Understanding Local Area Network Components and Structure
Elements of a Local Network
A local network requires several components, including:
File Server
This is typically a high-capacity computer that functions as the central repository for data and/or software applications on the network. It incorporates mass storage devices.
Workstations
These are the individual computers connected to the network. Computationally intensive tasks and the execution of application programs often occur at these stations.
Cabling
Connects the file server(s) and workstations. Also called transmission media, options include twisted pair, fiber optic, coaxial cables, and radio frequency (wireless).
Hubs
Also known as a wiring hub, it is used to connect devices in a local area network, typically older Ethernet standards like 10 Base-T or 100 Base-T. Hubs operate at the physical layer and broadcast data to all ports.
Switches
Segment a network, providing dedicated bandwidth to each connected station, unlike hubs. This improves performance by reducing collisions.
Routers
A device used to connect different networks or segment a larger network. Routers operate at the network layer, limiting broadcast traffic and providing security, control, and redundancy. They can connect a LAN to a Wide Area Network (WAN) and sometimes act as a basic firewall. A router performs two basic functions:
- Creates and maintains routing tables for network protocols.
- Selects the best path for data based on factors like hop count, delay, transmission cost, line speed, and traffic conditions.
Network Operating System (NOS)
Software that manages network resources, users, and security.
More complex networks may also include the following devices:
Gateways
Hardware and software combination that enables communication between dissimilar networks, such as a LAN and a mainframe, by translating protocols.
Bridges
Connect two separate LAN segments together, making them appear as a single network. They can be:
- Local: Connecting networks within the same building.
- Remote: Connecting networks in different locations, often over WAN links.
LAN Configuration Models
When configuring a LAN, three common models exist:
Peer-to-Peer
Each workstation can share some, all, or none of its resources directly with other workstations. There is no central server.
Shared Resources (Server-Based)
One or more centralized servers manage access to shared resources like files and printers. Workstations access these resources through the server(s).
Client-Server
Applications are distributed between client and server components. The client typically handles the user interface, accepts user input, prepares requests for the server, and presents results to the user. The server receives client requests, processes them (often involving database access), and provides the requested service or data back to the client. This model is prevalent today, leveraging client machines for application execution and powerful servers for data processing and storage.
The Network Interface Card (NIC)
The Network Interface Card (NIC), or network adapter, links a computer to the network, handling both physical and logical connections to enable communication with other devices.
Key components include:
Network Controller
A dedicated microcontroller or chipset that manages the reliable transmission and reception of data over the network.
Connectors (e.g., RJ-45, BNC)
Ports that allow the NIC to physically connect to the network cable. The RJ-45 connector is standard for modern Ethernet networks. BNC connectors were used for older coaxial cable networks.
Transceiver/Adapter Circuitry
Converts the digital signals used by the computer into the electrical or optical signals required for transmission over the network cable, and vice-versa.
Status LEDs
Indicator lights showing network activity (sending/receiving data), connection status, and link speed.
Physical Structure of a LAN
Modern Ethernet networks, such as 100 Base-T (Fast Ethernet) and 1000 Base-T (Gigabit Ethernet), typically use a star topology with devices connected to a central switch using twisted-pair cabling with RJ-45 connectors.
This cabling usually consists of 8 wires arranged in 4 twisted pairs. The specific wiring standard (e.g., T568A or T568B) dictates the pinout (connection order). While most connections go from a computer to a switch, a direct connection between two computers can be made using a crossover cable (though many modern NICs support auto-MDI/MDI-X, making crossover cables unnecessary).
Twisted pair cables come in several types:
- UTP (Unshielded Twisted Pair): Most common, no additional shielding.
- FTP (Foiled Twisted Pair) / ScTP (Screened Twisted Pair): Overall foil shield around all pairs.
- STP (Shielded Twisted Pair): Individual shielding for each pair and sometimes an overall shield.
Shielded cables (FTP, STP) offer better protection against electromagnetic interference but are more expensive and require proper grounding. Cables are also rated in categories (e.g., Cat 5e, Cat 6, Cat 6a) which define their performance characteristics and suitability for different network speeds.