real-time communication design

High end-to-end latency (>5 seconds) in real-time media streaming—such as live sports, news, surveillance, and linear TV—remains a critical bottleneck. Method: This paper systematically surveys the evolution of low-latency live streaming in the IP era, proposing the “Live Latency Continuum” model to unify and characterize latency coupling across the entire pipeline—from acquisition to playback. It comprehensively analyzes HTTP-based adaptive streaming extensions (e.g., LL-DASH, LL-HLS), integrates QUIC/HTTP/3 transport optimizations, and incorporates client-side adaptive buffering strategies, enabling rigorous latency modeling, root-cause attribution, and engineering validation. Contribution/Results: The study demonstrates the technical feasibility of sub-one-second end-to-end latency and delivers a reusable, modular low-latency optimization framework. This work establishes a critical benchmark for both academic research and industrial standardization efforts in ultra-low-latency streaming.

In Time-Sensitive Networking (TSN), multicast communication improves bandwidth efficiency but exacerbates scheduling complexity due to port contention and queue resource constraints. Method: This paper proposes a fine-grained multicast tree partitioning approach that dynamically decomposes large multicast trees into smaller multicast or unicast subtrees, integrating time-triggered planning, adaptive tree-splitting algorithms, and multi-strategy scheduling to jointly optimize resource utilization and schedulability under heterogeneous topologies. Contribution/Results: To the best of our knowledge, this is the first systematic incorporation of multicast partitioning into TSN time-triggered flow scheduling. The method guarantees end-to-end latency bounds and achieves load balancing across switches. Experimental evaluation shows a 5–15% reduction in flow rejection rate and a 5–125% increase in throughput compared to an unpartitioned baseline, significantly improving flow admission ratio and scheduling feasibility in dynamic TSN environments.

To address rebuffering, bitrate instability, and QoE fluctuations in video streaming over 6G networks, this paper proposes a behavior-aware adaptive streaming framework. Methodologically, it introduces— for the first time—a URLLC-inspired dynamic buffer management mechanism; constructs a reinforcement learning–driven edge–end collaborative bitrate decision model; integrates temporal modeling of multidimensional user viewing behaviors; jointly optimizes 6G channel state and QoE; and enables fine-grained adaptive encoding scheduling. Compared with conventional static strategies, the proposed framework achieves an 83% reduction in rebuffering rate, a 41% improvement in average QoE, and 99.2% streaming availability under high-mobility and dense-access scenarios—significantly enhancing seamless playback performance and user-experience robustness.

Delivering ultra-low-latency, high-quality video streaming over resource-constrained Open Radio Access Networks (O-RAN) remains challenging. Method: This paper proposes a semantics-aware collaborative architecture integrating Generative Artificial Intelligence (GAI) with Multi-access Edge Computing (MEC). It introduces a novel semantic control channel enabling tight coupling and closed-loop coordination among xApps, MEC, and RAN intelligent agents; and deploys, for the first time in a real O-RAN system, a GAI-driven dynamic joint optimization of radio resources and video quality. Contribution/Results: Experiments demonstrate a significant reduction in end-to-end latency and a 2.1 dB PSNR improvement over baseline methods, revealing a tunable trade-off between latency and reconstruction fidelity. The results validate the efficacy of fine-grained latency control for video quality enhancement and establish a new paradigm for O-RAN-native AI-enabled real-time media services.

To address the challenge of deterministic low-latency scheduling for periodic message transmission between antennas and remote processing units in Cloud-RAN—where strict protocol deadlines must be met while avoiding buffering and collision delays induced by statistical multiplexing—this paper pioneers the application of deterministic conflict-free scheduling to Cloud-RAN fronthaul networks. We propose two algorithms: (i) an analytical zero-buffer scheduling algorithm tailored for short-path or light-load scenarios, and (ii) PMLS (Periodic Message Latency Scheduling), a heuristic algorithm supporting bounded buffering. We theoretically prove that a zero-buffer feasible schedule always exists under short-path or low-load conditions. Experimental results demonstrate that PMLS achieves zero-delay deterministic schedules with high probability even under full load, significantly enhancing latency predictability and resource utilization.

This study addresses the critical issue of high end-to-end latency in real-time 3D voxel streaming, which severely degrades immersion, induces motion sickness, and impedes user interaction. The work presents the first hierarchical quantitative analysis of latency across the system stack, decomposing it into application, transport protocol, and network layers, and identifies bottleneck sources through empirical measurement. Building on these insights, the authors propose targeted system-level optimizations spanning protocol design, network scheduling, and rendering scheduling. Experimental evaluation demonstrates that the proposed approach substantially reduces end-to-end latency, thereby enhancing system responsiveness, scalability, and overall user experience.

High latency in interactive applications over cellular RAN stems from invisible RAN queue states and lagging end-to-end congestion control. Method: This paper proposes L4Span—a novel architecture that for the first time abstracts and explicitly feeds back RAN queue state to the transport layer, enabling millisecond-scale queue prediction and coordinated ECN marking. Fully compatible with 3GPP and O-RAN standards, L4Span requires only minimal protocol stack modifications. A C++ prototype is implemented on the srsRAN platform, supporting coexistence of ultra-low-latency flows and legacy traffic. Results: Experiments across diverse wireless channel conditions demonstrate up to 98% reduction in one-way latency while sustaining near-line-rate throughput. The core contribution is the first RAN-aware, end-to-end low-latency signaling framework—bridging a critical gap in real-time congestion signal generation within the wireless access network.

This work addresses the persistent stuttering in low-latency live streaming, which occurs even when the encoded bitrate remains below available bandwidth, due to the inability of traditional packet-level congestion control to accurately estimate available bandwidth amid video frame encoding fluctuations. To resolve this, the authors propose Camel—the first frame-level congestion control algorithm tailored for low-latency live streaming—that decouples encoding-induced variability from bandwidth estimation using frame-level network feedback and introduces a burst-length control mechanism to dynamically optimize both average sending rate and burst patterns. The system comprises three core modules: a bandwidth/delay estimator, a congestion detector, and a burst controller. Deployed on a platform serving hundreds of millions of users, Camel increased 1080p stream share by 70.8%, raised media bitrate by 14.4%, and reduced stuttering by 14.1%; simulations further demonstrated up to 93.0% stutter reduction and a 23.9% improvement in bandwidth estimation accuracy.

This work addresses the high transmission latency incurred by traditional ARQ protocols in high packet-loss networks, where successful delivery requires complete reception of all data packets. To overcome this limitation, the authors propose FLUID, a novel protocol that integrates fountain codes with receiver feedback and introduces a tunable slack parameter ε. FLUID employs a control strategy based on the Loss-Product Rule, enabling early termination of transmission before all original packets are received. This approach achieves bandwidth efficiency comparable to ideal ARQ while substantially reducing block delivery delay, demonstrating pronounced performance gains particularly under persistent multi-round packet loss conditions.

This work addresses playback stuttering in real-time streaming video generation caused by generation lag by proposing a dynamic serving mechanism that uses playout slack as a unified scheduling signal. The mechanism employs a three-level priority queue, re-homing, elastic sequence parallelism, and dual-mode Pareto routing to enable cross-stream resource reallocation and block-level quality adaptation, balancing timeliness and user experience without compromising generation quality. Experimental results on a 16-GPU H100 cluster demonstrate that the system improves continuous playback rate by 1.64–3.29× and reduces time-to-first-block by 1.61–9.65× compared to baseline approaches.