waveguide +ferrite+TLM

hello; I work on modeling a waveguide with ferrite bye transmission line matrix
My code diverge after some 4000 iterations
i used higdon s ABC
Did anyone can explain to me where can be the problem or i have to implement PML

can explen to me all detalles about yoer system work in wave gide

thankes

eng. ihssan

Added after 5 minutes:

i am wating to you
to discuss thes state

thank you

eng.ihssan
iraq-diyala university

I try to model the waveguide bellow

the waveguide is modeling as a PEC
I use a sinusoidal excitation in the plan XOY (Z=1)

and also higdon s absorbing boundary on (Z=1) and (Z=zmax)

the problem is after some iterations ; the transmission coefficient (in the time domain)of the structure oscilate and became sronger (dont converge to a finite value)
Is the problem is due to boundaries?
And What do you suggest