KR19980042346A - Digital tv signal receiver - Google Patents

Abstract

The first data slicer provides intermediate symbol decoding results responsive to multi-level symbols at the digital television receiver. Co-channel interference involving the multi-level symbols is suppressed by the first comb filter, so that the energy of the co-channel interference is reduced before the data slicing process is performed in the second data slicer. The first comb filter additionally re-encodes the symbols supplied to the second data slicer. The second comb filter is used as a postcoder to recode the data stream disposed behind the second data slicer and reconstructed by the second data slicer to produce a complementary symbol decoding result. The co-channel interference detection circuit determines the amount of co-interference energy of the intermediate symbol decoding result and the complementary symbol decoding result. The intermediate symbol decoding result indicates that the symbol stream reproduced at the DTV receiver has a sufficient energy level that can be suppressed in the first comb filter response and will not cause an incorrect error in the data of the first data slicer. Only when the co-channel interference detection circuit does not generate a current indication signal accompanying an artifact of the signal can be used as the final symbol decoding result.

Description

Digital tv signal receiver

The present invention relates to a digital television (DTV) system, such as a digital high definition television (HDTV) system used for US terrestrial broadcasting in accordance with the Advanced Television Sub-Committee (ATSC) standard. The present invention relates to a DTV receiver having an adaptive filter circuit for suppressing co-channel interference from an analog TV signal conforming to the standard of the Television System Committe.

The DTV standard, published by ATSC on September 16, 1995, includes residual sidebands, or VSBs (Vestigial), for transmitting DTV signals to 6 MHz bandwidth TV channels that are recently used for NTSC analog TV signal radio broadcasting in the United States. Sideband) signal is specified. The VSB DTV signal is designed such that its spectrum interleaves with the spectrum of the co-channel interfering NTSC analog TV signal. This means that NTSC analogs fall between 1/4 even multiples of the horizontal scan line rate of the NTSC analog TV signal, where the even multiples of the luminance and chrominance component energy of the co-channel interfering NTSC analog TV signal will belong. This is accomplished by locating the principal AM (Amplitude Modulation) sideband frequency and pilot carrier of the DTV signal in multiples of a quarter of the horizontal scan line rate of the TV signal. The video carrier of the NTSC analog TV signal is offset 1.25 MHz from the lower limit frequency of the TV channel. The carrier of the DTV signal is offset from the video carrier by 59.75 times the horizontal scanning line rate of the NTSC analog TV signal, positioning the carrier of the DTV signal at approximately 309,877.6 kHz from the low frequency limit of the television channel. Therefore, the carrier of the DTV signal is located about 2,690,122.4 kHz from the middle frequency of the TV channel.

The exact symbol rate of the DTV standard corresponds to 684/286 times the 4.5 MHz voice carrier offset from the video carrier of the NTSC analog TV signal. The number of symbols per horizontal scan line of an NTSC analog TV signal is 684, and 286 is a factor that allows the horizontal scan line rate of an NTSC analog TV signal to be multiplied to obtain 4.5 MHz offset from the video carrier of the NTSC analog TV signal. factor). The symbol rate is 10.762238 megasymbols / second that may be included in the VSB signal extended from 5.381119 MHz from the DTV signal carrier. That is, the VSB signal may be limited to a band extended from the low frequency limit frequency of the TV channel by 5.690997 MHz.

The ATSC standard for terrestrial broadcast of digital HDTV signals in the United States can transmit either of two HDTV formats with a 16: 9 aspect ratio. One HDTV display format uses 1920 effective horizontal scan lines per 30Hz frame with 1920 samples per scan line and a 2: 1 feedback interlace. The other HDTV display formats use 1280 luminance samples per scan line and 720 progressively scanned lines of TV video per 60 Hz frame. The ATSC standard also accommodates the transmission of DTV display formats other than the HDTV display format, ie the parallel transmission of four TV signals with normal resolution when compared to NTSC analog TV signals.

DTV transmitted by VSB AM during terrestrial broadcast in the United States includes a series of continuous data feeds each consisting of 313 consecutive-in-time data segments. The data feed may be regarded as successively numbered modulo-2, in which the odd-numbered data feed and the subsequent even-numbered data feed each form a data frame. The frame rate is 20.66 frames / second. Each data segment has a duration of 77.3 ms. The result is a symbol rate of 10.76 MHz, with 832 symbols per data segment. Each segment of data begins with a line synchronization code group consisting of four symbols with consecutive values of + S, -S, -S, + S. The + S value is one level below the maximum positive data excursion, and the -S value is one level above the maximum negative data excursion. The initial line of each data feed includes a group of code synchronization codes that encode a training signal for channel-equalization and multipath suppression. The training signal is a 511-sample PN sequence followed by three 63-sample pseudo-random (PN) sequences. The middle sequence of the 63-sample PN sequences of the feedback synchronization code is transmitted according to a first logic convention on the first line of each odd-numbered data feed, and on the first line of each even-numbered data feed. Transmitted according to the second logic protocol. The first and second logic protocols are complementary to each other.

Data in the data line is trellis coded using twelve interleaved trellis codes, each of the 2/3 rate trellis codes having one bit uncoded. The interleaved trellis code undergoes a Reed-Solomon forward error-correction encoding to prepare for the correction of a burst error caused by a noise source such as an automotive ignition system exposed in the immediate vicinity. The Res-Solomon coding result is transmitted as 8-level (3 bits / symbol) 1-dimensional constellation symbol coding for wireless transmission without symbol precoding distinct from the trellis coding process. . The Reed-Solomon coding result is transmitted as 16-level (4 bits / symbol) 1-dimensional constellation symbol coding for wired transmission without precoding. The VSB signal has a natural carrier wave whose amplitude may vary depending on the suppressed modulation ratio.

The inherent carrier is replaced with a pilot carrier of fixed amplitude corresponding to the specified modulation rate. These pilot carriers of fixed amplitude are generated by introducing a direct current component shift into the modulation voltage applied to the balanced modulator that generates the AM sideband supplied to the filter supplying the VSB signal as its response. If eight levels of 4-bit symbol coding have normalized values of -7, -5, -3, -1, +1, +3, +5, and +7 in the carrier modulated signal, the pilot The carrier has a normalization value of 1.25. The normalization value of + S is +5 and the normalization value of -S is -5.

In the early development of DTV technology, it is expected that a DTV broadcaster would have been asked to determine whether or not to use a symbol precoder for the transmitter, which provides a precoded symbol filtering subsequent to the symbol generator circuit. This decision at the broadcast device depended on whether interference at the co-channel NTSC station was expected. The symbol precoder is a symbol introduced incidentally to each DTV receiver by a comb filter used in front of the data slicer of the symbol decoder circuit to remove artifacts of the NTSC co-channel interference signal. Postcoding was expected. Symbol precoding would not have been used for data lines or data line synchronization code groups that would allow data feed synchronization data to be transmitted.

Co-channel interference is reduced at greater distances from NTSC stations, is more likely to occur when some ionosphere conditions are established, and is likely to occur in the summer when solar activity is active. Such interference will not occur at no co-channel NTSC stations. If there is a possibility of NTSC interference in the area of broadcast coverage, it is assumed that a symbol precoder is used in the HDTV broadcasting device to easily separate the HDTV signal from the NTSC interference. The comb filter is thus used as a symbol postcoder of the DTV receiver to perform matched filtering. If there is no chance of NTSC interference, or if there is a possibility of substantial interference, the DTV broadcaster may use a symbol predecoder to ensure that flat spectral noise does not cause a false decision about the symbol value of the trellis decoder. It is assumed that the operation can be stopped. Without recognizing the broadcaster's conditions, substantial co-channel NTSC interference can be achieved by varying skipping conditions, cabelcast leakage, improper IF (intermediate frequency) image suppression at NTSC receivers, and DTV including previous residual TV recordings. Due to the magnetic tape used for the recording, or some other exceptional condition, it may be substantially present in part of the broadcast receiving area. According to the current ATSC DTV standard, symbol precoding is not used at the transmitter. The suppression of the co-channel interfering analog TV signal is estimated to be performed in the trellis decoding process after the data slicing process related to symbol decoding. This avoids the problem of determining whether or not precoding is performed at the transmitter. However, co-channel interfering analog TV signals cause errors in the data-slicing process, making it difficult to perform error-correction decoding, trellis decoding, and Reed-Solomon decoding. These errors can reduce the broadcast coverage area, resulting in a loss of revenue for commercial DTV broadcasters. Thus, although the symbol precoding process is not currently permitted in DTV receivers by the ATSC DTV standard, it is desirable to prepare for suppression of co-channel interfering analog TV signals prior to data-slicing.

Meanwhile, the term linear combination generally means addition and subtraction performed according to a conventional operation or a modular operation. The term modular combination refers to a linear combination performed according to a modular operation. The encoding type for re-coding a digital symbol stream through a linear combination of differential delay and differential delay illustrated by symbol postcoding used in prior art HDTV receivers is described herein. Defined as symbol recoding. The encoding type for re-coding the digital symbol stream through the modular combination as a result of the delay of the modular combination illustrated by symbol precoding used in the HDTV transmitter of the prior art is defined herein as symbol recoding of the second type.

The problem of co-channel interference caused in the analog TV signal can be seen in terms of jamming problems in the receiver so that it can be solved by the adaptive filter circuit of the receiver. As long as the dynamic range of the system channel is not exceeded so that co-channel interference can capture the system channel by disabling the signal modulation capability for DTV modulation, the performance of the system is a matter of signal overlap. Can be considered. The filter circuit of the receiver is adapted to the analog TV signal according to the distinct correlation and anti-correlation characteristics of the analog TV signal in order to reduce its energy enough to capture system channels from the analog TV signal. It is adapted to select a digital signal from co-channel interference caused by the signal.

As far as the co-channel interference caused by the analog TV signal is concerned, co-channel interference is incorporated into the system channel after the DTV transmitter and before the DTV receiver. Co-channel interference by analog TV signals is unaffected, with or without symbol precoding in the DTV transmitter. In DTV receivers, to reduce errors that occur during data-slicing, unless co-channel interference is large enough to overload the receiver front-end and capture the system channel, It is advantageous to place the comb filter first to reduce the energy of the high energy spectral component before the data-slicing circuit. The DTV broadcaster is a carrier whose nominal frequency is 310 kHz above the low limit frequency of the TV channel so that the frequency can be optimally offset from the video carrier of the co-channel NTSC analog TV signal which may be interfering with the carrier frequency. You need to adjust the frequency. The optimal offset of the carrier frequency is exactly 59.75 times the horizontal scan line frequency f _H of the NTSC analog TV signal. The artifact of co-channel interference of the demodulated DTV signal is 59.75 times the horizontal scanning line frequency f _H of the NTSC analog TV signal generated by the heterodyne between the digital HDTV carrier and the video carrier of the co-channel interfering analog TV signal. 287.25 times the horizontal scan line frequency f _H of the NTSC analog TV signal generated by the heterodyne between the digital HDTV carrier and the chroma subcarrier of the co-channel interfering analog TV signal It will include a bite having a frequency very close to the fifth harmonic of the bead 59.75 times the line frequency f _H. Further artifacts digital HDTV carrier with the co-channel interfering analog horizontal scanning line frequency of about 345.75 times the non-root of f _H, that is the horizontal scanning line frequency of the NTSC analog TV signal that is generated by the heterodyne between the sound carrier of the TV signal f _H Will include a bead having a frequency very close to the sixth harmonic of the bead of 59.75 times. Due to the near harmonic relationship of these beads, the beads can be suppressed by a suitably designed single comb filter that includes very few symbol epochs of differential delay. By using an NTSC-rejection comb filter before data-slicing in the DTV receiver, a first type of symbol coding can be performed to change the symbols obtained by data-slicing.

The data-slicing operation following the first type of symbol recoding in the DTV receiver is designed from the first type of symbol recoding since the data quantizing level is designed to match the symbol level. This is a quantization process that makes useful symbols. However, the quantization process differentiates the residual signal of the co-channel interfering analog TV signal that remains after the filtering associated with the first type of symbol recoding and is somewhat smaller than the steps between symbol code levels. This is a kind of capture phenomenon where a stronger signal is obtained at the expense of a relatively weaker signal in the quantization process.

As far as data transmission is concerned, digital data symbol streams flow the system channels through their entire length. When a second type of symbol recoding is performed as symbol precoding in a DTV transmitter, the addition combination of differentially delayed data symbol streams does not increase the transmitter power or increase the average intersymbol distance to help overcome the jamming analog TV signal. Is made based on modularity Instead, the principal mechanism for overcoming jamming analog TV signals is provided by the comb filter in the DTV receiver and the residual analog TV signal in the comb filter response to be suppressed by the quantization effect of the data slicer immediately following the comb filter. Attenuation with the DTV signal, such as

The order of performing the symbol recoding process of the first and second types has little effect on the signal transmission through the system channel under the above circumstances, since any encoding method makes the signal transmission capability for the symbol stream useless. Do not. The order of performing the symbol re-encoding process of the first and second types may be performed in the same channel unless the symbol re-encoding of the second type is not intervened between the symbol re-encoding of the first type and the subsequent data-slicing process. It has no effect on the digital receiver's ability to suppress interfering analog TV signals. This technical knowledge provides a general basis upon which the present invention is based.

The first and second type of symbol re-encoding process performed by a comb filter to suppress NTSC co-channel interference and a comb filter with complementary partial response is performed by the inventors to determine whether the data is a synchronization code or information. Regardless of whether it is performed continuously. This is because the comb filtering technique for suppressing NTSC artifacts is mainly dependent on the periodic correlation and anti-correlation properties of NTSC co-channel interference, which is a continuous signal. Since the NTSC co-channel interference is a continuous signal with continuous analog characteristics, the comb filtering technique for suppressing NTSC artifacts should be continuous in that application, even though the results of these techniques can only be used selectively. do. Although the trellis coding steps have been carried out to raise the DTV signal energy of a portion of the spectrum that is relatively free of NTSC signal energy, if the data sampling is properly synchronized with the symbol rate, the data is combed rather than destroyed in some critical way. Will be re-encoded by filtering. Correlation and anti-correlation of data symbols do not fundamentally affect the process for suppressing the effects of NTSC artifacts in data slicing. The comb filter, which provides independence among interleaved trellis codes during the trellis decoding process after symbol decoding, is independent of the comb filtering performed during symbol decoding to prevent NTSC artifacts that cause errors in data slicing. Can be considered.

Accordingly, an object of the present invention is to provide a DTV signal detection apparatus for supplying a 2N-level symbol stream each having a symbol period of a specific time length and sensitive to carrying artifacts of a co-channel interfering analog TV signal, and an interleaved trellis. A DTV signal receiver having a trellis decoder for trellis decoding coded data, and a symbol decoding device for symbol decoding a 2N-level symbol stream to supply the interleaved trellis coded data to a trellis decoder In providing.

1 shows a digital television according to an embodiment of the present invention using an NTSC-removing comb filter before symbol decoding, a postcoding comb filter after symbol decoding, and a co-channel interference detector for comparing baseband energy. Block diagram of a signal receiver,

2 is a block diagram of an NTSC co-channel interference detector for use in the digital television signal receiver of FIG.

3 illustrates an embodiment of the present invention using an NTSC-removing comb filter before symbol decoding and a postcoding comb filter after symbol decoding and using a co-channel interference detector of the type disclosed in US patent application Ser. No. 08 / 821,945. A block diagram showing a part of a digital television signal receiver according to an example;

4 illustrates an embodiment of the present invention using an NTSC-removing comb filter before symbol decoding and a postcoding comb filter after symbol decoding and using a co-channel interference detector of the type disclosed in US patent application Ser. No. 08 / 821,944. A block diagram showing a part of a digital television signal receiver according to an example;

5 is selected from a symbol decoding result defined during a data synchronization interval, depending on whether or not the baseband symbol code is substantially free of NTSC co-channel interference, and a data slicer response to the received baseband symbol code or Part of the digital television signal receiver of FIG. 1, 3 or 4 with respect to the selection of the final symbol decoding result, which is selected at a different time from the postcoded data slicer response to the comb filter response of the received baseband symbol code Block diagram showing in detail,

FIG. 6 is a block diagram showing an alternative embodiment of the circuit configuration of the digital television signal receiver of FIG. 5; FIG.

FIG. 7 is a block diagram showing another alternative embodiment of some circuit configurations of the digital television signal receiver of FIG. 5;

8 is a block diagram showing in detail a part of the configuration of the digital television signal receiver of FIG. 1, FIG. 3 or FIG. 4 for generating a prescribed symbol decoding result during a data synchronization interval;

9 is a block diagram showing in detail a part of the circuit configuration of the digital television signal receiver of FIG.

FIG. 10 is a block diagram showing a partial circuit configuration of the digital television signal receiver of FIG. 1, 3, or 4 when the NTSC-removing comb filter uses a six-symbol delay circuit; FIG.

11 is a block diagram showing a partial circuit configuration of the digital television signal receiver of FIG. 1, 3, or 4 when the NTSC-rejection comb filter uses a 2-video-line delay circuit;

12 is a block diagram showing a partial circuit configuration of the digital television signal receiver of FIG. 1, 3, or 4 when the NTSC-rejection comb filter uses a 262-video-line delay circuit;

FIG. 13 is a block diagram showing some circuit configurations of the digital television signal receiver of FIG.

14 is a block diagram showing in detail the circuit configuration of a digital television signal receiver using a plurality of NTSC-removing comb filters for performing parallel symbol decoding;

15 is a detailed view of a symbol code selection circuit that can be used in the digital television signal receiver of the type shown in FIG.

FIG. 15A is a block diagram showing in detail the circuit configuration of the digital television signal receiver of FIG. 14 for generating a prescribed symbol decoding result during a data synchronization interval; FIG.

Fig. 15B is a block diagram showing in detail the circuit arrangement of the digital television signal receiver of Fig. 14 for generating a prescribed symbol decoding result between data synchronization intervals.

In order to achieve the above object, according to an aspect of the present invention, a symbol decoding apparatus of a DTV signal receiver includes a first data slicer, a first delay device, a first linear combiner, a second linear combiner, and a first A two data slicer, a multi-input multiplexer circuit and a second delay unit. The first data slicer decodes a 2N-level symbol stream for generating an intermediate symbol decoding result. A first delay for indicating a delay of the prescribed first number of symbol periods, a 2N-level symbol stream having a first delay stream of 2N-level symbols to generate a first differential delay stream pair of 2N-level symbols; Is connected to respond. The first linear combiner is sensitive to carry artifacts of co-channel interfering analog TV signals received as first and second input signals, so as to generate a first stream of (4N-1) -level symbols as output signals. Linearly combine the first differential delay stream pairs, and the first stream of (4N-1) -level symbols provides a first comb filter response in which artifacts of the co-channel interfering analog TV signal are suppressed. The second linear combiner linearly combines the received first and second input signals to supply respective output signals, and the first comb filter response is applied to the second linear combiner as the first input signal. One of the first linear coupler and the second linear coupler is an adder, and the other is a subtractor. The second data slicer is the first of the (4N-1) -level symbols supplied as respective output signals from the first linear combiner to produce a first complementary symbol decoding result applied to the second linear combiner as the first input signal. Decode one stream. The multi-input multiplexer circuit reproduces one of a plurality of input signals selected in response to a multiplexer control signal as an output signal, receives the intermediate symbol decoding result as one of a plurality of input signals, and outputs the output of the second linear combiner. Is received as another one of the plurality of input signals. The second delay unit is connected to delay the output signal of the multi-input multiplexer circuit by the first number of the prescribed symbol period to generate the second input signal of the second linear combiner, and the output signal of the multi-input multiplexer circuit. Is used for a minimum time as the final symbol decoding result including interleaved trellis coded data.

In the above description of the present invention, the first type of symbol recoding process performed on the first comb filter before performing the data-slicing process by the second data slicer is regarded as a precoding process. The second comb filter including the second linear combiner, the multi-input multiplexer circuit and the second delay unit performs a symbol re-encoding process of the second type after data-slicing. Post-coding process for compensating and generating accurate symbol decoding results.

DTV signal receivers implementing the present invention also use additional filtering to suppress co-channel interference. A third delay for indicating a delay of the first defined number of symbol periods, a 2N-level symbol stream having a second delay stream of 2N-level symbols to generate a second differential delay stream pair of 2N-level symbols; Is connected to respond. A third linear combiner sensitive to carry artifacts of the co-channel interfering analog TV signal received as the first and second input signals, so as to generate a second stream of (4N-1) -level symbols as an output signal; Linearly combine the differential delay stream pairs, and a second stream of (4N-1) -level symbols provides a second comb filter response in which artifacts of the co-channel interfering analog TV signal are suppressed. The fourth linear combiner linearly combines the received first and second input signals to supply each output signal applied to the multi-input multiplexer circuit as another input signal, and the third linear combiner and the fourth linear combiner. One of the linkers is an adder and the other is a subtractor. Wherein the third data slicer is a first of (4N-1) -level symbols supplied as respective output signals from the third linear combiner to generate a second complementary symbol decoding result applied to the fourth linear combiner as a first input signal; Decode two streams. The fourth delay unit is connected to delay the output signal of the multi-input multiplexer circuit by a second number of the prescribed symbol period to generate a second input signal of the fourth linear combiner.

The co-channel interference detection circuit determines the amount of co-channel interference energy of the intermediate symbol decoding result, the first complementary symbol decoding result, and the second complementary symbol decoding result, and generates a multiplexer control signal according to these co-channel interference energy amounts. It is included in the embodiment of the present invention to make. The multiple-input multiplexer circuit responsive to the multiplexer control signal is provided so that a stream of 2N-level symbols from a DTV signal detection device can be suppressed and uncorrectable by a stream of 2N-level symbols into one of the first and third comb filter responses. The intermediate symbol decoding result is the final symbol unless the co-channel interference detection circuit generates a current indication that it involves an artifact of a co-channel interfering analog TV signal having an energy level strong enough to interfere with the first data slicer that decodes the data. Select as decoding result. Otherwise, the multi-input multiplexer circuit responsive to the multiplexer control signal does not select one of the first and second complementary symbol decoding results having a large amount of co-channel interference energy.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, the same reference numerals will be used for the same parts throughout the drawings. In addition, a detailed description of functions of known components not related to the subject matter of the present invention will be omitted herein.

As will be appreciated by those skilled in the art of electronic design, in many of the circuits shown in the accompanying drawings, shimming delays are inserted so that the order of operation is correct. If there is anything unusual about the specific seaming delay requirement, it will not be explicitly mentioned here.

1 shows a DTV signal receiver used to recover error-correction data suitable for moving picture expert group (MPEG) -2 decoding and display in a TV set or recording by a digital video cassette recorder. Although the DTV signal receiver of FIG. 1 is shown as receiving a TV broadcast signal from a receiving antenna 8, it may receive a signal from a cable network instead of an antenna. The TV broadcast signal is supplied as an input signal to the DTV receiver front end 10. The DTV receiver front end 10 generally includes an RF for converting a radio-frequency (RF) TV signal into an IF TV signal supplied as an input signal to an IF amplifier chain 12 for obtaining a VSB DTV signal. A first detector is provided. The DTV receiver includes an IF amplifier for amplifying a DTV signal converted into a UHF (Ultra High Frequency) band by a first detector, a second detector for converting the amplified DTV signal into a Very High Frequency (VHF) band, and a VHF. It is advantageous to be of a multiple conversion type with the IF amplifier chain 12 having another IF amplifier for amplifying the band converted DTV signal. If demodulation to baseband is performed in a digital regime, IF amplifier chain 12 will further include a third detector for converting the amplified IF signal to the final IF band close to the baseband. .

Preferably a Surface Acoustic Wave (SAW) filter is used in the IF amplifier for the UHF band to form a channel select response and remove adjacent channels. The SAW filter is quickly cut off at least 5.38 MHz from the suppressed carrier frequency of the pilot carrier and the VSB DTV signal having a similar frequency and a fixed frequency. The SAW filter thus removes the frequency modulation (FM) voice carrier of the co-channel interfering analog TV signal. By eliminating the FM voice carrier of the co-channel interfering analog TV signal in the IF amplifier chain 12, artifacts of the carrier generated when the final IF signal is detected to recover the baseband symbol are prevented, and the base during symbol decoding Artifacts that interfere with data-slicing of band symbols can be predicted. The prevention of artifacts that interfere with data-slicing of baseband symbols during symbol decoding is better than that achieved by relying on comb filtering prior to data-slicing.

The final IF output signal generated by the IF amplifier chain 12 is a complex demodulator for a DTV signal that demodulates a VSB AM DTV signal in the final IF band to recover a real baseband signal and an image baseband signal. (complexer demodulator) 14 is supplied. Such demodulation is carried out by CB Patel et al., Entitled Digital VSB Detector with Phase Tracker, as for Inclusion in an HDTV Receiver, published December 26, 1995. It may be performed digitally after analog-to-digital conversion of the final IF band of the small megacycle range, such as described in US Pat. No. 5,479,449. Alternatively, demodulation may be performed in an analog regime, in which case the results typically undergo an analog-to-digital conversion process to facilitate another process. Complex demodulation is performed by in-phase (I) synchronous demodulation and quadrature-phase (Q) synchronous demodulation. The digital result of the demodulation process has an 8-bit accuracy and represents a 2N-level symbol that encodes N-bits of data. In general, 2N is 8 when the DTV signal receiver of FIG. 1 receives public broadcasting through the antenna 8, and 16 when the DTV signal receiver of FIG. 1 receives cable broadcasting. The present invention relates to terrestrial broadcast reception on the ground, and FIG. 1 does not show some circuits of a DTV receiver that provides symbol decoding and error-correction decoding for received wired transmissions.

The symbol synchronizer and equalizer 16 receives digitized real samples of the in-phase (I-channel) baseband signal of the complex demodulator 14. In the DTV receiver of FIG. 1, the symbol synchronizer and equalizer 16 also receive digitized imaginary samples of quadrature (Q-channel) baseband signals. The symbol synchronizer and equalizer 16 has a digital filter having an adjustable weighting coefficient that compensates for ghosts and tilts of the received signal. The symbol synchronizer and equalizer 16 provides for symbol synchronization or de-rotation as well as amplitude equalization and ghost cancellation. Symbol synchronizers and equalizers in which symbol synchronization is achieved prior to amplitude equalization are known from US Pat. No. 5,479,449. In such a design, the complex demodulator 14 will supply an oversampled demodulator response including the real baseband signal and the imagery baseband signal to the symbol synchronizer and equalizer 16. After symbol synchronization, the data is decimated to extract the baseband I-channel signal at the normal symbol rate and to reduce the sample rate through digital filtering used for amplitude equalization and ghost cancellation. In symbol synchronizers and equalizers where amplitude equalization precedes symbol synchronization, de-rotation or phase tracking is already known to those skilled in the art working on digital signal receiver design. Each sample of the symbol synchronizer and equalizer 16 is decomposed to about 10 bits and is in fact a digital representation of an analog symbol representing one of the levels (2N = 8).

The output signals of the symbol synchronizer and equalizer 16 are carefully gain-controlled by any of several known methods, and the ideal step level for the symbol is known. Since the response speed of the gain control is very fast, the direct component of the real baseband signal supplied from the complex demodulator 14 is adjusted to a normalization level of +1.25 by one method of gain control. This gain control method is generally described in U.S. Patent No. 5,479,454, and CB Patel et al., Automatic Gain Control of Radio Receiver for Receiving Digital High-Definition Television Signals, issued June 3, 1997. Is described in more detail in US Pat. No. 5,573,454, entitled " Automatic Gain Control of Wireless Receiver.

The output signal of the symbol synchronizer and equalizer 16 is input to a data synchronization detector 18 which recovers the data feed synchronization information F and the data segment synchronization information S from the equalized baseband I-channel signal. Supplied as a signal. Alternatively, the input signal of the data synchronization detector 18 can be obtained before equalization.

The normal symbol rate equalized I-channel signal samples supplied as output signals from the symbol synchronization and equalizer 16 are applied as input signals to the NTSC-rejection comb filter 20. The NTSC-rejection comb filter 20 is configured to generate a first delay combination 201 and a response of the NTSC-rejection comb filter 20 to generate a differential delay stream pair of 2N-level symbols. And a first linear combiner 202 for linearly combining the differential delay symbol streams to be generated as a coding filter response. As described in US Pat. No. 5,260,793, the first delay unit 201 provides a delay equal to the period of twelve 2N-level symbols, and the first linear combiner 202 can be a subtractor. The output signal of the NTSC-rejection comb filter 20 is decoded into 10 bits and is in fact a digital representation of an analog symbol representing one of (14N-1) = 15 levels.

The symbol synchronizer and equalizer 16 is thought to be designed to suppress the direct bias component of its input signal (represented as a digital sample), the DC bias component having a normal level of +1.25. And a real baseband signal supplied from the complex demodulator 14 by the detection of the pilot carrier. Thus, each sample of the output signal of the symbol synchronizer and equalizer 16, which is supplied as an input signal of the NTSC-removing comb filter 20, is, in fact, normalized at levels -7, -5, -3, -1, +1, Digital representation of an analog symbol representing one of +3, +5, or +7. These sample levels are named odd symbol levels and detected by the odd-level data slicer 22, resulting in interim symbol decoding results of 000, 001, 010, 011, 100, 101, 110, and 111, respectively. .

Each sample of the NTSC-removing comb filter 20 is in effect a normalization level of -14, -12, -10, -8, -6, -4, -2, 0, +2, +4, +6, + Digital representation of an analog symbol representing one of 8, +10, +12, +14. These sample levels are called even symbol levels and are detected by the even-level data slicer 24 to detect 001, 010, 011, 100, 101, 110, 111, 000, 001, 010, 011, 100, 101, 110. , A precoded symbol decoding result of 111 is generated, respectively.

The symbol synchronizer and equalizer 16 is thought to be designed to suppress the DC bias component of its input signal (represented as a digital sample), the DC bias component having a normal level of +1.25, Detection results in a real baseband signal supplied from the complex demodulator 14. Alternatively, the symbol synchronizer and equalizer 16 are designed to preserve the direct current bias component of the input signal, thereby facilitating the design of the equalizer filter of the symbol synchronizer and equalizer 16 somewhat. In such a case, the data-slicing level of the odd-level data slicer 22 is offset in consideration of the direct current bias component accompanying the data step with its input signal. If the first linear combiner 202 is a subtractor, whether the symbol synchronizer and equalizer 16 are designed to suppress or preserve the direct current bias component of its input signal is determined by the even-level data slicer 24. Not so important when it comes to data slicing. However, when the differential delay provided by the first delay unit 201 is selected so that the first linear combiner 202 becomes an adder, the data slicing level of the even-level data slicer 24 takes the data step into its input signal. It should be offset to account for the accompanying double direct current bias component.

The postcoding comb filter 26 is used after the data slicers 22, 24 to generate a postcoding filter response to the precoding filter response of the NTSC-removing comb filter 20. The postcoding comb filter 26 is a second having a delay equal to that of the 3-input multiplexer 261, the second linear combiner 262, and the first delay 201 of the NTSC-removing comb filter 20. And a retarder 263. The second linear coupler 262 becomes a modulo-8 adder if the first linear coupler 202 is a subtractor and a modulo-8 subtractor if the first linear coupler 202 is an adder. . The first linear coupler 202 and the second linear coupler 262 may be configured with respective read only memories (ROMs) to improve the linear combination operation to fully support the associated sample rate. The output signal of the multiplexer 261 provides the response of the postcoding comb filter 26 and is delayed by the second delay 263. The second linear combiner 262 combines the precoded symbol decoding result of the even-level data slicer 24 with the output signal of the second delay unit 263 to combine the second linear combination result with the second linear combiner ( 262 is generated as an output signal.

The output signal of the multiplexer 261 is selected in response to the multiplexer control signals in the first, second, and third states supplied from the controller 28 to the multiplexer 261, and is applied to the multiplexer 261. Play one of the input signals. The first input port of the multiplexer 261 is the controller 28 during the time that the data feed synchronization information F and the data segment synchronization information S from the equalized baseband I-channel signal are restored by the data synchronization detector 18. Receive the ideal symbol decoding result supplied from the memory in. The controller 28 controls the multiplexer 261 in a first state to control the multiplexer 261 to supply the ideal symbol decoding result supplied from the memory in the controller 28 as the final coding result as its input signal during the recovery time. Is supplied to the multiplexer 261. The odd-level data slicer 22 supplies the intermediate symbol decoding result as its output signal to the second input port of the multiplexer 261. The multiplexer 261 is adjusted by the multiplexer control signal in the second state to reproduce the intermediate symbol decoding result as the final coding result which is the output signal. The second linear combiner 262 supplies the postcoded symbol decoding result as its output signal to the third input port of the multiplexer 261. The multiplexer 261 is adjusted by the multiplexer control signal in the third state to reproduce the postcoded symbol decoding result as the final coding result which is the output signal.

The postcoding process performed subsequent to the comb filtering and data-slicing process has a fundamental problem that must be solved in order to properly perform this postcoding process. Looking at one aspect of this problem, once an error occurs in a partially-filtered symbol decoding result, the error is fed back with a delay and tends to propagate the error while a postcoded symbol decoding result is generated. Another aspect of the problem relates to a method of initializing conditions in a delayed feedback circuit and a method of reinitializing conditions in the delayed feedback circuit once error propagation occurs. These problems arise when the second type of recoding process is used for postcoding, since the feedback used in the second type of recoding is accumulated and provides some kind of integration. When the re-encoding process of the second type is performed during precoding and the re-encoding process of the first type is performed during postcoding, the re-encoding process of the first type is performed on the initial conditions of the re-encoding process of the second type. It provides a kind of differentiation that quickly suppresses the response. Initial conditions of accumulation or integration are not of interest. When the recoding process of the first type is performed during precoding and the recoding process of the second type is performed during postcoding, it is caused by an incorrect initial condition of accumulation or integration during recoding of the second type. Errors propagate themselves during the postcoding process. The running error in the final decoding result is a systematic error rather than a random error, that is, the running error will not be able to self-correct itself by accident.

The running error of the postcoded symbol decoding result generated by the postcoding comb filter 26 is determined by the controller 28 during the time during which the data feed synchronization information F and the data segment synchronization information S are restored by the data synchronization detector 18. It is reduced by feeding back the ideal symbol decoding results supplied from the memory within. The generation of the ideal symbol decoding result will be described later in detail. The data synchronization detector 18 feeds back the best estimate of the final symbol decoding result to the second linear combiner 262 via the second delay unit 263, thereby returning the data feed synchronization information F and the data segment synchronization information S. During the time of recovery, the running error of the postcoded symbol decoding result from the postcoding comb filter 26 can always be corrected. If a co-channel interfering NTSC signal has a high-luminance white or near white content, the intermediate symbol decoding result becomes insensitive to error and the best of the final symbol decoding result. It will be chosen to supply the evaluation value. As a result, the running error is corrected.

In the post-coding comb filter 26, the output signal of the multiplexer 261 is three-parallel-bits assembled by a data assembler 30 for application to a data interleaver 32. Contains the final symbol decoding results in groups. The data interleaver 32 converts the assembled data into a parallel data stream for application to a trellis decoder circuit 34. The trellis decoder circuit 34 generally uses twelve trellis decoders. The trellis decoding result is supplied from the trellis decoder circuit 34 to the data de-interleaver 36 for de-commutation. The byte parsing circuit 38 converts the output signal of the data deinterleaver 36 into bytes of Reed-Solomon decoder error-correction encoding for applying to the Reed-Solomon decoder circuit 40. Reed-Solomon decoding is performed to generate an error-corrected byte stream for application to the data de-randomizer 42. The data de-randomizer 42 provides playback data to the rest of the receiver (not shown). The remainder of a complete DTV receiver will include components such as a packet sorter, voice decoder, MPEG-2 decoder, and the like. The remainder of the DTV receiver included in the digital tape recorder / player will include circuitry for converting data into a format for recording.

NTSC co-channel interference detector 44 displays an indication of whether NTSC co-channel interference has sufficient strength to cause inaccurate errors in the data-slicing process performed by radix-level data slicer 22. Supply to the controller 28. If the NTSC co-channel interference detector 44 indicates that the NTSC co-channel interference does not have the strength described above, the controller 28 indicates that the data feed synchronization information F and the data segment synchronization information S are the data synchronization detector 18. The multiplexer control signal of the second state will be supplied to the multiplexer 261 at a time other than the period restored by the " Under this condition, the multiplexer 261 reproduces the intermediate symbol decoding result supplied from the odd-level data slicer 22 as an output signal. If the NTSC co-channel interference detector 44 indicates that the NTSC co-channel interference has sufficient strength to cause an incorrect error in the data-slicing process performed by the radix-level data slicer 22, the controller Numeral 28 will supply the multiplexer control signal of the third state to the multiplexer 261 at a time other than the period during which the data feed synchronization information F and the data segment synchronization information S are restored by the data synchronization detector 18. Under this condition, the multiplexer 261 reproduces the postcoded symbol decoding result provided as the second linear combination result from the second linear combiner 262 as its output signal.

2 illustrates a configuration form that the NTSC co-channel interference detector 44 described above may consider to be novel. The subtractor 441 differentially combines the intermediate symbol decoding result supplied from the odd-level data slicer 22 and the postcoded symbol decoding result supplied as the second linear combination result from the second linear combiner 262. If the amount of NTSC co-channel interference is negligible and the random noise of the baseband I-channel signal is negligible, the result of these intermediate and postcoded symbol decoding should be similar, so The difference output signal from 441 should be low. On the other hand, if the amount of NTSC co-channel interference is an appreciable amount, the difference output signal from subtractor 441 is not going low, but is often in a high state.

The measurement of the energy of the difference output signal generated in the subtractor 441 is performed by squarer 442 squares the difference output signal and at a short time interval defined by a mean averager 443. This is done by determining the average of the response of the squarer 442 over. The squarer 442 can be implemented using a ROM. The averager 443 may be implemented using a delay line memory for storing a plurality of consecutive digital samples and an adder for summing up the digital samples recently stored in the delay line memory. A short-term mean average of the energy of the difference output signal of the subtractor 441, as determined by the averager 443, is supplied to a digital comparator connected to provide to the threshold detector 444. The threshold of the threshold detector 444 is high enough not to exceed the short term average of the random noise accompanying the intermediate symbol decoding result and the postcoding symbol decoding result supplied to the subtractor 441. The threshold is exceeded if NTSC co-channel interference has sufficient strength to cause inaccurate errors during the data-slicing process performed by the odd-level data slicer 22. The threshold detector 444 provides an indication to the controller 28 indicating whether the threshold is exceeded.

FIG. 3 shows the configuration of a DTV receiver different from that of FIG. 1, in which NTSC co-channel interference will cause inaccurate errors during the data-slicing process performed by the radix-level data slicer 22 described above. The circuitry for determining whether it has sufficient strength is disclosed in US Patent Application No. 08 / 821,945 filed March 19, 1997, titled Using Video Signals from Auxiliary Analog TV Receivers for Detecting NTSC Interference in Digital TV Receivers. It differs from that of FIG. 1 in that it is a circuit of the type disclosed in the call. The DTV signal, such as converted to IF by the DTV receiver front end 10, is supplied to an IF amplifier chain 46 for an NTSC signal. The NTSC IF amplifier chain 46 is different from the IF amplifier chain used in conventional NTSC signal receivers. As far as the mid-band gain characteristics are concerned, the amplifier stage of the NTSC IF amplifier chain 46 corresponds to the amplifier stage of the IF amplifier chain 12 for the DTV signal, which is the same automatic gain control as the corresponding amplifier stage of the IF amplifier chain 12. And substantial linear gain. The residual sidebands of the NTSC signal are not suppressed in the NTSC IF amplifier chain 46. A portion of the entire sideband of the NTSC signal with a single-sideband match is preferably suppressed in the NTSC IF amplifier chain 46 in order for the co-channel DTV signal to reduce energy. As a result, the dynamic range of the NTSC IF amplifier chain 46 response is reduced, which facilitates further amplification of the video carrier for synchronizing the phase of the local video carrier oscillator used in the complex demodulator 48. The filtering process for establishing the bandwidth of the NTSC IF amplifier chain 46 may be performed through SAW filtering in the UHF IF amplifier when a plural-conversion receiver circuit is used. The amplified IF response of the NTSC IF amplifier chain 46 is fed directly or after some other amplification to the complex demodulator 48 for the NTSC video signal. The complex demodulator 48 supplies an in-phase I-channel response consisting of a sample of the NTSC signal and a real component accompanying the DTV artifact. The complex demodulator 48 also supplies a quadrature-phase Q-channel response consisting of samples of imaginary components involving DTV artifacts, which are applied to the Hilbert transform filter 50. The response of the Hilbert transform filter 50 is fed to a linear combiner 52. The linear combiner 52 combines the response of the Hilbert transform filter 50 with the appropriately delayed in-phase I-channel response to recover the samples of the NTSC signal that are substantially without DTV artifacts. The linear combiner 52 becomes an adder or subtractor according to relative image carrier tuning during the synchronous demodulation process used in the complex demodulator 48 to generate I-channel and Q-channel responses.

The NTSC signal, which is substantially not accompanied by the DTV artifacts supplied from the linear combiner 52, is applied to a low pass filter (LPF) 54 having a cutoff frequency of about 750 kHz. The estimate of the luminance signal energy of the co-channel interfering NTSC signal is squared by the squarer 56 to the response of the LPF 54 and squared 56 over a short time interval defined by the averager 58. Is generated by determining the average value of the response. This evaluation value is supplied to the threshold detector 60. The threshold of the threshold detector 60 is exceeded if the NTSC co-channel interference has sufficient strength to cause inaccurate errors during the data-slicing process performed by the odd-level data slicer 22. The threshold detector 58 supplies an indication to the controller 28 indicating whether the threshold is exceeded.

FIG. 4 shows the configuration of a DTV receiver different from the DTV receivers of FIGS. 1 and 3, wherein an NTSC co-channel interference is inaccurate during the data-slicing process performed by the radix-level data slicer 22 described above. The circuitry for determining whether or not it has sufficient strength to cause is described in US patent application Ser. No. 08 / 821,944, filed March 19, 1997, titled Using Intercarrier Signals for Detecting NTSC Interference in Digital TV Receivers. It differs from that of FIGS. 1 and 3 in that it is a circuit of the type disclosed. The DTV signal, such as converted to IF by the DTV receiver front end 10, is supplied to an IF amplifier chain 62 for a quasi-parallel type NTSC voice signal. The amplifier stage of the IF amplifier chain 62 for NTSC voice signals corresponds to the amplifier stage of the IF amplifier chain 12 for the DTV signal, which has the same automatic gain control and substantial linear gain as the corresponding amplifier stage of the IF amplifier chain 12. Have The frequency selectivity of the IF amplifier chain 12 is shown to emulate a response within an NTSC audio carrier of ± 250 Hz and an NTSC video carrier of approximately 250 Hz. The filtering process for establishing the frequency selectivity of the NTSC IF amplifier chain 62 may be performed through SAW filtering in the UHF IF amplifier when a multi-conversion receiver circuit is used. The response of the NTSC IF amplifier chain 62 uses an intercarrier speech IF signal having a 4.5 MHz carrier frequency using a modulated NTSC video carrier as an enhanced carrier for heterodinating an NTSC speech carrier. The intercarrier detector 64 to generate is supplied. This intercarrier speech IF signal is amplified by the intercarrier speech IF amplifier 66, and the 4.5 MHz IF amplifier 66 supplies the amplified intercarrier speech IF signal to the intercarrier amplitude detector 68. The response of the intercarrier amplitude detector 68 is averaged by the averager 70 over a defined short time interval, and the final average value is supplied to the threshold detector 72. The threshold of the threshold detector 72 is exceeded if the NTSC co-channel interference has sufficient strength to cause inaccurate errors during the data-slicing process performed by the odd-level data slicer 22. The threshold detector 72 supplies an indication to the controller 28 indicating whether the threshold is exceeded.

FIG. 5 illustrates a preferred type in which the operation of the multiplexer 261 of the postcoding comb filter 26 of FIG. 1 is performed. A three-input multiplexer 261 is shown that includes two two-input multiplexers 2611 and 2612. The controller 28 applies the output signal of the NTSC co-channel interference detector (eg 44) as a control signal to the two-input multiplexer 2611.

If the NTSC co-channel interference is of sufficient intensity to cause an incorrect error during the data-slicing process performed by the radix-level data slicer 22, the last one of the NTSC co-channel interference detector Due to the output signal, the multiplexer 2611 reproduces the post-coded symbol decoding result supplied by the second linear combiner 262 to the first input port of the multiplexer 2611 and applies it to the second input port of the multiplexer 2612. .

If the NTSC co-channel interference has sufficient strength to cause an incorrect error during the data-slicing process performed by the radix-level data slicer 22, the last zero (zero) of the NTSC co-channel interference detector. Due to the output signal, the multiplexer 2611 reproduces the intermediate symbol decoding result supplied by the odd-level data slicer 22 to the second input port of the multiplexer 2611. These reproduced intermediate symbol decoding results are applied to the second input port of the multiplexer 2612.

5, 6 and 7 show the OR gate 281 included in the controller 28, respectively. The OR gate 281 is configured to supply 1 in response to generation of the shield sync segment being detected by the feed segment sync detector 181 and generation of the data sync code being detected by the data segment sync detector 182. If you supply 1 in response, you supply a response of 1. In all other cases, OR gate 281 supplies a zero response.

In FIG. 5, the response of the OR gate 281 is applied to the multiplexer 2612 as a control signal. The response 0 of the OR gate 281 causes the multiplexer 2612 to send the output signal of the multiplexer 2611 to the data assembler 30 which is applied to the second input port of the multiplexer 2612 as an excellent evaluation value of the symbol decoding result. Play back as the final symbol decoding result to apply. The response 0 of the OR gate 281 causes the multiplexer 2612 to reproduce the ideal decoding result extracted from the memory of the controller 28 as the final symbol decoding result for applying to the data assembler 30, which is referred to in FIG. Will be described later in detail herein.

6 shows a postcoding comb filter 260 which is an alternative component of the postcoding comb filter 26 described above. The three-input multiplexer 261 comprising two two-input multiplexers 2611 and 2612 is replaced with a three-input multiplexer 2610 comprising three three-input multiplexers 26101, 26102 and 26103 in FIG. 6. do.

FIG. 7 shows a configuration according to a modified embodiment of the post-coding comb filter 26 where a three-input multiplexer 261 comprising two two-input multiplexers 2611 and 2612 is an OR gate 281. And two input multiplexers 26100 including two two input multiplexers 261001 and 261002 for receiving respective control signals from the NTSC co-channel interference detector. The postcoding comb filter 2600 provides slightly different computational results than the postcoding comb filters 26 and 260. The multiplexer 26001 replaces the postcoded symbol decoding result with an ideal symbol decoding result when the response of the OR gate 281 is one. If the NTSC co-channel interference detector is to supply 1 indicating that NTSC co-channel interference has sufficient strength to cause inaccurate errors during the data-slicing process performed by the radix-level data slicer 22. In this case, the multiplexer 261002 selects the last modified postcoded symbol decoding result as the final symbol decoding result for applying to the data assembler 30. If the NTSC co-channel interference detector has a zero indicating that the NTSC co-channel interference has insufficient strength to cause an incorrect error during the data-slicing process performed by the radix-level data slicer 22. When supplied, multiplexer 261002 selects the intermediate symbol decoding results of radix-level data slicer 22 as the final symbol decoding results for applying to data assembler 30, where these intermediate symbol decoding results are ideal symbol decoding results. Is not replaced by.

FIG. 8 is a block diagram illustrating in more detail the configuration of the multiplexer 2612 of FIG. 5 together with a circuit for generating an ideal symbol decoding result applied to the multiplexer 2612 of FIG. The multiplexer 2612 includes an output buffer register of ROMs 74, 76 and 78 for selectively reading from the multiplexer 2612 to the 3-bit-wide output bus 80. The multiplexer 2612 further includes a tri-state data buffer 82 for forward selection of the 3-bit-wide output of the multiplexer 2611. Circuits for generating an ideal symbol decoding result applied to the multiplexer 2612 include odd-numbered sync sync ROM ROM 74, even-feed sync sync ROM ROM 76, line sync code ROM 78, and symbols. Clock generator 84, 1-data-frame address counter 86 for addressing ROMs 74, 76, 78, and address counter jam reset circuit 88 for resetting address counter 86. The odd-numbered sync segment address decoder 94, the even-numbered sync segment address decoder 96, and the line sync address decoder 98 for generating a readable signal for the ROMs 74, 76, and 78; And a NOR gate 92 for controlling the three-state data buffer 82. The address counter 86 counts input pulses received from the symbol clock generator 84 at a symbol decoding rate and generates successive addresses that each represent symbols in one data frame. Appropriate portions of these addresses are applied to the ROMs 74, 76 and 78 as input addresses. The jam reset circuit 88 causes the address counter 86 to have an appropriate count in response to the data feed sync information F and the data segment sync information S restored by the data sync detector 18 of FIG. 1, 3 or 4. Reset to.

The address counter 86 is preferably configured such that the upper bit group counts the number of data segments per data frame and the lower bit group counts the number of symbols per data segment. As a result of this configuration, the design of the address counter jam reset circuit 88 is facilitated, the bit width of the input signal of the address decoders 94, 96, and 98 is reduced, and the partial counter from the address counter 86 The addressing operation of the ROMs 74, 76 and 78 becomes easy, thereby reducing the bit width at the addressing of the ROM.

The ROM 74 stores the ideal symbol decoding result for the odd-feed sync segment, receives 1 from the address decoder 94 and is optionally operational for reading. The ROM 74 is addressed by a lower bit group of the output of the address counter 86 which counted the number of symbols per data segment group, and the address decoder 94 counted the upper bit group which counted the number of data segments per data frame. Answer The address decoder 94 generates 1 only if the data segment portion of the address supplied by the address counter 86 corresponds to the address of the odd-numbered sync segment.

The ROM 76 stores the ideal symbol decoding result for the even-feed sync segment, receives 1 from the address decoder 96, and is selectively enabled for reading. The ROM 76 is addressed by a lower bit group of the output of the address counter 86 that counted the number of symbols per data segment group, and the address decoder 96 counted the number of higher data bits per data frame group. Answer The address decoder 96 generates 1 only when the data segment portion of the address supplied by the address counter 86 corresponds to the address of the even-order sync segment.

The ROM 78 stores the ideal symbol decoding result for the start code group at the beginning of each sync segment, receives 1 from the address decoder 98 and is optionally operational for reading. The ROM 78 responds to the two lower bits of the output of the address counter 86, and the address decoder 98 responds to the lower bit group of the output of the address counter 86, which counted the number of symbols per data segment group. do. The address decoder 98 generates 1 only if the data symbol per data segment count portion of the address supplied by the address counter 86 corresponds to the partial address of the start code group.

The NOR gate 92 receives the response of the address decoders 94, 96 and 98 at each of its three input connections. When an ideal symbol decoding result is available, one of the address decoders 94, 96, 98 provides a NOR gate 92 to supply 1 as its output signal and supply a 0 response to the tri-state data buffer 82. Is adjusted. As a result, the tri-state data buffer 82 is adjusted to exhibit a high source impedance on the data bus 80 such that the signal transmitted from the multiplexer 2611 is converted to a 3-bit-wide data bus ( 80) will not be identified. None of the address decoders 94, 96, and 98 supply 1 as its output signal during these data segment portions where the ideal symbol decoding result cannot be predicted, thus giving 1 to the 3-state data buffer 82. NOR gate 92 is adjusted to supply a response. As a result, the tri-state data buffer 82 is adjusted to exhibit a low source impedance on the data bus 80 such that the signal transmitted from the multiplexer 2611 is converted into a 3-bit-wide data bus ( 80).

The circuit of FIG. 8 for generating an ideal symbol decoding result applied to the multiplexer 2611 can be readily adapted by those skilled in the art working in digital circuit design for use in the circuit configurations shown in FIGS. 6 and 7. .

9 shows a NTSC-removing comb filter 120 of a modified configuration of the NTSC-removing comb filter 20 described above and a post-coding comb filter 126 of a modified configuration of the post-coding comb filter 26 described above. 1 illustrates a detailed configuration of some circuits of the DTV signal receiver of FIG. 1, 3, or 4. The subtractor 1202 performs the function of the first linear coupler of the NTSC-removing comb filter 120, and the modulo-8 adder 1262 performs the function of the second linear coupler of the postcoding comb filter 126. . In the NTSC-removing comb filter 120, a first delay 1201 indicating a 12 symbol time origin delay is used, and a second delay 1201 also representing a delay of a 12 symbol time origin in the postcoding comb filter 126. ) Is used. The 12-symbol delay represented by each of the delayers 1201 and 1263 is close to one cycle delay of the artifact of the analog TV video carrier at 59.75 times the analog TV horizontal scanning frequency f _H. The 12-symbol delay is close to the five cycle delay of the artifact of the analog TV luminance subcarrier 287.25 times the analog TV horizontal scanning frequency f _H. The 12-symbol delay is close to the six cycle delay of the artifact of an analog TV voice carrier 345.75 times the analog TV horizontal scanning frequency f _H. This is because the differential combined response of the subtractor 1202 to frequencies near the luminance subcarrier differentially delayed by the voice carrier, the image carrier and the first delayer 1201 tends to reduce co-channel interference. However, in the portion of the video signal having an edge crossing the horizontal scanning line, the correlation amount of the analog TV video signal spaced apart in the horizontal space direction is very small.

The multiplexer 1261, which is a variation of the multiplexer 261, has a second state in which most of the time points at which it is determined that there is insufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 are present. Most of the time when it is determined that there is sufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 is controlled by the multiplexer control signal which is in the third state. The multiplexer 1261 feeds the modulo-8 summation result of the adder 1262, which is the 12 symbol periods delayed by the delayer 1263, to the adder 1262 as an addend to the control signal in the third state. Is adjusted by This is a modular accumulation process in which a single error propagates as a running error, where the error reoccurs every 12 symbol periods. The running error of the postcoded symbol decoding result generated by the postcoding comb filter 126 is not only during the four symbol periods at the beginning of each data segment, but also in the first state throughout each data segment including the feedback synchronization. Which is reduced by the multiplexer 1261. When this control signal is in the first state, the multiplexer 1261 reproduces the ideal symbol decoding result supplied from the memory of the controller 28 as its output signal. As described above, the ideal symbol decoding result is introduced into the output signal of the multiplexer 1261 to stop the running error. Since there are 4 + 69 (12) symbols per data segment, the ideal symbol decoding result slips back the in-phase four symbol periods of each data segment so that running errors can last longer than three data segments. There will be no.

FIG. 10 uses a modified NTSC-removing comb filter 220 and a post-coding comb filter 26 of one modified configuration of the NTSC-removing comb filter 20 described above. 1, 3, or 4 show the configuration of some circuits of the DTV signal receiver in detail. In the NTSC-removing comb filter 220, a first delayer 2201 indicating a delay of 6 symbol periods is used, and a second delayer 2263 also indicating a delay of 6 symbol periods in the postcoding comb filter 226. Is used. The six-symbol delay represented by each of the delayers 2201 and 2263 is close to a 0.5 cycle delay of the artifact of the analog TV video carrier at 59.75 times the analog TV horizontal scanning frequency f _H. The 6-symbol delay is three cycles of the analog TV horizontal scan frequency f _H of 287.25 times the analog TV luminance near 2.5 cycle delay of the artifact of the sub-carrier, the analog TV horizontal scan frequency f _H of times the analog TV 345.75 artifacts of the audio carrier of the Close to delay. The subtractor 2202 performs the function of the first linear coupler of the NTSC-removing comb filter 220, and the modulo-8 adder 2262 performs the function of the second linear coupler of the postcoding comb filter 226. . Since the delay indicated by the delayers 2201 and 2263 is shorter than the delay indicated by the delayers 1201 and 1263 in FIG. 9, the null adjacent frequency converted from the analog TV carrier frequency. Even if is a narrow band, good anti-correlation is reduced in the combined signal subtracted by the subtractor 2202 rather than a good correlation in the differential combined signal by the subtractor 1202. More likely to be achieved. The voice carrier is less suppressed in the response of the NTSC-rejection comb filter 220 than in the response of the NTSC-rejection comb filter 120. However, if the voice carrier of the co-channel interfering analog TV signal is suppressed by SAW filtering or IF amplifier chain 12, insufficient speech cancellation of the comb filter 220 is not a problem. Responses to sync tips are continually reduced using NTSC-rejection comb filter 220 of FIG. 10 rather than NTSC-rejection comb filter 120 of FIG. 9, thereby trellis decoding and Reed-Solomon encoding. The tendency to overwhelm error-corrections is reduced.

The multiplexer 2221, which is a variation of the multiplexer 261, has a second state in which most of the time points at which it is determined that there is insufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 are present. Most of the time when it is determined that there is sufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 is controlled by the multiplexer control signal which is in the third state. The multiplexer 2221 feeds the modulo-8 addition result of the adder 2262, which is the six symbol periods delayed by the delayer 2263, to the adder 2262 as an addee to the control signal in the third state. Is adjusted by This is a modular accumulation process in which a single error propagates as a running error, where the error reoccurs every six symbol periods. The running error of the postcoded symbol decoding result generated by the postcoding comb filter 226 is in the first state not only during the four symbol periods at the beginning of each data segment, but also throughout each data segment including the feedback synchronization. Which is reduced by the multiplexer 2221. When this control signal is in the first state, the multiplexer 2261 reproduces the ideal symbol decoding result supplied from the memory of the controller 28 as its output signal. As described above, the ideal symbol decoding result is introduced into the output signal of the multiplexer 2261 so that the running error is stopped. Since there are 4 + 138 (6) symbols per data segment, the ideal symbol decoding result slips back the in-phase four symbol periods of each data segment so that running errors cannot last longer than two data segments. The possibility that the period of running error in the postcoding comb filter 226 is prolonged is substantially postcoding, although the running error recurs more frequently and affects twice the code of the 12 interleaved trellis codes. In the comb filter 126, the period of running error is less than likely to be extended.

11 shows an NTSC-removing comb filter 320 of one modified configuration of the NTSC-removing comb filter 20 described above and a post-coding comb filter 326 of a modified configuration of the post-coding comb filter 26. The configuration of some circuits of the DTV signal receiver of FIG. 1, FIG. 3, or FIG. 4 used is shown in detail. In the NTSC-removing comb filter 320, a first delayer 3201, which exhibits a delay of 1368 symbol periods substantially equal to the period of two horizontal scanning lines of an analog TV signal, is used, and also in the post-coding comb filter 326. Also used is a second delayer 3263 which represents the delay of the 1368 symbol period as described above. The first linear coupler of the NTSC-removing comb filter 320 is a subtractor 3202 and the second linear coupler of the postcoding comb filter 326 is a modulo-8 adder 3326.

In the multiplexer 361, which is a variation of the multiplexer 261, most of the time points at which it is determined that there is insufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 are in the second state. Most of the time when it is determined that there is sufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 is controlled by the multiplexer control signal which is in the third state. The DTV receiver may include circuitry for detecting a change between alternating scan lines of NTSC co-channel interference such that controller 28 can suppress the supply of the third state of multiplexer 3331 under the above conditions. desirable.

The multiplexer 3221 feeds the modulo-8 addition result of the adder 3326, which is a 1368 symbol period delayed by the delayer 3263, to the adder 3326 as an addend to the control signal in the third state. Is adjusted by This is a modular accumulation process in which a single error propagates as a running error, where the error reoccurs every 1368 symbol periods. Since this code span is longer than the span of the Rig-Solomon code single block, a single running error is easily corrected during the Lido-Solomon decoding process. The running error of the postcoded symbol decoding result generated by the postcoding comb filter 326 is in the first state not only during the four symbol periods at the beginning of each data segment, but also throughout each data segment including the feedback sync. Which is reduced by the multiplexer 3331. When this control signal is in the first state, the multiplexer 3301 reproduces the ideal symbol decoding result supplied from the memory of the controller 28 as its output signal. As described above, the ideal symbol decoding result is introduced into the output signal of the multiplexer 3331 so that the running error is stopped. Since the 16.67 ms duration of the NTSC video feed exhibits phase slippage for the 24.19 ms duration of the DTV data feed, the DTV data segment containing the feed sync eventually scans the entire NTSC frame raster. do. The 525 scan lines in the NTSC frame raster each comprise 684 symbol periods, thereby forming 359,100 symbol periods in total. This is slightly less than 432 times the 832 symbol period of the DTV data segment containing the feed sync, so running errors with a duration longer than 432 data feed result in ideal symbol decoding across the DTV data segment containing the feed sync. It can be assumed with confidence that it will be eliminated by the multiplexer 3221 generating. There is also a phase slippage between the data segments for the start code group and the NTSC image scanline where ideal symbol decoding results are available. We can estimate 359,100 symbol periods, which are 89,775 times the four symbol periods of the start code group, and are scanned over 89,775 consecutive data segments. Since there are 313 data segments per DTV data feed, we know that a running error with a duration longer than 287 data feeds will be eliminated by the multiplexer 3331 which produces an ideal symbol decoding result across the start code group. You can guess with confidence. The two sources of suppression of running errors are independent of each other, so running errors with a duration longer than about 200 data feeds are rarely similar. Furthermore, if the NTSC co-channel interference dips at the time of reoccurrence of a running error, in order to adjust the multiplexer 3301 to reproduce the response of one data slicer 22 as its output signal, this is not the case. The error may be corrected earlier.

NTSC-removing comb filter 320 of FIG. Very good. These artifacts are co-channel interference with the highest energy. The NTSC-rejection comb filter 320 does not exclude scan line-to-scan-line variations in the image content of an analog TV signal over a period of two scan lines. It provides a good suppression function for the image content regardless of color. The suppression on the FM video carrier of the analog TV signal is quite good if it is not suppressed by the symbol synchronization and tracking cancellation filter of the equalizer 16. Artifacts of most analog TV color bursts are also suppressed in response to NTSC-removing comb filter 320. The filtering provided by the NTSC-rejection comb filter 320 is also orthogonal to the NTSC-interference cancellation made during trellis decoding.

12 shows an NTSC-removing comb filter 420 of one modified configuration of the NTSC-removing comb filter 20 described above and a post-coding comb filter 426 of a modified configuration of the post-coding comb filter 26 described above. ) Shows the configuration of some circuits of the DTV signal receiver of FIG. In the NTSC-removing comb filter 420, a first delayer 4201 is used, which exhibits a delay of 179,208 symbol periods substantially equal to the period of 262 horizontal scan lines of the analog TV signal, and also in the post-coding comb filter 426. A second delayer 4403 is used which represents the delay of the above 179,208 symbol periods. The subtractor 4202 performs the function of the first linear combiner of the NTSC-removing comb filter 420, and the modulo-8 adder 4426 performs the function of the second linear combiner of the postcoding comb filter 426. .

The multiplexer 4451, which is a variation of the multiplexer 261, causes the majority of the time points at which it is determined that there is insufficient NTSC co-channel interference to cause inaccurate errors in the output signal of the data slicer 22 are in the second state. The majority of the time points at which it is determined that there is sufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 are controlled by the multiplexer control signal which is in the third state. The DTV receiver is a field-to-field change of NTSC co-channel interference such that controller 28 can suppress the supply of the third state of multiplexer 4421 under the conditions. It is preferable to include a circuit for detecting a.

The multiplexer 4421 feeds back the modulo-8 addition result of the adder 4402, which is 179,208 symbol periods delayed by the delayer 4403, to the control signal in the third state, as an addend. Is adjusted by This is a modular accumulation process in which a single error propagates as a running error, where the error reoccurs every 179,208 symbol periods. Since this symbol code span is longer than the span of a single block of Reed-Solomon code, a single running error is easily corrected during the Reed-Solomon decoding process. The running error of the postcoded symbol decoding result generated by the postcoding comb filter 426 is not only during the four symbol periods at the beginning of each data segment, but also in the first state throughout each data segment including the feedback sync. Which is reduced by the multiplexer 4401. When this control signal is in the first state, the multiplexer 4421 reproduces the ideal symbol decoding result supplied from the memory of the controller 28 as its output signal. As described above, the running error is stopped by the ideal symbol decoding result flowing into the output signal of the multiplexer 4421. The maximum number of data feeds required to remove a running error generated in the output signal of the multiplexer 4401 is substantially the same as the maximum number of data feeds required to remove a running error generated in the output signal of the multiplexer 4331. It seems to be. However, the number of times the error reoccurs in such a period is as small as factor 131.

The NTSC-rejection comb filter 420 of FIG. 12 suppresses not only most of the modulation artifacts generated in response to analog TV vertical sync pulses and equalization pulses, but also all of the modulation artifacts generated in response to analog TV horizontal sync pulses. These artifacts are the co-channel interference with the highest energy. The NTSC-rejection comb filter 420 is also caused by the video content of an analog TV signal that does not change from field-to-fiele or line-to-line. Artifacts are suppressed to remove stop patterns that are independent of their horizontal spatial frequency or color. Artifacts of most analog TV color bursts are also suppressed in NTSC-removing comb filter 420.

13 shows a NTSC-removing comb filter 520 of one modified configuration of the NTSC-removing comb filter 20 described above and a post-coding comb filter 526 of a modified configuration of the post-coding comb filter 26 described above. 1 illustrates a detailed configuration of some circuits of the DTV signal receiver of FIG. 1, 3, or 4. In the NTSC-removing comb filter 520, a first delay 5201 is used, which exhibits a delay of 718,200 symbol periods which is substantially equal to the period of two frames of the analog TV signal. A second delay 5526 is used which represents the same 718,200 symbol period delay. The subtractor 5202 performs the function of the first linear combiner of the NTSC-removing comb filter 520, and the modulo-8 adder 5262 performs the function of the second linear combiner of the postcoding comb filter 526. .

The multiplexer 5201, which is a variation of the multiplexer 261, causes the majority of the time points at which it is determined that there is insufficient NTSC co-channel interference to cause inaccurate errors in the output signal of the data slicer 22 are in the second state. The majority of the time points at which it is determined that there is sufficient NTSC co-channel interference to cause an incorrect error in the output signal of the data slicer 22 are controlled by the multiplexer control signal which is in the third state. The DTV receiver preferably includes circuitry for detecting changes between alternating frames of NTSC co-channel interference such that controller 28 can suppress the supply of the third state of multiplexer 5251 under the conditions. .

The multiplexer 5221 is fed back to the control signal in a third state to feed back the modulo-8 addition result of the adder 5262, which is a 718,200 symbol period delayed by the delayer 5203, to the adder 5262. Is adjusted by This is a modular accumulation process in which a single error propagates as a running error, where the error reoccurs every 718,200 symbol periods. Since this symbol code span is longer than the span of a single block of Reed-Solomon code, a single running error is easily corrected during the Reed-Solomon decoding process. The running error of the postcoded symbol decoding result generated by the postcoding comb filter 526 is not only during the four symbol periods at the beginning of each data segment, but also in the first state throughout each data segment including the feedback synchronization. Which is reduced by the multiplexer 4401. When this control signal is in the first state, the multiplexer 5251 reproduces the ideal symbol decoding result supplied from the memory of the controller 28 as its output signal. As described above, the ideal symbol decoding result is introduced into the output signal of the multiplexer 5261 so that the running error is stopped. The maximum number of data feeds required to remove a running error generated in the output signal of the multiplexer 5251 is substantially the same as the maximum number of data feeds required to remove a running error generated in the output signal of the multiplexer 3541. It seems to be. However, the number of times the error reoccurs in such a period is as small as factor 131.

The NTSC-rejection comb filter 520 of FIG. 13 suppresses not only all modulation artifacts generated in response to analog TV vertical sync pulses and equalization pulses, but also all modulation artifacts generated in response to analog TV horizontal sync pulses. These artifacts are the co-channel interference with the highest energy. The NTSC-removing comb filter 520 also suppresses artifacts caused in the video content of an analog TV signal that does not change over two frames, eliminating a stop pattern independent of its horizontal spatial frequency or color. Artifacts of all analog TV color bursts are also suppressed in NTSC-removing comb filter 520.

The skilled person in the field of TV system design can correlate and anti-correlation in analog TV signals that can be used to design other types of NTSC-rejection filters than those shown in FIGS. You will be clearly aware of the other characteristics of. By using NTSC-rejection filters connecting two NTSC-rejection filters of the type already disclosed, the 2N level of the baseband signal is increased to the (8N-1) data level. Such filters may be required to overcome the problem of particularly unfavorable co-channel interference, despite the disadvantage that the signal-to-noise ratio for random noise interference is reduced by the symbol decoding process.

14 shows the NTSC-removal comb filters of different types pre-deployed, post-coding comb filters respectively post-deployed to compensate for the precoding introduced by the pre-deployed NTSC-removal comb filters, and NTSC-removal comb filters; A variant embodiment of the above-described type of DTV signal receiver, constructed in accordance with another aspect of the present invention, for operating multiple symbol decoders in parallel using a good-level data slicer connected between post-coding comb filters, respectively. It is shown. The even-level data slicer A24 converts the response of the first type NTSC-rejection comb filter A20 into the first precoded symbol decoding result and applies it to the first type postcoding comb filter A26. The even-level data slicer B24 converts the response of the second type NTSC-rejection comb filter B20 into a second precoded symbol decoding result and applies it to the second type postcoding comb filter B26. The even-level data slicer C24 converts the response of the third type NTSC-rejection comb filter C20 into a third precoded symbol decoding result and applies it to the third type postcoding comb filter C26. The odd-level data slicer 22 supplies the intermediate symbol decoding result to the postcoding comb filters A26, B26, C26. The prefixes A, B, and C added to the reference numerals for the components of FIG. 14 are integers 1,2,3,4,5 when the receiver portion as shown in the components of FIGS. 9 to 13 is used. It is a different integer value that will correspond to either.

The symbol decoding selection circuit 90 of FIG. 14 selects from the intermediate symbol decoding results received from the data slicer 22 and the various postcoded symbol encoding results received from the postcoding comb filters A26, B26, and C26. The best estimate of the correct symbol decoding is formulated and applied to trellis decoder circuit 34. The best estimate of the decoding result is used to correct the summing process in postcoding comb filters A26, B26, C26.

Fig. 15, composed of Figs. 15A and 15B, is a detailed block diagram illustrating a preferred method of performing the function of the symbol decoding selection circuit 90 described above. FIG. 15A is a detailed block diagram showing the configuration of a circuit that generates a prescribed symbol decoding result for application to the 3-bit-wide output data bus 800 of the symbol decoding selection circuit 90 described above during a data synchronization interval. . The circuit of FIG. 15A operates similarly to the circuit described above with reference to FIG. 8.

FIG. 15B shows a configuration of a circuit embedded in the symbol decoding selection circuit 90 for selecting among intermediate symbol decoding results and various postcoded symbol encoding results to generate a final symbol decoding result during a time period between data synchronization intervals. It is a block diagram. In eliminating NTSC co-channel interference from the DTV signal, the efficiency of the NTSC-removing comb filters A20, B20, and C20 is converted to the baseband by the associated NTSC-rejection comb filters A100, B100, and C100. It is determined by observing how to reduce the energy of NTSC co-channel interference separated from the artifacts. Separation of NTSC co-channel interference generated from the DTV signal proceeds as described above with reference to FIG. The response of the LPF 54 to the baseband image synchronously detected from the NTSC co-channel interference is supplied as an input signal to the NTSC-rejection comb filters A100, B100 and C100. The NTSC-removing comb filter A100 is a type of linear combiner used, that is, the linear combiner of any one of the filters A20 and A100 is an adder, and the linear combiner of the other of the filters A20 and A100 is added. Only in the case of the subtractor, it is different from the NTSC-removing comb filter A20. This is because the NTSC-rejection comb filter A100 receives the baseband image, but the artifact of the NTSC image frequency of the DTV signal supplied to the NTSC-rejection comb filter A20 exists in the baseband for the image frequency. . For similar reasons, the NTSC-rejection comb filter B100 differs from the second type of NTSC-rejection comb filter B20 only in the type of linear combiner used, and the NTSC-rejection comb filter C100 is used linearly. Only the type of coupler is different from the third type NTSC-removing comb filter (C20). Each response of the NTSC-rejection comb filters A100, B100, C100 is squared by squarers A102, B102, C102, respectively, to determine the energy of these responses. The response of the LPF 54 is squared by the squarer 104 to determine its energy.

FIG. 15B shows a configuration of a circuit in which the multiplexer 1261 and the three-state data buffer 82 of FIG. 8 are replaced with four three-state data buffers 082, A82, B82, and C82. will be. The tri-state data buffer 082 is used to selectively verify the intermediate symbol decoding result generated from the data slicer 22 on the 3-bit-wide output data bus 800 of the symbol decoding selection circuit 90. . The three tri-state data buffers A82, B82, and C82 are used to selectively verify the postcoding symbol decoding result generated from the postcoding comb filters A26, B26 and C26 on the data bus 800. do.

When the response of the NOR gate 92 is 1, which of the three three-state data buffers A82, B82, and C82, rather than the three-state data buffer 082, will be adjusted to provide a low source impedance. To determine, it is determined which of the responses of the NTSC-rejection comb filters A100, B100, C100 has substantially less energy than the response of the LPF 54. If such a determination is made, when the response of the NOR gate 92 is 1, to determine which of the three-state data buffers 082, A82, B82, C82 will be adjusted to provide a low source impedance, It is determined whether which of the responses of the NTSC-removing comb filters A100, B100, C100 has a minimum residual energy therein. For this purpose, the responses of squarers 104, A102 are compared by comparator 106, the responses of squarers 104, B102 are compared by comparator 108, and squarers 104. The response of C102 is compared by comparator 110, the responses of squarers A102, B102 are compared by comparator 112, and the responses of squarers A102, C102 are comparator 114. Are compared, and the responses of squarers B102 and C102 are compared by comparator 116.

The three-input NOR gate 118 does not respond to any of the comparators 106, 108, 110, which means that the squarer 104 is responsible for any of the responses of the squarers A102, B102, C102 to supply the output signal 1. One response is exceeded, otherwise the output signal of the NOR gate 118 is zero. The two-input AND gate 120 adjusts the three-state data buffer 082 to provide a low source impedance when the response of the NOR gate 92 is one and at the same time the response of the NOR gate 118 is one. Supply 1 response.

The three-input AND gate 122 supplies one output signal responsive to the output 1 of the comparator 106, which indicates that the response of the squarer A102 has less energy than the response of the squarer 104. At the same time, the two complementary outputs of comparators 112 and 114 are equal to one, which indicates that squarer 104 has a response whose energy is not greater than that of squarers B102 and C102, if not. The output signal of the AND gate 122 becomes zero. The two-input AND gate 124 adjusts the three-state data buffer A82 to provide a low source impedance when the response of the NOR gate 92 is one and at the same time the response of the AND gate 122 is one. Supply 1 response.

The three-input AND gate 126 supplies one output signal responsive to the complementary output 1 of the comparator 116, which indicates that the response of squarer B102 has not more energy than the response of squarer C102. At the same time, the two outputs of the comparators 108 and 112 become 1, which indicates that the response of the squarer B102 has less energy than the response of the squarers 104, A102, and if not The output signal of the gate 126 is zero. The two-input AND gate 128 adjusts the three-state data buffer A82 to provide a low source impedance when the response of the NOR gate 92 is one and at the same time the response of the AND gate 126 is one. Supply 1 response.

The three-input AND gate 130 supplies one output signal when the outputs of the comparators 110, 114, and 116 are all one, which means that the response of the squarer C102 is greater than the response of the squarers 104, A102, B102. Indicates low energy, and if not, the output signal of AND gate 130 is zero. The two-input AND gate 132 adjusts the three-state data buffer C82 to provide a low source impedance when the response of the NOR gate 92 is one and at the same time the response of the AND gate 130 is one. Supply 1 response.

Referring back to FIG. 14, the circuitry of the NTSC-removing comb filter A20 and the post-coding comb filter A26 described above is based on the NTSC-removing comb filter 520 and the post-coding comb filter of FIG. 13. It is advantageously selected to be of a type similar to the circuit of 526. This is true despite the significant cost in memory, since 718,200 symbols must be stored in each of the two-picture-frame delayers 5201 and 5263 described above. ) Provides the storage required by the co-channel interference detector A44 of FIG. Moreover, the same memory can be used to realize the short delays of the short delay delays 4201, 3201, 2201, 1201 and other co-channel interference detectors of FIG. 15. The storage of the two-picture-frame delayer 5203 also provides the storage required by the short-term delayers 4403,3262, 2263, 1263.

The high-energy demodulation artifacts generated in response to analog TV synchronization pulses, equalization pulses and color bursts are all suppressed when the NTSC-rejection comb filter A20 adds and combines alternating picture frames. In addition, artifacts resulting from the video content of an analog TV signal that does not change over two frames are suppressed, so that a still pattern is removed regardless of its spatial frequency or color. When the NTSC-removing comb filter A20 adds and combines alternating image frames, the NTSC-rejection comb filter A100 of FIG. 15B differentially combines these alternating image frames, and the squarer A102 described above. And detectors for detecting changes between alternating frames of NTSC co-channel interference.

The remaining problem of suppressing demodulation artifacts is mainly related to the suppression of these demodulation artifacts resulting from the frame-to-frame difference of certain pixel positions in the analog TV signal raster. These demodulation artifacts can be suppressed by intraframe filtering techniques. The circuit of the NTSC-removing comb filter B20 and the post-coding comb filter B26 can be selected to suppress residual demodulation artifacts depending on the correlation in the horizontal direction, and the NTSC-rejection comb filter C20 And the circuit of the postcoding comb filter C26 may be selected to suppress residual demodulation artifacts depending on the correlation in the vertical direction. Consider how such design decisions are made.

If the voice carrier of the co-channel interfering analog TV signal is not suppressed by voice trap or SAW filtering of the IF amplifier chain 12 described above, the NTSC-rejection comb filter B20 and the post-coding comb described above. The circuit of filter B26 is advantageously selected to be of a type similar to the circuit of the NTSC-removing comb filter 120 and the post-coding comb filter 126 described above in FIG. 9. If the voice carrier of the co-channel interfering analog TV signal is suppressed by the voice trap or SAW filtering of the IF amplifier chain 12 described above, the NTSC-removing comb filter B20 and the post-coding comb filter B26 described above. The circuit of is advantageously selected to be of a type similar to the circuit of the NTSC-removing comb filter 220 and the postcoding comb filter 226 described above in FIG. The reason is that the anti-correlation between video components spaced apart every 6 symbol periods is better than the correlation between video components spaced apart every 12 symbol periods.

The optimal selection of the circuitry of the NTSC-removing comb filter C20 and the post-coding comb filter C26 described above takes into account the Due to the decision of whether to select the spatially proximal scan line of the previous feed or to temporarily select the proximate scan line of the same feed to be combined with the current scan line, it is not simple. Choosing the same near-temporal near scan line may generally be the best choice, since jump cuts between the feeds may adversely affect NTSC removal by the comb filter C20. Because this is less. As a result of such a selection, the circuitry of NTSC-removing comb filter C20 and postcoding comb filter C26 is of a type similar to the circuit of NTSC-removing comb filter 320 and postcoding comb filter 326 of FIG. It is composed. When the NTSC-removing comb filter C20 adds and combines alternating image scanning lines, the NTSC-removing comb filter C100 of FIG. 15B differentially combines these alternating image scanning lines, and squarer C102 and Together, a detector is provided for detecting changes between alternating scan lines of NTSC co-channel interference.

On the other hand, as a result of the above selection and other selections, the circuitry of the NTSC-removing comb filter C20 and the post-coding comb filter C26 is the circuit of the NTSC-removing comb filter 420 and the post-coding comb filter 426 of FIG. It is composed of a similar type. The NTSC-rejection comb filter C20 and squarer C102 together provide a detector for detecting changes between the feeds of NTSC co-channel interference.

The digital TV receiver device of FIG. 14 further comprises additional parallel data-slicing performed respectively through the serial connection of each NTSC-rejection comb filter, each good-level data slicer subsequently connected thereto, and each post-coding comb filter subsequently connected thereto. It is modified as another embodiment of the present invention for using the operation. While two additional parallel data-slicing operations are shown in FIG. 14, a variant to use another additional parallel data-slicing operation may provide the ability to establish the best estimate of the correct symbol decoding result. Can be.

The trellis decoder circuit 34 can be repeated and the relative success of the various symbol decoding decisions can be compared to further establish the best estimate of the symbol decoding result. Nevertheless, this means that the number of digital hardware is quite large.

Co-channel interference by analog TV signals of standards other than NTSC, such as the Phase Alternation by Line (PAL) standard, can be caused by DTV systems that modify the DTV system used for terrestrial broadcasting in the United States. The present invention can be easily modified in terms of design only to accommodate such co-channel interference.

As described above, according to the present invention, before performing the data-slicing process by the second data slicer, a symbol re-encoding process of the first type is performed by a first comb filter called a precoding process. Re-signing a symbol of the first type by performing a second type symbol re-encoding process, called post-coding, after data-slicing by a second comb filter comprising a bilinear combiner, a multi-input multiplexer circuit and a second delay unit. The coding process can be compensated and an accurate symbol decoding result can be generated. The DTV signal receiver embodying the present invention can suppress co-channel interference using additional filtering.

On the other hand, those skilled in the design of digital communication receivers and who are familiar with the present specification and the accompanying drawings will be able to design various embodiments of the present invention other than those described above. And while the present invention has been described with reference to specific embodiments, the disclosure of the present invention is merely an application of the present invention, and is limited to the specific embodiments disclosed herein as the best mode for carrying out the present invention. It is not. In addition, it will be readily understood by those skilled in the art that the present invention may be variously modified and changed without departing from the spirit or the field of the present invention provided by the following claims. .

Claims (15)

In a digital television signal receiver, A digital television signal detection apparatus for supplying 2N-level symbol streams each having a symbol period of a specific time length and sensitive to carrying artifacts of a co-channel interfering analog television signal; A trellis decoder for trellis decoding interleaved trellis coded data; And a symbol decoding device for symbol decoding the 2N-level symbol stream to supply the interleaved trellis coded data to the trellis decoder, The symbol decoding apparatus, A first data slicer for decoding the 2N-level symbol stream to produce an intermediate symbol decoding result; In response to the 2N-level symbol stream having a first delay stream of the 2N-level symbol, representing a delay of a defined first number of symbol periods, to generate a first differential delay stream pair of the 2N-level symbol; A first delay connected to the Received as first and second input signals to generate as output signals a first stream of (4N-1) -level symbols that provide a first comb filter response where artifacts of the co-channel interfering analog television signal are suppressed A first linear combiner that linearly combines the first differential delay stream pair sensitive to carrying artifacts of the co-channel interfering analog television signal; Linearly combine the received first and second input signals to supply respective output signals, a first comb filter response is applied to the second linear combiner as the first input signal, and the first linear combiner and the A second linear coupler, wherein one of the second linear couplers is an adder and the other is a subtractor; A first of the (4N-1) -level symbols supplied as respective output signals from the first linear combiner to generate a first complementary symbol decoding result applied to the second linear combiner as the first input signal; A second data slicer for decoding the stream; Reproduces one of a plurality of input signals selected in response to a multiplexer control signal as an output signal, receives the intermediate symbol decoding result as one of the plurality of input signals, and outputs an output signal of the second linear combiner among the plurality of input signals A plurality of input multiplexer circuits to receive as another one; Coupled to delay the output signals of the plurality of input multiplexer circuits used for a minimum time as a result of the last symbol decoding by the first number of the prescribed symbol period to generate the second input signal of the second linear combiner. And a second delay unit. The digital television signal receiver as claimed in claim 1, wherein the first linear combiner is a subtractor and the second linear combiner is a modulo-2N adder. The digital television signal receiver as claimed in claim 2, wherein the first number of the prescribed symbol periods is twelve. The digital television signal receiver as claimed in claim 1, wherein the first linear combiner is an adder and the second linear combiner is a modulo-2N subtractor. The digital television signal receiver as claimed in claim 4, wherein the first number of the prescribed symbol periods is six. 5. A digital television signal receiver as claimed in claim 4, wherein the first number of defined symbol periods is substantially equal to the number of symbol periods of two horizontal scanning lines of said co-channel interfering analog television signal. 7. The digital television signal receiver as claimed in claim 6, wherein the first number of the prescribed symbol periods is 1361. 5. A digital television signal receiver as claimed in claim 4, wherein the first number of defined symbol periods is substantially equal to the number of symbol periods of 62 horizontal scanning lines of said co-channel interfering analog television signal. 9. The digital television signal receiver as claimed in claim 8, wherein the first number of the prescribed symbol periods is 179,208. 5. A digital television signal receiver as claimed in claim 4, wherein the first number of defined symbol periods is substantially equal to the number of symbol periods of two video frames of said co-channel interfering analog television signal. 11. A digital television signal receiver as claimed in claim 10, wherein the first number of defined symbol periods is 718,200. The method of claim 1, The 2N-level symbol stream supplied from the digital television signal detection device interferes with the first data slicer which can be suppressed in the first comb filter response and decode the 2N-level symbol stream without inaccurate error. Further comprising a co-channel interference detector connected to generate an indication signal at a time that involves the artifact of the co-channel interfering analog television signal having a sufficiently strong significant energy level, The display signal is applied to the multi-input multiplexer circuit as the minimum portion of the multiplexer control signal, and the multi-input multiplexer circuit is configured to store the 2N-level symbol stream supplied from the digital television signal detection apparatus in the first comb. The intermediate symbol decoding result is output to the final symbol only if the co-channel interference detector does not generate a current indication signal that can be suppressed in a filter response and involves an artifact of the co-channel interfering analog television signal having the significant energy level. Digital television signal receiver characterized in that it is selected as a result of decoding. The method of claim 12, Respond to the 2N-level symbol stream having a second delay stream of the 2N-level symbol, indicating a delay of a defined second number of symbol periods, to generate a first differential delay stream pair of the 2N-level symbol. With a third delay connected to Received as first and second input signals to generate a second stream of (4N-1) -level symbols as an output signal providing a second comb filter response where artifacts of the co-channel interfering analog television signal are suppressed A third linear combiner for linearly combining said second differential delay stream pair sensitive to carrying artifacts of said co-channel interfering analog television signal; A fourth linear combiner for linearly combining the received first and second input signals for supplying respective output signals, which are applied as another input signal to the multi-input multiplexer circuit, wherein the third linear combiner And a fourth linear coupler, wherein one of the fourth linear couplers is an adder and the other is a subtractor; A second of the (4N-1) -level symbols supplied as respective output signals from the third linear combiner to generate a second complementary symbol decoding result applied to the fourth linear combiner as the first input signal; A third data slicer for decoding the stream, And a fourth delay unit coupled to delay the output signal of the multi-input multiplexer circuit by a second number of the prescribed symbol period to generate the second input signal of the fourth linear combiner. Digital television signal receiver. The method of claim 1, Respond to the 2N-level symbol stream having a second delay stream of the 2N-level symbol, indicating a delay of a defined second number of symbol periods, to generate a first differential delay stream pair of the 2N-level symbol. With a third delay connected to Received as first and second input signals to generate a second stream of (4N-1) -level symbols as an output signal providing a second comb filter response where artifacts of the co-channel interfering analog television signal are suppressed A third linear combiner for linearly combining said second differential delay stream pair sensitive to carrying artifacts of said co-channel interfering analog television signal; A fourth linear combiner for linearly combining the received first and second input signals for supplying respective output signals, which are applied as another input signal to the multi-input multiplexer circuit, wherein the third linear combiner And a fourth linear coupler, wherein one of the fourth linear couplers is an adder and the other is a subtractor; A second of the (4N-1) -level symbols supplied as respective output signals from the third linear combiner to generate a second complementary symbol decoding result applied to the fourth linear combiner as the first input signal; A third data slicer for decoding the stream, And a fourth delay unit coupled to delay the output signal of the multi-input multiplexer circuit by a second number of the prescribed symbol period to generate the second input signal of the fourth linear combiner. Digital television signal receiver. The method of claim 1, Determine an amount of co-channel interference energy of the intermediate symbol decoding result, the first complementary symbol decoding result and the second complementary symbol decoding result, and generate the multiplexer control signal based on the amount of the co-channel interference energy Further comprising a co-channel interference detection circuit, The multi-input multiplexer circuitry allows the 2N-level symbol stream supplied from the digital television signal detection apparatus to be suppressed in either of the first and third comb filter responses and the 2N-level without inaccurate error. Only if the co-channel interference detection circuit does not generate a current indication signal that involves an artifact of a co-channel interfering analog television signal having a significant energy level strong enough to interfere with the first data slicer decoding the symbol stream. Select the intermediate symbol decoding result as the final symbol decoding result, and if not, out of the first complementary symbol decoding result and the second complementary symbol decoding result having a greater amount of the co-channel interference energy A digital television signal receiver, characterized in that one is selected.