Consider a legacy 10BASE-T Ethernet hub (shared half-duplex). Do all attached ports share one collision domain so that the aggregate bandwidth is shared, and is it therefore correct that “the fewer ports a hub has, the less bandwidth available per port”?

Introduction / Context:
Hubs create a single shared collision domain where one station transmits at a time. While this means bandwidth is shared, the statement that “the fewer ports, the less bandwidth per port” is logically flawed. The number of physical ports does not mechanically reduce the per-port line rate; rather, effective throughput per host depends on actual traffic contention among active transmitters.

Given Data / Assumptions:

• Device type: classic Ethernet hub (multiport repeater), not a switch.
• Duplex mode: half-duplex with CSMA/CD.
• All ports participate in one collision domain.

Concept / Approach:
On a hub, all stations share the medium. The nominal line rate per port (e.g., 10 Mbps for 10BASE-T) does not change with the number of ports. What changes is the probability of collisions and the time each host must wait to transmit as the number of active talkers increases. Thus, while more active hosts can reduce observed throughput, the blanket claim tying per-port bandwidth to the mere count of hub ports is incorrect.

Step-by-Step Solution:

Verification / Alternative check:
Measurements on lightly loaded hubs show hosts can achieve near line rate when alone; as simultaneous transmitters increase, aggregate throughput approaches but does not exceed the hub’s total capacity, divided by contention—independent of how many unused ports exist.

Why Other Options Are Wrong:

• Correct: would wrongly imply port count sets per-port bandwidth.
• Only true on full-duplex/STP/jumbo frames: irrelevant to hub operation; hubs are half-duplex and do not run STP.

Common Pitfalls:
Confusing hubs with switches (which provide dedicated bandwidth per port); assuming idle ports reduce available bandwidth to used ports.

Final Answer:
Incorrect