1: \begin{abstract}

2: Motivated by reduction of computational complexity,

3: this work develops sign-error adaptive filtering algorithms for estimating time-varying system parameters.

4: Different from the previous work on sign-error algorithms,

5: the parameters are time-varying and their dynamics are modeled by a discrete-time

6: Markov chain.

7: A distinctive  feature of the algorithms  is the

8: multi-time-scale framework for characterizing parameter variations and algorithm updating speeds.

9: This is realized by considering the stepsize of the estimation algorithms and

10: a scaling parameter that defines the transition rates of the Markov

11: jump process. Depending on the relative time scales of these two processes, suitably scaled sequences

12: of the estimates

13: are shown to converge to either an

14: ordinary differential equation, or a set of ordinary differential equations modulated by random switching, or

15:  a stochastic differential equation,

16: or stochastic differential equations with random switching. Using weak convergence methods, convergence and rates of convergence of the algorithms

17: are obtained for all these cases.

18: 

19: \vskip 0.1 true in

20: \nd{\bf Key Words.}

21: Sign-error algorithms,

22: regime-switching models,

23: stochastic approximation, mean squares errors,

24: convergence, tracking properties.

25: 

26: \vskip 0.1 true in

27: \nd{\bf EDICS.} ASP-ANAL

28: 

29: \vskip 0.1 true in

30: \nd{\bf Brief Title.} Sign-Error  Algorithms

31: for Markovian Parameters

32: 

33: \end{abstract}

34: