Topic 7 - Network Layer

What is the primary function of the network layer?

It provides inter-network communication by forwarding packets between devices on separate networks and enabling end-to-end communication.

What is a Local Area Network (LAN)?

A group of computers interconnected within a single physical network allowing easy internal communication.

What separates networks from each other?

Routers act as gateways between networks.

What are network bridges?

Devices that connect two or more LANs and enable communication between them.

When can communication be done solely through the data link layer?

When all routers and networks involved are under the same administrative control.

Why is the network layer necessary for Internet communication?

"Because it allows communication between networks we don’t control

How does the network layer deliver messages?

"It forwards packets hop-by-hop through routers

What is IPv4?

A network layer protocol that enables communication between previously unknown networks.

What is the most widely used network layer protocol?

The Internet Protocol (IP).

What are the two versions of IP?

IPv4 and IPv6.

What is ICMP used for?

Internet Control Message Protocol is used for error reporting and diagnostics.

How does the IP header behave along the packet's journey?

It mostly remains the same except for fields like TTL which decrease at each hop.

How does this differ from the Data Link layer Protocols?

In Data Link layer protocols like Ethernet or 802.11, Frame Headers are removed at each hop

What is the most important information in the IP header?

The source and destination IP addresses.

What is an IP address?

A 32-bit number that uniquely identifies a device on the internet. Each device on the internet has one

Who assigns IP addresses?

Internet Service Providers (ISPs) temporarily lease them to users from their available address pools in return for payment.

Do you always have the same IP address?

No, unless you pay for business-class internet with a static IP address

What is the format of an IPv4 address?

Four numbers from 0–255 separated by dots (dotted decimal notation). Example: 172.16.254.1.

How many unique IPv4 addresses exist?

Roughly 4.2 billion (2^32)

What is the IPv4 address exhaustion problem?

"All IPv4 addresses have been allocated globally

What are the two parts of an IP address?

The network portion and the host portion.

What does the network portion identify?

The specific network the device belongs to.

What does the host portion identify?

An individual device within a network, allowing them to have the same prefix

What happens when the network portion is smaller?

More devices are allocated to a single host, allowing them to have more connected devices

What are the implications of fewer bits in the Network Portion?

There are more bits allocated for the host component allowing for more devices, but it reduces the total number of unique networks (contributing to the IPv4 Exhaustion Problem

What was the historical IP addressing system?

Classful addressing

How were addresses divided in classful addressing?

At every 8 bits (1 octet)

What range of first octets defines Class A addresses?

0–126. (1 Network Octet, 3 Host Octets)

How many devices can a Class A network support?

2^24 - 2 ≈ 16 million devices.

What range of first octets defines Class B addresses?

128–191. (2 Network Octets, 2 Host Octets)

How many devices can a Class B network support?

2^16 - 2 = 65,534 Devices

What range of first octets defines Class C addresses?

192–223 (3 Network Octets, 1 Host Octets)

How many devices can a Class C network support?

2^8 - 2 = 254 hosts.

What was the main problem with classful addressing?

It had only three fixed network sizes

What problem did small companies face under classful addressing?

They could outgrow their Class C allocation and need multiple blocks, either needing another Class C address or upgrade to a Class B address

Why was upgrading between classes inefficient?

It created huge jumps in available addresses (256× more)

How many total Class A networks could exist?

Only 126 Class A networks

What replaced classful addressing?

Classless Inter-Domain Routing (CIDR).

What is CIDR?

A system that allocates IP addresses with variable-length network prefixes for more efficient use of address space.

How does CIDR improve flexibility?

Instead of changing entire octets, a single bit is moved from the Networks to the Hosts

What are the two forms of IP addresses in CIDR?

CIDR network addresses (with Network and Host portions and a subnet identifier) and individual host IP addresses.

What does the number after the slash in a CIDR address represent?

The number of bits used for the network portion (subnet identifier).

Example of a CIDR Address: 64.64.0.0/20, what does /20 mean?

20 bits are allocated for the network, while the remaining 12 bits are allocated for the hosts

How many hosts can a /20 network support?

2^12 - 2 = 4094 hosts (subtracting network and broadcast addresses).

What address identifies the network itself?

The one where the host bits are all zeros.

What address is used for broadcasting within the network?

The one where the host bits are all ones.

How are CIDR and regular IP Host addresses different?

CIDR addresses have a subnet identifier while IP hosts don’t, as uniqueness is the only thing that matters for IP addresses

What is the key difference between CIDR and regular IP addresses?

IP addresses identify individual machines while CIDR addresses identify entire networks (ranges of IPs)

How does CIDR fix allocation problems?

It allows resizing networks by shifting bits to expand or shrink host capacity without wasting address space. (Making network upscaling/downscaling easier)

Example: What happens when changing from /20 to /19?

The network portion shrinks by one bit, allows for twice as many hosts while ensuring that all devices share the same network prefix

What is the protocol data unit (PDU) of IPv4 called?

An IPv4 packet or IPv4 datagram.

What does the IPv4 datagram contain?

32 bits. The header includes control information like version, IHL,Type of Se3rvice

What happens to the IPv4 datagram before transmission?

It’s payload is encapsulated with a Data Link layer header before being sent over for transmission.

What do the first four bits of the IP header represent?

The IP version, Internet Head Length (IHL), Type of Service, and Total Length

What does the Internet Header Length (IHL) field specify?

It indicates the size of the IP header. The header length is variable because of the options field

What is the minimum and most common IP header size?

20 bytes

How can we derive the 20-byte header size?

"Each row in the IP header diagram is 32 bits (4 bytes)

What does the Type of Service field represent?

It specifies information for Quality of Service technologies

What does the 'Total Length' field represent?

The total size of the entire IP packet

What is the maximum size of an IP packet?

65,535 bytes

What does the Identification field do?

It collects and groups together fragments of a packet so they can be reassembled correctly at the destination.

What triggers IP packet fragmentation?

When a packet exceeds the Maximum Transmission Unit (MTU) of a link.

What are the three flags in the IP header’s fragmentation section?

Reserved bit (always 0), Don’t Fragment bit (indicates whether datagram can be fragmented, More Fragments (indicates more fragments)

What is the Fragment Offset field?

13 bit field that helps sequence fragmented packets, indicates where in the message each fragment should be placed

What does the TTL (Time to Live) field do?

8 bit field that represent how many router hops a datagram can traverse before its discarded

Why is TTL important?

It prevents routing loops and endlessly circulating packets caused by misconfigured routers by decrementing after each hop and dropping at 0

What is the default TTL value?

64 hops

What is the Protocol field?

8 bit field that determines the transport layer protocol used in the header above

What does is the Header Checksum?

16 bit field that provides Basic data integrity verification. Not found in IPv6 (handled by DLL)

Why is the header checksum less important today?

Because Ethernet and 802.11 already handle integrity through Frame Check Sequences

What is a routing table?

A list of destination networks and corresponding next-hop addresses or interfaces.

What does a router do when it receives an IP packet?

It checks if the destination is on its own network and makes the final delivery. If not it will forward the message to the next router on a specified interface

What is the default gateway?

A special 'catch-all' route used when no other routing rule matches, usually the ISP’s IP.

What happens to the TTL field as a packet passes through routers?

It decreases by 1 at each hop.

When is fragmentation performed by routers?

Before sending data to the next hop on a specified interface

What are reserved IP ranges used for?

Addresses that are not globally routable and exist only for local use.

What is 127.0.0.0/8 used for?

Loopback address range — used for a host to send messages to itself (Used in Netcode dev)

What is 192.0.0.0/24 used for?

A typical local area network (LAN) address block supporting up to 256 devices

What is 10.0.0.0/8 used for?

Another private LAN range for larger networks

What is 224.0.0.0/4 used for?

IP multicast

What is IP Multicasting

Uses a one-to-many association to route datagrams simultaneously in a single transmission to multiple recipients, message is sent once and received on multiple machines through specified IPs

What is an example use of IP multicast?

CCTVs, PC hosting, Software Updates, Conference Calls

Why doesn’t multicast traffic leave your network?

It uses local IP ranges that are non-routable on the public internet.

What is DHCP used for?

Automatically assigns a temporary (dynamic) IP addresses to devices on a local network.

What are the four stages of the DHCP process?

Discover, Offer, Request, Acknowledgment

What happens during DHCP Discover?

The new device broadcasts a request containing its MAC address to find available IPs

What happens during DHCP Offer?

Servers with available IPs respond with offers to lease an address

What happens during DHCP Request?

The client accepts the first offer received, assuming the first has the least latency

What happens during DHCP Acknowledgment?

The server logs the lease in the DHCP Table to keep track of in use local IP addresses and sends an acknowledgement containing exp. dates

What are dynamic IPs?

Temporary IPs assigned via DHCP that revert back to the server once expired

What devices typically use dynamic IPs?

Devices that don’t require consistent local addresses (Laptops)

What is a static IP address?

A manually assigned address that doesn’t expire unless revoked

Why are static IPs useful?

DHCP automatically connects a device to an IP address in its pool, so it’s good for devices that others connect to (Network Printers)

What problem can occur with DHCP and network printers?

If a printer’s IP address changes because DHCP gives it a new one, the print job may be sent to the wrong location

Do IP addresses belong to devices or networks?

They belong to the network, devices lease them temporarily

How does a MAC address differ from an IP address?

The MAC address is permanent and hardware-specific, it doesn’t change across networks.

What is ARP (Address Resolution Protocol) used for?

DLL Protocol used to map a known IPv4 address to a device’s MAC address on the local network, cannot be used outside Local Networks

What is stored in an ARP table?

A list of IP and MAC addresses associated with each device

Do ARP table entries expire?

Yes, they expire after a short window of time to account for network changes

How can you view your ARP table on Windows?

By typing 'arp -a' in Command Prompt.