Hello everyone!
This post explains the SMC2.0 rate adaptation rules. When a conference is created on the SMC2.0, the system adjusts the conference and participant rates and audio and video parameters based on the bandwidth of the conference and participant. In this way, MCU resources can be properly allocated.
PREREQUISITES
When the conference is created, the conference and participant parameters for a conference have been set to Auto.
INTRODUCTION TO THE RANGE SETTING POLICY
When scheduling a conference, SMC2.0 adds the participants to the conference based on the site requirements. The conference parameters of the added sites may not be completely adapted, including the adaptation protocol, video format and audio protocol. SMC2.0 functions as the coordinator, as it can adjust the parameters to ensure that conferences can be scheduled properly.
During the actual scheduling, SMC2.0 schedules the MCU to encode and decode the audio and video data of a conference. Before scheduling a conference, the SMC2.0 delivers the global ranking information and conference/participant parameters to the MCU. The MCU allocates the resources to the conference and processes the audio and video data based on the information delivered by the SMC2.0 (the following uses the ViewPoint 8000 series MCU as an example - during conference scheduling and site addition, audio and video resources are allocated).
Audio resources
The audio resources allocated by the MCU during conference scheduling and participant adding are reflected in the number of occupied audio DSPs (Digital Signal Processor). The number of occupied audio DSPs can be calculated based on the number of AAC-LD (Advanced Audio Coding With Low Delay) in the MCU license and the number of audio channels. The DSPs occupied by different audio protocols are as follows:
Table 1. DSP occupied by different audio protocols
Audio protocols | G.711 | G.722 | G.728 | AAC-LD (mono) | AAC-LD (Bino-channel) | AAC-LC (mono) | AAC-LC (Bino-channel) |
DSP occupied | 4 | 4 or 5 | 8 | 16 | 32 | 16 | 32 |
Video resources
The video resources allocated by the MCU during conference scheduling and site addition include video adaptation resources and continuous presence codec resources:
video adaptation resources - allocate resources based on the N rule.
That is, if a conference template contains N sites of different levels (different rates, protocols, and formats), N channels of adaptation resources are required.
Adaptation resources = Number of adaptation channels x Adaptation resources required for the maximum conference capability.
Continuous presence codes resources - allocate resources based on video sending and receiving modes.
In symmetric mode, all codecs are calculated based on the conference video capability (N decoding and 1 encoding).
In asymmetric mode, one codec is allocated based on the conference video capability. If the conference video protocol is lower than H.264, N channels of decoding are allocated based on the conference video capability. Otherwise, N channels of H.264 CIF decoding are allocated.
To properly allocate MCU resources and minimize resource waste and scheduling failures, the SMC2.0 delivers information to the MCU based on certain policies, that is, range setting policy.
Currently, the SMC2.0 has a new rate range adaptation policy and an old rate range adaptation policy. The SMC2.0 uses the new rate range adaptation policy. You can modify the policy as required.
HOW TO CONFIGURE THE RANGE SETTING POLICY
To save MCU transrating resources, divide the conference rate range (128 kbit/s to 8 Mbit/s) into several subranges and specify audiovisual parameters for each subrange. Any two neighboring subranges must be adjacent but not overlapping. When a conference is to be held, the SMC2.0 determines the conference parameters (including the video protocol, video format and audio protocol) based on the conference rate and participant parameters (including the video protocol, video format and audio protocol) based on the number of participants and rate range thresholds.
Log in to the SMC2.0 web interface.
Choose System > Settings.
Choose Range Setting Policy, set the parameters shown in the following picture and click Save.
Table 2. Parameter description of range setting policy
Parameter | Configuring the range setting policy |
Range No | Identifies different rate ranges. |
Audio and video parameter | Includes the video protocol, video format, and audio protocol. Default audio and video parameters are also provided in the default rate range of the SMC2.0. You are advised to retain the default settings. |
Threshold | In the old policy, the parameter adjustment procedure is determined based on the comparison between the threshold and the number of participants. |
Rate Range | Set the minimum and maximum rates for each range.You can set the rate range by adjusting the minimum rate in each range. The maximum rate in the next range automatically matches the minimum rate in the previous range.(The default maximum rate in the first range is 8 Mbit/s, and the default minimum rate in the last range is 64 kbit/s.) |
After the partitioning policy is configured, the SMC2.0 matches the participant into the rate range based on the rate specified by the participant and adjusts the participant parameters.
OLD RATE ADAPTATION RULES
Rate Adaptation Rules
Determine the rate ranges for a conference and its participants based on their specified rates.
Adjust the conference rate and parameters. Use the conference rate specified beforehand and audiovisual parameters in the corresponding rate range.
Adjust the participant rate and parameters. Check whether the number of participants in rate ranges exceed the thresholds and determine the participant rate and parameters accordingly.
If the number of participants is no smaller than the threshold, check whether all participants have the same rate. If yes, the participants still use the rate. If no, all participants decrease their rates to the minimum in their rate ranges.
If the number of participants is smaller than the threshold and participants exist in the next rate range, the participants within the current rate range will use the maximum rate in the next rate range.
If the number of participants is smaller than the threshold and the next rate range does not have any participants, the participants within the current rate range will use the minimum rate in the next rate range.
Participant parameters use audiovisual parameters adopted after participant rate is adjusted. If the value of an audiovisual parameter goes beyond the value range allowed the participant, the SMC2.0 automatically adjusts the audiovisual parameter to the maximum value allowed by the participant.
Examples
This section assumes four rate ranges and describes how old rate adaption rules determine the participant rate.
Rates are divided into six ranges: [4 Mbit/s, 8 Mbit/s], [1536 kbit/s, 4 Mbit/s), [512 kbit/s, 1536 kbit/s), [192 kbit/s, 512 kbit/s).
The maximum numbers of participants allowed in ranges 2 and 3 are 4 and 2 respectively.
The conference has three participants: participant 1 (1920 kbit/s), participant 2 (1152 kbit/s), and participant 3 (1216 kbit/s).
Based on the participants' transmission rates, the SMC2.0 recognizes that participant 1 belongs to range 2, and participants 2 and 3 belong to range 3.
Following the rule for determining the participant rate, the SMC calculates the participant rate as follows:
The number of participants with rates falling into range 2 is smaller than the threshold set for range 2, so participant 1 in range 2 will decrease its rate to the maximum rate of range 3 that is, 536 kbit/s.
Now three participants reside in range 3. However, only two participants are allowed in range 3 at most.
The three participants have different rates (that is, 1152 kbit/s, 1216 kbit/s, and 1919 kbit/s), so they finally use the lowest one among the three rates, that is, 1152 kbit/s.
As a result, the three participants join the conference all at 1152 kbit/s.
NEW RATE ADAPTATION RULES
Rate Adaptation Rules
Determine the rate ranges for a conference and its participants based on their specified rates.
Adjust the conference rate and parameters. Adopt the lowest rate in a rate range as the conference rate and audiovisual parameters in the rate range.
Adjust the rates and parameters for participants.
If the adjusted conference rate and audiovisual parameters are supported by participants, they are adopted by the participants.
If the adjusted conference rate and audiovisual parameters are not supported by participants, adjust the rates and parameters for the participants as follows:
If a participant supports audiovisual parameters of the range where the participant resides, the participant adopts the audiovisual parameters and the minimum rate of this range.
If a participant does not support audiovisual parameters of the range where the participant resides, the SMC2.0 decreases the participant's rate to the next range and determines whether the participant supports audiovisual parameter of this range. If supported, the participant adopts the audiovisual parameters and the minimum rate of this range.
If a participant does not support audiovisual parameters in any range, the participant adopts the rate and audiovisual parameters lower than the conference's.
Reserve resources for the rate range that does not contain any participants and all its lower ranges.
During a conference, the conference's capabilities, participants' capabilities, and rate ranges' capabilities are sorted. The participant that joins the conference later will adopt the maximum capabilities among the sorted capabilities.
Examples
This section assumes four rate ranges and describes how new rate adaption rules adjust the rates and parameters of a conference and its participants.
Rates are divided into six ranges: [4 Mbit/s, 8 Mbit/s] H.264 1080p, [1536 kbit/s, 4 Mbit/s) H.264 720p, [512 kbit/s, 1536 kbit/s) H.264 4CIF, [192 kbit/s, 512 kbit/s) H.264 CIF.
The conference has a rate of 2 Mbit/s and contains four participants: participant 1 (2 Mbit/s H.264 1080p/720p/CIF), participant 2 (1 Mbit/s H.264 1080p/720p/4CIF), participant 3 (5 Mbit/s H.264 1080p60/1080p/720p), and participant 4 (5 Mbit/s H.264 720p).
The adaptation procedure is as follows:
Participant 1 falls into range 2 and supports the adjusted conference parameters, so participant 1 adopts the rate of 1536 kbit/s and the parameter of H.264 1080p.
Participant 2 falls into range 2 and supports the adjusted conference parameters, so participant 2 adopts the rate of 1536 kbit/s and the parameter of H.264 1080p.
Participant 3 falls into range 1 and supports the adjusted conference parameters, so participant 3 adopts the rate of 4 Mbit/s and the parameter of H.264 1080p.
Participant 4 falls into range 1 but does not support the adjusted conference parameters. As participant 4 supports audiovisual parameters in range 1, participant 4 is moved to range 2 but still does not support audiovisual parameters in range 2. Then participant 4 is moved to range 3, and it supports audiovisual parameters in range 3. Finally, participant 4 adopts the rate of 512 kbit/s and the parameter of H.264 4CIF.
As a result, the conference's rate is 2 Mbit/s H.264 1080p, participant 1's rate is 1536 kbit/s H.264 1080p, participant 2's rate is 1536 kbit/s H264 720p, participant 3's rate is 4 Mbit/s H.264 1080p, and participant 4's rate is 512 kbit/s H.264 4CIF.
That is all, thanks for reading!
Nice 
