GB2275148A

GB2275148A - Transmitting enhanced digital video signal on compatible waveform

Info

Publication number: GB2275148A
Application number: GB9402528A
Authority: GB
Inventors: Christopher Keith Perry Clarke
Original assignee: British Broadcasting Corp
Current assignee: British Broadcasting Corp
Priority date: 1993-02-09
Filing date: 1994-02-09
Publication date: 1994-08-17
Also published as: GB9402528D0; WO1994018795A1; GB9302528D0

Abstract

An enhanced digital television signal is transmitted within a conventional television waveform by modulating 22 the digital signals onto active picture area lines of the conventional waveform (such as IAL SECAM, NSTC). This may be achieved by first modulating a set of orthogonal waveforms, derived e.g. by the Walsh TransForm, with the digital signals and then transmitting the modulating orthogonal waveforms on the active picture area lines. A sync pulse generator 24 adds conventional Field/line sync pulses and optional colour burst to the transmitted signal. These may be used at the receiver, but are not necessary. <IMAGE>

Description

TELEVISION SYSTEMS Field of the Invention This invention relates to television systems and in particular to digital television systems which can be transmitted using the same transmission format as existing transmissions such as PAL NSTC or SECAM.

A degree of compatibility with existing television standards is widely recognised as a great advantage in simplifying the introduction of digital and enhanced television systems. Initially, many proposals for enhanced television systems were made to be signal compatible, so that the new television signals could produce pictures on existing sets, although generally without the enhanced features of the new system. The enhancements would only have been available by buying a new set capable of decoding the full content of the enhanced signal. Examples of such systems are the Japanese EDTV1 and EDTV2 systems and the European Enhanced PAL and HD-MAC systems. However, the constraint of signal compatibility has tended to limit the performance that such systems can offer, both in the degree of enhancement of the new signal and in the acceptability of the picture displayed on existing sets.

More recently, several proposals have been made in the United States of America for the development of Advanced Television (ATV). These introduced the concept of enhanced television systems in which the broadcast signals are spectrum compatible, that is, they can fit into similar frequency allocations to the existing standards and, to an extent, co-exist with them. The most attractive of these systems use digital encoding of the television picture and sound information, the digital bitstream then being modulated on to one or two carriers to occupy a normal television channel.

The power levels needed for such broadcasts and the immunity to interference of the digital signals allows the possibility of such signals being broadcast at the same time as conventional analogue transmissions, by re-using the so-called "taboo" channels.

More advanced spectrum-compatible systems based on multiple-carrier Orthogonal Frequency Division Multiplexing have since been proposed, initially in Europe. All the spectrum-compatible systems give greater freedom to introduce enhancements than the signal-compatible systems, but require completely new transmitter systems and new receiving equipment in addition to that for the existing broadcasts.

Summarv of the Invention The system proposed here and embodying the invention is based on a different definition of compatibility, that is waveform compatibility. According to this approach, the waveforms produced by the encoder would be of the same video format as existing television standards, that is, containing line and field pulses and, perhaps, standard colour synchronising bursts and blanking intervals.

However the television signal information (picture and sound) would be encoded digitally, to occupy the activepicture area of the waveform. It should be noted that there is no requirement for a relationship, in terms of number of lines per picture, number of fields per second, etc., between the signal format of the, digitally-encoded signal and that of the analogue synchronising waveforms that are used to carry it. The source signal may be a conventional television signal or one of the so called enhanced signals for HDTV or the like.

The amplitudes and frequency ranges of the wave form- compatible signals would be identical to those of existing PAL, NTSC or SECAM signals. Thus all existing broadcast equipment that needs to take no account of the signal information would have at least some degree of compatibility. Existing receivers would produce a locked display, although the picture information would be completely meaningless without a decoder.

Such a system would have the following advantages: 1. Fully digital picture coding, resulting in improved quality and greater freedom in choosing the scanning standard of the enhanced picture.

2. Existing television distribution links and transmitters, including satellites, could carry the signals without modification.

3. Existing television receivers could receive and display pictures by the addition of a SCART-connected decoder, if we assume reasonable compatibility in the display standards.

4. It is possible that existing domestic VCRs may be able to record and replay the signals, albeit at reduced quality.

5. The system could be introduced rapidly. The most immediate application of a waveform-compatible system would be as a research tool, to gain experience with broadcasting digital signals to a wide range of reception sites. In the longer term, such a system might form the basis of a means of introducing digital broadcasting of enhanced television signals which, when established, could compatibly omit the analogue synchronising waveforms and thus gain improvements of coding efficiency. This could result in an evolutionary approach to full digital broadcasting.

Normal television waveforms already contain digital signals, although not to carry the television signal information itself. For example, teletext signals can be operated in full-field mode, whereby all active lines of the picture are used in addition to the normal lines of the field blanking interval. However, such a system only produces a data capacity of about SMbit/s, currently inadequate for a good quality digital picture signal.

It is also vulnerable to short-duration echoes, such as might be caused by multipath propagation or by reflections in cables. D2-MAC data would provide a slightly higher capacity in full-field mode, but again would be- vulnerable to multipath. Similarly, a wider bandwidth variant of the QPSK NICAM, digital sound signal could be accommodated in the active-picture area, but would suffer the same problems.

It is recognised that, particularly for terrestrial broadcasting, the uhf spectrum is a very hostile environment, with multipath distortion, poor linearity in transmitters, interference, both from other television broadcasts and from electrical equipment, and, to a lesser extent, noise.

Ideally, a system for terrestrial broadcasting should allow the possibility of gradual failure, rather than the abrupt go/no-go failure characteristics of many digital systems.

Such a system would then be in a better position to accommodate distorted signals and might be able to support the operation of portable sets at reduced resolution, while providing enhanced pictures on larger sets where a strong signal can be received, such as from a roof aerial.

The invention is defined in its various aspects in the appended claims to which reference should now be made.

Brief Descrintion of the Drawings A preferred embodiment of the invention will now be described in detail by way of example with reference to the accompanying drawings in which: Figure 1 is a block diagram of an encoder, transmission channel, and decoder embodying the invention; and Figure 2 shows the encoder and decoder of figure 1 in more detail.

Detailed DescriPtion of Preferred Embodiments It will be appreciated that the technique of wave form- compatible coding could be implemented in many different forms using various digital encoding systems to carry the television information. One approach is to encode the digital information by amplitude modulation of a set of orthogonal functions, a process known as code-division multiplexing. The following description is based particularly on the use of Walsh functions.

Code-division multiplexing has several advantages: first, the duration of the individual symbols is long, which gives good resistance to distortion and to impulsive interference and, secondly, the nature of the carriers will mean that some, being made up of predominantly low-frequency information, are more rugged than others. Thus the system includes the basic requirement for a gradual failure system, that is, it provides a multi-channel transmission capability with different degrees of ruggedness. The most vital information of the digital signal can then be conveyed through the most rugged channel, while less important information is encoded less ruggedly and will therefore fail first if the signals are subject to distortion, interference or noise.

Another advantage of using the Walsh functions is that the modulation process can be accomplished by a fast transform algorithm, akin to the Fast Fourier Transform, but significantly simpler to implement.

The basis of the system is therefore as shown in Figure 1, consisting of an encoder (20) receiving and encoding for digital data (10), a transmission channel (30) for the encoded data and a decoder (40) which retrieves the digital data (50) after transmission. The digital data (10) represents the picture and sound content of a television signal in a digitally-encoded form. This can optionally consist of several channels some of which contain, predominantly, data of great importance and others less so, possibly multiplexed for convenience, but with identifiable time-slots containing the important information. The demodulated data (50) can be received into the same format of channels of varying importance as was used in the encoder.

The data stream is used to modulate the amplitudes of orthogonal functions by the modulator (22). The modulated data is then added by the adder (26) to a standard television composite video waveform, produced by a generator (24) and optionally containing a colour burst.

The amplitudes are chosen so that the modulated signal (28) resembles a standard colour television signal. The principal purpose of the syncs and colour burst waveforms is to ensure compatibility with existing equipment, although they may additionally play a part in synchronising the decoder (40) and the colour burst may be used by an enhanced receiver.

At a receiver the decoder receives the video format signals (42) from the channel, perhaps after conventional radio frequency and intermediate frequency carrier modulation and demodulation used to locate the transmitted signals at a particular point in the rf spectrum. The orthogonal functions contained in the active-picture period of the video waveform are demodulated by the demodulator (44) to retrieve the digital data (50), which can then be demultiplexed as necessary. The sync separator (46) can optionally play a part in the synchronisation of the demodulator, but this is not essential, the modulated waveform containing sufficient information for self-synchronisation.

The processes of modulation and demodulation of the orthogonal functions by the input data will now be described in more detail, by way of a simplified example based on a set of four Walsh functions, as shown in Figure 2. The four functions (100-103) are modulated by individual parts of the incoming data stream (104-107) by individual modulators during the active video period of a standard television signal. Each of the resulting waveforms (110-113) represents a weighted version of the corresponding Walsh function with its polarity determined by whether the modulating data bit was a '0' or a w Walsh functions of value 1 and 0, rather than +1 and -1, can also be used with similar results. The relative ruggednesses of the individual channels can be adjusted conveniently by using data waveforms (104-107) of different amplitudes.The waveforms 110-113 are then added together by an adder (120) to produce the modulated signal (130), to which a composite video sync waveform (109) is added by adder 26. This produces the output signal (28).

At the demodulator, the received signal (42) is demodulated by locally-generated versions (140-143) of each of the four Walsh functions in parallel by demodulators (150-153) locked in phase to the incoming waveform either by analysing the results of the demodulation process or by using the incoming video sync pulses. As the functions are orthogonal, the waveforms (160-163) produced by demodulation are not affected by the presence of the other Walsh functions and resemble the individual modulating waveform (104-107). The individual received bit values ('0 or '1') can then be determined by slicing waveforms 160-163 in the conventional manner or by alternative techniques.

In practice, it would be advantageous to use a much larger set of Walsh functions, perhaps 512 or more, so that the individual symbols occupy a large part of a line, a whole line, or more than one line. This would have the dual advantages of increasing the length of the individual modulated symbols to reduce the effects of interference and distortion, and increasing the number of parallel channels so that a greater variety of ruggedness levels could be accommodated. With such an arrangement, it may be convenient to position the orthogonal functions to be locked to the line structure of the television waveform and many suitable arrangements could be devised for this.

Furthermore this would enable a plurality of television signals to be transmitted over a single conventional channel.

As so far described, the modulator has the capacity for a number of input bits equal to the number of Walsh function symbols that can be fitted on to each line of the television waveform, consistent with the bandwidth capacity of the video signal. Clearly, if a Walsh function consists of a large number of transitions in a short period of time, such that the spectral content is significantly beyond the nominal signal bandwidth (5.5MHz for System I), this would exceed the capacity of the video signal. If 512 functions were used, that would give a data capacity of 512 bits/line or about 7.4Mbit/s.Greater capacity could be provided by using multilevel modulation of the orthogonal functions, so using four-level data waveforms at 104-107 (2-bit data) would give 14.8Mbit/s, eight-level data would give 22.4Mbit/s, and so on, although at each stage the noise and distortion immunity of the signal would be degraded. Such proposals could be contemplated in this case, whereas they would not be possible with, for example, teletext data, because the orthogonal symbols can be expected to have greater immunity to distortion, thus causing the signals to be limited by noise rather than distortion.

Another feature of the modulation process that needs to be taken into account is that the peak signal amplitude will tend to grow as the number of orthogonal functions is increased. The amplitude can be normalised by dividing the modulated signal by vN where N is the number of functions.

This results in the modulated signals having the same noise immunity as the modulating data (104-107). However, this does not completely solve the problem because the modulated waveform (130) is noise-like and thus has a large peak-tomean ratio. Peaks of the waveform would therefore still extend beyond the normal video range. The signal could be clipped without much distortion, simply tending to degrade the eye-height of all the functions together. However, it may be preferable to non-linearly compress the waveform 130 before transmission and to use a corresponding expansion of the received signal 42 before demodulation. This would tend to reduce the distortion of the signal peaks. The result would however, slightly distort the orthogonality of the signal.

While the system of individual modulators shown in Figure 2 is perfectly feasible, the amount of hardware required becomes unmanageable when long symbols are used. Under these circumstances it is an advantage to use a set of orthogonal functions for which a fast transform algorithm exists. This is true for the Walsh functions and for several other orthogonal sets. The digital bit values 104-107 are then presented serially, although not necessarily in that order, to the transform'processor, which then produces the output waveform (130) directly.

This can still support multi-level waveforms at the input and individual weighting of the functions for ruggedness profiling can be applied by a single multiplier placed before the fast transform processor. In the same way, a fast transform processor can be used in place of the individual demodulators 150-153 of the demodulator unit 44.

Other orthogonal modulation schemes such as Quadrature Amplitude Modulation (QAM) may be used. In this the bits are modulated on carriers with different phases. A Fast Fourier Transform (FFT) of the carriers then yields a 'symbol' for transmission. At a receiver this is demodulated using a further FFT to separate the carriers.

The individual bits may then be derived.

Any audio signal required may be included in a separate channel such as that used for NICAM audio transmission.

Claims

1. A transmission system for digital television signals comprising a source of enhanced digital television signals, a source of conventional television transmission waveforms, and means for modulating the digital television signals onto the active picture area lines of the conventional television waveforms

2. A transmission system for digital television signals according to claim 1 in which the digital television signals are modulated onto substantially all the active picture area lines of the conventional television waveform.

3. A transmission system for digital television signals according to claim 1 or 2 wherein the conventional television waveform contains a colour reference burst in a blanking interval.

4. A transmission system for digital television signals according to claim 1, 2, or 3 wherein an audio signal associated with the digital television signal is included in a separate transmission channel.

5. A transmission system for digital television signals according to claim 1, 2, 3 or 4 wherein the bandwidth and amplitude of the digital signals modulated onto the conventional waveforms are maintained within the bandwidth and amplitude levels of the conventional waveforms.

6. A transmission system for digital television signals according to claim 4 including non-linear compression means to restrict the amplitude of the digital signals.

7. A transmission system for digital television signals according to any of claims 1 to 6 in which the digital television signals are modulated onto lines in the active picture area of the conventional television waveforms.

8. A transmission system for digital television signals according to any of claims 1 to 7 in which the digital television signals are modulated onto lines in the field blanking intervals of the conventional television waveforms.

9. A transmission system for digital television signals according to any of claims 1 to 8 in which the modulating means modulates the most important data in the digital television signals with the greatest immunity to interference and distortion.

10. A transmission system for digital television signals according to any preceding claim in which the modulating means comprises a code-division multiplexing means.

11. A transmission system for digital televsion signals according to claim 10 in which the code-division multiplexing means comprises means for amplitude modulating each data bit of the digital television signals by one of a plurality of orthogonal functions.

12. A transmission system for digital television signals according to claim 11 in which the orthogonal functions comprise Walsh functions.

13. A transmission system for digital television signals according to claim 11 or 12 in which each orthogonal function has a duration of less than one line.

14. A transmission system for digital television signals according to claim 11 or 12 in which each orthogonal function has a duration of one line.

15. A transmission system for digital television signals according to claim 11 or 12 in which each orthogonal function has a duration greater than one line.

16. A transmission system for digital television signals according to any of claims 11 to 15 in which the relative amplitudes of the orthogonal functions are different.

17. A transmission system for digital television signals according to any of claims 11 to 16 in which at least some of the orthogonal functions can take more than two amplitude levels.

18. A transmission system for digital television signals according to any of claims 11 to 17 comprising non-linear compression means acting on the modulated enhanced television signals.

19. An digital television system receiver comprising means for receiving a conventional television waveform and means for demodulating digital television signals included in lines of the active picture area of the conventional television waveform.

20. A digital television system receiver according to claim 19 in which the demodulating means comprises means for demodulating a plurality of orthogonal functions modulated with data bits of the digital television signals.

21. A digital television system receiver according to claim 19 or 20 in which the demodulating means is locked in phase to the incoming waveform by synchronising signals provided in the incoming waveform.

22. A digital television system receiver according to any of claims 19 to 21 including means for non-linearly expanding the received waveform.

23. A transmission television system receiver according to claims 19 to 22 in which the digital signals are included in substantially all of the lines of the active picture area of the conventional waveform.

24. A transmission television system receiver according to claims 19 to 23 wherein the receiver is responsive to a colour reference burst included in a blanking interval of the conventional waveform.