How to combine Poisson arrival with Exponential service time in m/m/n/n queue process?

297 Views Asked by At

I have a discrete-time process that follows a queueing m/m/n/n discipline (Earling-B Model).

The system can at most serve $n$ requests simultaneously as there are $n$ servers.

In this case,

The time-window is divided into slots. Lets say, there are $T$ time-slots, each time-slot is of duration $\Delta=10$ milli-seconds. The length of the time-window is $T\times\Delta$ milli-second.

The traffic arrival at each time-slot is a Poisson process with mean, $\lambda=0.2$/time-slot. The service time follows an exponential distribution with mean, $V=2$ time-slots.

How to get the number of users/requests being served in each time-slot, Lets say, $T=100$.

Lets say, I have a grid of $5 \times 100$. each request with 1 time-slot service time fills $1\times1$ grid space. Likewise, each request with 2 time-slot service time fills $1 \times 2$ grid space.