Media Bridge Active Speaker Naming in ICN RTC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Real-Time Communications (RTC) architectures using point-to-point transport protocols like TCP or UDP are inefficient due to lack of awareness of multi-point and real-time nature, leading to inefficient content replication and congestion issues.
Innovation Solution
The implementation of Hybrid Information-Centric Networking (hICN) with named-based communication and a pull-based connectionless transport approach, where each participant's media is mapped to a network name, allowing the media bridge to pull media from active speakers instead of pushing from all participants, reducing overall traffic and improving media quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If point-to-point transport protocols (TCP/UDP) are used in RTC architectures, then communication between participants is established, but inefficient content replication and congestion occur due to lack of awareness of multi-point and real-time nature
Solution Approach 1:
The patent merges multiple unicast streams into a single multicast stream at the network layer. The media bridge consolidates duplicate media packets from multiple participants into one multicast flow, eliminating redundant transmissions across the network while maintaining efficient content delivery to all recipients.
Solution Approach 2:
The patent implements a dual naming scheme where a single media bridge infrastructure serves multiple functions: it acts as a multicast router, a name resolution service, and a traffic consolidation point. This universal approach allows the same system to handle both efficient multicast distribution and individual participant requests without requiring separate specialized components.
2Reliability
If all participants push media to the media bridge, then complete media distribution is achieved, but overall traffic and congestion increase
Solution Approach 1:
The patent inverts the traditional push-based architecture by implementing a pull-based model. Instead of all participants pushing media to the bridge, the media bridge proactively pulls media only from active speakers using named-based requests. This inversion dramatically reduces network traffic while ensuring complete media distribution for relevant participants.
Solution Approach 2:
The patent extracts and filters only the necessary media flows from the full participant set. The media bridge identifies active speakers and extracts only their media streams for distribution, eliminating unnecessary traffic from non-speaking participants while maintaining complete distribution for those who need the content.
3Adaptability or versatility
If traditional RTC architectures are used, then communication is established, but scaling is limited due to congestion and inefficiency
Solution Approach 1:
The patent introduces a named-based intermediary layer at the network layer that mediates between participants and the media bridge. This intermediary uses content-based names to identify and route media flows, simplifying the scaling mechanism while reducing congestion management complexity through intelligent name-based routing and filtering.
Data Source
AI summary
Presented herein are methods for dynamic management of names and of named-based requests for media flows in an Information-Centric Network (ICN)-based real-time communication system. The use of different naming schemes on contribution segments and on distribution segments in such architectures results in an effective low latency and scalable communication model based on the notion of active speakers regardless of the underlying mapping between active speakers and participants controlled by the media bridge. Operations are performed both at client side and at media bridge side to dynamically manage binding of names to speakers, while minimizing overhead and complexity of operations involved by a change of active speaker/quality/layout.