Networks and Communications
Networks and Communications tests layered protocols, IPv4 addressing and subnetting, transport behavior, and link-level error handling. Numeric subnet questions are common and are usually lost through mask or host-count errors.
Layering and transport
The OSI reference model has seven layers; TCP/IP collapses them into application, transport, internet, and link. TCP provides connection setup with a three-way handshake, sequencing, acknowledgments, flow control (receive window), and congestion control; UDP is connectionless with no delivery guarantee. Ports identify processes, not hosts.
IPv4 addressing
CIDR /n leaves 32 − n host bits. For a subnet, usable hosts = 2^h − 2, excluding network and broadcast addresses. Private ranges are 10.0.0.0/8, 172.16.0.0/12, and 192.168.0.0/16. A /26 mask is 255.255.255.192, with blocks of 64 addresses.
Link-level topics
CRC detects burst errors using polynomial division; parity detects odd-bit errors only. ARQ uses acknowledgments and timeouts for reliability. Legacy Ethernet uses CSMA/CD; modern switched full-duplex Ethernet does not collide in the same way.
Worked example
Subnet 192.168.10.0/26. Mask: 255.255.255.192. Block size 64; subnets begin at .0, .64, .128, .192. For the first subnet, network = 192.168.10.0, first host = .1, last host = .62, broadcast = .63, giving 62 usable hosts.
Traps
- Counting the network and broadcast addresses as usable hosts.
- Confusing /26 (192) with /27 (224).
- Mixing TCP’s receive window with the congestion window.
- Treating CRC as encryption or authentication.