Traditionally, the video used to be a point to point technology wherein one person used to make the video call to another person, with the need for any additional infrastructure. Now think about a scenario, if they want to make a multi-party video call to have a video conference meeting with few additional participants, this is where the MCU comes into the picture.
The MCU provides controlled video and audio connections for each one of these conference participants who will dial into the MCU to conduct their virtual meeting. Each one of these participants sends a unique video stream so the MCU can integrate them all together but so each participant can also then customize their display according to what they want to see. For example, each conference attendee can control whether they want to see themselves or not see themselves on their display monitor, or if they want to see only one speaker or another speaker or a lot of the virtual meeting attendees at the same time or whatever composed image like a presentation is being shown to the group and so on.
Describing technically, MCU’s are DSP (digital signal processing) based hardware whose primary functionality is to decode all incoming media streams, compose a single stream for each far-end participant and finally re-encode that stream before sending it out; needless to say all this requires a huge amount of processing power. Moreover, hardware-based MCUs define scalability on a per-port basis, which means if we want to have more participants connected in a Videoconference at the same time, we need MCUs with more video ports i.e. more DSPs and DSP hardware does not come cheap.
With fast-changing consumer behavior, stiff competition & higher costs, few Video Conferencing industry vendors have started looking for alternate choices to reduce the video infrastructure cost. Because of the change brought by this revolution in the industry, Software based MCU are being evolved.
These are called virtual or soft MCU’s, they eliminates the need for DSP hardware and can run on virtualized servers on private or public clouds, therefore it is significantly cheaper to deploy them in comparison to the expensive hardware MCU. They serve the same purpose as the hardware-based, except that all the transcoding and signal processing is done in the software which introduces big advantages in terms of cost, scalability, and flexibility.
In terms of scalability, customers can often just purchase the number of ports they require at the time and then scale up or down easily as the need changes. Let us discuss few important feature & functionalities of an MCU.
- Video Processing: Video and audio processing are carried out per user rather than per conference. Each user connects using unique, optimized audio and video settings to enjoy the best audio and video quality supported by their endpoint and network, without affecting the other participants in a conference
- Video Switching/Transcoding: MCU supports multiple video codecs, resolutions. When using switched video, all endpoints in the conference must support the same resolution. If a network experiences high packet loss, switched video might not be displayed properly for all endpoints in the conference. While during transcoding video, this limit does imply.
- Interoperability: MCU supports various video & voice codec including H.264, H.265, H.264 High profile, SVC, H.323, and SIP etc. More of this codec support, better the MCU interop.
- Video quality: The MCU delivers high-quality video and audio processing, using latest industry standards and upgradeable DSP chip software.
- Video & Audio Error concealment mechanism
- Error Resiliency algorithms by reducing video freezes in lossy network conditions
- Bandwidth Estimation to improve call quality
- SIP and H.323-based content sharing: The MCU supports sharing presentations and other content via SIP (using the BFCP standard) and H.323 (using the H.239 standard). A user can connect to a meeting from a SIP or H.323 endpoint to share content such as presentations, spreadsheets, documents, and movies.
- Security: To achieve secure communication with endpoints, the MCU uses H.235-based encryption for H.323 endpoints and SRTP and TLS encryption for SIP endpoints. It also features optional PIN protection for joining a conference and web access, and additional PIN protection for conference Moderator Control.
- Management: Intuitive web-based management and control
- Controls: Interactive Voice Response (IVR) messages, In-conference control, Personal layouts per participant, In-meeting indicators etc.