Queuing Delay Calculator and Router Queue Simulation

Queuing delay is the time a packet waits in the router's buffer until the link is free. It varies packet by packet and explodes as the traffic intensity La/R approaches 1.

About this simulator

Queuing delay is the time a packet waits in a router's buffer before it can be transmitted. Unlike the other delays it changes from packet to packet and depends on traffic. This simulation runs a real first-in-first-out queue so you can compare a burst of packets with steady and random arrivals and see how traffic intensity decides everything.

How it works

  1. Packets arrive at the router and join a FIFO queue if the link is busy transmitting another packet.
  2. Each packet waits for every packet ahead of it to finish transmitting (L/RL / R each).
  3. If the arrival rate times the packet length exceeds the link rate (La/R>1La/R > 1) the queue grows without bound.

Formulas used

  • Traffic intensity=L aR\text{Traffic intensity} = \dfrac{L\,a}{R}
  • Burst of N packets: average wait=(N−1) L2R\text{Burst of } N \text{ packets: average wait} = \dfrac{(N-1)\,L}{2R}
  • dqueue(k)=max⁡(0, link free time−arrival time)d_{\text{queue}}(k) = \max(0,\ \text{link free time} - \text{arrival time})

Worked example

Five 1,500-byte packets arrive together at a 10 Mbps link (each takes 1.2 ms1.2\ \text{ms} to transmit). The first waits 00, the second 1.2 ms1.2\ \text{ms}, then 2.42.4, 3.63.6 and 4.8 ms4.8\ \text{ms}. The average is 2.4 ms2.4\ \text{ms}, exactly (N−1) L2R=4×1.22 ms\dfrac{(N-1)\,L}{2R} = \dfrac{4 \times 1.2}{2}\ \text{ms}. With traffic intensity La/RLa/R above 1 the waits keep growing without limit.

Enter these numbers in the simulator above to see every step.

When to use it

  • Practising the average queuing delay for a burst of N packets.
  • Seeing why buffers fill when arrival rate times packet size exceeds the link rate.
  • Comparing steady traffic with random (Poisson) arrivals.

Frequently asked questions

What is traffic intensity?

Traffic intensity is L aR\dfrac{L\,a}{R}: the average packet arrival rate aa times the packet length LL, divided by the link rate RR. When it is below 1 the queue stays bounded; at or above 1 the queue grows without bound.

How do you calculate queuing delay for N packets arriving at once?

The kk-th packet waits (k−1) LR(k-1)\,\dfrac{L}{R}. The average over NN packets is (N−1) L2R\dfrac{(N-1)\,L}{2R}.

Why does queuing delay change from packet to packet?

Each packet finds a different number of packets ahead of it, so unlike transmission or propagation delay it depends on traffic, not just on the packet and the link.