Virtual Lab
Log InSign Up

Packet Switching Simulator: Store-and-Forward and Pipelining

In packet switching a message is cut into packets that are forwarded hop by hop. Because later packets can use earlier links while earlier packets travel on, the packets are pipelined and the total time is much shorter than sending the whole message at once.

About this simulator

In packet switching a message is split into packets that travel hop by hop. Each router must receive a whole packet before forwarding it (store-and-forward), but different packets can be on different links at the same time. The Gantt-style diagram shows this pipelining and the formula that predicts the total delivery time.

How it works

  1. Each router must receive a whole packet before it forwards it (store-and-forward).
  2. While packet 2 is crossing link 1, packet 1 is already crossing link 2 — the links work in parallel.
  3. NN packets over KK links of rate RR finish after (N+K−1)×L/R(N + K - 1) \times L / R (ignoring propagation).

Formulas used

  • Tpackets=(N+K−1) LRT_{\text{packets}} = (N + K - 1)\,\dfrac{L}{R}
  • Tmessage=K MRT_{\text{message}} = K\,\dfrac{M}{R} (one big packet)
  • N=⌈M/L⌉N = \lceil M / L \rceil

Worked example

A 15,000-byte message is cut into N=10N = 10 packets of 1,500 bytes and sent over K=3K = 3 links of 10 Mbps. Each packet needs 1.2 ms1.2\ \text{ms} per link, so the total is (N+K−1) LR=(10+3−1)×1.2=14.4 ms(N + K - 1)\,\dfrac{L}{R} = (10 + 3 - 1) \times 1.2 = 14.4\ \text{ms}. Sending the whole message as a single packet would take K MR=3×12 ms=36 msK\,\dfrac{M}{R} = 3 \times 12\ \text{ms} = 36\ \text{ms} — packet switching is 2.5×2.5\times faster here.

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

When to use it

  • Solving store-and-forward delay problems with N packets and K links.
  • Understanding why smaller packets reduce end-to-end delay on multi-hop paths.
  • Comparing packet switching with circuit switching and message switching.

Frequently asked questions

What is the formula for packet switching delay?

For NN equal packets of length LL crossing KK links of rate RR (ignoring propagation and queuing), the total time is (N+K−1) LR(N + K - 1)\,\dfrac{L}{R}.

What does store-and-forward mean?

A router receives the entire packet before it starts transmitting it on the next link. This adds one packet transmission time per hop but allows error checking.

Why is packet switching faster than sending one big message?

Because pipelining lets several links work at once: while one packet crosses link 2, the next is already crossing link 1.