Network optimization

Real-time video encoding processes are designed to compress large amounts of video data to ensure video can efficiently be transmitted over a network. Magor TeleCollaboration is designed from a network perspective to maintain a consistent user experience while optimizing use of network resources.

A set of powerful processes runs continuously on each endpoint to evaluate the network status, adapt traffic to enhance the quality of the experience, condition the traffic to create a predictable flow and, when necessary, recreate the missing information. The interaction between these processes minimizes bandwidth requirements, simplifies network engineering and delivers consistent, optimal video quality.

 

Evaluate the network environment

The encoding processes are optimized independently at each endpoint to enhance the user experience based on each endpoint’s capabilities and the network environment. The evaluation process continuously monitors each user’s actions and conveys the status in real time to the adaptation process. Because Magor TeleCollaboration considers both the user’s actions and the condition of the network between endpoints, each user enjoys a consistently high-quality video experience.

Adapt to change

Magor’s SVC++, a codec-agnostic peer-to-peer video architecture, adapts the encoding rate based on the real-time information provided by the evaluation process. The encoding is adapted to maintain maximum HD video quality, while responding to the varying network conditions at each endpoint.

Condition the data streams

Encoded 1080p video is naturally extremely bursty. Although the average rate does not need to exceed 3 Mb/s, a single encoded video stream can burst up to 64,000 bytes at 100 Mb/s1. The collaboration traffic creates further unpredictable, and potentially very large, bursts in addition to the video and voice traffic.

To deliver consistently high quality of performance, Magor TeleCollaboration combines the burst reduction features of SVC++ with comprehensive traffic conditioning. This further reduces the size of bandwidth bursts and optimizes collaboration responsiveness, while protecting the video and voice quality. The traffic conditioning technologies, which include hierarchical weighted scheduling, combined with aggregate shaping at each endpoint, ensure that the aggregate data generated by a Magor TeleCollaboration session is sent at a near-steady rate. This strengthens the ability to behave as a good network citizen, while preserving the required performance characteristics (delay, jitter, loss) and providing a consistent user experience — even over a best effort network, such as the Internet.

Correct errors

To work ubiquitously over both private networks and the Internet, Magor TeleCollaboration is highly resilient to loss, delay jitter and latency. Magor has advanced the state-of-the-art in video networking by implementing jitter-tolerant voice-to-image synchronization combined with on-demand Forward Error Correction (FEC). This unique approach minimizes overhead while correcting errors. Each Magor TeleCollaboration session can determine whether FEC is required on certain parts of the video.

1 Cisco TelePresence Fundamentals (Chapter 10), by Szigeti Tim et al., ©Ciscopress.com, 2009.