Please note: This master’s thesis presentation will take place online.
Joohan Lee, Master’s candidate
David R. Cheriton School of Computer Science
Supervisors: Professors Bernard Wong, Khuzaima Daudjee
The demand for low latency video streaming has greatly increased as live video streaming applications, such as Twitch and YouTube Live, are becoming more popular these days. According to the 2021 Bitmovin video developer report, the biggest challenge that video developers are experiencing today is providing low latency video streaming. One common form of live streaming is using a wireless LTE network.
There have been many approaches to characterizing wireless links and accurately measuring available bandwidth to provide low latency streaming over a wireless LTE network link. However, even with fine-grained bandwidth estimation, video streaming on a single LTE link is still susceptible to unexpected congestion from a sudden drop in available bandwidth or temporal disconnection. Link congestion increases the network delay of a frame, and it is detrimental to the quality of service for real-time or live video streaming.
To overcome the limitations of using a single LTE link for low latency video streaming, we utilize multiple LTE links. Using multiple links enhances video quality through increased bandwidth and resilience. However, multi-homed low latency video streaming protocols may achieve lower video quality than single-homed protocols when a frame is split and sent over more than one link. If one of the links becomes congested or gets disconnected, the part of the frame that was sent must wait until the packets sent on the problematic link are re-transmitted through another link. Re-transmission requires at least one extra round trip time. A video player may end up skipping the late frame or serving only the received part of the frame due to the re-transmission delay. Ferlin et al. suggests using forward error correction (FEC) on MPTCP to reduce re-transmission delay, but FEC is not helpful under a significant bandwidth drop. If more data than recovery is late, the encoded frame cannot be decoded before its deadline. FEC requires using a large portion of the network bandwidth for recovery to handle significant bandwidth drops.
In this thesis, I present Squash, a low latency video transport protocol that encodes the same frame at multiple bitrates and sends them across different links to minimize video stream disruption when unexpected bandwidth drops occur in LTE networks. Squash does not require additional overhead in re-transmission latency or in bandwidth for sending recovery data for erasure coding. To accomplish this, Squash utilizes a multi-bitrate encoder and multiple links. The encoder encodes a frame into two different bitrates, high-bitrate and low-bitrate. When a high-bitrate frame cannot arrive on time due to a congestion from an unexpected drop in available bandwidth, the low-bitrate frame is used to replace the missing frame because the low-bitrate frame is smaller and is sent on the links that are disjoint from those used by the high-bitrate frame. To the best of my knowledge, Squash is the first architecture to use multi-bitrate frames to increase resilience on multiple links. The SSIM of the video streamed on Squash is 13–58% higher than streamed on the baseline protocol which is designed in the same manner as Squash except that it employs single-frame encoding.