3rd International Turkic World Congress on Science and Engineering , Niğde, Turkey, 14 - 15 June 2021, pp.1-11, (Full Text)
This
paper introduces an alternative mixing parameter update rule for complex-valued
collaborative adaptive filter structures, which is based on the minimization of
the kurtosis cost function defined by the instantaneous error signal. With the
help of this alternative mixing parameter, the advantages of the CLMS and ACLMS
algorithms over circular and noncircular complex-valued signals are integrated
into the proposed hybrid adaptive filtering algorithm. The simulation results
on system identification settings demonstrate that the proposed hybrid
algorithm with an alternative mixing parameter update rule is outperformed in a
collaborative manner of the CLMS and ACLMS algorithms when compared to the
individual algorithms.