawback, we present a new approach based on the Filtered-x Least Mean Square (FxLMS) algorithm that introduces collaboration between nodes following a diffusion strategy. Moreover,
and trying to simulate the data acquisition performed in real-time applications, the adaptive algorithm has been implemented in the frequency domain, which involves working with blocks of samples, and it has been also partitioned. The simulation results show that, when the nodes of the network have a certain acoustic coupling between their acoustic channels, the diffusion algorithm outperforms the non-collaborative strategy thanks to the network information exchanges among the nodes.
