Laboratory Report 96/22
INVESTIGATION OF HD-DIVINE COFDM TRANSMISSION EQUIPMENT.
By Neil Pickford
Abstract:
This report describes a series of laboratory tests that were performed on a set of engineering prototype Digital Terrestrial Television Broadcasting (DTTB) hardware. The equipment was able to be configured to produce a range of different COFDM modulation signals. (COFDM is the modulation format that has been adopted in European DTTB standards). The hardware was produced by Swedish company HD-Divine and was loaned to the Laboratory for testing over a two week period in September 1996.
This testing provided a valuable opportunity to gain practical experience with the operation of a COFDM DTTB system and to validate earlier simulated DTTB measurements made using noise-based test signals.
The report comprises five major sections. These are: an Introduction; a description of the equipment; a description of test procedures; results; and conclusions.
Important results included in this report were:
- a 7 MHz DTTB system capable of providing a useful data rate of 18.2 Mbit/s was found to have a C/N threshold (for 2e-4 errors) at approximately 17.9 dB. The same system was ruggedised against loss of synchronisation down to -16 dB
- Trellis and Reed-Solomon coding are extremely powerful techniques for overcoming the effects of transmission errors and interference. Testing on a 64-QAM COFDM system showed that Trellis coding offered an improvement in C/N threshold of 9 dB. Reed Solomon coding further improved the system noise performance, the improvement being especially marked at low BER.
- Protection ratio measurement for PAL into COFDM interference (using BER as the metric of COFDM performance) indicates that the COFDM system fails at a protection ratio of +7 dB for systems operating well above threshold. For systems close to threshold, co-channel PAL signals could cause performance degradation to the BER of the COFDM signal for protection ratios between +15 and +7 dB.
- the PAL into COFDM protection ratio was degraded by up to 3 dB if the relative frequencies of the COFDM signal and the interfering PAL signal were slightly offset such that the first sound carrier of the PAL signal overlapped with the COFDM signal. This occurs with a relatively small frequency offset of 0.1 MHz.
- Protection ratio measurements for COFDM into PAL intereference (using a "limit of perceptibility" subjective viewing method) were generally consistent with earlier measurements using a shaped noise spectrum to simulate the DTTB signal. The following COFDM into PAL protection ratios were obtained.
Channel Position
|
Limit of Percept P/R |
|---|
Lower Adjacent Channel |
+3 dB |
Co-Channel |
+54 dB |
Upper Adjacent Channel |
+0 dB |
- Additional Protection Ratio measurements for other frequency offsets between the co-channel and adjacent channel cases are shown in graphs within the report.
- these protection ratio measurements were performed using a "limit of perceptibility" method because a proposed alternative method, Subjective Comparison Method (SCM), was unsuitable except for co-channel or near co-channel conditions.
- Observation of the COFDM signal shows that the synchronising component of the signal has a significant role in determining the peak power of the COFDM signal (it is 10 dB above the "signal" part of the COFDM signal).
- The transmitter linearity measurements showed that , for levels of up to -10 dB below PAL operating levels, existing solid state UHF PAL transmitters can handle COFDM signals.
- Some existing power measurement procedures for PAL transmission equipment overestimate actual COFDM power levels by approximately 2.3 dB.
Return to 1996 Reports List
or
Return to Lab Home Page