Rate Shaping Element for Adaptive Video Streaming QoS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for rate shaping in adaptive video streaming fail to accurately and effectively manage application layer throughput, leading to inconsistent video playback due to dynamic network conditions, particularly due to round trip time (RTT) and bandwidth fluctuations, which results in rebuffering and poor Quality of Service (QoS).
Innovation Solution
A Rate Shaping Element (RSE) is deployed at the customer's premise to dynamically measure application layer throughput, estimate the current playback rate, and adjust the video playback rate by holding or releasing HTTP responses to match the desired playback rate, ensuring consistent video playback without requiring modifications to content servers or players.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If rate shaping is performed at IP layer, then rate shaping capability is provided, but the shaped rate differs from targeted video playback rate due to RTT factors and the device is too far from client to accurately measure RTT
Solution Approach 1:
The patent introduces a rate shaping device positioned at the customer premises equipment (CPE) or gateway, which acts as an intermediary between the network and the client playback device. This intermediary location enables accurate RTT measurement by being proximal to the client, while still providing rate shaping capabilities for video streaming traffic.
2Adaptability or versatility
If rate shaping is performed at IP layer, then rate shaping is provided, but it is not reactive in response to changing network conditions and drives playback rate lower than it should be
Solution Approach 1:
The patent implements a feedback mechanism where the rate shaping device continuously monitors application layer throughput metrics and dynamically adjusts rate shaping parameters in response to changing network conditions. This reactive approach prevents unnecessary playback rate reductions while maintaining smooth video playback under varying bandwidth conditions.
3Measurement precision
If TCP congestion window clamp is used, then rate shaping can be done at TCP layer, but it offers only low granularity of control at low RTT and does not take into account actual RTT between player and content server
Solution Approach 1:
The patent changes the operational parameters of rate shaping by moving from TCP layer congestion window clamping to application layer throughput shaping. This enables fine-grained control of video playback rates by directly manipulating application layer metrics and incorporating actual RTT measurements between the player and content server, rather than using coarse TCP-level controls.
4Reliability
If video is streamed over WAN from remote content server, then video streaming service is provided, but link conditions change dynamically causing playback buffer to run out and video playback to stop or become jittery
Solution Approach 1:
The patent applies preliminary rate shaping to video streaming traffic before it reaches the client playback device. By proactively controlling the throughput rate at the CPE/gateway level, the system prevents bandwidth fluctuations from causing playback buffer depletion, thereby ensuring consistent video playback without requiring reactive buffer management at the client device.
Data Source
AI summary
Application layer throughput (ALT) shaping is provided for a streaming media session is provided. A request for a content segment is received. The receipt time of the request is recorded and the request is forwarded to a content server. A response message is received from the server. The receipt time of the response and a segment size is determined. An ALT for the current segment is determined based on the receipt times of the request and response messages and the segment size. The ALT rate is compared to a target ALT rate required for a desired session PBR. When the ALT rate is higher than the target ALT rate, the content response is held for a time period determined to lower the current ALT rate to the target ALT rate, and then released to the client device. Otherwise, the content response is promptly provided to the client device.