Challenges of Livecast Computing Network: A Contemporary Survey

With the rise of Internet services in recent years, especially the flourishing development of new social platforms such as Taobao Live, Tiktok, Huya, YouTube, and Twitch, they have led to a significant increase in Internet video services. For example, YouTube Live stats reveal that 2020 was the biggest year ever for Gaming Live, with 100 billion watch time hours. As a result, the entire network communication model has gradually evolved into a Livecast computing network (LCN) architecture, where various network units involved in Internet live video services, with their computing, storage, distribution, and transmission capabilities, form an architecture that can meet high concurrency and low latency requirements. This article investigates, sorts, and analyzes LCN architecture related standards, protocols, and network model solutions from the perspective of Internet live video services. Firstly, it outlines the standards, system architecture, and main streaming media protocols of Internet live video. Secondly, it analyzes the video algorithm network architecture models, including cloud-based, edge-based, and cloud-network-edge collaborative models. Finally, it summarizes the main challenges and research directions faced by video algorithm networks in the future.