Canadian digital terrestrial television system technical parameters

IEEE TRANSACTIONS ON BROADCASTING, VOL. 45, NO. 4, DECEMBER 1999

353

Canadian Digital Terrestrial Television System Technical Parameters Yiyan Wu, Pierre Bouchard, and Bernard Caron Communications Research Centre Canada Ottawa, Canada Donald Tyrie and Royce Trenholm Spectrum Engineering, Industry Canada Ottawa, Canada 1. Abstract

On November 8, 1997, Canada formally adopted the ATSC Digital Television Standard for terrestrial transmission [l] as defined in document A/53 of the Advanced Television Systems Committee of the USA [2] and as modified by the FCC document MM Docket No. 87-268 [3]. This paper will discuss the Canadian digital terrestrial television system technical parameters for implementation and frequency planning: interference protection ratios, reference system parameters, and radio frequency emission masks, as well as their design methodology. First, the Canadian terrestrial DTV design guidelines are presented, followed by the introduction of DTV reference receiving system set up and the parameter selection of the interference protection ratios. Last, the design methodology of the emission masks is discussed.

2.

Canadian DTV Service Objectives and Protection Ratios

contour. This implies that reliable service be provided to all locations presently served and that the quality of service be maintained throughout the coverage area. The protected or Grade B contour of NTSC service is defined as the outer geographic limits within which the field strength at 10 metres above ground is sufficient to provide an acceptable picture at least 90 percent of the time at the best 50 percent of receiving locations, i.e., F(50, 90) availability. During the NTSC to DTV transition period, the desired service availability for DTV is F(50, 90). The interference criteria is F(50,lO). Considerations for the future DTV-only plan is incorporated by providing DTV allotments using separation distances based on DTV service availability of F(90, 90) and DTV-DTV interference determination at F(10, 10) [4,5]. Table 1 lists the Canadian [4,5] and the USA [6, 141 DTV system protection ratios. The values in bracket are the measurement results from the Advanced Television Test Center (ATTC) [6]. In Table 1, it can be seen by comparing the first row with the fourth, that the co-channel DTV to DTV interference behaves very much like random noise [17].

The basic objective of the DTV service' is to provide a coverage area replicating as closely as possible the existing NTSC service areas out to the protected or Grade B

Table 1: DTV and NTSC system protection ratios Publisher Item Identifier

S 0018-93 16(99)10440-2

00 18-93 16/99$10.00 0 1999 IEEE

354

Canadian DTV Planning Parameters

2.1 NTSC Co-channel Interference into DTV

3.

The required Carrier-to-Interference ratio (C/I) for NTSC interfering into DTV depends on whether or not a comb filter is implemented in the DTV receiver. Based on laboratory test results on combined noise and co-channel NTSC interference test [6], with the comb filter switched on, the worst C/I that the receiver can withstand is 7.2 dB (DTV/NTSC) at a carrier-to-noise ratio of about 19.5 dB. Without the comb filter, the worst C/I that the receiver can withstand is 20 dB, corresponding to a Carrier-to-Noise ratio (C/N) of 16 dB. This implies that the comb filter degrades the C/N performance by approximately 3.5 dB. The comb filter will only be required during the transition period when both DTV and NTSC are in operation. After the phaseout of the NTSC service, there will be a 3.5 dB improvement in the DTV carrier-to-noise ratio, which means that either an improvement in service availability or a reduction in transmitter power is possible.

The minimum required field strength for the three TV bands using the parameters proposed for the Canadian DTV allotment planning is presented in Table 2 [4, 51.

Assuming that the co-channel interference will either be DTV or NTSC, but not both in the same instance, and that the comb filter will be used in areas where performance is constrained by NTSC interference and switched off in those areas where performance is limited by noise and/or DTV interference, a value of co-channel C/I of 7.2 dB for NTSC into DTV will be appropriate. Meanwhile, the required C/N is at least 19.5 dB.

The receiver noise figure is determined by design. The values recommended for the proposed DTV receiver are 5 dB in the VHF frequency band and 10 dB in the UHF frequency band. The use of a low noise amplifier mounted at the antenna gives a 5 dB noise figure for both VHF and UHF frequency bands, provides gain to compensate for the downlead loss and minimizes the effect of the TV receiver noise figure. External noise is the noise received by the antenna from the environment where it is installed. The external noise results from both natural and man-made sources. In the VHF TV bands, external man-made noise is significant and needs to be included in the determination of the planning factors. In the UHF TV band, external noise is much lower and its effect is negligible compared to the noise of the receiver. The effect of the external noise is included by assuming an equivalent antenna noise, T,. The typical dB values used are 4 and 1 dB in the United States and 6 and 1 dB in the United Kingdom for the low VHF and high VHF TV bands respectively. In Canada, the method described in ITU-R Report 258-5 is used to estimate man-made noise in the receiving environment and to determine an effective antenna noise temperature.

The required field strength is the value of signal strength necessary at the antenna input to produce the minimum Interference carrier-to-noise ratio, which results in the desired quality of Laboratory tests [ 6 ] show that the ATSC system C/N at the service. In the Canadian DTV plan, the required field subjective Threshold Of Visibility (TOV) is 15.28 dB (C/N strengths for Low VHF, High VHF and UHF bands are 35, = 15.19 dB for BER = 3 ~ 1 0 - ~ However, ). this value 33 and 39 dBpV/m, respectively. represents the minimum C/N without additional impairments resulting from co-channel and adjacent channel interference, nonlinear distortion and multipath distortion. Laboratory 4. Adjacent Channel Interference and Canadian DTV R F Emission Masks tests also show that, for typical multipath distortion, an additional headroom of 1.2 to 3.6 dB is required [6]. Assuming the minimum recommended headroom of 1.2 dB An emission mask is used to limit the transmitter spectral results in a minimum C/(N + I) requirement of 16.5 dB at out-of-band spillover to control the interference into the adjacent channels. During the NTSC to DTV transition TOV. period, the predominant interference is the DTV signal Considering the minimum required C/N of 19.5 dB for the interference into the first adjacent NTSC signal. This is the operation of the NTSC rejection filter (comb filter), as limiting factor in the design of the DTV emission mask, discussed in the previous section, an equal partition between because the NTSC threshold of visibility (TOV) is around noise and co-channel DTV to DTV interference is proposed, C/N = 50 dB, while the ATSC DTV threshold is about C/N i.e., C/N = C/I = 16.5 + 3 = 19.5 dB. In the Canadian DTV = 15 dB [6]. This section presents the proposed Canadian channel allotment process, the major limiting factor for DTV planning is co-channel DTV interference.

2.2 Carrier-to-Noise Ratio and Co-channel DTV to DTV

Planning Parameter Frequency (MHz) C/N (dB) K (dB) B (dB) . _(6. MHz) G,(lm2) (dB) Gdipole (dB) Gisotropic (dB)

Line Loss (dB) a (numeric) Balun 300/75 Loss (dB)

Low VHF 69 19.5 -228.6 67.78 -1.77 6 8.15 1.05 0.786 0.5

High VHF 195 19.5 -228.6 67.78 7.25 8 10.15 1.81 0.659 0.5

UHF LNA 615 19.5 -228.6 67.78 17.23 10 12.15 3.29

9552.6 39.8 7.65 35

1169.6 30.68 9.65 33

717.8 28.56 11.65 39

Te 10 log,dTe) GA (dB) Erequtred (dBFV/rn)

0*468 0.5

I

Table 2: Canadian DTV planning parameters.

DTV emission masks and the design methodology used to determine them. Two emission masks are proposed: a “relaxed mask” designed for exactly co-located DTV and NTSC transmission antennas; and a “tight mask” recommended for transmitter separation up to 8 km based on the F(50,90), location and time availability, as the desired signal and F(50,lO) as the undesired signal. An NTSC spectrum weighting function is used in the interference calculation.

4.1 DTV and NTSC Thresholds and Transmission Power

As discussed in Section 2.2, the DTV system TOV for noise and co-channel DTV interference are C/N = C/I = 19.5 dB. The NTSC AWGN channel TOV is about 47 to 50 dB for continuously injected noise [7, 171. A C/N = 50 dB is used in the calculation of the emission mask.

From a coverage point of view, DTV is expected to provide equal or better service than the NTSC Grade B contour using the CCIR F(50, 90) curve [8]. The calculated DTV emission power should be 12 dB below the corresponding NTSC power [9], at frequencies of operation in the same portion of the band. It should be pointed out that the NTSC power level is measured as the peak synchronization signal power, while the DTV power is measured as an average quantity. The ATSC DTV signal peak-to-average power ratio is 7 dB for 99.99% of the time [6]. On the other hand, the peak-toaverage power ratio for the NTSC signal depends on the program content and is usually between 4 and 7 dB [19].

356

4.2 The Exact CO-Location Case and Relaxed Mask A typical broadcast transmitter high power amplifier (HPA) output spectrum (unfiltered), carrying an ATSC DTV signal, is shown in Figure 1 (resolution bandwidth: 500 kHz) [lo]. The Output Back-Off (OBO) of the amplifier is about 6-7 dB and no attempt has been made to linearize the amplifier. At the edge of the band, the spectrum is about 35 dB down relative to the in-band flat portion of the DTV spectrum. An NTSC visual signal RF subjective weighting curve, Figure 2 [10,11,12], is used in the interference calculation. Figure 2 shows that the NTSC visual signal is most sensitive to interference between 1.5 and 2.5 MHz from the low channel edge. In Figure 1, it is assumed that the NTSC peak sync power is 0 dB and the DTV signal average power is -12 dB, i.e., D/N = - 12 dB, where (D/N) dB = 10 log,, (DTV average power / Peak sync.power of adjacent NTSC channel). A 500 kHz measurement bandwidth is used. The DTV spectrum in-band flat portion should be 10 log,,(O.5/6) E -1 1 dB below the DTV power level, or -12 - 11 = -23 dB below the NTSC peak sync level. The DTV band edge spectrum is -23 - 35 = -58 dB. This value is used in the subsequent spectrum mask calculations. It is assumed that the NTSC signal TOV is 50 dB for AWGN uniformly distributed across the 6 MHz channel [7]. This ,value, when expressed as a noise density in twelve 500 kHz bands, is -50 + 10 log,,(0.5/6) -61 dB. Applying the NTSC signal weighting curve, Figure 2, to this flat -61 dB noise level and summing up the twelve bands, the weighted NTSC noise interference TOV is -55dB, or 5 dB below the unweighted case.

=

Similarly, using the upper and lower adjacent channel DTV spillover (unfiltered) curves in Figure 1 and the NTSC weighting curve in Figure 2, the calculated N+l and N-1 weighted interference into the adjacent NTSC channel is -56 dB and -58 dB respectively. It should be noted that, although the DTV spillover is symmetrical, the impact onto the two adjacent NTSC channels are different, due to the non-symmetrical nature of the NTSC weighting curve. The interference into upper adjacent, or N+1, channel is the limiting factor, because the NTSC visual carrier is very close to the interfering DTV signal. Since the weighted NTSC TOV is -55 dB, there is a 1 dB margin for the N+l channel and a 3 dB margin for the N-1 channel. These values have been c o n f i i e d by the laboratory tests [ 131.

It should be pointed out that the TOV for NTSC is very “soft”. Even if the DTV to NTSC interference temporarily exceeds the threshold, the viewer will only see a “just noticeable” disturbance on the NTSC screen. This is due to the generous assumption of C/N = 50 dB for NTSC TOV. It is equivalent to the ITU-R Grade 4.5 picture quality. Although the above calculations demonstrate that, with a proper OBO, the unfiltered DTV spectrum (Figure 1) meets the adjacent channel DTV to NTSC interference requirement, there should be a limit for the spurious emission in the second adjacent channel. ITU-R SG-1 recommends an attenuation of 60 dB for the spurious emission limit for the broadcast television service 15 MHz away from the channel edges. Since the DTV spectrum band edge has an attenuation of 35 dB, an additional 25-dB attenuation should be achieved at the edge of the second adjacent channel. Based on this assumption, an emission mask is defined [14]. With this definition, the above mentioned 60-dB requirement is met on the first adjacent channel. For a DTV spectrum mask, the out-of-band emission measured in a 500 kHz resolution bandwidth shall be attenuated below the average DTV transmitted power according to the following schedule: For 0.25 MHz < Af 5 6 MHz: Attenuation in dB 2 46 + [(AA’ / 1.441

(1)

In Equation 1, Af is the deviation in MHz from the center of the 500 kHz measurement bandwidth to the edge of the assigned 6 MHz DTV channel. The effects of the 500 kHz measurement bandwidth and the “smearing” effect of the measurement bandwidth at the channel edge are considered. The emission mask, as defined by Equation 1, is plotted in Figure 1. For Af 2 6 MHz: Attenuation in dB 2 71

(2)

It should be pointed out that the emission mask described above, or relaxed mask as it is called, is only for the scenario where DTV and NTSC adjacent channel transmission antennas are exactly co-located. If there is a separation between the transmission towers, a tighter emission mask is required, as described in the following section. The relaxed mask requires limited filtering at the far-side of the adjacent channel edges. It does not contribute much to the reduction of the total adjacent channel spillover power. Also appearing in Figure 1 is the NTSC Threshold of Audibility (TOA) for the upper and lower adjacent channel.

351

Assuming the NTSC audio carrier is 13 dB below the visual carrier and the required audio C/N = 35 dB (resolution bandwidth: 500 kHz), which provides an audio S I N = 69 dB [ll], the required TOA is -13-35 = -48 dB. From Figure 1, the N-1 TOA margin is 10 dB and the N+l margin is 33 dB, when the relaxed spectrum mask is used. When NTSC audio-to-video ratio is less than 13 dB, the margins will be higher.

4.3 Tight Mask for an 8-km Separation

When the adjacent channel DTV and NTSC signals are transmitted from separate towers, the desired signal and undesired signal are likely to behave differently due to separate propagation paths. The desired signal could experience heavier loss than the undesired signal, which might result in adjacent channel interference. For the Canadian DTV service [4,5], the F(50,90) curve is used for the desired signal, while F(50,lO) curve is used for the undesired signal. Table 3 lists the differences, in dB, between the F(50,lO) and F(50,90) curves at the end of the Grade B coverage for different classes of NTSC stations and frequency bands (see Table 4 for details) [4,5]. The CCIR 370 propagation model is used [SI. The assumed transmitter antenna separation distances are 8 km and 16 km (5 and 10 miles), respectively. It should be pointed out that the differences listed in Table 3 could be over-pessimistic when used for adjacent channels, since their signal strengths might exhibit high correlation. From Table 3, for an 8-km separation, an UHF Class C station requires a protection ratio of 12.67 dB and an UHF Low Power (LP) station requires 27.39 dB. The LP case might not be relevant, since its coverage radius is only 12 km and it is, therefore, unlikely to require a separation of 8 km. From the field strength differences listed in Table 3, the required attenuation factor a (dB/MHz), can be calculated. From Figure 1, the DTV band edge spectrum is -58 dB, while the NTSC weighting curve flat top is centered at 2 MHz from the edge, where the DTV unfiltered spectrum level is -61 dB. Thus, the a factor can be calculated as: [Field Strength Difference + (-58+61)] x 0.5 (dB/MHz) (3) The calculated a factors are listed in Table 5. For an 8-km separation, the low and high VHF ct factors are 6.89 dB/MHz and 6.46 dB/MHz. An a factor of 6.9 is recommended. For UHF Class A, B and C, CL = 7.8 meets the requirement. Considering the UHF adjacent channels

are likely to exhibit high correlation, the a factor is trimmed to 7.5. For the 16-km separation case, a = 7.8 dBMHz can be used for VHF and a = 9.8 dBMHz for UHF. Again, UHF Class A is unlikely to have a 10-mile (16-km) separation, since its coverage radius is only 25 km. From Table 6, a = 7.5 meets all the requirements for up to an 8-km separation. It is also quite close to the 16-km separation case for VHF. A “tight” emission mask is developed using a = 7.5 dB/MHz (see Figure 3 where the resolution bandwidth is 500 kHz). From Figure 2, the 0.5 MHz at both ends of the spectrum is the least sensitive part to the interference. The roll-off can, therefore, be slow in that range. The critical point is at 2 MHz from the upper band-edge, where a 12 dB filter attenuation must be achieved (this is 4 dB higher than the 8 dB filtering proposed in [I 11 for antenna pattern differences). After 3 MHz from the band edge, the mask can level off at a slower roll-off rate to reach -88 dB, or 4 5 dB from the in-band flat portion.

The emission mask can be expressed mathematically as in Equations (4) to (6). The out-of-band emission measured in a 500 kHz resolution bandwidth shall be attenuated below the average DTV transmitted power according to the following schedule: For 0.25 MHz I Af < 3 MHz: Attenuation in dB 2 46 + Af x 7.5

(4)

For 3 MHz I Af < 6 MHz: Attenuation in dB 2 61 + Af x 2.5

(5)

For Af 2 6 MHz: Attenuation in dB 2 76

(6)

In Equations (4) and (9,Af is the deviation in MHz from the edge of the assigned 6 MHz DTV channel. A filtered DTV spectrum is presented in Figure 3, using the Shorder Tchebychev band-pass filter shown in Figure 4 [ 10,111. A similar performance high power filter has been implemented in the USA DTV field trials [9,10]. The filter has about 15-dB attenuation at 2 MHz from the band edges, which is 3 dB better than the 12-dB requirement.

358

~~

Table 3: Field strength differences between F(50,90) and F(50,lO) curves with transmitter separations of 8 km and 16 km.

1

Class of Station

1

VHF

1

VHF

)I Class A 1 Class B 1

Class c

UHF LP

5 30

Table 4: Maximum ERP values and standard Effective Heights Above Average Terrain (EHAAT) for various classes of NTSC stations in the VHF and UHF band.

VHF Attenuation Factor a 8-km Separation (dB/MHz) 16-km Separation (dB/MHz) UHF Attenuation Factor a 8-km Separation (dB/MHz) 16-km Separation (dB/MHz)

Low VHF 6.89 7.66 Class C Class B 7.84 6.81 9.83 9.64

High VHF 6.46 7.83 Class A LP 7.78 14.74 -

Table 5: Required attenuation factor a.

VHF Attenuation Factor a 8-km Separation (dB/MHz) 16-km Separation (dB/MHz) UHF Attenuation Factor a 8-km Separation (dB/MHz) 16-km Separation (dBNHz)

Low VHF

High VHF 6.9

Class C

I

7.8 Class B I Class A

Table 6: Suggested attenuation factor a.

7.5

9.8

I

LP

359

The “tight mask” can also be used for co-located mixed class of stations. Since the “tight mask” provides an additional 12-dB attenuation on the upper adjacent channel (N+l) NTSC visual carrier, the DTV power level can be increased by 12 dB, or the NTSC power can be reduced by 12 dB. This will allow DTV and NTSC transmitting at the same power level, that is D/N = 0 dB. Based on the calculations made in Section 3, the limiting factor for the lower adjacent channel (N-1) interference into NTSC is the interference into the audio carrier. Since the mask does not provide much filtering on the (N-I) audio carrier location, the margin remains close to the 10 dB range. This means that the lower adjacent channel NTSC station can reduce power by up to 10 dB, or that the DTV power can be increased by 10 dB. In order to prevent DTV pilot interference into NTSC, an exact carrier frequency offset might have to be used, where the DTV pilot should be offset 5.082139 MHz above the NTSC visual carrier [12].

For 0.5 MHz 5 Af < 6 MHz: Attenuation in dB 2 -1 1.5 x (Af

+ 3.6)

(8)

For Af 2 6 MHz: Attenuation in dB 2 110

(9)

The FCC emission mask requires more attenuation at the far-end of the first adjacent channel, -99 dB relative to the DTV in-band spectrum flat portion. However, studies show that after 2-3 MHz from the authorized channel edge, the spillover power is negligible. Meanwhile, it is difficult to measure the spillover power in the -99 dB range. For low power DTV stations, there is no need to attenuate the out of band emission to the -99 dB level. It should be pointed out that the DTV threshold is very “sharp”. A few tenths of a dB below the threshold will result in picture freeze and loss of audio. Every effort should be made to maintain the DTV interference level below the threshold.

4.4 Adjacent DTV-DTV Interference Table 7 lists the calculated first adjacent channel DTV-DTV interference thresholds for the two spectrum masks. The relaxed mask spillover into the first adjacent channel is -39.8 dB, relative to the DTV in-band spectrum flat top. The tight mask spillover is -45.6 dB. Considering that the interference could come from two adjacent channels and there will also be noise distortion, the co-channel DTV-DTV C/I = 19.5 dB is used in the Canadian DTV allotment plan [4,5,18]. The calculated first adjacent channel DTV-DTV interference tolerance levels are, then, -20.3 and -26.1 dB for the relaxed and the tight emission mask, respectively [ 181. These levels are much higher than the 12.67 dB protection ratio required for 8-km (5-mile) transmitter separation (see Table 3). Therefore, larger DTV-DTV tower separation or mixed class operation is possible for the adjacent channel DTV-DTV case. If the adjacent DTV signals are transmitted at the same power level, the -39.8 dB spillover will only cause a DTV threshold (19.5 dB) loss of 0.04 dB. Table 7 also lists the (N+l) and (N-1) DTV-DTV interference threshold values indicated by the FCC [15, 161, -26 and -28 dB, where a much tighter emission mask is specified [ 161 (see Figure 5). The FCC mask can be formulated as: For 0 MHz < Af < 0.5 MHz: Attenuation in dB 2 47

(7)

5. Conclusions The Canadian digital terrestrial television system technical parameters for implementation and frequency planning have been presented. Due to different design guidelines and service quality assumptions, the Canadian terrestrial DTV system planning parameters and emission masks are different from the ones used in the USA. During the NTSC to DTV transition period, each NTSC station in Canada will be paired with a DTV channel. The service area of the paired DTV channel replicates the coverage of the existing Grade B contour of the NTSC station. During the transition period, a service availability for DTV of F(50, 90) is assumed. The interference criteria is F(50,lO). Considerations for the future DTV-only plan are incorporated by providing DTV allotments using separation distances based on DTV service availability of F(90, 90) and DTV-DTV interference determination at F(10, 10). The receiving parameters selection include equal partitioning between noise and interference for a C/N = CA = 19.5 dB and a co-channel NTSC-DTV interference of 7.2 dB. Two DTV emission masks have been proposed, as shown in Figure 5. A “relaxed mask” is used for the DTV/NTSC exact co-location case. The other, the “tight mask”, is designed for up to 8-km separation between the transmitters. The tight mask might also be used for mixed class operations.

360

Emission mask used (1) (2)

(31

(4)

DTV first adjacent channel spillover relative to the in-band spectrum flat top Assuming co-channel DTV-DTV threshold of 16.5 dB, calculated fnst adjacent channel DTV interference below DTV threshold: (1) + 16.5dB (ATTC measured level in bracket [ 151) Using C/I = C/N = 19.5dB, calculated first adjacent channel DTV interference below DTV threshold: (1) + 19.5dB Protection ratios used in DTV planning (where N is the desired channel)

Canadian Relaxed mask

Canadian Tight mask

FCC Mask [I61

-39.8 dB

-45.6 dB

-45.7 dB

-23.3 dB

-29.1 dB

-29.2 dB

(-23/-2 1dB)

WA)

-20.3 dB

-26.1 dB

-26.2 dB

NIA

-27 dB

-26dB for N+l -28dB for N-1

Table 7: First adjacent channel DTV-DTV interference calculation. No antenna polarization discrimination has been assumed in the development of the masks and the DTV signal average power is presumed to be 12 dB below the NTSC peak sync power. Linearizing HPAs might provide additional reduction of the adjacent channel spillover.

[SI “VHF and UHF Propagation Curves for the Frequency Range from 30 MHz to 1000 MHz”, ITU-R P.370-7, 1995.

References

[ 101 Carl G. Eilers, “The Development of a High Definition Television (HDTV) Terrestrial Broadcasting Emission Mask”, IEEE Transactions on Broadcasting, vol. 4 1 , no. 4, Dec. 1995.

[l] Canada Gazette, Part I, Notice No. SMBR-004-97 “Adoption of a Standard for Digital Television (DTV) Broadcasting in Canada”, Nov. 22, 1997.

[2] ATSC, “ATSC Digital Television Standard”, ATSC Doc. Al53, Sept 16, 1995. [3] FCC, “Fourth Report and Order”, FCC Document MM Docket No. 87-268, December 24, 1996. [4] “Planning Parameters and Allotment Plan for DTV Service in Canada”, ITU-R WP 11C Document llC/19, March 6, 1998. [5] “Digital Television Service Considerations and Allotment Principles”, JTCAB Ad Hoc Group on DTV Planning Parameters, Doc. AHGDTV003K, Aug. 1997. [6] “Digital HDTV Grand Alliance System Record of Test Results, October, 1995. Submitted to Advisory Committee on Advanced Television Service, to the Federal Communications Commission”. [7] Y. Wu, B. Caron, B. Ledoux and M. Guillet, “Evaluation of COFDM for ATV transmission over 6 MHz channels”, Proceedings of the I996 International Broadcasting Convention, Amsterdam, Sept. 1996.

[9] W. Y. Zou, Y. Wu and M. Guillet, “Analysis of ATV Transmission Subsystem Field Test Data”, IEEE Transactions on Broadcasting, vol. 42, No. 1, March 1996.

[ 1I] Carl Eilers and Gary Sgrignoli, “Analyzing the FCC’s DTV Spectral Emission Mask and Potential Degradation to Adjacent Channels Due to Antenna Pattern Differences”, IEEE Transactions on Broadcasting, vol. 44, No. 1, March 1998. [ 121 ATSC A/64, “Transmission Measurement and Compliance Standard for Digital Television”, ATSC, 17 November 1997. [ 131 Stanley J. Salamon, “Adjacent Channel Interference Revisited”, ATTC Report, 1996.

[14] FCC, “Sixth Report and Order,” MM Docket No. 87268, adopted April 3, 1997, FCC 97-1 15 (released April 10, 1997), pp. 90. [ 151 “An Evaluation of the FCC F W Mask for the Protection of DTV Signal from Adjacent Channel DTV Interference”, ATTC Document #97-06, July 17, 1997.

36 I

[19] J. J. Giardina et al., “True APL Picture Power of a TV Transmitter,” Proceedings of the 1987 SBE & Broadcast Engineering Conference, Indianapolis, IN.

E161 FCC, “Memorandum Opinion and Order on Reconsideration of the Sixth Report and Order”, MM Docket No. 87-268, adopted February 17, 1998, FCC 98-24 (released February 23, 1998), pp. 38. [17] Y . Wu, B. Ledoux and B. Caron, “Evaluation of Channel Coding, Modulation and Interference of Digital ATV Terrestrial Transmission Systems”, IEEE Transactions on Broadcasting, vol. 40, No. 2, June 1994. [ 181 “Addendum Digital Television Service Considerations and Allotment Principles”, JTCAB Ad Hoc Group on DTV planning Parameters, Doc. AHG_ADD003K, Issue 1, May 26, 1998.

0

e

e

NTSC peak power 0 dB

e DTV average power -12 dB T I 1 dB

N+lTOA 1

N-1 TOA

-

RES BW 500 kHz

-----

0

Relaxed mask for exact co-location

1

1

1

1

1

1

1

1

1

1

1

1

I

1

1

I

I

1

2

3

4

5

6

1

2

3

4

5

6

1

2

3

4

5

6

Frequency (MHz) Figure 1: Unfiltered DTV spectrum and “relaxed mask” in the case of exact co-location.

0

-8

dB -16

-24 -32

3

2

I

0

5

(Mdz)

Frequency

6

Figure 2: NTSC visual signal RF subjective weighting curve.

0

e NTSC peak power 0 dB

10 B,

DTV average power -12 dB

20

3 .

30

IdB

40

dB 50

60

,*

70

-

0

0

1

RES BW 500 kHz

80

- - - - - Unfiltered spectrum Filtered spectrum Mask based on (50,90)/(50,10)\ (5-mile separation)

90

I

0

1

2

I

I

I

I

I

I

I

I

I

I

I

I

I

3

4

5

6

1

2

3

4

5

6

1

2

3

I

4

5

6

Frequency (MHz) Figure 3: “Tight mask” for an S-km separation and filtered DTV spectrum.

363

Frequency

M M

Figure 4: Transmitter filter response: fifth order (N=5) Tchebychev.

0

DTV average power 0 dB

4I lldB

@

+

10 20

30 35 dB 40

50

dB

BW 500 kHz 60 \

70

\

80

FCC mask Canadian relaxed mask Canadian tight mask

90

100

110

0

1

2

3

4

5

6

1

2

3

4

5

6

1

Frequency (MHz) Figure 5: Proposed emission masks.

2

3

4

5

6

364

Dr. Yiyan Wu is a senior research scientist with the Communications Research Centre Canada, in Ottawa. His research interests include digital video compression and transmission, high definition television (HDTV), signal and image processing, satellite and mobile communications. He is actively involved in the ATSC technical and standard activities and ITU-R digital television and data broadcasting studies. He is an adjunct professor at Carleton University, Ottawa, Canada and a member of the IEEE Broadcast Technology Society Administrative Committee.

Mr. Bernard Caron received a B. Sc. in Electrical Engineering from Lava1 University, Quebec in 1978 and a M. Sc. from University of Ottawa in 1984. Since 1979, he has been with the Communications Research Centre Canada and has worked on Teletext, mobile data transmission, video channel characterization and simulation. He is now the manager of the Television Systems and Transmission program in the Broadcast Technologies branch. He is currently involved in the introduction of digital television terrestrial broadcasting in Canada. Donald Tyrie, Royce Trenholm: Biography and photograph not available at the time of publication.

Dr. Pierre Bouchard is a research engineer in the Television Systems and Transmission group at the Communications Research Centre Canada, in Ottawa. He is currently involved in research on DTV transmission, DTV coverage, and on broadband broadcasting technologies (MMDS and LMCS/LMDS).