yeah guys, I came up with an explanation. So, I wanted to post it here to be useful for others who search.
From Sklar's paper: r(t)=hc(t) * s(t)
'received signal' can be partitioned in terms of two component random variables, as follows:
r(t)=m(t) x ro(t)
where m(t) is called the large scale fading component, ro(t) is called the small-scale fading component.
m(t) is lognormal distributed or if measured in dB, gaussian distributed.
ro(t) is Rayleigh fading or if an LOS dominant path is present, Rician distributed.
For LOS (line-of-sight) case, only direct path is valid, no multipath, no motion. So only large-scale fading is present. It is lognormal distribution.
For NLOS case, both large-scale and small-scale fading is considered. Large scale fading random variable is still lognormal. Small scale is Rayleigh.
I have still implementation complexity though. Those theoretical information will be applied in MATLAB. I will be happy if you have any ideas. Thanks