Simulation :
Divergence of RLS Algorithm
(a) Simulate the divergence in the RLS algorithm. One way to achieve this is to is to round-off every variable
to, say, the third or second significant digit, at the end of each iteration (if divergence still does not occur,
you may have to round-off to first significant digit). This way, we artificially introduce round-off errors and
examine its accumulation effect.
(b) For the the procedure of part (a) which simulates divergence, now implement the reinitialization scheme
for the RLS algorithm and check whether it is able to overcome the problem of divergence. Thus, in the
reinitialization algorithm too, you will be rounding off each variable to third or fourth significant digit.
Problem solution
The first three significant digits were considered for divergence analysis of RLS algorithm.
