Abstract:

In this paper, a low communication parallel distributed adaptive signal processing (LC-PDASP) architecturefor a group of computationally incapable and inexpensive small platforms is introduced. The proposed architectureis capable of running computationally high adaptive filtering algorithms parallely with minimally low communicationoverhead. A recursive least square (RLS) adaptive algorithm based on the application of multiple-input multiple-output(MIMO) channel estimation is implemented on the proposed LC-PDASP architecture. Complexity and Communicationburden of proposed LC-PDASP architecture are compared with that of conventional PDASP architecture. The compar-ative analysis shows that the proposed LC-PDASP architecture exhibits low computational complexity and provides animprovement more than of85%reduced communication burden than the conventional PDASP architecture. Moreover,the proposed LC-PDASP architecture provides fast convergence performance in terms of mean square error (MSE) thanthe PDASP architecture.

Keywords: