What are CU’s?
The capacity of a MUX ensemble is expressed in Capacity Units (CU’s). A CU stands for a capacity of 64 bits.
A total of 864 CU’s are available to transport the digital audio and the error correction. 864 CU’s corresponds to a CIF (Common Interleave Frame) which has a duration of 24ms.
864 CU’s of 64 bits means a total capacity of 55296 bit per 24ms are available as total MUX-capacity.
The duration of a CIF-frame is 24 ms!
This corresponds also to the duration of an audio frame of 24ms!
The origin is to be found in the early DAB development where bit compression was based on MP2 frames of 24ms.
This was one of the problems to be solved for DAB+ who uses HE-AAC with frames of different durations (20, 30, 40 and 60 ms).
In the CU’s , the audio bits together with the error-correction bits are sent. The error code has to be seen as a convolution code in which audio bits and error bits are entangled. The final number of bits is the audio bit rate divided by the error correction factor. One can now easily calculate the number of CU’s for an audio sub-channel in the MUX. This is based on the audio bitrate and the error code used.
Example:
Take the bitrate of the audio signal and divide it by the error correction factor. Bitrate 96kbps and EEP-3A (factor ½) then gives 96kbps divided by ½ or 192kbps total bits to be sent for this audio channel.
- We transform this bitrate from 1 second to 0,024s (24ms). 192kbps * 0,024 = 4,608 kb per 24ms.
- We divide the number of obtained kbits per 24ms by 0,064 kb (64-bit) to get the number of CUs.
- 4,608 kb per 24ms dividing by 0,064kb (64 bit) gives 72 CUs So it is important to know that one CU is 64 bit large and is calculated per 24 ms.
Keep in mind that the maximum number of CUs in a MUX is 864.
Also keep in mind that the transmitter coverage area will decrease slightly when going from EEP-1A to EEP-4A, while the number of stations that fit in the MUX will increase.
| Audio bit rate | EEP-1A | EEP-2A | EEP-3A | EEP-4A |
| 48 kbps | 72 CU’s | 48 CU’s | 36 CU’s | 24 CU’s |
| 56 kbps | 84 CU’s | 56 CU’s | 42 CU’s | 28 CU’s |
| 64 kbps | 96 CU’s | 64 CU’s | 48 CU’s | 32 CU’s |
| 72 kbps | 108 CU’s | 72 CU’s | 54 CU’s | 36 CU’s |
| 80 kbps | 120 CU’s | 80 CU’s | 60 CU’s | 40 CU’s |
| 88 kbps | 132 CU’s | 88 CU’s | 66 CU’s | 44 CU’s |
| 96 kbps | 144 CU’s | 96 CU’s | 72 CU’s | 48 CU’s |
In DAB literature you will find two options for EEP-error correction named A and B. Each option consists of 4 different protection levels ranging from 1 to 4. Level 1 stands for the highest protection level and level 4 for the lowest protection level.
Option A assumes a division of the audio sub-channels in multiples of 8kbps. These formats give the maximum flexibility for organizing audio channels and are used by providers for DAB+.
Option B assumes a division of the sub-channels in multiples of 32kbps and is therefore less flexible for audio services. This option was mainly defined for DMB video services. As an example an 80kbps audio channel is not possible with option B because 80 is not a multiple of 32! So option B will only allow 32kbps, 64kbps, 96kbps as audio bit rates! Therefore option B is not really useful for DAB+ and mixing options A and B in the same MUX is not an option when rolling out DAB+.
More information can be found on page 16 (2.4. Protection levels, coding and net bit rates) in this document.