Video streaming performance is strongly influenced by transport-layer behavior under bandwidth constraints. This paper provides a protocol-oriented comparative analysis of four video transmission approaches: DASH streaming, TCP-based transmission, UDP-based transmission, and Big Packet Protocol (BPP) transmission with trimming support. Using a controlled experimental framework, we evaluate delivery continuity, quality variability, and low-latency playback behavior through throughput/bitrate dynamics, frame/layer reception patterns, and total pause duration. The results show that DASH achieves the most consistent playback continuity in moderate and time-varying conditions through segment-level adaptation, while BPP approaches DASH-level continuity in dynamic scenarios by enabling graceful degradation via trimming. UDP is highly sensitive to bandwidth limitation and loss, leading to unstable reception and persistent interruptions. TCP preserves constant decoded quality without switching due to retransmissions, but incurs the largest stall durations, making it less suitable for low-latency streaming targets. Overall, the findings highlight a fundamental trade-off between continuity and reliability and motivate practical guidance for selecting transmission mechanisms under both fixed and dynamic bottleneck conditions.
Tüker et al. (2026) studied this question.