CEN610S Core Engineering Networks - IPv4 Addressing and Prefixes

IPv4 Addressing Fundamentals

• IP version 4 (IPv4) uses $32\text{-}bit$ addresses; total address space = $2^{32}$ addresses.

• Each host or router must have a unique address.

• Two key elements: Address and Subnet mask.

• Addressing written in dotted quad notation $A.B.C.D$ with each octet being $8$ bits (example: $18.31.0.1$).

IP Prefixes – Modern

• Addresses are allocated in blocks called prefixes; all hosts in an $L\text{-bit}$ prefix share the same top $L$ bits.

• Number of addresses in a prefix: $2^{32-L}$.

• CIDR notation: $A.B.C.D/L$;

  • E.g., $128.13.0.0/16$ covers $128.13.0.0$ to $128.13.255.255$.

  • E.g., /24 yields $2^8 = 256$ addresses; /32 yields 1 address.

• Binary representations (optional): e.g. a /24 prefix defines the first 24 bits as the network; the remaining 8 bits are host bits.

IP Prefixes – More Specific / Less Specific

• More specific prefix → longer prefix → smaller set of addresses.

• Less specific prefix → shorter prefix → larger set of addresses.

• Common prefix lengths: $/8, /16, /24, /32, /0$.

IP Address Classes – Historical

• Classful addressing used fixed-size blocks with the class encoded in high-order bits; networks/hosts split by class:

  • Class A: 8-bit network portion

  • Class B: 16-bit network portion

  • Class C: 24-bit network portion

• These class distinctions are historical and largely ignored in favor of CIDR.

Public / Private IP Addresses

• Public IP addresses: globally routable on the Internet; must be allocated; space is finite.

• Private IP addresses: internal use, not routable on the public Internet; NAT used to reach the Internet.

• Private ranges (examples):

  • $10.0.0.0/8$

  • $172.16.0.0/12$

  • $192.168.0.0/16$

Allocating Public IP Addresses

• Allocation follows a hierarchy: IANA → RIRs (Regional Internet Registries) → ISPs/enterprises → customers/computers (via DHCP).

• RIRs include: ARIN, APNIC, RIPE, LACNIC, AfriNIC.

IP Prefix in Practice / Subnet Mask

• Subnet mask is often described by prefix length (e.g., /16) rather than decimal; default for Class B is /16 ≡ $255.255.0.0$.

• Subnet mask helps decide if the destination is local to the sender's network.

• Process (routing decision): devices perform ANDing between source/destination addresses and the subnet mask to determine locality.

ANDing and Locality (Subnetting)

• Definition: local if $(S \& M) = (D \& M)$, where

  • $S$ = Source address, $D$ = Destination address, $M$ = Subnet mask.

• Steps:

  • Compute source network ID: $S_{net} = S \& M$

  • Compute destination network ID: $D_{net} = D \& M$

  • If $S<em>{net} = D</em>{net}$, destination is local; otherwise, route via a gateway.1.1