Bus Topology — How It Works, Advantages and Why It Died Out

Bus topology connects every device on the network to a single shared cable, the bus, with a terminating resistor at each end. There is no central device. Every signal placed on the cable travels the full length of it and reaches every station.

It was the layout of the original Ethernet standards, and understanding it explains a lot of networking vocabulary that still exists today, collision domain, CSMA/CD, half duplex. In production networks it is effectively extinct, replaced by star topology.

The Physical Parts of a Bus Network

The terminator is the detail most explanations skip, and it is the single most important component. Remove one terminator, or break the cable anywhere, which creates two unterminated ends, and the entire segment stops working, not just the devices past the break.

How Bus Topology Works

1. A station’s NIC listens to the cable first. If another station is already transmitting, it waits. This is the carrier sense in CSMA/CD.
2. When the cable is quiet, the NIC puts the frame onto the bus as an electrical signal.
3. The signal propagates in both directions from the tap point and reaches every station on the segment.
4. Every NIC reads the destination MAC address. The station it matches processes the frame; every other station discards it.
5. The terminators at both ends absorb the signal so it does not bounce back.

Collisions and CSMA/CD

Because the cable is shared, two stations can start transmitting at almost the same instant, neither hears the other yet because the signal has not reached it. The signals overlap and both frames are destroyed. That is a collision.

Ethernet handled this with CSMA/CD (Carrier Sense Multiple Access with Collision Detection): a station that detects a collision sends a jam signal, then both parties wait a random backoff interval before retrying, doubling the possible wait after each successive collision.

The consequence is the defining weakness of bus topology: every station shares one collision domain. Ten megabits is not ten megabits per device, it is ten megabits split across everyone, and it degrades faster than linearly as stations are added, because more stations means more collisions means more retransmissions. This is exactly what switches fixed by giving each port its own collision domain.

Advantages of Bus Topology

• Minimal cable. One run serves the whole segment, versus a dedicated cable per device in a star.
• No central hardware. No switch or hub to buy, power or fail.
• Cheap for a short, small deployment. A few machines in one room could be joined for the price of cable and connectors.
• Simple to understand. There is one path and one shared medium.

Disadvantages of Bus Topology

• One break kills everything. A cut cable or missing terminator takes down the whole segment, not just one host.
• Shared bandwidth. Capacity is divided across all stations, and collisions waste more of it as the station count rises.
• Half duplex only. A station cannot send and receive at the same time.
• Painful troubleshooting. A fault anywhere affects everyone, so isolating it means physically splitting the cable and testing halves.
• Strict distance limits. 10BASE2 allowed 185 m per segment and 30 stations; 10BASE5 allowed 500 m and 100 stations. Exceeding either broke collision detection.
• No security separation. Every station physically receives every frame, so any NIC in promiscuous mode could read all traffic on the segment.

Redundancy in a Bus Network

A single bus has no redundancy by design. The standard workaround was a dual-bus layout, two independent cables, each station tapped into both, so a break on one bus leaves the other carrying traffic. FDDI’s dual counter-rotating ring used the same idea in ring form. In practice, dual bus doubled the cabling cost of a topology whose only real advantage was cheap cabling, which is part of why it never caught on.

Where Bus Topology Still Exists

Bus is gone from office and home LANs, but the electrical model survives where a shared, low-cost, low-speed medium is genuinely the right choice:

• CAN bus in vehicles and industrial control
• RS-485 multidrop links in building automation and SCADA
• I²C and SPI on circuit boards
• Legacy coaxial cable TV distribution

Bus vs Star, Why Star Won

Once twisted-pair cabling and switching became cheap, every one of bus topology’s advantages disappeared while its failure modes stayed. Compare all five layouts in our network topology guide.

Frequently Asked Questions

Is bus topology still used today?

Not in computer networks. It survives in embedded and industrial buses like CAN and RS-485, where a shared low-speed medium is appropriate.

What happens if a terminator is missing in a bus network?

The signal reflects off the open end and travels back down the cable, interfering with live transmissions. The whole segment becomes unreliable or stops working entirely.

Is bus topology unidirectional or bidirectional?

The cable is bidirectional, a signal injected at any tap propagates toward both ends. But communication is half duplex: only one station can transmit at a time without a collision.

How many devices can a bus network support?

10BASE2 allowed 30 stations per 185 m segment; 10BASE5 allowed 100 per 500 m. Those were hard limits set by signal timing, and performance degraded well before reaching them.

Why does a break in a bus cable take down the entire network?

A break creates two unterminated cable ends. Both halves then suffer signal reflection, so neither half works, even the stations that are still physically connected to each other.