Network Topologies

Network topology refers to the layout or arrangement of devices and communication links in a computer network. It defines how computers, servers, switches, routers, and other devices are connected and how data moves between them — essentially, the blueprint of a network.

Think of a city's road system: roads represent network cables or wireless links, vehicles represent data packets, and intersections represent network devices such as switches and routers. Just as the arrangement of roads determines how efficiently traffic moves, network topology determines how efficiently data travels across a network.

Physical Topology vs. Logical Topology

Network topology is described from two different perspectives, and it's important not to confuse them.

A network can look like one topology physically while behaving like a different one logically. For example, an Ethernet LAN is almost always wired as a physical star — every device cables back to a central switch — but older, hub-based Ethernet networks behaved like a logical bus, because data was broadcast to every device sharing the medium. Keeping this distinction in mind matters later in the lesson, when "logical bus," "logical ring," and "logical star" behavior are discussed separately from the cabling layout.

Why Network Topology Matters

The topology chosen for a network directly affects:

• Performance — how much data can move without bottlenecks
• Scalability — how easily new devices can be added
• Reliability and fault tolerance — how much the network is affected when one link or device fails
• Security — how easily traffic can be intercepted or isolated
• Cost — how much cabling and hardware is required
• Maintenance — how easy it is to locate and fix problems

Choosing the right topology up front can meaningfully improve network efficiency and reduce long-term operational cost, which is why it's one of the first decisions made when designing a network.

Major Types of Network Topology

There are six commonly recognized network topologies:

1. Bus Topology

In a bus topology, every device connects to a single shared cable called the backbone. When a device transmits, the signal travels in both directions along the backbone, passing every other device connected to it. Each device checks the destination address carried in the signal; only the intended recipient accepts the data, and every other device ignores it. A terminator is placed at each end of the backbone to absorb the signal and stop it from reflecting back and interfering with new transmissions.

Advantages

• Inexpensive — requires the least cabling of any topology
• Simple to set up for small or temporary networks
• A single failed device usually doesn't affect the rest of the network

Disadvantages

• The backbone cable is a single point of failure — if it breaks, the entire segment goes down
• Faults are hard to isolate, since the break could be anywhere along the cable
• Because the medium is shared, two devices transmitting at once cause a collision
• Performance degrades as more devices are added, since they all compete for the same cable

Bus topology was common in early Ethernet (coaxial-cable networks such as 10BASE2 and 10BASE5) but has been almost entirely replaced by star topologies built around switches.

2. Ring Topology

In a ring topology, each device connects to exactly two neighboring devices, forming a closed loop. Data passes from device to device around the ring until it reaches its destination.

Many ring networks avoid collisions using token passing: a special frame called a token continuously circulates around the ring, and a device may transmit only while it holds the token. This guarantees that only one device transmits at a time, so collisions can't occur — unlike the shared medium of a bus topology. Token Ring (IEEE 802.5) is the best-known implementation of this idea.

Advantages

• No collisions, thanks to token passing
• Predictable performance, since access to the medium is orderly rather than contested
• Every device gets an equal opportunity to transmit

Disadvantages

• A single failed cable or device can disrupt an entire simple ring
• Adding or removing a device can interrupt the network while the ring reconfigures
• Data may pass through several intermediate devices before reaching its destination, adding latency

3. Star Topology

Star topology is the most widely used topology today. Every device connects directly to a central device — typically a switch, or historically a hub — and all communication passes through it.

Advantages

• Easy to troubleshoot, since faults can usually be traced to one cable or device
• Easy to expand — a new device just needs a cable to the central switch
• A failed device doesn't affect any other device on the network
• Centralized administration

Disadvantages

• The central switch is a single point of failure: if it goes down, the whole network goes down
• Requires more total cabling, since every device needs its own run back to the center

Almost every modern Ethernet LAN and home Wi-Fi network is physically wired as a star, with a switch or router at the center.

4. Tree Topology

Tree topology combines bus and star characteristics: groups of devices are wired as stars around local switches, and those local switches connect up to a central "root" switch or backbone, forming a hierarchy similar to branches on a tree.

Advantages

• Scales well, since new branches can be added without redesigning the rest of the network
• Each branch can be managed and expanded somewhat independently
• Problems are often isolated to a single branch

Disadvantages

• The backbone connecting the branches is critical — if it fails, multiple branches lose connectivity
• Requires more equipment, and therefore more cost, than a simple star
• Design and configuration require more planning

Tree topology is common in enterprise and campus networks, where individual departments or buildings are wired as stars and then tied together through a core switch or fiber backbone.

5. Mesh Topology

In a mesh topology, devices are interconnected through multiple paths rather than a single central point, which provides redundancy: if one link fails, traffic can be rerouted through another.

• Full mesh — every device connects directly to every other device.
• Partial mesh — only the most critical devices have multiple connections; others connect through fewer links.

For a full mesh, the number of direct connections required grows quickly with the number of devices, following the formula:

Number of connections = n(n − 1) / 2

where n is the number of devices. For example, with 5 devices:

5 × (5 − 1) / 2 = 10 connections

This is why full mesh is rarely used at large scale: the number of cables needed grows much faster than the number of devices.

Advantages

• Highly reliable, since alternative paths exist if one link fails
• High fault tolerance — the network keeps operating despite individual failures
• Multiple simultaneous paths can improve overall throughput

Disadvantages

• Expensive — requires far more cabling and network interfaces than other topologies
• Complex to manage at scale
• The redundancy can be excessive, and wasteful, for networks that don't need that level of fault tolerance

Internet service provider backbones are a real-world example of mesh topology: multiple redundant paths connect major network nodes so traffic keeps flowing even when individual links fail.

6. Hybrid Topology

A hybrid topology combines two or more different topologies into a single network. For example, a company might use a star topology inside each department — devices connecting to a local switch — and a mesh topology between branch offices, for redundancy over long-distance links. Together, these form one hybrid network.

Advantages

• Flexible — can be tailored to the specific needs of different parts of an organization
• Scalable, and able to combine the strengths of its component topologies
• Failures are often isolated to one segment

Disadvantages

• Design is more complex, since it has to account for multiple topology types
• Typically the most expensive option, due to the mix of equipment required
• Needs more specialized administration

Logical Topologies in Practice

As noted earlier, logical topology describes how data actually flows, independent of the physical wiring. Three patterns are common:

• Logical bus — data is broadcast to every device on the segment, as in older Ethernet networks that used CSMA/CD (Carrier Sense Multiple Access with Collision Detection) over a shared medium.
• Logical ring — data moves sequentially from one device to the next, as in Token Ring and FDDI (Fiber Distributed Data Interface).
• Logical star — data passes through a central switching device that forwards it only to the intended recipient. Most modern Ethernet networks behave this way: a switch forwards each frame directly to its destination port instead of broadcasting it everywhere, giving better traffic control, improved security, and more efficient use of bandwidth.

Factors Affecting the Choice of Topology

Topology and Networking Devices

The devices used to build a network are closely tied to its topology:

Where Each Topology Is Used in Practice

• Home networks typically use a star topology, with phones, laptops, smart TVs, and other devices all connecting through a central Wi-Fi router. This keeps installation simple, cost low, and troubleshooting easy.
• Office networks commonly use star or tree topologies, with departments connecting through switches for centralized management and easy expansion.
• Campus networks (universities, large institutions) often use tree or hybrid topologies, connecting multiple buildings through high-speed fiber-optic backbones.
• Data centers favor mesh or hybrid topologies, since high availability, redundancy, and fault tolerance are critical.
• The internet itself is one of the largest real-world examples of mesh topology, with many redundant paths between networks ensuring communication continues even when individual connections fail.