In this paper the problem of block recursive least squares (RLS) adaptive filtering is formulated in the context of block fast transversal filter (FTF) algorithm. A 'modified BFTF algorithm' is derived by modifying the constrained block-LS cost function to guarantee global optimality. In addition to all the benefits of the original block FTF , this new soft-constrained algorithm provides more efficient way of transferring information between blocks of data. The tracking ability of this algorithm can be controlled by varying the block length. The effectiveness of the new method is tested on a real-life problem dealing with underwater target identification from acoustic backscattered signal.
|Original language||English (US)|
|Number of pages||5|
|Journal||Proceedings - IEEE International Symposium on Circuits and Systems|
|State||Published - Jan 1 1996|
|Event||Proceedings of the 1996 IEEE International Symposium on Circuits and Systems, ISCAS. Part 1 (of 4) - Atlanta, GA, USA|
Duration: May 12 1996 → May 15 1996