Abstract
During the last decade, there has been growing interest in developing active noise cancellation (ANC) systems since they have emerged as a potential solution in the noise reduction of indoor or outdoor sources. Regarding the last point, if the users need to reduce the noise in some extended areas, a very-large amount of microphones and loudspeakers are required. As a consequence, the ANC system demands a large amount of computation. One potential solution can be given if the information is partitioned and distributed into several computing systems since the use of a centralized computing system could be insufficient. However, current distributed strategies demand a huge computationally cost. Therefore, the development of an efficient distributed ANC system to be applied in practical real-time ANC applications is a challenging task. Here, we present two contributions, which involve the development of a new variant of the filtered-x set membership affine projection-like algorithm to save a large amount of computational cost and the design of a FPGA-based distributed neural processor to efficiently simulate the proposed algorithm. Specifically, we improve two aspects to create a compact and high-performance distributed neural processor. The first aspect is linked to the improvement of the processing system. In particular, we make extraordinary efforts to optimize existing spiking neural arithmetic circuits, which are highly demanded in the computation of the proposed algorithm. The second improvement is related to the development of a new communication scheme based on cutting-edge variants of spiking neural P (SN P) systems to efficiently perform the data distribution between multiple FPGAs. To demonstrate the computational capabilities of the proposed FPGA-based distributed neural processor, we develop an acoustic sensor network as proof-of-concept. Our results have demonstrated that the proposed distributed FPGA-based neural processor can be used in practical real-time ANC applications.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.