Load-balanced implementation of a delayless partitioned block frequency-domain adaptive filter

Abstract

In this paper, a load balanced implementation of a delayless FxLMS algorithm for the purpose of active noise cancellation is proposed. Frequency-domain adaption algorithms using FFT's are well-known for their efficiency. However, their block-based character will lead to a delay in the order of the used block length. A hybrid adaption approach is chosen in this study to combine the advantages of frequencyand time-domain processing. To allow the usage of the proposed system on various platforms, including single-threaded embedded and real-time systems, a manual yet flexible approach of load-balancing the block-based operations is demonstrated. This successfully reduces the peak-to-average computational load and allows an optimized utilization of the available processing power.