 |
|
|
||
 |
||||
|
||||
|
|
||||||||||||||||||
 |
|||||||||||||||||||
BandplanThe current broadcast bandplan for medium wave in Europe is based on 9 kHz channel spacing - starting at 531 kHz and ending 1611 kHz. This give 120 potential radio channels but to avoid interference not all channels can be used in each country. In the UK about 76 radio channels are used. Long wave also uses a 9 kHz bandplan (153 to 279 kHz) - however there are some existing long wave radio stations that are unlikely to convert to DRM as their AM carrier is used for data transmissions. [see LF Radio Data for more information] Short wave is divided up into many bands allocated to broadcasting and uses a 10 kHz interleaved channel spacing. Radio stations broadcast on a frequency that ends in either nnnn0 or nnnn5 kHz. Double channelOne of the ‘killer features’ of DRM is the ability to broadcast multimedia but any data stream transmitted has an adverse impact on the quality of the audio as the bit rate available to the audio has to be reduced accordingly. To overcome this I would be in favour of allowing medium wave stations to broadcast using 18 kHz (double channel), this is defined within the ETSI DRM specification so a DRM radio showing the DRM logo should already incorporate appropriate filters. Some radio stations could continue to broadcast in 9 kHz channel such as sports or news stations. Short wave stations that are broadcasting mainly speech (which constitutes the majority of broadcasts) or news would find the 10 kHz channel adequate but any station transmitting music (such as RTL) or multimedia should be allowed to broadcast in 20 kHz. Obviously using a double channel (18 kHz instead of 9 kHz) halves the number of medium wave radio channels but more than doubles the available data rate per radio station. As already explained Single Frequency Networks makes more efficient use of the radio spectrum so this increases the available frequencies when compared to AM. Only a DRM station broadcasting in mode A or mode B using 18/20 kHz spectrum can deliver ‘FM quality’ audio. The audio quality will be greatly improved and the DRM station can broadcasts other services without having a detrimental effect on the quality of the main audio. Bi-lingual broadcasts become a possibility as well as transmitting text (news, weather, etc.) or a second audio voice giving traffic news, weather, etc. This would be an attractive proposition particular for the ‘drive-time’ shows where commercial radio makes most of its money. I would prefer to have fewer medium wave DRM radio stations broadcasting good quality audio than many stations broadcasting mediocre quality (i.e. DAB in the UK). More stations does not necessarily mean more choice just more duplication of the same radio format. At night time medium wave reception can deteriorate, this is due to changes in the ionosphere that allow sky wave propagation. As a result more distant radio stations become audible, these stations are not normally heard during the day. A medium wave radio station broadcasting DRM will probably need to increase the protection level at night to compensate for increased reception errors caused by co-channel/skywave interference. However this results in a reduced bit rate being available. Consequently the quality of the audio will be far from the promised ‘near FM quality’ if single channel occupancy is being used. (See modes page for more examples of typical bit rates).
Comparison of bit rates for different modes and spectrum occupancy (MSC = 64-QAM, protection level 1). OFDMWith a OFDM signal it is not necessary to have a frequency guard band between DRM stations. On medium wave there is a 9 kHz gap between stations so that the analogue IF filters in the receiver can demodulate an AM station without the risk of interference from any adjacent co-channel AM stations. With COFDM the demodulation is done in the ‘frequency domain’ using Digital Signal Processing software were a brick wall type filter can be implemented in software (i.e. a shape factor of 1:1). It is possible for DRM stations to be broadcasting right next to each other. The actual bandwidth of a ‘10 kHz’ Mode B DRM station is nominally 9.657 kHz so there is no overlap or interference between two Mode B stations if they were broadcasting on x and x+10 kHz. From the DRM specification - Mode A Tu = 24.00 mS = 41.666 Hz spacing. Tu = useful part of symbol Spectral occupancy #2 ( 9 kHz) = 204 OFDM carriers. 9 kHz spectral occupancy = 41.666 x 204 = 8499.86 Hz (mode A) X bandAt present in the UK frequencies between 1642 - 1782 kHz are allocated to analogue cordless telephones (CT1). From ‘Responses to the Consultative document on the Future of Analogue and CT2 Cordless Telephony in the United Kingdom’ These analogue phones are being withdrawn in favour of digital DECT phones and cannot be sold after April 2005 when these frequencies could be re-allocated. This process started in the year 2000 with a five year notice that these changes will happen. The allocation of these frequencies to AM broadcasting is already happening in the USA. 1. 1642 kHz Channel number with the associated CT1 base station frequency. If the radio authorities (Ofcom in the UK) want to promote DRM as an alternative to AM then they should consider allocating this eXtended band exclusively to new DRM radio stations. This way DRM could be introduced without the need to replace existing AM stations. A DRM radio station broadcasting using 18 kHz radio bandwidth will easily demonstrate the superiority of DRM over AM in terms of audio quality and data services. This will encourage sales of DRM receivers which in turn will convince the owners of AM radio stations to convert to DRM. When broadcasting in a 9 kHz channel the advantages of DRM over AM is not so overwhelming with the audio, particularly rock and pop music, sounding so heavily compressed. |
|||||||||||||||||||