Pulse or digital communications – Receivers – Amplitude modulation
Reexamination Certificate
1999-09-15
2003-01-14
Le, Amanda T. (Department: 2734)
Pulse or digital communications
Receivers
Amplitude modulation
C375S355000, C438S537000, C438S726000, C438S735000, C329S357000, C329S361000
Reexamination Certificate
active
06507626
ABSTRACT:
The invention relates to digital television receivers for vestigial-sideband (VSB) digital television (DTV) signals and, more particularly, to the portions of such receivers used for recovering baseband symbol coding proceeding from intermediate-frequency signals.
BACKGROUND OF THE INVENTION
The digitization of intermediate-frequency VSB DTV signal and its subsequent demodulation in the digital regime are described in U.S. Pat. No. 5,479,449. This patent entitled “DIGITAL VSB DETECTOR WITH BANDPASS PHASE TRACKER, AS FOR INCLUSION IN AN HDTV RECEIVER” issued Dec. 26, 1995 to C. B. Patel and A. L. R. Limberg. Demodulation in the digital regime is performed in U.S. Pat. No. 5,479,449 by converting the digitized intermediate-frequency VSB DTV signal to complex form to be multiplied in a complex digital multiplier by a complex digital carrier signal supplied from look-up tables stored in read-only memory (ROM). To facilitate converting the digitized I-F signal to complex form using a digital Hilbert transform filter, the final intermediate-frequency band is offset a megahertz (MHz) or so from zero frequency, but its uppermost frequency is kept lower than 10 MHz.
Equalization of the digitized baseband symbol coding that results from demodulation is facilitated by choosing a sampling clock of a rate that is a rational multiple of the symbol rate (i. e., is related to symbol rate by a whole number ratio) and that will satisfy the Nyquist criterion. Supplying the complex digital carrier signal from ROM is facilitated by choosing the carrier in the digitized I-F signal to be a submultiple of the system clock signal rate as described by C. B. Patel and A. L. R. Limberg in U.S. Pat. No. 5,606,579 issued Feb. 25, 1997 and entitled “DIGITAL VSB DETECTOR WITH FINAL I-F CARRIER AT SUBMULTIPLE OF SYMBOL RATE, AS FOR HDTV RECEIVER”. Such choice of carrier permits perfect wrap-around of cycles of digital carrier when they are conceived as being mapped to the surface of a cylinder with circumference measured by ROM addresses according to a modular arithmetic.
C. B. Patel and A. L. R. Limberg advocate the digital carrier being located at the upper-frequency end of the final I-F signal band in U.S. Pat. No. 5,731,848 issued Mar. 24, 1998 and entitled “DIGITAL VSB DETECTOR WITH BANDPASS PHASE TRACKER USING NG FILTERS, AS FOR USE IN AN HDTV RECEIVER”. That is, the vestigial sideband is above full sideband in frequency in the final I-F signal that is digitized. U.S. Pat. No. 5,731,848 discloses there is reason for this choice of carrier, aside from facilitating the use of Ng filters for converting the real final I-F signal to complex form after its digitization. Fast changes in symbol values are converted to lower-frequency variations in the final I-F signal offered for digitization, which alleviates problems of accurately sampling the final I-F signal as the initial step in digitization. Small changes in sampling phase result in larger changes in the zero-frequency demodulated carrier, so there is tighter automatic frequency and phase control (AFPC) of a local oscillator used in converting the radio-frequency (R-F) VSB DTV signal to the final I-F signal.
The equalization of quadrature amplitude-modulation (QAM) digital signals had been a topic of extensive study before digital television broadcasting standards were being formulated in the early 1990's. The problems of carrier acquisition and of symbol synchronization were separately considered. If synchrodyning is performed in the analog regime and symbol synchronization is performed in the digital regime, this is a natural dichotomy. If the carrier frequency is not locked to symbol rate, this is a natural dichotomy as well. One general approach to receiver design was to demodulate two mutually orthogonal phases of received QAM using a beat frequency oscillator with highly stabilized frequency and with symbol synchronization being performed after demodulation using adaptive equalization techniques based on errors detected from decision results.
An alternative general approach to receiver design was to demodulate two mutually orthogonal phases of received QAM using an oscillator subject to automatic frequency and phase control based on errors detected from decision results. This latter approach was adapted in the Patel et alii patents described above, the phase of digital carrier being adjusted by a method similar to that described by S. U. H. Qureshi in his paper “Timing Recovery For Equalized Partial-Response Systems”, IEEE Transactions On Communications, December 1976, pp.1326-1330.
When automatic frequency and phase control (AFPC) of the local oscillator is based on errors detected from decision results, passband equalization, rather than baseband equalization, is preferred by those skilled in the art of digital communications receiver design. Baseband equalization is carried out after demodulation, so the equalization filters are included in the feedback loop used for AFPC of the local oscillator based on errors detected from decision results. Changes in the filter coefficients of an adaptive equalization filter affect loop delay and can cause jitter in carrier synchronization unless the loop bandwidth is kept narrow. Narrow loop bandwidth compromises the ability to acquire carrier rapidly when first tuning to a transmission channel and the ability to synchronize to shifts in carrier phasing during multipath reception. Passband equalization is carried out previous to demodulation, which avoids the equalization filters being included in the feedback loop used for AFPC of the local oscillator based on errors detected from decision results. A problem with passband equalization is encountered when adaptive equalization filters are used. The error signal based on errors detected from decision results must be modulated onto the digital carrier to convert it from the lowpass filter regime into the bandpass filter regime in which filter coefficient adjustments are to be made. This problem is merely an inconvenience with QAM, which has a symmetrical passband structure. But the problem is more difficult with VSB, which has an asymmetrical passband structure.
The extraction from ROM of digital carrier of a frequency that is a submultiple of symbol rate or is a submultiple of symbol rate of a multiple of symbol rate ties together the problems of carrier synchronization and symbol synchronization in a way that previously has not been fully appreciated. The digital carrier is sampled in a prescribed way in the ROM, and the AFPC of the local oscillator so the final I-F signal can be synchrodyned to baseband automatically controls the timing of sampling vis-à-vis the carrier of the analog final I-F signal. So, adjustment of symbol synchronization after the digital complex demodulation of baseband symbol coding is not required, there being no need to overcome randomness of sampling phase versus the phasing of the carrier used in the synchrodyning procedure. However, symbol synchronization may be imperfect because of lack of uniform group delay across the receiver passband, which will be the case when multipath reception occurs. Any adjustment of symbol synchronization that is actually required in practice can be taken care of by the equalization filtering. In order to keep equalization filtering from being included in the AFPC loop in the bandpass tracker, baseband equalization is employed in the invention described below, rather than passband equalization being employed.
There is still need for stabilization of sample clock rate to lock its phase to symbol rate or a multiple thereof. This is required for maintaining perfect wrap-around of cycles of digital carrier when conceptually they are mapped to the surface of a cylinder with circumference measured by ROM addresses sequentially addressed at sample clock rate. Sample clock rate stabilization can be done by extracting half-symbol-frequency component from symbol coding and multiplying up in frequency to generate sampling frequency. Such procedures for stabilization of sample clock rate are per se already
Le Amanda T.
Samsung Electronics Co,. Ltd.
Sughrue & Mion, PLLC
LandOfFree
Bandpass phase tracker that automatically samples at... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Bandpass phase tracker that automatically samples at..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Bandpass phase tracker that automatically samples at... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3001308