Diffusion recursive total least square algorithm over adaptive networks and performance analysis

In most existing distributed estimation algorithms, many scholars adopted the assumption that output signal of system is noisy and input sufficiently accurate. However, in real applications, filter vector usually contains noise. such situations, it has been proved adaptive filtering algorithm using total least squares (TLS) method shown better performance than classical (LS) method. So this paper, we propose a diffusion recursive (DRTLS) by single inverse power iterations network. Then, analyze mean square behaviors proposed DRTLS algorithm. addition, to reduce computational complexity algorithm,an improved called DCD-DRTLS also dichotomous coordinate descent (DCD) iterations. At last, simulation results illustrate superiority algorithms validity theoretical analysis results.