Pulse or digital communications – Receivers – Interference or noise reduction
Reexamination Certificate
2000-03-08
2004-02-03
Chin, Stephen (Department: 2634)
Pulse or digital communications
Receivers
Interference or noise reduction
C375S346000
Reexamination Certificate
active
06687317
ABSTRACT:
BACKGROUND OF THE INVENTION
In such high data rate AM digital transmissions, the intersymbol interference (ISI) caused by echoes occurring in the transmission path can sometimes be so high that the forward error correction codes (FEC) which are included within the transmitted digital data streams can no longer protect the integrity of the digital information. In the case of digital television, ISI causes a loss of FEC information which in turn results in a complete collapse of the channel and complete loss of the video and/or audio. As a result, TV broadcasters and manufacturers are justifiably concerned that the 8 VSB signals will not be robust enough to overcome the ISI (also known and alternatively referred to herein as “echoes” and “multipath interference”) which will occur for example in a city environment where there are many tall buildings which can cause such interference. Some industry researchers have suggested that the 8 VSB modulation system has been shown under actual use conditions, to be particularly susceptible to echo interference. As a result, interested parties have suggested changing the modulation system to one which might arguably be more resistant to ISI.
Advocates of ATSC modulation approach argue that alternative methods of digital modulation do not demonstrate superior resistance to ISI. More importantly, many broadcasters have already invested money in building transmitter facilities for the 8 VSB digital transmission, and television receivers are already being designed, and sold, which are ATSC (8 VSB) compliant. Therefore, for practical reasons, the manufacturers of digital television receivers will have to find a solution to the ISI problem by improving the performance of 8 VSB receivers in the face of ISI. One way of doing this is through the use of improved channel equalization at the receiver.
Adaptive filters are well known in the art. The coefficients of such filters are adjusted in dependence upon the characteristics of the received signal. This filtering is also termed equalization, because it has the effect of reducing, or equalizing, the effects of those environmental sources which caused the observed errors. After the adjustment, or training, of the receiver, the processing of message symbols commences. The underlying assumption in this scenario is that the environmental conditions which caused detected differences in the received training symbols as compared to the transmitted training symbols, would affect the subsequent received message symbols as well, and, therefore, an adjustment to the filters which minimized errors in the received training symbols would also minimize errors in the received message symbols.
In order for a receiver to perform the necessary channel equalization to minimize or eliminate the ISI caused by echoes, the time domain processing technique of correlation can be used to locate the presence of echoes in time. Once the echoes are so located, an echo canceling digital filter or equalizer (EQ) having the necessary high performance to suppress the echoes may be designed with proper hardware and algorithms. But the performance and the hardware cost becomes a measure of efficiency. Generally the time range of the echo-spread can be very large, for example from about −3 microseconds to about +40 microseconds in terrestrial broadcasting/communications, and sometimes it can even have the range of from about −20 microseconds to about +60 microseconds and beyond, and therefore it is not, very practical or efficient to provide static multiplier branches (also called taps) in the EQ to cover all of the possible echo time positions.
Co-pending U.S. patent application Ser. No. 08/880,378 describes an example of an adaptive digital filter which features movable clustered taps. A number of taps are combined together to form a cluster in the equalizer and those clusters can be flexibly moved to any time/delay location to perform echo cancellation when needed. This type of movable cluster tap equalizer can greatly reduce or cancel the echoes on demand if the clusters can be positioned on the time axis precisely. The present application and the '378 application are owned by a common Assignee and the '378 application is specifically incorporated by reference herein. Details of about the ATSC signal structure can be found in the document “ATSC Digital Television Standard” (Doc. A/53) which is incorporated by reference herein.
SUMMARY OF THE INVENTION
Current methods for locating echoes in time have not demonstrated the necessary sensitivity to effectively deal with the presence of mid-range and long-range echoes. Without an effective way to locate all of the source of the echoes, equalization to mitigate ISI will be insufficient to adequately improve the performance of digital television receivers in high echo environments. The instant invention comprises an improved method for locating echoes present in a digital data transmission. The method and apparatus are especially effective for locating in time, mid-range and long-range echoes. The invention also comprises a novel correlation process to precisely find the cluster location information needed to effectively position adjustable taps of an adaptive filter in the time axis. In a preferred embodiment the invention comprises a method for substantially reducing, at a receiver, inter-symbol interference (ISI) caused by multipath in a digital television signal.
The ATSC signal is received at a digital television receiver containing a demodulator, an adaptive filter with moveable tap clusters and a decoder. The receiver separates from the ATSC signal a portion of the signal which is then correlated with a first vector of data samples (Kernel
1
) as listed, for example, in Appendix A and shown graphically, for example, in
FIG. 3
, to obtain a correlated signal as shown graphically in
FIG. 4. A
second vector of data samples (Kernel
2
), as listed for example, in Appendix B and shown graphically for example, in
FIG. 5
, is then added to the correlated signal to obtain a correction signal, as shown in
FIG. 6
, which provides information about multipath, for example, the location in time of detected echoes.
Using the echo location information provided by the correction signal, the moveable filter tap clusters of the EQ are positioned to the appropriate locations in time of the most significant echoes thereby compensating for the ISI interference caused by the echoes to achieve substantial echo reduction or elimination and preserving of the integrity of the data transmission.
REFERENCES:
patent: 5371471 (1994-12-01), Chennakeshu et al.
patent: 5687198 (1997-11-01), Sexton et al.
patent: 6014407 (2000-01-01), Hunsinger et al.
patent: 6233295 (2001-05-01), Wang
patent: 6366629 (2002-04-01), Chen et al.
patent: 0425458 (1991-05-01), None
patent: 0 425 458 (1991-05-01), None
patent: 9844655 (1998-10-01), None
“ATSC 8-VSB demodulator and decoder” Philips Data Sheet TDA 8960, Jun. 14, 1999.
“ATSC Digital Television Standard” Advanced Television Systems Committe, Doc. A/53. Apr. 12, 1995 Sep. 16, 1995.
Chin Stephen
Williams Lawrence
LandOfFree
Method for correlation directed echo cancellation and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for correlation directed echo cancellation and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for correlation directed echo cancellation and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3277151