Q13Principle of Electronic Communication

Ethernet vs Token Ring LANs

Ethernet uses a bus or star physical topology (modern implementations almost universally use a star topology with a central switch) and employs CSMA/CD (Carrier Sense Multiple Access with Collision Detection) as its access method in legacy shared-medium/hub-based configurations: any station wishing to transmit first listens to check the medium is idle, transmits if clear, and if a collision is detected (two stations transmitting simultaneously), both stations back off for a random time before retrying. Modern switched Ethernet largely eliminates collisions altogether by providing each device its own dedicated, full-duplex switch port.

Token Ring uses a logical (and often physical) ring topology, in which a special control frame called a 'token' circulates continuously around the ring; a station may transmit data only when it possesses the token, guaranteeing collision-free, deterministic access and predictable maximum access delay, in contrast to Ethernet's probabilistic, collision-based access. Token Ring's deterministic access made it historically attractive for time-sensitive industrial applications, but its higher hardware cost and lower peak throughput compared to switched Ethernet led to Ethernet becoming the dominant LAN technology, with Token Ring now largely obsolete in modern network deployments.

In practice, virtually all Token Ring installations, even those nominally 'ring' topology, actually used a physical star wiring pattern (via a Multistation Access Unit, MAU) with the logical ring implemented internally, so both technologies commonly appear physically as star-wired networks despite their very different logical access methods - the crucial distinguishing factor remains the access method itself (probabilistic collision-based CSMA/CD versus deterministic token-passing) rather than the outward physical cabling layout.