Orthogonal Block Adaptive Line Enhancing Based Onmusic Algorithm Aided

In this paper an alternative adaptation of the Orthogonal Block Adaptive Line Enhancer (OBALE) is presented. The adaptive lter within the OBALE has an auto-regressive-moving-average (ARMA) form which is based on classical Laguerre orthogonal functions. This has the advantages of short lter length and, with a lattice structure, stability. The frequencies and radii of the poles within the orthogonal blocks admit tuning. Tuning of these terms in this contribution is performed using a fourth-order cumulant-based version of the MUSIC algorithm. This exploits the properties of the cumulants to aid harmonic retrieval of narrowband signals embedded in broadband noise. Furthermore, adaptation of the feedforward lter parameters is straightforward since they are linearly related to the lter output. Simulations are presented to show some of the performance capabilities of the technique. It is shown to improve MUSIC frequency estimates at little extra computational cost.