DRM to Extend up to 120 MHz

DRM (Digital Radio Mondiale, not to be confused with digital rights management), the new digital radio standard that currently targets the LW, MW and SW frequency bands, is to be extended from its current maximum operating frequency of 30 MHz up to 120 MHz, with work expected to be completed by 2008 to 2010.

Currently, DRM is optimised for the lower frequency bands it was originally designed for, and because broadcasting channels in these frequency bands are very narrow, the widest channel that DRM can currently use is 20 kHz, which only allows DRM multiplexes to carry very low bit rates, typically between 25 kbps and 55 kbps.

After enquiring about what changes might happen to the DRM specification, I received a reply which made the following points (all made with the caveat that it is very early days, so these things will change):

• a 40 kHz channel is very likely to be an option and "maybe even 50kHz, as
  that fits into the current Band II frequency grid"
• it makes sense to use wider-channels at higher frequencies
• 16-QAM is likely to be preferred over 64-QAM, because 16-QAM is more robust and requires lower transmission powers
• the extended specification is likely to use as much of the current specification as possible
• it is likely that DRM multiplexes will be able to carry 4 services per multiplex, as they do now
• MPEG-4 5.1 surround sound might become an option

The following table shows approximate bit rate levels that could be expected for different channel frequencies where the transmission frequency is 120 MHz, the receiver is moving at 100 mph, and DRM Mode A and FEC code rate = 0.6 are being used:

To avoid the crucial mistakes made when DAB was designed (i.e. insufficient data capacity for the actual requirements of a digital radio system for the 21st century), the extensions to DRM should take into consideration how DRM multiplexes are likely to be actually used in practice, and make sure that DRM can cater for -- and preferably surpass -- the most demanding requirements.

The most demanding case is when the maximum number of services are carried on a multiplex, which in DRM's case is 4 services. DRM can use the most efficient audio codec that exists today, HE AAC, but with a bandwidth of 50 kHz then using 16-QAM the bit rate per service would only be 26 kbps, and for 64-QAM 37kbps. HE AAC may be very efficient, but it requires a bit rate of 64 kbps to provide good audio quality, so 50 kHz channels are not sufficient. 

100 kHz Channels are Required

As can be seen in the above table, 100 kHz channels would allow bit rates of between 208 kbps and 300 kbps, depending on whether 16-QAM or 64-QAM is used, and depending on the FEC code rate. This would allow bit rates of between 52 kbps and 75 kbps per service if 4 services shared a multiplex, and using the HE AAC codec then this would be sufficient to provide good audio quality (significantly better than on DAB).

New FEC Coding Required

Despite the fact that the DRM specification was not ratified until April 2003, DRM does not take advantage of the major advances made in error correction coding during the early- to mid-1990s and uses the relatively weak multi-level coded modulation (MLCL) forward error correction (FEC) scheme (which is nowhere near as weak as DAB's FEC coding scheme, but DAB's is extremely weak). Therefore, with DRM extending to 120 MHz, it would be a mistake not to take advantage of the newer, near-optimal FEC coding schemes such as turbo coding (which will be used for the new HSDPA 3G system upgrade to allow higher download bit rates), or LDPC (low density parity check) coding (LDPC is used in the new DVB-S2 specification). Alternatively, if they want to minimise the differences between the new and old DRM specification, then a turbo coding version of multi-level coding has been developed where turbo encoders simply replace the current, standard convolutional encoders.

The advantages that a stronger FEC coding scheme would bring are as follows:

• far more robust reception

• increased multiplex data capacity

• lower transmission powers

It should be interesting to see how much the engineers involved with DRM want to improve the system, because many of the engineers were also involved with DAB, and they might choose not to improve DRM by much so as to avoid showing-up the dreadful, out-dated DAB system.

Bookmark with:
Digg	del.icio.us	Reddit	Facebook
Google	Stumbleupon	Slashdot