On the Performance of Linear Adaptive Filters driven by the Ergodic Chaotic Logistic Map

TL;DR

This paper demonstrates that sequences generated by the logistic map at λ=4 are ideal for driving adaptive filters in channel estimation, providing optimal performance and stability through analytical and simulation validation.

Contribution

It analytically proves the suitability of logistic map sequences at λ=4 for adaptive filter-based channel estimation and derives optimal parameters for enhanced performance.

Findings

Sequences at λ=4 are uncorrelated and zero mean.

Using these sequences maximizes FIR adaptive filter performance.

An optimal damping parameter for LMS is derived for stability.

Abstract

Chaotic dynamical systems are increasingly considered for use in coding and transmission systems. This stems from their parameter sensitivity and spectral characteristics. The latter are relevant for channel estimation methods. In particular the logistic map $f_\lambda =\lambda x\left( 1-x\right) $ has been employed in chaotic coding and spread spectrum transmission systems. For $\lambda =4$ the statistical properties of sequences generated by $f_4$ are considered as ideal drive signals for channel estimation schemes. This assumption is proven in the present paper. To this end the higher order statistical moments and the autocorrelation of time series generated by $f_4$ are derived. It is shown that for $\lambda =4$ the zero mean time series is uncorrelated. The adaptation performance of finite impulse response (FIR) digital adaptive filters (DAF) used for channel estimation is analyzed. It is shown that using zero mean sequences of $f_4$ leads to the maximal possible FIR DAF performance. An optimal value for the damping parameter in the LMS scheme is derived that leads to the maximal performance and ensures stability. The analytic considerations are confirmed by simulation results.