Explore our research white paper on the evaluation of TCP Prague technology. Read more about how it reduces internet queue delays and enhances latency through scalable Congestion Control (CC). Discover how implementations like Google’s BBRv2 and Linux TCP Prague are striving to fulfill this promise while maintaining compatibility with legacy systems, and how ongoing research seeks to align them more closely with the Prague standard for even better results.

Low Latency for the Future Internet (L4S) technology is poised to address the persistent queue delay issues in our modern internet infrastructure. It leverages scalable congestion control (CC) and active queue management to revolutionize latency reduction. At the heart of this transformative approach lies TCP Prague, a specialized protocol designed to make scalable CC practical for internet use.

Numerous TCP implementations have rallied behind the Prague standard, including Google’s BBRv2 and Linux TCP Prague. These implementations share a common goal: enhancing latency without sacrificing bandwidth, all while maintaining compatibility and fairness with legacy counterparts. This is the promise of L4S.

Our research journey embarks on a thorough exploration of these existing implementations, scrutinizing them against the requirements outlined by the Prague standard. We meticulously evaluate their adherence to the standard’s principles and specifications. Along the way, we make a crucial discovery: BBRv2 diverges somewhat from the standard’s guidelines.

Armed with insights gained from our evaluation, we endeavor to enhance the implementation of BBRv2. We delve into the kernel, adapting its variables, and striving for optimization. Our efforts yield success, marking a step towards aligning BBRv2 more closely with the Prague standard.