Week 7: Network Hardware, Media, and IP Foundations

A network is not the Wi-Fi icon.

It is not the router box on the shelf either.

A useful network model starts with a path:

endpoint
  ↓
NIC / radio
  ↓
local media
  ↓
switch or access point
  ↓
local IP network
  ↓
default gateway
  ↓
provider edge
  ↓
remote networks

Each boundary can work while the next one fails.

That is why a device can reach a local printer while the internet is unavailable. It is why an access point can have Ethernet link but remain powered off. It is why a computer can show a valid-looking IP address and still have the wrong gateway.

This week builds the physical and addressing model underneath all of that.

The system question

How far can an endpoint communicate, and which physical or addressing boundary decides what happens next?

How the chapter moves

Lesson 1 separates NICs, switches, routers, access points, firewalls, patch panels, PoE, and provider-edge devices by the job each one performs.

Lesson 2 treats cabling as a compatibility chain. Copper category, shielding, T568 wiring, environmental rating, coax, fiber, connector family, and installation purpose all matter independently.

Lesson 3 connects IPv4, IPv6, private/public addressing, APIPA, static/dynamic assignment, subnet masks or prefixes, and default gateways into one endpoint configuration.

Lesson 4 builds a fictional SOHO network from the provider edge inward and places satellite, fiber, cable, DSL, cellular, WISP, LAN, WAN, WLAN, PAN, MAN, and SAN into real contexts.

Lesson 5 uses the model for local fault isolation. You will prove link, address acquisition, same-subnet reachability, and gateway access before saying "the internet is down."

The rule that matters

Do not troubleshoot the internet before proving the local path.

If the NIC has no link, DNS is not your first problem.

If the endpoint has APIPA when the design expects DHCP, changing the browser will not fix address acquisition.

Networking gets easier when you move through the path in order.