Skip to main content

The Internet & Performance

The vocabulary for talking about any network. This module starts with what the Internet physically and commercially is, moves on to how packets share links and who pays to carry them, then names the layers every later module hangs on. It ends with the numbers that describe a path: delay, jitter, loss, throughput and the bandwidth-delay product.

What you will be able to answer​

  • What is the Internet, and what is the one thing every network on it has to agree on?
  • What is the difference between the network edge and the core, and which devices sit where?
  • Why do packets beat circuits for bursty traffic, and what does store-and-forward add to the delay over N links?
  • How does statistical multiplexing let a shared link carry more users than its peak demand would suggest?
  • Who pays whom between ISPs, and how do transit, peering, IXPs and autonomous systems differ?
  • Which five TCP/IP layers and seven OSI layers exist, and what changes at each hop when a request crosses a router?
  • What are the four nodal delays, why does queuing blow up near full utilization, and how do latency and bandwidth differ?
  • What is the difference between bandwidth, throughput and goodput, and how does the bottleneck link cap a path?
  • Why does bandwidth times RTT decide whether one TCP flow can fill a fast, long pipe?
On this page