RTP Header Extensions for Low-Latency User Interaction Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems struggle to efficiently transport time-sensitive user interaction data, such as pose tracking and hand gesture data, for immersive applications like XR and cloud gaming, due to the lack of a flexible and encoding-agnostic method that can handle rapidly evolving interaction data formats.
Innovation Solution
The use of real-time transport protocol (RTP) header extensions to transmit user interaction data, encrypted and authenticated using TLS/DTLS, allowing for fast and reliable communication of interaction data in a time-critical path, bypassing the need for standard profile definition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If user interaction data is transmitted using traditional wireless communication protocols, then the data can be transported, but the transmission latency is high and cannot meet time-sensitive requirements for immersive applications
Solution Approach 1:
The patent segments user interaction data into two categories: time-critical interaction data (e.g., pose tracking, hand gestures) and non-time-critical data (e.g., media content). Time-critical data is transmitted through RTP header extensions along the time-critical path, while non-time-critical data uses traditional payload transmission. This segmentation enables differentiated handling that reduces latency for interaction data while maintaining overall system reliability.
Solution Approach 2:
The patent introduces RTP header extensions as an intermediary mechanism to carry interaction data directly in the protocol header rather than through the traditional payload path. This intermediary structure allows interaction data to bypass standard processing queues and be transmitted with higher priority, reducing transmission latency while maintaining protocol integrity through authentication tags.
2Adaptability or versatility
If a flexible and encoding-agnostic method is used to handle rapidly evolving interaction data formats, then adaptability is improved, but system complexity increases due to lack of standard profile definition
Solution Approach 1:
The patent makes the RTP header extension mechanism universal by designing it to accommodate multiple interaction data formats (pose tracking, hand gestures, eye tracking, etc.) through a common structure. The header extension uses standardized fields (interaction short message type, timestamp, raw data length, raw data payload) that can represent various interaction formats without requiring separate protocol definitions, thereby reducing implementation complexity while maintaining high adaptability.
Solution Approach 2:
The patent enables flexible handling of evolving interaction data formats by allowing dynamic parameter changes within the header extension structure. The interaction short message type field can be updated to reflect new interaction formats, and the raw data payload can accommodate varying data structures without changing the underlying protocol, allowing the system to adapt to new formats while maintaining a stable protocol implementation.
3Productivity
If interaction data is transmitted with high frequency (250-1000 times per second), then real-time responsiveness is improved, but network bandwidth consumption increases
Solution Approach 1:
The patent applies partial action by transmitting only the essential interaction data fields (interaction short message type, timestamp, and compact raw data payload) in the RTP header extension, rather than transmitting complete interaction data packets. This selective transmission approach enables high-frequency updates (250-1000 times per second) while consuming minimal network bandwidth, as only the most time-critical parameters are included in each header extension.
Data Source
AI summary
Various aspects of the present disclosure relate to user interaction data transportation using real-time transport protocol header extension. One or more physical device interactions by a user are sampled to obtain interaction data inputs, such as for a user playing cloud gaming or using extended reality on their user equipment. These interaction data inputs are generated frequently, such as approximately 250 to 1000 times per second. One or more interaction short messages (e.g., less than 100 bytes), which are time-sensitive, are generated from the user interaction data and included in one or more transport header extensions of a real-time transport protocol data unit. The real-time transport protocol data unit is transmitted to a remote device, such as a network entity.