3.4Networks (OSI/TCP-IP Layers, IP Addressing, DNS/DHCP, Protocols, Bandwidth Calculation)

Networking is an area of the FE exam where bandwidth/time calculation problems and understanding of layered structures are tested together. Building on a solid grasp of how the OSI reference model maps to TCP/IP layers, how IP addresses and subnet masks divide a network, and how DNS and DHCP divide responsibilities, you can then tackle the frequently tested protocols and calculation problems with more stable results.

3.4.1The OSI reference model and TCP/IP layers

• The OSI reference model divides the roles of communication into seven layers (from the top: application, presentation, session, transport, network, data link, physical). Each layer builds only its own responsibility on top of the functions of the layer below it.
• The TCP/IP stack is simplified in practice to four layers (application, transport, internet, network interface). A classic distinction at the transport layer: TCP prioritizes reliability (retransmission, ordering guarantees), while UDP prioritizes speed (no retransmission—used for video streaming, etc.). The internet layer corresponds to the OSI network layer and handles path selection (routing) via IP addresses.

3.4.2IP addressing, subnetting, routing, and DNS/DHCP

• An IP address identifies a device on a network (IPv4 is 32 bits, written as four decimal groups of 8 bits each). A subnet mask divides the address into a "network portion" and a "host portion," determining whether two addresses are on the same network (e.g., 255.255.255.0 means the first 24 bits are the network portion and the remaining 8 bits are the host portion, accommodating up to 254 hosts).
• Routing is the process of choosing a forwarding path for packets between different networks (a router looks at the network portion of the destination IP address to decide the next hop). DNS (Domain Name System) is the mechanism that translates human-readable domain names into IP addresses (e.g., translating example.com to 203.0.113.1). DHCP is the protocol that automatically assigns IP addresses to network devices (avoiding the effort of manual configuration and the risk of address duplication).

3.4.3Key protocols and bandwidth calculation

• HTTP is the protocol for retrieving web pages (plaintext). HTTPS is HTTP encrypted with TLS, preventing eavesdropping and tampering. SMTP is the protocol for sending mail. POP (POP3) downloads mail from the server for local management on the device. IMAP keeps mail managed on the server, letting multiple devices view the same state (the POP-vs-IMAP distinction is frequently tested).
• A LAN is a network confined to a limited area (an office, a home, etc.). Wired LANs use Ethernet (classified by cable standard and transfer speed); wireless LANs use Wi-Fi (no cabling needed, but interference and security configuration are concerns).
• Bandwidth calculation: transfer time = data size / (line speed x transfer efficiency). The most important step is aligning units to bits (1 byte = 8 bits). Example: sending a 10MB file over a 100Mbps line at 50% efficiency—the data size is 10 x 8 = 80 Mbits, the effective speed is 100 x 0.5 = 50 Mbit/s, so the transfer time is 80 / 50 = 1.6 seconds.

Consider a scenario where a branch office is connected to headquarters over a network to access a business system. When a branch PC accesses the HQ web server, the domain name typed into the browser (e.g., portal.example.com) is first translated into an actual IP address by DNS. Next, if the branch and HQ belong to different networks, packets are forwarded along an appropriate path via routing. In doing so, a subnet mask is used to determine whether a branch PC is on the same subnet as another device—for example, on the network 192.168.1.0/24, the first 24 bits form the network portion and the remaining 8 bits the host portion, accommodating up to 254 devices. Manually configuring an IP address every time a new employee's PC is added is inefficient and risks duplication, so a DHCP server assigns addresses automatically instead. If the traffic carries confidential customer data, the design must use encrypted HTTPS rather than plaintext HTTP. Now consider a practical calculation problem: transferring a 600MB business data file from the branch to HQ over an 80Mbps line at 75% effective transfer efficiency—how do you find the time required? First align the data size to bits: 600MB x 8 = 4800 Mbits. The effective speed is 80 x 0.75 = 60 Mbit/s. So the time required is 4800 / 60 = 80 seconds—carefully working through the unit conversion and the efficiency multiplication step by step gets you to the correct answer. Leaving the calculation in bytes, or forgetting to multiply by the transfer efficiency, are the classic ways to lose points here. As for mail, if the requirement is to share the same inbox state (read/unread, folders) across multiple laptops and a smartphone, IMAP, which manages mail centrally on the server, is the right fit; conversely, if everything is meant to be self-contained on a single device with limited storage, POP, which downloads mail for local management, is the better fit.

3.4.4Section summary

• OSI = 7 layers; TCP/IP = 4 layers (a simplified practical version). TCP prioritizes reliability, UDP prioritizes speed. A subnet mask splits an address into network/host portions
• DNS = name resolution; DHCP = automatic address assignment. HTTPS = TLS-encrypted HTTP. POP = local device management; IMAP = server-side management
• Bandwidth calculation = data size / (line speed x transfer efficiency). Keep units in bits (1 byte = 8 bits) and watch for forgetting the efficiency multiplier