I have the simulation setup as below.

A source of data generator with source impedance 35 Ohm. DataRate 5 GHz with rise and fall time specified as below.
The characteristic impedance of the line is 60 in this setup which matches the 60 Ohm load impedance.


The result voltage at node "V" is shown below, even though the line impedance is matched to the load impedance, there is still ringing.


However if I decrease the length of the line, the ringing is significantly reduced. (the line impedance is not changed, or little if any if I change the length)
Why is this happening? The shorter the line, the less it rings? considering the load and the line are impedance matched already.

 

This is the difference between "lumped" component values and distributed component values.

 

If the transmission line impedance is truly 60 ohms, then the length shouldn't change the waveform.

How are you determining the line impedance?

 

Replace the microstripline with an ideal 60 ohm transmission line. If it still rings then it is the simulation, if not, then it is likely the dispersion in the microstrip (the impedance of the microstrip varies with frequency).