Network Topologies
What Is a Network Topology?
A network topology describes how the devices in a network are connected and how data travels between them. Choosing the right topology involves trade-offs: a design that minimises cabling cost may create a single point of failure; a design that is highly resilient may be expensive to install and maintain. The three topologies you must know are bus, star and mesh.
Understanding topologies matters for designing reliable networks. A small business setting up a new office will consider cost and ease of adding new machines. A hospital designing its patient-records network will prioritise reliability and fault tolerance. The “best” topology depends entirely on what the network is for.
Bus Topology
In a bus topology, all devices share one central cable — called the backbone or bus. When any device sends data, the signal travels along the backbone in both directions. Every device on the network receives the signal, but only the device with the matching address actually processes it. Terminators at each end of the backbone absorb signals so they do not bounce back.
Advantages:
- Cheap to install — uses the least amount of cable of the three topologies, because all devices share one backbone.
- Easy to add new devices — a new computer is simply connected to the backbone.
- Simple to set up — no central switch or hub is required.
Disadvantages:
- A single point of failure — if the backbone cable breaks, the entire network fails because there is no alternative path for data.
- Performance degrades as more devices are added — all devices share the same bandwidth, so collisions become more frequent on a busy network.
- Difficult to fault-find — a fault anywhere on the backbone is hard to isolate.
- Not suitable for large networks because of bandwidth sharing and reliability issues.
In a bus topology, the backbone cable is a single point of failure. If it breaks, the entire network stops because there is no alternative path for data. This makes bus topology unsuitable for environments where continuous network access is critical.
Star Topology
In a star topology, each device has its own dedicated cable connecting it to a central switch (or hub). Data from any device travels to the switch, which reads the destination address and forwards the data only to the correct device. Modern networks almost universally use star topology for wired LANs.
Advantages:
- Failure of one cable only affects the device on that cable — the rest of the network continues to function normally.
- Fast performance — each device has a dedicated connection to the switch, so bandwidth is not shared in the same way as a bus.
- Easy to add new devices — simply run a new cable from the device to the switch, without disrupting any existing connections.
- Easy to identify faults — if a specific device loses connection, the problem is isolated to that device’s cable or connection.
Disadvantages:
- The central switch is a single point of failure — if it breaks down, the entire network is disrupted because all communication passes through it.
- More cable required than bus topology — each device needs its own run of cable to the central switch, which increases installation cost.
- Cost of the central switch — a high-quality managed switch is an additional expense.
Mesh Topology
In a mesh topology, devices are interconnected with multiple paths between them. In a full mesh, every device connects to every other device directly. In a partial mesh, most devices have multiple connections but not every possible connection exists. The internet’s core infrastructure uses a partial mesh, which is why traffic can be re-routed around a failed link automatically.
Advantages:
- Highly resilient — if one link or device fails, data can be re-routed via alternative paths. There is no single point of failure (in a full mesh).
- Fast data transfer — data can take the shortest or least congested path between any two devices.
- Well suited to high-reliability environments such as core internet infrastructure, hospital networks or military communications.
Disadvantages:
- Very expensive to install — a full mesh of 10 devices requires 45 separate connections (the formula is n(n−1)÷2 connections for n devices).
- Complicated to manage and configure — every device must be set up to handle multiple connections and routing decisions.
- Uses a large amount of cable (for wired mesh) or radio spectrum (for wireless mesh), which adds to cost and complexity.
Topology comparison at a glance: Bus — cheapest, backbone is a single point of failure, poor scalability. Star — moderate cost, switch is a single point of failure but individual cable faults are isolated, most common for modern LANs. Mesh — most expensive, no single point of failure (full mesh), used for critical infrastructure requiring high resilience.
Scenario: A small café wants to connect five computers and a printer with the lowest possible installation cost. A hospital wants to connect its patient-records servers so that no single cable break can interrupt access.
Café: Bus topology is the cheapest option — one backbone cable serves all five computers and the printer. The café accepts the risk that a backbone fault will take down everything, because the cost saving is the priority and the network is small and not mission-critical.
Hospital: Mesh (or star with redundant links) is appropriate. Patient records are mission-critical: any network downtime could affect patient safety. The higher cost of multiple links is justified by the need for resilience. A full mesh of five servers requires only 10 connections: 5 × 4 ÷ 2 = 10 — manageable for a small cluster of critical servers.
Typical school LAN: Star topology. Each classroom computer has its own cable run to a central switch. If one cable is damaged, only that one computer is affected. Adding new computers is straightforward — run a new cable to the switch. The central switch is the only point of failure but can be replaced without rewiring the whole building.