The current situation in IP networks shows the need for new congestion control algorithms that can be flexible, scalable, and capable of avoiding additional queue delays caused by loading the bottleneck buffers. Most common internet flows use loss-based congestion controls, which can achieve high bottleneck bandwidth utilization and fair resource sharing but cause overload bottleneck buffers. In this paper we present an investigation of the performance of a PID-based congestion control solution for high-speed IP networks. It uses measurements of a round trip time and receiver delivery rate to reach and keep maximum available bottleneck performance and constant node buffer load caused by bottleneck queue on some low level. This algorithm can be effective in high-speed IP networks and delay-sensitive applications. It is designed to be flexible and scalable for different connectivity cases. This algorithm then is investigated on the example of RMDT.