Latency and Round-Trip Time

What Latency Actually Measures

Latency is the delay between sending data and it arriving at its destination - not how much data can be sent (that's bandwidth/throughput, covered in the previous note), but purely how long any single piece of data takes to get there.

One-Way Latency vs Round-Trip Time (RTT)

One-way latency:  Sender -----------> Receiver           (time for one direction)
Round-Trip Time:  Sender -----------> Receiver -----------> Sender  (there and back)

One-way latency measures a single direction, which is genuinely difficult to measure precisely without perfectly synchronized clocks on both ends. Round-Trip Time (RTT) - measuring the full round trip - is much easier to measure accurately, which is exactly what tools like ping (covered earlier in this project) actually report.

What Actually Contributes to Latency

• Propagation delay - the physical time for a signal to travel the actual distance, ultimately limited by the speed of light.
• Processing delay - time each router along the path spends examining a packet and deciding where to forward it.
• Queuing delay - time a packet spends waiting in a router's buffer before being forwarded, related to the buffer bloat issue covered earlier in this project.

Why Latency Matters More for Some Applications Than Others

A large file download mostly cares about throughput - total speed over time. A live video call or competitive online game cares intensely about low latency - even a connection with excellent throughput feels sluggish and unresponsive if latency is high.

Why Latency Has a Hard Physical Floor

No amount of better equipment or clever engineering can make a signal travel faster than the speed of light allows - which is exactly why a connection between two very distant locations always has a certain minimum, unavoidable latency, no matter how good the equipment on either end is.