round-trip time

Round-trip time, or RTT, is the elapsed duration from when a device sends a signal or data packet to a remote location until it receives the corresponding response. In network diagnostics, this is typically measured in milliseconds and forms the basis of fundamental performance assessment across telecommunications systems, data networks, and control circuits. RTT encompasses both the forward propagation delay and the return path delay, plus any processing time at the receiving end.

The measurement is most commonly performed using ping utilities, which send ICMP echo request packets to a target device and record the time taken for echo replies to return. A typical RTT on a local area network spans 1 to 10 milliseconds, while internet connections to distant servers may exhibit RTTs of 50 to 200 milliseconds depending on geographic distance and routing. Satellite communications, by contrast, suffer inherent RTT of 500 to 600 milliseconds due to the physics of signal propagation to orbit and back.

RTT becomes a critical design constraint in real-time control systems where response time directly affects safety or function. Industrial automation networks, video conferencing systems, and online multiplayer applications all impose maximum tolerable RTT thresholds. A control loop waiting for sensor feedback cannot respond faster than the RTT allows; high RTT introduces lag that degrades precision and responsiveness. In voice calls, round-trip delays exceeding 150 milliseconds produce noticeable echo and conversational awkwardness.

Measurement and interpretation

RTT measurements vary due to network congestion, packet loss, and routing changes. A single ping result is unreliable; engineers instead capture mean RTT over dozens or hundreds of samples and report standard deviation as well. High variance signals an unstable path. Asymmetric routing, where outbound and return paths differ, can produce misleading results if only aggregated RTT is examined.

Round-trip time differs from one-way delay, which is harder to measure accurately because it requires synchronized clocks at both endpoints. Network jitter, the variation in RTT from packet to packet, matters as much as the absolute value for real-time applications; stable 100 ms RTT is preferable to unstable 50 ms RTT with high variance. In system commissioning, RTT testing reveals bottlenecks, misconfigured routes, and hardware faults before they degrade production performance.