In a Switch queuing model, should we consider a Poisson process ( Time intervals between two arrivals and two departures follows exponential distributions ) OR Should we say only the input follows an exponential distribution, and the output distribution is uniform? This assumption is made because it seems in real life packets are sent concurrently at different times and no packets are sent between the two units of time.

To clarify more, I am modeling a switch queuing system in Arena and the distribution of Input and Output affect the different time functions like delay of packets. I was wondering if I can use M/M/1/K queuing models to simulate a switch or not?

About the Poisson process, we can simply say that the time between 2 packet arrivals and also 2 packet departures follow exponential distribution (which is kind of random) and they happen continuously in time. Is this possible in a real switch? Or should we consider departure happen only at time units together and nothing happens between two-time units? For more details take a look at the following paragraph from Wikipedia.

Includes applications in wide area network design [11], where a single central processor to read the headers of the packets arriving in an exponential fashion, then computes the next adapter to which each packet should go and dispatch the packets accordingly. Here the service time is the processing of the packet header and cyclic redundancy check, which are independent of the length of each arriving packets. Hence, it can be modeled as an M/D/1 queue.[12]