PlayInfoBox

Home /Network /Network Topology Types Explained: Star, Bus, Ring, Mesh, and Tree

Network Topology Types Explained: Star, Bus, Ring, Mesh, and Tree

The shape a network is wired in affects its speed, cost, and how badly a single failure hurts -- here's how the main topology types compare.

What a network topology actually is

A network topology is the physical or logical arrangement of how devices (nodes) like computers, switches, and routers connect to each other. The same number of devices wired in different shapes can behave very differently in cost, speed, and how far a single failure spreads.

Star topology: everything connects to one hub

Every device connects with its own cable to a central switch or router. If one device's cable fails, nobody else is affected, which makes problems easy to isolate -- but if the central hub itself fails, every connected device loses connectivity at once, a weakness known as a single point of failure (SPOF).

Bus topology: one shared cable for everyone

All devices tap into a single shared backbone cable, and every device receives all the data traveling along it. It was cheap to build with early Ethernet, but a break anywhere on that one cable takes the whole network down, and this design has been almost entirely replaced by star topology.

Ring topology: a circular relay

Each device connects to exactly two neighbors, forming a loop that passes data around it in one direction (or both, with a dual ring). Older technologies like Token Ring and FDDI used this design; a single break can take down the whole ring, which is why some implementations added a second, redundant ring.

Mesh topology: multiple overlapping paths

Devices connect through more than one path, so traffic can reroute around a broken link. A full mesh, where every node connects directly to every other node, is the most resilient but grows expensive fast since connections increase roughly with the square of the node count -- a partial mesh, connecting only some nodes redundantly, is the common compromise.

Tree and hybrid topologies: star networks stacked in layers

A tree topology links multiple star networks in a hierarchy -- for example, per-floor switches (access layer) connecting up to a building switch (distribution layer), which connects up to a core switch. Mixing different topologies to fit different needs like this is called a hybrid topology, and it is how most real corporate and campus networks are actually built.

Your home Wi-Fi is a star topology too

When every phone, laptop, and smart TV in your home connects wirelessly to a single router, that is structurally the same as a wired star topology -- just using radio instead of cables. It's also why the whole house loses internet access at once when the router loses power: that's the star topology's single point of failure showing up at home.

Physical topology versus logical topology

The physical topology is how cables and devices are actually laid out; the logical topology is how data flows across them. A classic example is early Token Ring, which was wired physically as a star around a central unit while data still passed logically around a ring. Modern Wi-Fi is another case: devices are logically on one shared segment but physically scattered. Keeping the two views apart helps when diagnosing problems.

Wi-Fi mesh systems in the home

Consumer mesh kits use several units that talk to each other wirelessly, or over cable, so coverage reaches the whole house. Despite the name, many of them are arranged like a tree or a star of access points, with traffic ultimately passing to one gateway. They reroute around some failures, but the router connected to the internet is still a single point of failure.

Loops, spanning tree, and why redundancy needs care

Extra links add resilience but create loops in Ethernet, where broadcast traffic can circulate endlessly and bring down a network. The Spanning Tree Protocol solves this by blocking some redundant links until they are needed. Newer designs, such as leaf-spine fabrics in data centers, use routing between layers so that many paths can carry traffic at once.

Point-to-point and wide-area links

The simplest topology is a direct link between two devices, such as a leased line between two offices or a cable between two switches. Wide-area networks are often built from many such links arranged in a hub-and-spoke pattern (branches to a headquarters) or a partial mesh between major sites. Hub-and-spoke is cheap but makes the headquarters link a bottleneck.

Why star topology quietly won

Star topology dominates modern networks not because it's the most failure-resistant on paper -- a full mesh technically beats it there -- but because it offers the best practical balance: cheap to wire, easy to add or remove a device without disrupting anyone else, and simple to troubleshoot since a technician can isolate a fault to one cable or port. Its one real weakness, the central hub as a single point of failure, is manageable by simply buying a more reliable switch or router rather than wiring an entire redundant network.

How real networks blend more than one shape

Very few networks stick to one pure topology in practice. A typical office runs star-wired access switches in a tree hierarchy up to a core, then adds mesh-like redundant links between core switches so that no single cable failure can split the network in two. Understanding the basic shapes is what lets you recognize which trade-offs a given design is making.

Choosing a topology: the questions that matter

When planning a network, start with the failures you can tolerate. If one switch failing may stop a whole floor, accept a star; if it must not, add a second path. Count the cost of cable runs and ports, including the labor of installation. Consider growth: a star lets you add a device by connecting one cable, while a ring or bus forces you to disturb existing connections. Finally, consider troubleshooting skill: simple structures make faults easier to find, and complexity needs monitoring tools and documentation. The best choice is rarely the most resilient design on paper; it is the one that matches the budget and the real consequences of downtime.

Where each topology still shows up

Bus-like behavior survives in shared media such as CAN bus networks in vehicles and some industrial fieldbuses. Ring designs live on in metro fiber rings and industrial Ethernet rings, where a break triggers fast failover to the other direction. Full or partial meshes appear in backbone links between data centers and in wireless mesh networks. Stars and trees dominate offices, homes, and campuses. Learning to recognize these shapes in diagrams helps you read a network design quickly and ask the right questions about it.

Frequently Asked Questions

Which topology do real companies actually use?

Most mid-size and large organizations don't use a single pure topology. They typically build a hierarchical tree with access, distribution, and core layers, then add extra ('meshed') links between core devices for redundancy -- a hybrid design.

How do you fix star topology's single-point-of-failure problem?

By adding redundancy at the hub itself: duplicate core switches or routers configured to fail over automatically, backup power (a UPS) to survive outages, and, for critical networks, mesh-style extra links so traffic has more than one path to take.

Is a bus topology still used for Ethernet?

Not in wired office or home networks. Early 10BASE2 and 10BASE5 Ethernet used a shared coaxial bus, and it was replaced by twisted-pair cabling to a hub or switch. The bus idea survives in other technologies such as vehicle and industrial buses.

How many devices can a star network support?

It depends on the number of ports on the switch, and switches can be chained or stacked to add more. Practical limits come from bandwidth, broadcast traffic, and management, not from the star shape itself.

Is my home router network a mesh?

Only if you use a mesh kit, and even then the traffic usually heads to one gateway. A single router with devices attached by cable or Wi-Fi is a star. The word mesh in product names is not always the same as the topology term.