Network Throughput Measurement via Round Trip Time Analysis

The patent applies preliminary action by pre-identifying bottleneck sections in the network before throughput measurement. The measurement target selection unit identifies sections where bandwidth is fully utilized (bottleneck sections) based on round-trip time variations, then selects measurement targets within these identified sections. This preliminary identification ensures that throughput is measured under conditions where maximum effective throughput can be accurately captured, resolving the contradiction between measurement precision and method complexity.

The patent employs feedback by using round-trip time measurements as an indirect indicator of bandwidth utilization. The round-trip time variation detection unit monitors changes in round-trip time, which feedback indicates whether the network has reached its bandwidth limit. When round-trip time increases beyond a threshold, the system infers that the bottleneck section is fully utilized. This feedback mechanism allows the system to detect maximum effective throughput conditions without continuously monitoring all bandwidth usage, reducing energy consumption while maintaining measurement precision.