A Multi-User 360-Video Streaming System for Wireless Network.

Haowei Cao,Jialiang Lu,Nan Zong

DOI: https://doi.org/10.1145/3359997.3365713

2019-01-01

Abstract:With the rapid development of Virtual Reality technology and its hardware, 360-degree video is becoming a new form of media which arouses the interest of public. In the past few years, many 360-degree video delivery schemes are proposed, but there hasn’t been a standard solution which can perfectly overcome the difficulties caused by network latency and bandwidth limit. In this paper, we consider the context of a wireless network consisting of a base station and several users. We proposed a 360-degree video delivery and streaming scheme which serves multiple users simultaneously while optimizing the global bandwidth consumption. The system will predict the head movement of users using machine learning algorithm, and extract the visible portion of the video frame for transmission. The core contribution of the scheme is that it will recognize the conjoint viewport of multiple users, and then optimize the global bandwidth consumption by arranging the transmission of conjoint viewport over the public channel of the wireless network. The results prove that the proposed scheme can effectively reduce global bandwidth consumption of the network with relatively simple configuration.