Computer Networks
Computer networks tests layered architecture, framing and error control, MAC and Ethernet, IP addressing and routing, transport protocols, and application protocols. Subnetting and TCP behavior are frequent calculation areas.
Core rules and concepts
- OSI layers: physical, data link, network, transport, session, presentation, application. TCP/IP collapses to link, internet, transport, application.
- CRC detects burst errors using polynomial division over GF(2); checksums are weaker. Stop-and-wait efficiency depends on propagation delay and frame transmission time.
- Ethernet uses CSMA/CD historically; switched full-duplex Ethernet avoids collisions. MAC addresses are link-local; IP addresses are routed.
- IPv4 header has a 20-byte minimum; fragmentation uses identification, flags, and fragment offset in 8-byte units. NAT rewrites addresses and ports.
- TCP uses sequence numbers, cumulative ACKs, sliding window, and congestion control (slow start, congestion avoidance, fast retransmit and recovery). UDP has no reliability or ordering.
Worked example
Subnet 192.168.10.0/26. A /26 mask is 255.255.255.192, leaving 6 host bits. Each subnet has 2^6=64 addresses, of which 62 are usable. The subnet ranges are .0–.63, .64–.127, .128–.191, and .192–.255. The first usable address in the second subnet is 192.168.10.65, the broadcast is 192.168.10.127, and the network address is 192.168.10.64.
Traps that cost marks
Counting the network and broadcast addresses as usable hosts; using the wrong mask for a prefix; confusing port numbers with IP addresses; and assuming TCP guarantees message boundaries-it is a byte stream. In routing, longest-prefix match wins even if another route is more specific administratively.