How would you measure latency and jitter in the EPD Pilot, and why are these metrics important?

• A. Track one-way latency only
• B. Track round-trip times and variation; important for meeting real-time constraints and user experience
• C. Ignore jitter and assume fixed latency
• D. Measure only CPU cycles

Answer: (B)

Measuring latency and jitter centers on understanding how responsive and stable the system is during real use. The best approach is to track round-trip times and how much they vary. Round-trip latency captures the total time for a message to go to the remote end and back, which aligns with what users experience in interactive apps and doesn’t require perfectly synchronized clocks. Jitter, the variation in those times, matters because even small fluctuations can degrade real-time performance and user experience. Together, these metrics reveal both typical delay and the stability of that delay, helping you diagnose congestion, queueing, or path issues and verify you meet real-time constraints in the EPD Pilot environment.

Why not the other options? One-way latency relies on clock synchronization and can miss asymmetries or the actual end-user experience. Ignoring jitter assumes constant delays, which isn’t realistic in most networks. Measuring only CPU cycles misses the network transit time entirely and won’t reflect how the system performs in practice.