Internet Architecture and IP Addresses

Introduction of TCP/IP Internet

v      Internet Architecture

Ø      Physical network: computers on the same physical network are physically connected.

Ø      Computers on different physical networks are not physically connected.

Ø      IP router (or IP gateway): dedicated systems that connect two or more networks.

Ø      Host: end-user system. It connects to physical networks, and there are possibly many hosts per network

v      The two view of a TCP/IP Internet

v     Packet Transmission

Ø      Source Host:

t         If the destination is on the same physical network, deliver it directly

t         Otherwise, send it to a router

Ø      Intermediate Routers:

t         The destination is not on the same physical network, forward the packet to another router

Ø      Final Router

t         The destination is physically connected to this final router, so send the packet directly to the destination.

v     How do routers work?

Ø      Routers need to find the right routes when forwarding packets.

Ø      Routerst decision is based on the routing information they have

t         Routing table: use destination network, not the destination host; otherwise, the table will be huge.

IP Address

v      Overview

Ø      32 bit binary value

Ø      Unique value assigned to each host

Ø      Values chosen to make routing efficient

v      Dotted Decimal Notation:

Ø      Binary:

Ø      Dotted decimal notation:

v      Classful Addressing Scheme (The original scheme, didntt last long)

v      Classes

Ø      A: 1.0.0.0 --- 126.0.0.0

Ø      B: 128.1.0.0 --- 191.255.0.0

Ø      C: 192.0.1.0 --- 223.255.255.0

Ø      D: 224.0.0.0 --- 239.255.255.255

Ø       E: 240.0.0.0 --- 255.255.255.254

Ø       Example: IBM (9.0.0.0), AT&T (12.0.0.0), Syracuse University (128.230.0.0)

v      Properties of the classful addressing scheme?

Ø      They are self-identifying: the boundary between netid and hostid is self-explained from the address. This can benefit routing because the entries of routing tables store mainly use netid, not the entire IP address.

v      Special Addresses

Ø      255.255.255.255: Limited broadcast (local net)

Ø      0.0.0.0: this host. Can only be used as source address. It is used during bootstrap before a computer knows its IP address. t0t means THIS.

Ø      net + all 1s: directed broadcast for net

Ø      127.anything (often 1): loopback.

v      Classless Addressing Scheme (Devised in 1990s)

Ø      Allow the division between prefix and suffix to occur at an arbitrary point.

Ø      Allow more complete utilization of the address space.

CIDR: Classless Inter-Domain Routing

a)      Internet Part + Local Part

b)      Internet Part + Physical Network + Host

i)        Example: IP:128.230.211.195. Netmask FFFFF800

ii)     

iii)    What is the CIDR representation? What are the lowest IP and highest IP addresses?

iv)    Is Apollo (128.230.208.46) on the same subnet? 208 = 1101 0000

Reserved address prefixes

a)      10.0.0.0 - 10.255.255.255

b)      172.16.0.0 - 172.31.255.255

c)      192.168.0.0 - 192.168.255.255

d)      169.254.0.0 - 169.254.255.255