Mobile user equipment performance evaluation of QUIC transport protocol over 5G new radio environment

Abstract

The rapid increase in the number of IoT devices within 5G wireless networks has resulted in the need for efficient and low-latency transport protocols that can support heterogeneous application requirements. This study provides a proportional ns-3 discrete-event simulation analysis of QUIC-like UDP transport with TCP NewReno for 5G New Radio (NR) mmWave channel communications over IoT nodes. A scenario with 100 mobile user equipment (50 using TCP and 50 using QUIC-like protocols) is evaluated under Gauss-Markov mobility for five performance metrics: aggregate throughput, per-user goodput, packet loss ratio, end-to-end delay, and congestion window dynamics. The results show TCP NewReno achieving a much higher aggregate throughput of 89.50 Mbps than QUIC-like UDP’s 8.50 Mbps. Furthermore, QUIC achieved a 43.6% reduction in the average end-to-end latency (23.8 ms vs. 42.2 ms) and a 47.2% reduction in the P95 latency (42.6 ms vs. 80.6 ms). QUIC exhibits improved fairness across users per equipment, with a coefficient of variation in user throughput 4.1× lower than that of TCP. These results define specific protocol-use-case associations within an integrated 5G IoT landscape, streamlining real-world guidance for network dimensioning by illustrating how QUIC’s performance improvements can be leveraged across IoT applications, such as smart cities and connected vehicles. It is emphasized that the UDP flow studied here is a rate-limited “QUIC-like” stand-in rather than an RFC 9000 implementation; the throughput gap therefore reflects the offered-load configuration (a 200 kbps/UE cap versus a greedy BulkSend source) and the latency and loss figures reflect an open-loop flow without QUIC’s loss recovery or congestion control, so they must not be read as a measure of QUIC-versus-TCP protocol efficiency.