ADJACENT AND CO-CHANNEL INTERFERENCE
DISTURBANCES FROM A DIGITAL TERRESTRIAL
TELEVISION SIGNAL(COFDM-8K System) ON
ANALOGUE PAL SYSTEMS
A. Arrinda (jtparsaa@bi.ehu.es), M. Mª Vélez, P. Angueira, D. de la Vega, J. L. Ordiales
UNIVERSITY OF THE BASQUE COUNTRYETSII y IT de Bilbao Alda. Urkijo s/n 48013 BILBAO, Spain
Abstract
This paper presents some early results from a measurement campaign in the service area of the digital terrestrial television (COFDM-8K system) experimental network of Madrid (Spain). The purpose of this survey has been to determine the degradation suffered by analogue PAL-G signals in presence of co-channel and adjacent channel digital emissions. The result from this study is a curve that estimates the protection ratios between analogue and digital received power levels.
1. INTRODUCTION
The introduction of the digital terrestrial television, standardized in Europe by the Digital Video Broadcasting (DVB) consortium [1], will be gradually carried out in most European countries. This progressive development will allow the consumers to make a gradual access to the new services, taking into account that they will need to purchase a new IRD. In the same way, broadcasters will have a reasonable time margin to be able to update their infrastructures towards the digital terrestrial technologies.
These facts lead to a simulcast environment during a time period of 8-10 years depending on different countries and estimations, in which analogue and digital services will be sharing the same spectrum frequency bands. As an example, Spain will be one of the first countries launching digital terrestrial television services by January 2000 and the shut-down of analogue PAL-G services is expected to be completed by the year 2010. During the transition period, due to the co-existence of both technologies, the minimization of analogue quality of service degradation caused by digital transmitters in the same coverage areas will be one of the main factors to take into account when planning digital services.
2. TARGETS
This work has focused on the study of interference disturbances on an analogue PAL-G television signal caused by the presence of a digital terrestrial television [2][3] broadcast inside the same coverage area. Some protection ratio values have been specified by the ITU-R [4]. The study presented has considered two different types of interference:
Degradation of a PAL-G signal caused by a COFDM-8K broadcast on the adjacent frequency channel.
Degradation suffered by a PAL-G signal interfered by a COFDM-8K broadcast on the same frequency channel (co-channel interference)
These two scenarios will be quite common during the co-existence period of digital and analogue services. Channel protection ratios will be necessary to plan digital services without interfering analogue coverage areas. Up to now, when planning analogue broadcast networks there have been tight restrictions for the selection of frequency channels. It has not been possible to use two adjacent channels in the same coverage
area, so these adjacent channels at each transmitter site could become suitable candidates in order to allocate digital services [5][6].
3. EXPERIMENTAL NETWORK
Measurements presented here have been taken inside the coverage area of the digital terrestrial television experimental network located in Madrid (Spain). This network has been designed for test purposes and allows broadcasting according to the European DVB-T standards [1]. This work has been the result of the co-operative research and development project VIDITER, developed by the Spanish Government, the telecommunications operator RETEVISION, Spanish broadcast and radiofrecuency equipment manufacturers and the University of the Basque Country.
Two main sites compose the broadcast infrastructure, one in the middle of Madrid, at the Torrespaña communications tower and the second at the Navacerrada mountain range, 50 Km away from the city at 2256 meters above sea level. The network architecture has been described in detail in [7] and [8].
The interference measurements have been carried out broadcasting DVB-T signals from the Navacerrada transmitter. Four dipole panels with horizontal polarisation and oriented 120 degrees from North direction compose the radiating system installed on this location. The tower height is 20 meters a.g.l. and the gain pattern has been designed to ensure adequate coverage on the north-west urban and suburban areas of Madrid.
From the same site, using an adjacent channel, a PAL-G signal is also being broadcasted with a different radiation pattern. Despite of that difference, the field strength of the analogue signal within the area where measurements have been taken is higher than the minimum values recommended by the ITU-R to achieve a good quality of service [9]. Figure 2 shows the spectra of both analogue and digital signals from the Navacerrada transmitter captured at one of the measurement points.
There is still another PAL-G signal, within the measurement area, broadcasted from a transmitter site located at the southern suburbs of Madrid on the same channel as the digital signal broadcasted from the Navacerrada transmitter.
Figure 1. Measurement area and studied signals.
510 MHz
518 MHz
526 MHz
CHANNEL 26
CHANNEL 27
f = 519.25
vf = 524.75
sThe mentioned set of analogue and digital signals with relevant power levels in different points within the measurement area has allowed to study interference caused by digital co-channel and adjacent channel signals on PAL-G services. Figure 1 shows the geographic location of the transmitter sites and the measurement area. Table I shows the main transmission parameters for all the surveyed signals within the measurement area.
Transmitter Signal Channel Power (w)
Navacerrada COFDM-8K 26 250 Navacerrada PAL-G 27 5000
Madrid PAL-G 26 5000
Table I. Parameters of the transmitter in the
area near Madrid.
The DVB-T standards describe two different modes of operation (8K and 2K) for digital terrestrial television broadcasting. The VIDITER project has focused his efforts on the 8K system, as this is the choice for future networks in Spain. Other transmission parameters like carrier modulation, Viterbi coding and guard interval are shown on table II [10].
COFDM 8K System
Number of carriers 6817Bandwidth 7.61 MHz Carrier Modulation 64 QAM
Inner code rate 2/3 Guard interval 1/4
Table II. COFDM transmitter operating mode.
4. MEASUREMENT TECHNIQUES
First step before the measurement campaign was to obtain computer aided field strength distribution simulations over the whole measurement area for the three signals considered in the interference analysis (PAL-G on channels 26, 27 and COFDM on channel 26). These simulations allowed a suitable selection of the measurement locations.
Measurement locations were divided in two different groups. The first group was composed by points where the received signal power was only significant for those broadcasted in the channel 26, i.e., COFDM-8K from Navacerrada and PAL-G from the transmitter site at the south of Madrid. Data obtained from such points was
useful to survey co-channel interference disturbances. Locations where the only relevant received signals were the ones broadcasted from the Navacerrada transmitter site, i.e., PAL-G on channel 27 and COFDM-8K on channel 26, were included in the second group. Measurements taken at these points were used to study adjacent channel interference disturbances.
All the measurement locations have been selected with the restriction of having the Fresnel first ellipsoid free of obstacles from the transmitter (or transmitters) that was being analyzed at that point. Also, the analogue PAL-G received field strength had to be higher than the minimum values recommended by the ITU-R to ensure proper signal quality.
Two different measurements have been made at each location. First one consisted on analyzing all the baseband parameters that characterize the PAL-G image quality. When making this first measurement, the COFDM-8K transmitter was switched off, so the deviations from ideal values were caused by transmission impairments and propagation path disturbances. The same parameters were captured on the second measurement but this time with the digital transmitter switched on. The nature and amount of degradation caused by the presence of the digital signal (on the same channel at some points and on adjacent channel at the rest) was analyzed by means of comparing the mentioned two data sets.
5. MEASUREMENT SYSTEM
The whole measurement campaign was carried out using a mobile unit specially designed for that purpose. It consisted on a van equipped with the following:
Directive log-periodic antenna with a gain of 7 dBi and flat response in the frequency margin where digital and analogue channels were allocated (510-526 MHz). This antenna was situated at the top of a telescopic mast that allowed taking data at 10 metres above ground level.
Professional PAL-G Signal Demodulator (Rhode & Schwarz EMFP).
Baseband Video Parameter Analyzer (Tektronix VM700A)
TV Monitor (SONY BVM-2010P)
Spectrum Analyzer (Hewlett Packard HP8591A)
All the measuring equipment was controlled by means of data capture and storage software
running on a laptop. Figure 3 shows the block diagram of the whole measuring system inside the mobile unit.
Figure 3. Measurement system.
6. RESULTS AND DISCUSSION
The digital terrestrial television signal COFDM-8K is composed by 6817 carriers equally spaced 1116 Hz from each other. This carrier set occupies a spectrum range which bandwidth is 7.6 MHz. The shape of this signal, due to the closeness of the mentioned carriers, is very similar to a flat response noise spectrum inside that 7.6 MHz bandwidth. Because of that, the expected interference caused onto an analogue PAL-G signal broadcasted on the same or adjacent channel could be an increase of the noise rms value.
Although the highest variation (with and without the digital COFDM-8K signal presence) was expected for the parameters related to the baseband S/N ratio, linear and non-linear distortion parameters were measured during this campaign. Obtained results have corroborated former suppositions so the only parameter included in the analysis is the baseband S/N ratio.
The parameters measured corresponding to the PAL-G signal at each point were the following: Baseband Unweighted Signal to Noise Ratio. Subjective Image Quality.
Received Power.
The only data analyzed for the COFDM-8K signal was the received power measured in the 7.6 MHz bandwidth[11].
In order to obtain the Baseband Unweighted
Signal to Noise Ratio the luminance bar
amplitude (which is inserted in one of the test lines commonly known as Vertical Interval Test
lines VIT) is measured and compared with the
noise rms value of an empty line. The theoretical value of this parameter should be infinite as the ideal bar amplitude is 700 mV and the theoretical rms value of noise should be zero.
As mentioned before, two different measurements of the Baseband Unweighted Signal to Noise
Ratio have been obtained, one in presence of the
digital signal and the other one in absence of it. The COFDM-8K signal interference causes degradation on the analogue baseband signal quality increasing the noise and therefore decreasing the unweighted S/N ratio. The difference between this ratio in presence and absence of the COFDM-8K signal will be referred to as S/N degradation. If this value is high enough the interference will produce subjective image quality degradation, i. e., from grade 4 to 3 in a scale of 1 to 5 levels of image quality.
Another considered value at each measurement location is the difference between the received power of the PAL-G signal and the received power from the COFDM-8K transmitter. The spectra of both signals have been captured using a resolution bandwidth of 100 KHz. The PAL-G received power is (due to the narrowband nature of this signal) approximately equal to the video carrier received power.
The digital received power is more complex to obtain with a standard spectrum analyzer and should be measured with a specifically designed 8 MHz very sharp filter and a field strength meter
[12][13]. In this measurement campaign no field
strength meter was available so the way to obtain the received power was made by means of adding a correction factor to the measurement done by the spectrum analyzer. First, the medium spectral power density was obtained using a 100 KHz resolution bandwidth filter. The correction factor (k) added to this value was:
k = 10 log (7600/100)
This assumption was possible because the COFDM-8K was approximately flat over the 7.6 MHz bandwidth at the selected measurement locations. The ratio of both measured power levels ANTENNA RF Switch Spectrum Analyzer Video Analyzer PAL-G Demodulator Baseband Video
CONTROL & DATA ADQUISITION RF
RF Signal
RF
(analogue signal power to digital signal power) will be directly related to the amount of S/N degradation. As this ratio increases the S/N degradation will also increase, until it reaches the zero value. This relationship will provide some reference values for digital terrestrial television planning in order to avoid disturbances on existing analogue services [14].
ITU-R grade
Image Quality
5 Excellent
4 Good
3 Acceptable
2 Poor
1 Bad
Table III. Image quality scale.
The degradation of the PAL-G image quality parameters can be appreciated by evaluating the subjective image quality. The ITU-R establishes five different grades for image quality depending on the impairments observed on a picture displayed by a professional TV monitor. This quality scale is shown in table III. At each of the measurement locations three different persons formerly trained to detect common TV image impairments observed the subjective image quality. The reduced number of observers was imposed by obvious lack of space at the mobile unit.
6.1 Adjacent Channel Interference
On figure 4 the horizontal axis shows the difference between the received power on channel 27 (analogue signal) and 26 (digital signal). The vertical axis represents the S/N degradation. As the power ratio of both signals increases, the S/N degradation diminishes.
At measurement locations where the power received from the COFDM-8K transmitter was higher than the one received from the analogue one, the S/N degradation appeared to be less than 10 dB. The tendency curve shows that higher power differences (around 30 dB) the S/N degradation is unappreciable.
As far as the subjective quality of the analogue television signal is concerned, and at most points, degradation in image quality is observed when comparing pictures in presence and absence of the digital signal. However, there have been locations where no differences were appreciated between both situations.
Table IV shows image subjective qualities at some measurement points compared. The relationship between power ratios and S/N degradation can also be seen on the same table.
Figure 4. Relationship between S/N degradation and PAL/COFDM power ratio.
0
5
10
15
20
-20
-10
0
10
20
30
40
50
60
70
Protection Ratio PAL/DTV (dB)
S
/N
d
e
g
ra
d
a
ti
o
n
Measurements
Power Ratio PAL COFDM S/N Degradation (dB) ITU-R Quality grade ITU-R Quality grade (interfered) -4,8 dB 12,3 3 3 -2,8 dB 14,3 3 2 -1,3 dB 13,1 4 3 -0,8 dB 14,1 4 3 -0,3 dB 6,8 2 2 1,2 dB 5,7 3 2 2,2 dB 7,7 4 3 9,2 dB 2,9 4 4 14,2 dB 8,9 3 3 14,2 dB 6,5 3 3 22,7 dB 7,1 4 3 39,2 dB 2,2 4 4
Table IV. Image quality in presence and absence
of the digital signal.
Recently, a digital terrestrial television COFDM-8K system transmitter has been installed in Bilbao (SPAIN) so a few measurements have been made on this transmitter service area to analyze the impact of this emission on adjacent analogue PAL-G channels. Data resulting from this survey have been added to the curve obtained from the measurement campaign in Madrid. Data obtained in Bilbao match the expected results. Figure 5 shows measurements taken from both networks.
6.2 Co-channel Interference
The co-channel interference analysis has been carried out following the same methodology. It should be remarked that calculation of digital
received power is more complex, as there are two signals on the same frequency at the same time, and the error will also be higher. By now, few data have been obtained but much higher S/N degradation values are expected. First results have shown S/N degradations of 15 dB when the power ratio was 25 dB.
7. CONCLUSIONS
As a result from the work carried out during the measurement campaign in the service area of the VIDITER project experimental network some curves have been obtained. These curves show the effects caused by a digital terrestrial television COFDM-8K on an analogue television PAL-G signal broadcasted on an adjacent upper channel. The relationship between the ratio of PAL and COFDM signal powers and the PAL-G baseband unweighted signal to noise degradation has been obtained as a result of field measurements in a real broadcast environment.
The obtained results are preliminary and more extensive measurement campaigns are needed to obtain definitive curves to use when digital terrestrial television planning becomes a reality over the whole Europe.
Actually, another experimental network is being developed in Bilbao (Spain). The channel assigned for that emission is the lower adjacent one to one used to broadcast a PAL-G signal.
0
5
10
15
20
-20
-10
0
10
20
30
40
50
60
70
80
Protection Ratio PAL/DTV (dB)
S/
N
de
g
ra
da
tio
n
Measurements Madrid
Measurements Bilbao
Figure 5. Relationship between S/N degradation and PAL/COFDM power ratio.
A more detailed measurement campaign is now under design to obtain more data and as a result, to be able to accurately define the protection ratios to be used in digital terrestrial television planning.
8. ACKNOWLEDGEMENTS
The authors wish to thank all the partners of the VIDITER project and specially RETEVISION for facilitating this work.
9. REFERENCES
[1] ETSI (European Telecommunications Standards Institute) “Digital Video Broadcasting (DVB); Framing structure, channel coding and modulation for digital Terrestrial television (DVB-T)” ETS 300 744 March 1997.
[2] J.H. Stott “Explaining some of the magic of COFDM” Proceedings of 20th International Television Symposium, Montreux, 13-17th June 1997.
[3] W. Y. Zou, Y. Wu. “COFDM: An Overview” IEEE Transactions on Broadcasting, Vol. 41, No. 1, March 1995.
[4] Rec. ITU-R BT. 1368-1 “Plannig Criteria for Digital Terrestrial Television Services in the VHF/UHF Bands”
[5] A. Vahlin, N. Holte “OFDM for Broadcasting in Presence of Analogue Co-Channel Interference” IEEE Transactions on Broadcasting, Vol. 41, No. 3, September 1995.
[6] S. O’Leary “Digital/Analogue Co-channel Protection Ratio Field Mesaurements” IEEE Transactions on Broadcasting, Vol. 44, No. 4, December 1998.
[7] P. Cañizares, J.L. Torres, J.A. Martínez “VIDITER: Spanish experience on DTT” Proceedings of 20th International Television Symposium, Montreux , 13-17th June 1997. [8] P. Cañizares, J.L. Torres, S. Mata. “The first Spanish experience on digital terrestrial television broadcasting” IBC (International Broadcasting Convention), IEE No. 428, 1996. [9] ITU-R BT 417-4 Minimum field strengths for which protection may be sought in planning a television service”. Volume 1997-BT series. [10] Lis Grete Moller. “COFDM and the choice of parameters for DVB-T” Proceedings of 20th
International Television Symposium, Montreux, 13-17th June 1997.
[11] ETSI (European Telecommunications Standards Institute] “Digital Video Broadcasting (DVB); Measurement guidelines for DVB systems” ETR 290, May 1997.
[12] C. Weck, R. Schramm. “RECEIVING DVB-T: Results of field trials and coverage considerations” Proceedings of 20th International Television Symposium, Montreux, 13-17th June 1997.
[13] T. de Couasnon et al. “Results of the first digital terrestrial television broadcasting field-tests in Germany” IEEE Transactions on Consumer Electronics, Vol. 39, No. 3, August 1993.
[14] ITU-R. Rec. 500-3, “Method for subjective quality evaluation of TV images” XVI Plenary Session, Dubrovnik, 1986.
