Foundations of Computer Networks

Introduction to Computer Networks What is a Computer Network? A computer network is fundamentally a set of interconnected computers that share resources and communicate with one another. When you use the internet, send emails, or stream videos, you're relying on computer networks to transport your data from one point to another. Networks work through a combination of three essential elements: Communication protocols are standardized sets of rules that allow computers to understand each other. Without these protocols, a computer wouldn't know how to interpret the data it receives. The most important protocol suite for modern networks is TCP/IP (Transmission Control Protocol/Internet Protocol), which forms the foundation of the Internet. Network nodes are the actual devices connected to the network. These can be personal computers, smartphones, servers (computers that provide services to other computers), networking hardware like routers, or specialized devices. Each node plays a role in the network ecosystem. Physical interconnections are the actual pathways that data travels through. These can be wired (using copper cables or fiber optic cables), optical (using light pulses through fiber), or wireless (using radio frequencies through the air). The choice of interconnection technology affects how fast and reliably data can travel. How Nodes Identify Each Other For computers in a network to communicate, they need a way to find and identify each other—much like how you need an address to send a letter. This is where network addresses come in. A network address is a unique identifier assigned to each node on the network. The most common type is an IP address (Internet Protocol address), which looks like 192.168.1.1. IP addresses allow the Internet Protocol to route data to the correct destination. Think of it this way: when you want to visit a website, your browser needs to find the server hosting that website. The server has an IP address, and the network uses this address to deliver your request and send back the response. Without addresses, data would have no way of reaching its intended destination. How Networks Are Classified Computer networks vary greatly in their characteristics and purposes. Networks can be categorized using several criteria: Transmission medium: The physical method used (wired, optical, or wireless) Bandwidth: How much data the network can transmit per unit time Communications protocols: The rules the network follows Network size: From small personal networks (PAN) to global networks (WAN) Topology: The physical or logical arrangement of nodes Organizational intent: Whether it's for research, business, public use, etc. Understanding these classification criteria helps you appreciate why different networks are built differently. A hospital network prioritizes reliability and security differently than a public WiFi network, for example. <extrainfo> Historical Context: The Origins of Modern Networking The computer networks we use today didn't emerge overnight. Understanding their development helps explain why networks work the way they do. Packet switching was a revolutionary concept invented independently during the 1960s by Paul Baran and Donald Davies. Instead of sending data as one continuous stream (like a phone call), packet switching breaks data into small chunks called packets and sends them across the network independently. Baran's work focused on how these packets could be routed adaptively across a distributed network, while Davies designed hierarchical network structures with high-speed routers. TCP/IP, the protocol suite that powers the modern Internet, was formalized by Vinton Cerf and Robert Kahn in their 1974 publication "A Protocol for Packet Network Intercommunication." This framework became the foundation for all Internet communication. Ethernet, introduced by Robert M. Metcalfe and David R. Boggs in 1976, was a practical implementation of packet switching for local area networks (LANs). Ethernet's success and standardization by the IEEE led to the widespread adoption of local computer networks we see today. These historical developments explain why modern networks use packet switching and why TCP/IP is the universal standard for Internet communication. </extrainfo> Understanding Network Packets What is a Network Packet? The core mechanism that makes packet switching work is the network packet. A network packet is a formatted unit of data that travels across a packet-switched network. Every time you load a webpage, send a message, or download a file, that data is broken into packets before being transmitted. This image illustrates how packets can be transmitted along different paths through a network from source A to destination B, potentially through intermediate routing points like R. A network packet has a clear structure with three main components: Headers contain control information—data about the packet itself rather than the payload. Headers include: The source network address (where the packet came from) The destination network address (where it's going) Error detection codes (used to check if the packet arrived uncorrupted) Sequencing data (to reassemble packets in the correct order if they arrive out of sequence) Payload (also called user data) is the actual information being transmitted—the text of an email, the image you're viewing, or part of a video you're watching. This is what the packet is fundamentally trying to deliver. Trailers (also called footers) are placed at the end and typically contain additional error-checking information to ensure data integrity. Why Packet Switching Matters Packet switching offers significant advantages over circuit switching (the older telephone network model where a dedicated path was established for the entire duration of a call). With packet switching, the network's bandwidth is shared among many users much more efficiently. Imagine a highway with multiple cars: in circuit switching, one user gets an entire lane for as long as they need it, even if they're stopped. In packet switching, many users' packets share the same lanes—when one user's data isn't actively traveling, other users' packets can use that bandwidth. This approach dramatically reduces overall cost and network interference because no bandwidth is wasted on idle connections. What Happens When Networks Get Busy: Queuing and Fragmentation Two important phenomena occur in real networks that you should understand: Queuing happens when a packet arrives at a network node but the outgoing path isn't immediately available. Instead of disappearing, the packet waits in a queue until a link becomes free. This is why networks can experience delays during heavy traffic—packets are waiting their turn, much like cars waiting at a traffic light. Fragmentation occurs due to a limit called the maximum transmission unit (MTU). Each physical link in a network has a maximum size packet it can transmit. If your data is larger than this limit, it gets broken into smaller fragments that are transmitted separately, then reassembled at the destination. This might seem inefficient, but it's necessary because different network technologies have different physical constraints. <extrainfo> Building on Packet Networks: Additional Protocols Once packet switching was established, additional protocols were layered on top to provide specific services. The Simple Mail Transfer Protocol (SMTP), standardized in RFC 5321 (2008), specifically handles email transmission over networks. RFC stands for "Request for Comments"—the standard format for Internet protocol specifications. The Domain Name System (DNS), specified in RFC 1035 (November 1987), solves a fundamental usability problem: humans remember names (like google.com) much better than IP addresses. DNS is the system that maps human-readable domain names to the IP addresses that computers actually use for routing. </extrainfo> Distributed Computing: Leveraging Network Resources One powerful use of computer networks is distributed computing, where resources from multiple computers across a network are combined to perform tasks collaboratively. Instead of one powerful computer doing all the work, many computers work together on the same problem. This approach is used for everything from scientific simulations to cloud computing services.