CA2111170C - Technique for determining signal dispersion characteristics in communications systems - Google Patents
Technique for determining signal dispersion characteristics in communications systemsInfo
- Publication number
- CA2111170C CA2111170C CA 2111170 CA2111170A CA2111170C CA 2111170 C CA2111170 C CA 2111170C CA 2111170 CA2111170 CA 2111170 CA 2111170 A CA2111170 A CA 2111170A CA 2111170 C CA2111170 C CA 2111170C
- Authority
- CA
- Canada
- Prior art keywords
- signal
- sub
- correlators
- training sequence
- received
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 239000006185 dispersion Substances 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000004891 communication Methods 0.000 title claims abstract description 13
- 125000004122 cyclic group Chemical group 0.000 claims abstract description 28
- 230000002596 correlated effect Effects 0.000 claims abstract description 19
- 230000008878 coupling Effects 0.000 claims abstract description 8
- 238000010168 coupling process Methods 0.000 claims abstract description 8
- 238000005859 coupling reaction Methods 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 105
- 238000012545 processing Methods 0.000 claims description 16
- 230000010076 replication Effects 0.000 claims description 9
- 230000000875 corresponding effect Effects 0.000 claims description 3
- 230000003362 replicative effect Effects 0.000 claims description 3
- 230000001276 controlling effect Effects 0.000 claims 1
- 239000002243 precursor Substances 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 150000002500 ions Chemical class 0.000 description 5
- GWUSZQUVEVMBPI-UHFFFAOYSA-N nimetazepam Chemical compound N=1CC(=O)N(C)C2=CC=C([N+]([O-])=O)C=C2C=1C1=CC=CC=C1 GWUSZQUVEVMBPI-UHFFFAOYSA-N 0.000 description 4
- 230000003111 delayed effect Effects 0.000 description 3
- 238000010420 art technique Methods 0.000 description 2
- 230000000903 blocking effect Effects 0.000 description 2
- 230000001934 delay Effects 0.000 description 2
- 238000002592 echocardiography Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- YFONKFDEZLYQDH-OPQQBVKSSA-N N-[(1R,2S)-2,6-dimethyindan-1-yl]-6-[(1R)-1-fluoroethyl]-1,3,5-triazine-2,4-diamine Chemical compound C[C@@H](F)C1=NC(N)=NC(N[C@H]2C3=CC(C)=CC=C3C[C@@H]2C)=N1 YFONKFDEZLYQDH-OPQQBVKSSA-N 0.000 description 1
- 241000282320 Panthera leo Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000006735 deficit Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010348 incorporation Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000010363 phase shift Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
- Time-Division Multiplex Systems (AREA)
Abstract
The present invention relates to an apparatus and a method for use in a communications system wherein a received signal is sampled. The apparatus is comprised of a unit for demultiplexing the received signal samples onto a plurality of output terminals in a cyclic fashion. A plurality of sub-correlators are provided, each connected to a different one of the output terminals and providing and output. A multiplexing unit are provided for multiplexing the outputs of the plurality of sub-correlators in the cyclic fashion. Finally, a unit is provided which is responsive to the multiplexed outputs of the plurality of correlators for determining at least one signal dispersion characteristic. The method, therefore is comprised of the steps of demultiplexing samples of a signal onto a plurality of output terminals in a cyclic fashion; coupling each output terminal to an associated correlator in a plurality of correlators, each correlator receiving the samples from a differentoutput terminal and providing a correlated output signal; and multiplexing the outputs of the plurality of correlators in the cyclic fashion. The method also includes the step of determining at least one signal dispersion characteristic to respond to the multiplexed outputs of the plurality of correlators.
Description
211117 ~
TECHNIQUE FOR DETERMINING SIGNAL DISPERSION
CIIARACTERISTICS IN CO~fMUNICATIONS SYSTEMS
Technical Field The present invention relates to a technique for determining the characteristics of signal dispersion and, more particularly, to a technique in which a training sequence is transmitted and processed at a receiver so as to create a time interval in which the char~sten~tics of the signal dispersion can be accurately quantified.
Background of the Invention A training sequence is a series of a priori known symbols which are tla~ls~ led at predetermined times from a signal llallsllliller to a signal receiver.
Such sequences have long been used to adjust the operation of signal dispersion compensation a~paralu;, in the receiver, such as equalizers, echo cancellers and the like.
Signal dispersion, such as echoes or signal "ghosts", is an inherent problem in communications systems and the severity of the problem can vary with the system application. For example, signal ghosts or echoes frequently exist in co,lventional television l,alls,llisi,ion systems which are highly objectional to the viewer and which can render a high definition television (HDTV) signal unintelligible. Accordingly, compensation for signal dispersion is highly desirable in conventional television systems and, indeed, is required in HDTV and in many other co,l,lllunications applications. Moreover, while signal dispersion compensation apparatus, such as equalizers and cancellers, provides satisfactorycompensation, knowledge of the signal dispersion characteristics, e.g., amplitude, delay and phase, is useful for the adjustment of such apparatus during system startup and operation.
Prior art techniques exist which can determine the characteristics of signal dispersion. Such techniques typically transmit specific signals, such as pulses or training sequences, which are detected and analyzed in the receiver. These techniques provide satisfactory estimates of the signal dispersion characteristics in applications wherein the dispersion is large in amplitude and short in duration.However, the prior art techniques provide inaccurate results when the signal 211117~
dispersion is small in amplitude, particularly in the presence of channel noise, jitter or similar illlpaillllents, and provide ambiguous results when the signal dispersion is long in duration. These shortcomings have hindered development of co~ ,ullications systems requiring ever-more precise signal dispersion S compensation and more precise determination of the signal dispersion characteristics .
Most recently, as disclosed in the prior art, a technique which overcomes the aforesaid problems of the prior art has been developed which transmits two different training sequences wherein each such training sequence includes a 10 different number of symbols. This technique, while providing greater accuracythan those of the previous prior art, requires a time duration for the tr~n~mis~ion of the two different training sequences which is not always available. Therefore, a technique which provides greater accuracy in the determination of small-amplitude and long-duration signal dispersion and which requires less time duration would be 15 desirable.
Summary of the Invention The present invention covers the notion of determining the amplitude, phase and delay of signal dispersion by tr~ ",i~ ,g a training sequence including a plurality of apriori known symbols and having a time duration at least equal to the 20 expected range of signal dispersion delay. At a receiver, the training sequence is recovered and then processed using correlation and signal replication techniques.
This processed training sequence is then utilized to determine at least one of the aforesaid signal dispersion characteristics.
A feature of the present invention is that it is applicable to a variety of 25 different colllnlullications systems and is particularly adaptable for television signal tr~n~mi~cion wherein the training sequence can be advantageously ll~ llPilled during the vertical blanking interval. Another feature of the present invention is that it significantly illlpro~es the accuracy of dete.,llillillg the characteristics of signal dispersion when such dispersion is of long duration, has a small amplitude, or 30 is accompanied by channel noise, jitter and similar impairments.
In accordance with one aspect of the invention there is provided an appard~us for use in a co~ r~tions system wherein a received signal is sampled, said ayyaldLus comprising: processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a 5 prede~ uled symbol rate, for processing said received training sequence using correlation and signal replication, the processing means including; means for cleml-ltirlexing the received signal samples onto a plurality of output terminals in a cyclic fashion; a plurality of sub-correlators each connected to a different one of said output terminals and providing an output; multiplexing means for multiplexing 10 the outputs of said plurality of sub-correlators in said cyclic fashion; and means resporlsive to the multiplexed outputs of said plurality of said sub-correlators for de~lll~inillg at least one signal dispersion characteristic.
In accordance with another aspect of the invention there is provided in a col,-"""-ir~tions system having a signal dispersion with an expected range of delay, 15 a method of correlating comprising the steps of: generating a training sequence, said training sequence having a duration at least equal to said range of delay;
dem--ltirlexing samples of a signal onto a plurality of output terminals in a cyclic fashion; coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a different 20 output terminal and providing a correlated output signal; multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion; and del~llllinillg at least one signal dispersion characteristic to respond to the multiplexed outputs of said plurality of sub-correlators.
In accordance with yet another aspect of the invention there is provided an 25 ayyalaLus for use in a comm--nir,~tions system wherein an information signal is IlA~ lecl at a prede~;lllli~ed rate, said information signal being received after propagation through a co",llllll~ir~tions link and sampled at an integer multiple of said predetermined rate, said appardLus comprising: processing means responsive to a received training sequence including a plurality of predeLellllilled symbols 30 occurring at a predetermined symbol rate, for processing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the samples of the received il~o~ d~ion signal onto a plurality of terminals in a cyclic fashion; a plurality of sub-correlators each - 3a -connected to a dirrelent one of said terminals and providing a correlated output;
and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
In accordance with still yet another aspect of the present invention there is S provided a method of correlating in a co""",l,-ir~tions system wherein an h~lllla~ion signal is L~ d at a predetermined rate, said information signal being received after propagation through a colllllllll~ir~tions link and sampled at an integer multiple of said prede~llllilled rate, said method comprising the steps of:
geneldLillg a training sequence, said training sequence having a duration at least 10 equal to an expected range of delay of signal dispersion; deml~ltiplexing thesamples of the received information signal onto a plurality of output terminals in a cyclic fashion; coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a dirÇ~lelll output terminal and providing a correlated output signal; and multiplexing 15 the outputs of said plurality of sub-correlators in said cyclic fashion.
In accordance with still yet another aspect of the present invention there is provided an al)pdld~us for use in a col~ lll-icdlions system wherein a received signal is sampled, said apl)ald~us comprising: processing means responsive to a received training sequence including a plurality of predetermined symbols occurring 20 at a predeL~lll~illed symbol rate, for processing said received training sequence using correlation and signal replication, the processing means including; means for demultiplexing the received signal samples onto a plurality of terminals in a cyclic fashion; a plurality of sub-correlators, each receiving a common sequence and received signal samples from a dirr~lenl terminal, each sub-correlator correlating 25 its received sample using said common sequence to provide a correlated output;
and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
In accordance with still yet another aspect of the present invention there is provided a method of correlating in a collllllll~ir~tions system wherein a received 30 signal is sampled, said method comprising the steps of: generating a trainingsequence, said training sequence having a duration at least equal to an expectedrange of delay of signal dispersion; demultiplexing said received signal samples onto a plurality of output terminals in a cyclic fashion; coupling the received signal - 3b -samples on each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator being supplied with a common sequence and correlating its received signal samples with a said common sequence to provide acorrelated output; and multiplexing the outputs of said plurality of sub-correlators 5 in said cyclic fashion.
Brief Description of the Drawin~s FIGs. 1 and 2 are representations of an illustrative signal format wherein a training sequence is Ll~n.~llliLIed in accordance with the present invention;
FIG. 3 is a representation of illustrative signals, including signals which 10 precede and succeed the Ll~lsll~ission of the training sequence of FIG. 1, utilized by the disclosed embodiment of the present invention;
FIG. 4 is a block-schematic diagram ~epresellL~lion of an embodiment of a tr~n.cmitt~r which incorporates the present invention;
FIGs. S and 6 are alternate block-schematic diagram representations of 15 embodiments of a baseband receiver which incorporates the present invention;
FIG. 7 is a representation of an illustrative output signal of the inphase correlator 515 of FIG. 5;
FIGs. 8 and 9 are alternate block-schematic diagram representations of embodiments of a passband receiver which incorporates the present invention;
FIG. 10 is a block-schematic diagram represent~tion of an embodiment of the correlator in FlGs. 5, 6, 8 or 9;
FIG. 11 is a block-schematic diagram representation of an alternate embodiment of the correlator in FlGs. 5, 6, 8 or 9;
FIG. 12 is a block-schematic diagram representation of an alternate embodiment of a ~ s,l.iL~er which incorporates the present invention; and FIG. 13 and 14 are each a block-schematic diagram representation of an alternate embodiment of a rece;vel for use with the llal~ Ler of FIG. 11.
Detailed Description FIG. 1 shows a portion 100 of an illu~ ivt; enhanced National Television System Committee (NTSC) television signal format which incorporates a training sequence 101 during time interval 102 pursuant to the present invention. At thisjuncture, it will, of course, be understood that the present invention is not restricted to use in such a television signal format and can be utilized with any television signal format, e.g., Sequence Electronique Couleur Avec Memoire (SECAM), Phase Alternation Line (PAL), and HDTV, as well as signal formats for col,.lllunications systems other than television ~ ;c~;on. In this illustrative format, sequence 101 includes 255 symbols which are ll~ ."iL~ecl at a rate of 7.160 x 10C symbols/second so that time interval 102 has a duration of 35.6 x 10~
seconds (,us). Preferably, sequence 101 is a pseudorandom sequence. In addition,as shown, sequence 101 is advantageously ll;~ ed during the 63.56 ,us horizontal line period 103 which extends between holi~onl~l synchronization pulses 104 and 105. Horizontal period 103 is one of a plurality of line periods in a vertical blanking interval in a video frame. Incorporation of the training sequence in the unused time interval in horizontal line period 103 advantageously does not require any alteration of the television signals tr~n~mitte~l before and after sequence 101. Envelopes 106 and 107 represent the color bursts in an NTSC
format which respectively follow pulses 104 and 105. Within intervals 109 and 110, the signals in the NTSC format are known and provide well-known functions unconnected with the present invention. However, as will be discussed hereinbelow, the known nature of the signals in these intervals will be advantageously utilized by the present invention to permit the accurate replication of the ~lall~,lli~led training sequence 101 by a store-and-subtract process. Interval 109 must be at least as long as the m~illlulll delay of any postcursor ghosts and interval 110 must be at least as long as the system of the m~ill-ulll postcursorghost delay and precursor ghost delay. The former delay is typically represented by a positive time value and the latter delay is typically represented by a negative time S value. The sum of the m~xi,llulll postcursor ghost delay and precursor ghost delay time values, ignoring any algebraic sign, is hereinafter referred to as the expected range of signal dispersion delay.
FIG. 2 shows a portion 200 of the illu,Ll~tive NTSC television signal format wherein portion 200 succeeds portion 100 and includes a horizontal line period 201 within a vertical blanking interval. Horizontal line period 103 is disposed between horizontal synchlollizalion pulses 202 and 203. Envelopes 204 and 205 represent color bursts and are identical to envelopes 106 and 107 of FIG. 1. It should be noted that portions 100 and 200 are identical except that the training sequence 101 is not ll~l-c~--illed during horizontal line period 201.
FIG. 3 shows the periodic nature of the tr~n~mi~sion of training sequence 101 within eight portions 300-307 of an NTSC television signal format. Each of portions 300 through 307 has the same time duration and respectively includes one holi~onlal line period 308 through 315 in eight successive vertical blanking intervals. Each such holi~ollLal line period is disposed between a pair of horizontal synchloni~aLion pulses 104 and 105.
We shall define the term training sequence period as a time interval within one horizontal line period of each vertical blanking interval. During this training sequence period, either the training sequence 101 or a zero dc signal is "c,.,ille~ Accordingly, the d~cign~tions 308 through 315 in FIG. 3 ~le~ign~te this training sequence period.
It should be noted that in the NTSC signal format, shown in FIG. 3, the signal surrounding any training sequence period are different from those surrounding an immediately following training sequence period. For example, the signals in intervals 316 and 317, which immediately precede and succeed trainingsequence period 308, are different from those in intervals 318 and 319, which immediately precede and succeed training sequence period 309. Training sequence period 309 occurs during a vertical blanking interval immediately following the vertical blanking interval that includes training sequence period 308. However, the 2 ~ rl fJ
signals surrounding any given training sequence period are identical to those surrounding the fourth next training sequence period after the given training sequence period. In this regard, note that portions 300 and 304 are identical but for the training sequence 101. This identity also exists between portions 301 and 305, portions 302 and 306, and between portions 303 and 307. Accordingly, if thesignals in the foregoing paired signal portions were stored and then one stored signal in each pair was subtracted from the other, an accurate replica of the L~ ed training sequence could be recovered. For example, training sequence 101 could be recovered in the receiver by subtracting portion 304 from portion 300, subtracting portion 301 from portion 305, subtracting portion 306 from portion 302 and subtracting portion 303 from porion 307. It is this process of pairing the above-described signal portions, each including a training sequence period, which is used in a receiver constructed pu"uallt to the present invention.
Refer now to FIG. 4 which shows a ~ ,l"iUer 400 incorporating the present invention. The tr~n~mittec~ information signal, which in the illustrative embodiment is a conventional NTSC formatted signal, is generated by information source 401. This signal is then coupled to signal gateway 405 under the control of timing signals from timing generator 403. Training sequence generator 402 generates training sequence 101 within horizontal line periods 308, 310, 313 and315 under the control of timing signals from timing signal generator 403 and couples these sequences to signal gateway 405. Advantageously, each of these pseudorandom sequences is spectrally shaped by raised cosine transmit filter 404before being coupled to signal gateway 405. Gateway 405, illustratively replescnted by a single-pole, multi-throw switch, sclc.,Livt;ly couples the signals from source 401 or generator 402 to modulator 406 which provides vestigial sideband (VSB) amplitude modulation in the illustrative NTSC television application, or quadrature amplitude modulation (QAM) in other communication systems.
The ~lansuliL filter 404 advantageously includes tapped delay line 407, digital-to-analog (D/A) converter with a sample hold circuit 408 and an analog low-pass filter (LPF) 409 connected in series. Tapped delay line 407 and D/A
collvelLer 408 are clocked at four times the symbol rate (Fc). Analog LPF 409 advantageously elimin~tes out-of-band signal energy.
FIGs. S and 6 respectively show alternate embodiments of a baseband receiver 500 and 600 which incorporate the present invention. As shown in FIG. 5, the received modulated signal incorporating training sequence 101 is coupled via lead 501 through automatic gain control (AGC) circuit 502 to S demodulators 503 and 504. Each demodulator multiplies the received signal by carrier signals generated by carrier source 505. One of these carrier signals isrepresented by cos (<~)ct + ~e) while the other is represented by sin (~')ct + ~e)~
where ~c is the carrier frequency used in the L~ ll,iLler modulator of FIG. 4 and ~e is the phase angle between the carrier signals supplied by carrier source 504 and those utilized by the l-dlls.lliLLer modulator.
Each demodulator output is coupled through low-pass filters (LPF) 506 and 507 to analog-to-digital collvelL~ (A/D) 508 and 509. The digital outputs of these converts are supplied to a complex equalizer 510, which removes distortion in the received signal so as to accurately recover the Ll~ .lliL(e~i information signal and provide a pair of output signals 511 for further conventional television processing.
The term "complex", with reference to equalizer 510, refers to the fact that this equalizer utilizes coefficients which are complex numbers, i.e., they have a real and an im~gin~ry component. The digital outputs of converts 508 and 509 are also coupled to store-and-subtraction circuits 512 and 513.
The function of store-and-subtraction circuits 512 and 513 is to recover the training sequence 101 from the received signal by storing the paired waveforms of F~G. 3, i.e, 300 and 304, 301 and 305, 302 and 306, and 303 and 307, and then subtracting one waveform in each pair from the other. To extract the incoming signal waveform at the applop.iate time so as to recover each of these paired waveforms, circuits 512 and 513 are periodically enabled by timing circuitry 514.
The recovered training sequences rec~vered by circuits 512 and 513 are respectively de~ tecl as Sl and SQ.
Correlators 515 and 516 are designed to produce an output pulse upon ~letecting the training sequence. These output pulses, along with the correlatoroutput at other times, are replicated three times within shift-and-add ~ ,UiLly 517 and 518 at a spacing equal to the length of training sequence. The correlator 21lll70 output pulses and their replications at the training sequence duration produces signals P, and PQ. Combiner 519 receives the signal P~ and PQ and produces the sum of the square of these signals. This sum is dç~ n~ted as P. Logic cil~uiLIy 520 determines the amplitude and delay characteristics of the signal dispersion in 5 response to signal P.
It should be noted that the use of the shift-and-add circuitry has the effect of advantageously reproducing the correlator output pulses as if training sequence 101 were L~ ed four times in s~lcce~inn. While this could be done, such s~-cces~ive Ll~ l;csion of the training sequence whose duration is at least equal to 10 the range of signal dispersion delay can require a greater time duration than is available. This is particularly true in the case of a conventionally Lla~ liLLedtelevision signal, such as NTSC, PAL or SECAM wherein the intervals not used for other signal ~ lll;csion are not long enough for such succçs~ive training sequence L~ "i ~,ion.
While the amplitude and delay characteristics determined by logic ~ ;uiLly 520 can be made using only a single correlator, e.g., correlator 515 and its associated single shift-and-add circuit 517, and, therefore, use of correlator 516, shift-and-add Cil-;uiLly 518 and combiner 519 can be elimin~ted the use of both correlators and the combiner advantageously provides a meçh~ni~m for adjusting 20 the carrier phase produced by carrier source 505 and the automatic gain control provided by AGC circuit 502. In addition, use of both signals P~ and PQ allows the estim~tion of ghost characteristics while carrier phase adjustment is still taking place.
Fctim:~tion of the correct carrier phase can be determined by using P, and 25 PQ by well-known means within phase estim~tor 521 which approxi",~te the arc tangent of PQ/P,. Adjustment of the carrier phase can then be readily provided by coupling the estim~tecl carrier phase, designed as ~e~ appearing at the output of the phase estim~tor to carrier source 505. Similarly, by using the signal P, it is well-known that the amplitude of the received signal can be çstim~te~l so as to correct 30 the gain of the receiver. In FIG. 5, signal P is coupled to AGC circuit 502 to provide this gain adjustment.
21~117~
g The amplitude and delay characteristics provided by logic circuitry 520 along with signals P, and PQ are coupled to complex equalizer 510 wherein they are used to adjust the operation thereof. Specifically, in well-known fashion, the amplitude and delay characteristics are used to determine which equalizer coefficients arenon-zero and signals P~ and PQ are used to adjust the values of these non-zero coefficients.
The receiver embodiment shown in FIG. 6 is identical in function and operation to that already described in reference to FIG. 5 except that the serial connection of the correlators and shift-and-add Cii~;uiLly is reversed. As a result, the training sequences are replicated in FIG. 6 as opposed to replicating the correlator output in FIG. 5. This reversal is equivalent.
An illustrative output of the inphase shift-and-add circuitry 517 of FIG. 5, de~ign~ted as P" is shown in FIG. 7. The output of the quadrature shift-and-add circuitry 518 and dçsign~ted as PQ is related to P~ by a function of the phase shift of the co~ llunications medium through which the received signal has propagated.As is well known, a correlator is designed to provide a large amplitude output pulse upon detectin~ a training sequence. At other times, the correlator output may beambiguous and the correlator output amplitude is subst~nti~lly less than when the training sequence is detecte-l Signal P~ includes four signal peaks 701 through 704 wherein each peak represents a training sequence or its replicas generated by shift-and-add circuitry of FIG. 5. These peaks are spaced apart by 35.6 ,~Ls time intervals. It should be noted that the intervals between peaks 701 and 702 and between peaks 703 and 704 are respectively corrupted by incompletely correlated postcursor and precursor ghosts of the training sequence. These incompletely correlated postcursor and precursor ghosts are respectively represented by waveforms 705 and 706. The presence of such incompletely correlated ghosts lessens the ability to clearly discern ghost characteristics in those portions of intervals 707 and 708 cont~ining waveforms 705 and 706. However, either one of these intervals could be used to determine signal dispersion characteristics and, consequently, the shift-and-addcircuitry in the receiver need only provide at least one replica of its input. It is however, advantageous, for such ~ ;uilly to provide three replicas of its input signal so as to provide the four illustrated peaks 701 through 704 and thereby 21~117~
provide buffer intervals 707 and 708 for an innermost interval 709, dc~ign~ted as a "quiet zone". Such buffer intervals sufficiently separate quite zone 709 in timefrom incompletely correlated ghosts so that it is devoid of all ghost signals except those of the transmitted training sequence that undergo full correlation. Such ghosts can be either postcursor ghosts, i.e., they are received after the training sequence, or precursor ghosts, i.e., they arrive at the receiver prior to the c~"illed training sequence. F~IG. 7 depicts one such postcursor ghost as signal peak 710 which is located at a time interval 711 following peak 702 and one precursor ghost by peak 712 which is located at a time interval 713 preceding peak 703. Accordingly, the methodology used by the present invention and provided within logic cil~uiLly 519 is to determine the amplitude of any ghost by using amplitude comparators which determine the presence of any signals in the quiet zone above a predetermined level and the magnitude of such signals. The delay ofany such ghost is then determined by the location of any detected signal above this threshold relative to peaks 702 or 703. Note that there is virtually no ambiguity in discerning between postcursor and precursor ghosts as the duration of quiet zone709 extends for at least the range extending from the m~ "u", delayed precursor ghost, typically dçsign~ted by a negative time value, to the maximum delayed postcursor ghost, which is typically de~ign~te-l by a positive time value.
FIGs. 8 and 9 show alternate embodiments of a passband receiver which incorporates the present invention. In these receivers the received QAM signal on lead 801 is coupled through AGC cil~;uiLly 802 to A/D converter 803 and thence to passband Hilbert filter pair 804 and 805. Filters 804 and 805 provide outputs which are coupled to complex equalizer 810 and to store-and-subtraction circuits812 and 813. The outputs of complex equalizer 810 are connected to rotator 825 to produce the tr~ncmitte-l information signals on outputs 826. The operation ofstore-and-subtraction circuit 812 and 813, correlators 815 and 816, shift-and-add ;uiLly 817 and 818, combiner 819, logic cil~;uiLly 820 and phase estim~tor 821 are functional equivalents to their counterparts in the baseband receiver. FIG. 9 is similar in operation to FIG. 8 but for the reversal in the series connection of shift-and-add circuitry 817 and 818 and correlators 815 and 816.
In FIGs 5, 6, 8 and 9, it is preferable that the correlator filter coefficients be binary numbers corresponding to those in the training sequence since such coefficients can have only one of two values, such as +1 and -1, and implementation is thereby simplified. However, the use of such binary coefficients in both the correlator and the l~ liLLer training sequence generator can producedc bias in the quiet zones. Such dc bias, which is inherent due to the autocorrelation nature of a pseudorandom sequence, can be çlimin~te~ using well-known techniques. For example, in one such technique the dc bias can be measured and then subtracted from P, and PQ or from the signal combiner output signal P. In another technique, the dc bias can be elimin~tcd by passing Pl and PQ
or P through a p,op~ chosen dc blocking filter (not shown), such as those which subtract a delayed version of the signal from the signal itself.
A modified pseudorandom sequence can be tr~n~mitte(l in lieu of a pseudorandom sequence. In a modified pseudorandom sequence, a dc constant is added to each symbol of a pseudorandom sequence so that the transmitted symbols are either zero or a positive value rather than the equal positive or negative values of the original pseudorandom sequence. The advantage of using a modified pseudorandom sequence is that, unlike the original pseudorandom sequence, there is no dc bias in any quiet zone at the output of a correlator. Accordingly, it is preferable to utilize both binary correlators and modified pseudorandom trainingsequences in the preferred embodiment of the present invention. For another arrangement, a pseudorandom sequence having +1 and -1 symbol values is transmitted and the coefficients of the binary correlator take values of either +1 or 0 instead of the +1 and -1 values.
In FIGs. 5, 6, 8 and 9, it is advantageous for A/D converter 508, store-and-subtract 512, correlator 515, and shift-and-add ~ Ui~ly 517 to operate at a ratewhich is a multiple, n, of the symbol rate, where n is an integer > 1, in order to provide higher resolution of the signal dispersion delay. With such an arrangement, it is preferable to implement the correlator as shown in FIG. 10. As illustrated in FIG. 10, the correlator, e.g., 515, is implemented by ~ltili~ing a tapped delay line 1001, clocked at twice the symbol rate, 2FC, multipliers 1002 and summer 1003. Each multiplier utilizes a fLl~ed correlation coefficient which multiplies the symbol value stored in one of the taps of delay line 1001. The delay line taps 21111 ~ O
which are combined by summer 1003 are spaced apart by the symbol rate. The intermediate taps are not associated with a multiplier and therefore are not inputted to summer 1003. In the preferred embodiments, where a binary correlator is lltili7.ori the correlation coefficients are either +1, -1 or +1, 0. With 5 such coefficients, the multipliers 1002 can be realised by pass through elements for +1 multiplication, sign inversion elements for a -1 multiplication, and signal blocking elements for a 0 multiplication.
In many applications, the sampling rate at the receiver is generally chosen to be an integer multiple, I, of the training sequence symbol rate. Therefore, each 10 active correlator tap is separated by I-1 zero or null taps. Each active tap multiplies its associated symbol value by a non-zero value while each null tap, in effect, multiplies its associated symbol value by zero. One computationally-simple implementation of a correlator has been shown in FIG. 10. Another alternative isto use the sub-correlator structure 1100 shown in FIG. 11. Such a structure 15 elimin~tes the I-1 null taps separating each active tap. In FIG. 11, a demultiplexer 1101 sequentially distributes s~lcc~ssive received samples on lead 1102 to a different one of I sub-correlators 1103-1 through 1103-I. Each sub-correlator has an ordered sequence of tap settings which is equal to the ordered sequence of symbol values in the training sequence. The output of each sub-correlator is then 20 sequentially multiplexed via multiplexer 1104 to provide a correlator output on lead 1105.
In the foregoing embodiments of the present invention, a training sequence extending for a predetermined time interval, i.e., interval 102 of FIG. 1, was utili7e~1 This time interval is a function of the symbol rate and the number of 25 symbols in the sequence. Of course, either of these two parameters can be varied in different applications so long as time interval 102 is at least as long as the sum of the m~Ailllulll postcursor and precursor ghost delays. Each of these delays will vary with different system applications and is dependent upon the geographical terrain through which the ll~"~",iLIecl signal propagates. FIG. 12 shows an 30 alternate embodiment of a LlallsllliLLer 1200 which may be used in lieu of the one shown in FIG. 4 to provide any of a plurality of training sequences wherein eachsuch training sequence includes a different number of symbols or a different symbol rate so that each sequence has a different associated time interval 102.
7 n For a television application, the symbol rate is generally fixed and the number of symbols in the training sequence provided by any training sequence generator is different from that provided by any other generator. As shown in FIG.12, to provide any such training sequence, transmitter 1200 includes a plurality of training S sequence generators 1201-1 through 1201-K, where K is a predetermined integer, and single-pole, multi-throw switch 1202. Switch 1202, whose switch position is set for a particular system application, permits any of these training sequence generators to be selected and couples the output of the selected generator to transmit raised cosine filter 404. Each of the training sequence generators is clocked by timing generator 403. Filter 404, information source 401, timing generator 403, signal gateway 405 and modulator 406 function in an identical fashion as described in reference to FIG. 4. Advantageously, the time intervals of the training sequences provided by generators 1201-1 through 1201-K should be selected to provide an appropriate range of different training sequence durations to meet the range of expected ghost delay characteristics for a number of dirreLellt expected system applications.
FIG. 13 shows an embodiment of a receiver 1300 for use with the Ll~lsllli~Ler of FIG. 12. Receiver 1300 operates in similar fashion to receiver 500 of FIG. 5 and shares a number of components with receiver 500. Such components in FIG. 13 bear the same reference numerals as their counterparts in F~G. 5. Receiver 1300 also includes a number of components which have been added to provide operation co.llpa~ible with that provided by transmitter 1200. As shown, receiver 1300 includes a plurality of correlators 1301-1 through 1301-K
coupled to the output of store and subtraction circuit 512 in the "I" or inphasechannel and a plurality of correlators 1301-1 through 1301-K coupled to the output of store and subtraction circuit 513 in the "Q" or quadrature channel. Each of these correlators is associated with a different one of the training sequences provided by generators 1201-1 through 1201-K in FIG. 12. The function of each correlator is to provide an output pulse upon detectin~ the associated training sequence. A pair of decision circuits in the "Q" channel, examines such outputs and determines which of the K training sequences has been transmitted. The determination is then used to control the delay provided within shift-and-add circuitry 1305 and 1306 via a signal on leads 1307 and 1308. In this regard it should be noted that Cil- uilly 1305 and 1306 is identical to shift-and-add circuitry 517 and 518 in FIG. 5 except that programmable delay devices are used in lieu offixed delay devices. Such fixed delay devices provide a delay of N symbols or the duration of the training sequence. Now, however, since this duration can be any on of K different values, the control signal on leads 1207 and 1208 sets the delay of each programmable delay element to equal the duration of the detected training sequence.
FIG. 14 shows an alternate embodiment of a receiver 1400 which can be used with the LlallslllilLer 1200 shown in FIG. 12. Receiver 1400 is analogous to receiver 600 of FIG. 6 and shares a number of components which bear the same reference numerals. To be compatible with tr~n~mitter 1200, receiver 1400 includes a plurality of shift-and-add circuits 1401-1 through 1401-K in each of the I
and Q channels. Each such circuit incorporates the same number of delay elements and all of the delay elements in any one shift-and-add circuit provide the same delay. In addition, each of the delay elements in different ones of the K
shift-and-add circuits 1301-1 through 1301-K provides a delay corresponding to the duration of a different one of the K training sequences provided by ~lal~ er 1200 of FIG. 12. That is, the delay provided by each delay element in shift-and-add circuit 1301-1 is equal to the duration of the training sequence provided bytraining sequence generator 1201-1, and the delay provided by each delay elementin shift-and-add circuit 1301-K is equal to the duration of the training sequence provided by training sequence generator 1201-K. The outputs of the shift-and-addcircuits are coupled to correlator 1301-1 through 1301-K in the I channel and correlators 1302-1 through 1302-K in the Q channel and thence to decision circuits 1303 and 1304. The correlators and decision circuits operate in the same manner as described in reference to FIG. 13.
Receivers of the passband variety for use with tr~n~mitter 1200 can also be devised. Such passband receivers can be provided by modifying the receivers of FIGs. 8 and 9 in a manner analogous to that provided to the circuitry of FIGs. Sand 6.
It should, of course, be understood that while the present invention has been disclosed in reference to specific embodiments, numerous other arrangementsshould be apparent to those of ordinary skill in the art. For example, while thepresent invention has been disclosed in reference to a VSB or QAM signal, the 5 invention is applicable to any analog or digital communication system regardless of the presence, absence or type of modulation. In addition, the present invention is not limited to pseudorandom or modified pseudorandom training sequences and, indeed, can employ any type of training sequence. Furthermore, the complex equalizer shown in FIGs. 5, 6, 8 or 9 can be positioned so that its outputs are 10 coupled to each correlator input. In such an arrangement, the equalizer serves as a unit impulse fixed filter during system initi~ tion and serves as an adaptive equalizer thereafter.
TECHNIQUE FOR DETERMINING SIGNAL DISPERSION
CIIARACTERISTICS IN CO~fMUNICATIONS SYSTEMS
Technical Field The present invention relates to a technique for determining the characteristics of signal dispersion and, more particularly, to a technique in which a training sequence is transmitted and processed at a receiver so as to create a time interval in which the char~sten~tics of the signal dispersion can be accurately quantified.
Background of the Invention A training sequence is a series of a priori known symbols which are tla~ls~ led at predetermined times from a signal llallsllliller to a signal receiver.
Such sequences have long been used to adjust the operation of signal dispersion compensation a~paralu;, in the receiver, such as equalizers, echo cancellers and the like.
Signal dispersion, such as echoes or signal "ghosts", is an inherent problem in communications systems and the severity of the problem can vary with the system application. For example, signal ghosts or echoes frequently exist in co,lventional television l,alls,llisi,ion systems which are highly objectional to the viewer and which can render a high definition television (HDTV) signal unintelligible. Accordingly, compensation for signal dispersion is highly desirable in conventional television systems and, indeed, is required in HDTV and in many other co,l,lllunications applications. Moreover, while signal dispersion compensation apparatus, such as equalizers and cancellers, provides satisfactorycompensation, knowledge of the signal dispersion characteristics, e.g., amplitude, delay and phase, is useful for the adjustment of such apparatus during system startup and operation.
Prior art techniques exist which can determine the characteristics of signal dispersion. Such techniques typically transmit specific signals, such as pulses or training sequences, which are detected and analyzed in the receiver. These techniques provide satisfactory estimates of the signal dispersion characteristics in applications wherein the dispersion is large in amplitude and short in duration.However, the prior art techniques provide inaccurate results when the signal 211117~
dispersion is small in amplitude, particularly in the presence of channel noise, jitter or similar illlpaillllents, and provide ambiguous results when the signal dispersion is long in duration. These shortcomings have hindered development of co~ ,ullications systems requiring ever-more precise signal dispersion S compensation and more precise determination of the signal dispersion characteristics .
Most recently, as disclosed in the prior art, a technique which overcomes the aforesaid problems of the prior art has been developed which transmits two different training sequences wherein each such training sequence includes a 10 different number of symbols. This technique, while providing greater accuracythan those of the previous prior art, requires a time duration for the tr~n~mis~ion of the two different training sequences which is not always available. Therefore, a technique which provides greater accuracy in the determination of small-amplitude and long-duration signal dispersion and which requires less time duration would be 15 desirable.
Summary of the Invention The present invention covers the notion of determining the amplitude, phase and delay of signal dispersion by tr~ ",i~ ,g a training sequence including a plurality of apriori known symbols and having a time duration at least equal to the 20 expected range of signal dispersion delay. At a receiver, the training sequence is recovered and then processed using correlation and signal replication techniques.
This processed training sequence is then utilized to determine at least one of the aforesaid signal dispersion characteristics.
A feature of the present invention is that it is applicable to a variety of 25 different colllnlullications systems and is particularly adaptable for television signal tr~n~mi~cion wherein the training sequence can be advantageously ll~ llPilled during the vertical blanking interval. Another feature of the present invention is that it significantly illlpro~es the accuracy of dete.,llillillg the characteristics of signal dispersion when such dispersion is of long duration, has a small amplitude, or 30 is accompanied by channel noise, jitter and similar impairments.
In accordance with one aspect of the invention there is provided an appard~us for use in a co~ r~tions system wherein a received signal is sampled, said ayyaldLus comprising: processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a 5 prede~ uled symbol rate, for processing said received training sequence using correlation and signal replication, the processing means including; means for cleml-ltirlexing the received signal samples onto a plurality of output terminals in a cyclic fashion; a plurality of sub-correlators each connected to a different one of said output terminals and providing an output; multiplexing means for multiplexing 10 the outputs of said plurality of sub-correlators in said cyclic fashion; and means resporlsive to the multiplexed outputs of said plurality of said sub-correlators for de~lll~inillg at least one signal dispersion characteristic.
In accordance with another aspect of the invention there is provided in a col,-"""-ir~tions system having a signal dispersion with an expected range of delay, 15 a method of correlating comprising the steps of: generating a training sequence, said training sequence having a duration at least equal to said range of delay;
dem--ltirlexing samples of a signal onto a plurality of output terminals in a cyclic fashion; coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a different 20 output terminal and providing a correlated output signal; multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion; and del~llllinillg at least one signal dispersion characteristic to respond to the multiplexed outputs of said plurality of sub-correlators.
In accordance with yet another aspect of the invention there is provided an 25 ayyalaLus for use in a comm--nir,~tions system wherein an information signal is IlA~ lecl at a prede~;lllli~ed rate, said information signal being received after propagation through a co",llllll~ir~tions link and sampled at an integer multiple of said predetermined rate, said appardLus comprising: processing means responsive to a received training sequence including a plurality of predeLellllilled symbols 30 occurring at a predetermined symbol rate, for processing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the samples of the received il~o~ d~ion signal onto a plurality of terminals in a cyclic fashion; a plurality of sub-correlators each - 3a -connected to a dirrelent one of said terminals and providing a correlated output;
and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
In accordance with still yet another aspect of the present invention there is S provided a method of correlating in a co""",l,-ir~tions system wherein an h~lllla~ion signal is L~ d at a predetermined rate, said information signal being received after propagation through a colllllllll~ir~tions link and sampled at an integer multiple of said prede~llllilled rate, said method comprising the steps of:
geneldLillg a training sequence, said training sequence having a duration at least 10 equal to an expected range of delay of signal dispersion; deml~ltiplexing thesamples of the received information signal onto a plurality of output terminals in a cyclic fashion; coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a dirÇ~lelll output terminal and providing a correlated output signal; and multiplexing 15 the outputs of said plurality of sub-correlators in said cyclic fashion.
In accordance with still yet another aspect of the present invention there is provided an al)pdld~us for use in a col~ lll-icdlions system wherein a received signal is sampled, said apl)ald~us comprising: processing means responsive to a received training sequence including a plurality of predetermined symbols occurring 20 at a predeL~lll~illed symbol rate, for processing said received training sequence using correlation and signal replication, the processing means including; means for demultiplexing the received signal samples onto a plurality of terminals in a cyclic fashion; a plurality of sub-correlators, each receiving a common sequence and received signal samples from a dirr~lenl terminal, each sub-correlator correlating 25 its received sample using said common sequence to provide a correlated output;
and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
In accordance with still yet another aspect of the present invention there is provided a method of correlating in a collllllll~ir~tions system wherein a received 30 signal is sampled, said method comprising the steps of: generating a trainingsequence, said training sequence having a duration at least equal to an expectedrange of delay of signal dispersion; demultiplexing said received signal samples onto a plurality of output terminals in a cyclic fashion; coupling the received signal - 3b -samples on each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator being supplied with a common sequence and correlating its received signal samples with a said common sequence to provide acorrelated output; and multiplexing the outputs of said plurality of sub-correlators 5 in said cyclic fashion.
Brief Description of the Drawin~s FIGs. 1 and 2 are representations of an illustrative signal format wherein a training sequence is Ll~n.~llliLIed in accordance with the present invention;
FIG. 3 is a representation of illustrative signals, including signals which 10 precede and succeed the Ll~lsll~ission of the training sequence of FIG. 1, utilized by the disclosed embodiment of the present invention;
FIG. 4 is a block-schematic diagram ~epresellL~lion of an embodiment of a tr~n.cmitt~r which incorporates the present invention;
FIGs. S and 6 are alternate block-schematic diagram representations of 15 embodiments of a baseband receiver which incorporates the present invention;
FIG. 7 is a representation of an illustrative output signal of the inphase correlator 515 of FIG. 5;
FIGs. 8 and 9 are alternate block-schematic diagram representations of embodiments of a passband receiver which incorporates the present invention;
FIG. 10 is a block-schematic diagram represent~tion of an embodiment of the correlator in FlGs. 5, 6, 8 or 9;
FIG. 11 is a block-schematic diagram representation of an alternate embodiment of the correlator in FlGs. 5, 6, 8 or 9;
FIG. 12 is a block-schematic diagram representation of an alternate embodiment of a ~ s,l.iL~er which incorporates the present invention; and FIG. 13 and 14 are each a block-schematic diagram representation of an alternate embodiment of a rece;vel for use with the llal~ Ler of FIG. 11.
Detailed Description FIG. 1 shows a portion 100 of an illu~ ivt; enhanced National Television System Committee (NTSC) television signal format which incorporates a training sequence 101 during time interval 102 pursuant to the present invention. At thisjuncture, it will, of course, be understood that the present invention is not restricted to use in such a television signal format and can be utilized with any television signal format, e.g., Sequence Electronique Couleur Avec Memoire (SECAM), Phase Alternation Line (PAL), and HDTV, as well as signal formats for col,.lllunications systems other than television ~ ;c~;on. In this illustrative format, sequence 101 includes 255 symbols which are ll~ ."iL~ecl at a rate of 7.160 x 10C symbols/second so that time interval 102 has a duration of 35.6 x 10~
seconds (,us). Preferably, sequence 101 is a pseudorandom sequence. In addition,as shown, sequence 101 is advantageously ll;~ ed during the 63.56 ,us horizontal line period 103 which extends between holi~onl~l synchronization pulses 104 and 105. Horizontal period 103 is one of a plurality of line periods in a vertical blanking interval in a video frame. Incorporation of the training sequence in the unused time interval in horizontal line period 103 advantageously does not require any alteration of the television signals tr~n~mitte~l before and after sequence 101. Envelopes 106 and 107 represent the color bursts in an NTSC
format which respectively follow pulses 104 and 105. Within intervals 109 and 110, the signals in the NTSC format are known and provide well-known functions unconnected with the present invention. However, as will be discussed hereinbelow, the known nature of the signals in these intervals will be advantageously utilized by the present invention to permit the accurate replication of the ~lall~,lli~led training sequence 101 by a store-and-subtract process. Interval 109 must be at least as long as the m~illlulll delay of any postcursor ghosts and interval 110 must be at least as long as the system of the m~ill-ulll postcursorghost delay and precursor ghost delay. The former delay is typically represented by a positive time value and the latter delay is typically represented by a negative time S value. The sum of the m~xi,llulll postcursor ghost delay and precursor ghost delay time values, ignoring any algebraic sign, is hereinafter referred to as the expected range of signal dispersion delay.
FIG. 2 shows a portion 200 of the illu,Ll~tive NTSC television signal format wherein portion 200 succeeds portion 100 and includes a horizontal line period 201 within a vertical blanking interval. Horizontal line period 103 is disposed between horizontal synchlollizalion pulses 202 and 203. Envelopes 204 and 205 represent color bursts and are identical to envelopes 106 and 107 of FIG. 1. It should be noted that portions 100 and 200 are identical except that the training sequence 101 is not ll~l-c~--illed during horizontal line period 201.
FIG. 3 shows the periodic nature of the tr~n~mi~sion of training sequence 101 within eight portions 300-307 of an NTSC television signal format. Each of portions 300 through 307 has the same time duration and respectively includes one holi~onlal line period 308 through 315 in eight successive vertical blanking intervals. Each such holi~ollLal line period is disposed between a pair of horizontal synchloni~aLion pulses 104 and 105.
We shall define the term training sequence period as a time interval within one horizontal line period of each vertical blanking interval. During this training sequence period, either the training sequence 101 or a zero dc signal is "c,.,ille~ Accordingly, the d~cign~tions 308 through 315 in FIG. 3 ~le~ign~te this training sequence period.
It should be noted that in the NTSC signal format, shown in FIG. 3, the signal surrounding any training sequence period are different from those surrounding an immediately following training sequence period. For example, the signals in intervals 316 and 317, which immediately precede and succeed trainingsequence period 308, are different from those in intervals 318 and 319, which immediately precede and succeed training sequence period 309. Training sequence period 309 occurs during a vertical blanking interval immediately following the vertical blanking interval that includes training sequence period 308. However, the 2 ~ rl fJ
signals surrounding any given training sequence period are identical to those surrounding the fourth next training sequence period after the given training sequence period. In this regard, note that portions 300 and 304 are identical but for the training sequence 101. This identity also exists between portions 301 and 305, portions 302 and 306, and between portions 303 and 307. Accordingly, if thesignals in the foregoing paired signal portions were stored and then one stored signal in each pair was subtracted from the other, an accurate replica of the L~ ed training sequence could be recovered. For example, training sequence 101 could be recovered in the receiver by subtracting portion 304 from portion 300, subtracting portion 301 from portion 305, subtracting portion 306 from portion 302 and subtracting portion 303 from porion 307. It is this process of pairing the above-described signal portions, each including a training sequence period, which is used in a receiver constructed pu"uallt to the present invention.
Refer now to FIG. 4 which shows a ~ ,l"iUer 400 incorporating the present invention. The tr~n~mittec~ information signal, which in the illustrative embodiment is a conventional NTSC formatted signal, is generated by information source 401. This signal is then coupled to signal gateway 405 under the control of timing signals from timing generator 403. Training sequence generator 402 generates training sequence 101 within horizontal line periods 308, 310, 313 and315 under the control of timing signals from timing signal generator 403 and couples these sequences to signal gateway 405. Advantageously, each of these pseudorandom sequences is spectrally shaped by raised cosine transmit filter 404before being coupled to signal gateway 405. Gateway 405, illustratively replescnted by a single-pole, multi-throw switch, sclc.,Livt;ly couples the signals from source 401 or generator 402 to modulator 406 which provides vestigial sideband (VSB) amplitude modulation in the illustrative NTSC television application, or quadrature amplitude modulation (QAM) in other communication systems.
The ~lansuliL filter 404 advantageously includes tapped delay line 407, digital-to-analog (D/A) converter with a sample hold circuit 408 and an analog low-pass filter (LPF) 409 connected in series. Tapped delay line 407 and D/A
collvelLer 408 are clocked at four times the symbol rate (Fc). Analog LPF 409 advantageously elimin~tes out-of-band signal energy.
FIGs. S and 6 respectively show alternate embodiments of a baseband receiver 500 and 600 which incorporate the present invention. As shown in FIG. 5, the received modulated signal incorporating training sequence 101 is coupled via lead 501 through automatic gain control (AGC) circuit 502 to S demodulators 503 and 504. Each demodulator multiplies the received signal by carrier signals generated by carrier source 505. One of these carrier signals isrepresented by cos (<~)ct + ~e) while the other is represented by sin (~')ct + ~e)~
where ~c is the carrier frequency used in the L~ ll,iLler modulator of FIG. 4 and ~e is the phase angle between the carrier signals supplied by carrier source 504 and those utilized by the l-dlls.lliLLer modulator.
Each demodulator output is coupled through low-pass filters (LPF) 506 and 507 to analog-to-digital collvelL~ (A/D) 508 and 509. The digital outputs of these converts are supplied to a complex equalizer 510, which removes distortion in the received signal so as to accurately recover the Ll~ .lliL(e~i information signal and provide a pair of output signals 511 for further conventional television processing.
The term "complex", with reference to equalizer 510, refers to the fact that this equalizer utilizes coefficients which are complex numbers, i.e., they have a real and an im~gin~ry component. The digital outputs of converts 508 and 509 are also coupled to store-and-subtraction circuits 512 and 513.
The function of store-and-subtraction circuits 512 and 513 is to recover the training sequence 101 from the received signal by storing the paired waveforms of F~G. 3, i.e, 300 and 304, 301 and 305, 302 and 306, and 303 and 307, and then subtracting one waveform in each pair from the other. To extract the incoming signal waveform at the applop.iate time so as to recover each of these paired waveforms, circuits 512 and 513 are periodically enabled by timing circuitry 514.
The recovered training sequences rec~vered by circuits 512 and 513 are respectively de~ tecl as Sl and SQ.
Correlators 515 and 516 are designed to produce an output pulse upon ~letecting the training sequence. These output pulses, along with the correlatoroutput at other times, are replicated three times within shift-and-add ~ ,UiLly 517 and 518 at a spacing equal to the length of training sequence. The correlator 21lll70 output pulses and their replications at the training sequence duration produces signals P, and PQ. Combiner 519 receives the signal P~ and PQ and produces the sum of the square of these signals. This sum is dç~ n~ted as P. Logic cil~uiLIy 520 determines the amplitude and delay characteristics of the signal dispersion in 5 response to signal P.
It should be noted that the use of the shift-and-add circuitry has the effect of advantageously reproducing the correlator output pulses as if training sequence 101 were L~ ed four times in s~lcce~inn. While this could be done, such s~-cces~ive Ll~ l;csion of the training sequence whose duration is at least equal to 10 the range of signal dispersion delay can require a greater time duration than is available. This is particularly true in the case of a conventionally Lla~ liLLedtelevision signal, such as NTSC, PAL or SECAM wherein the intervals not used for other signal ~ lll;csion are not long enough for such succçs~ive training sequence L~ "i ~,ion.
While the amplitude and delay characteristics determined by logic ~ ;uiLly 520 can be made using only a single correlator, e.g., correlator 515 and its associated single shift-and-add circuit 517, and, therefore, use of correlator 516, shift-and-add Cil-;uiLly 518 and combiner 519 can be elimin~ted the use of both correlators and the combiner advantageously provides a meçh~ni~m for adjusting 20 the carrier phase produced by carrier source 505 and the automatic gain control provided by AGC circuit 502. In addition, use of both signals P~ and PQ allows the estim~tion of ghost characteristics while carrier phase adjustment is still taking place.
Fctim:~tion of the correct carrier phase can be determined by using P, and 25 PQ by well-known means within phase estim~tor 521 which approxi",~te the arc tangent of PQ/P,. Adjustment of the carrier phase can then be readily provided by coupling the estim~tecl carrier phase, designed as ~e~ appearing at the output of the phase estim~tor to carrier source 505. Similarly, by using the signal P, it is well-known that the amplitude of the received signal can be çstim~te~l so as to correct 30 the gain of the receiver. In FIG. 5, signal P is coupled to AGC circuit 502 to provide this gain adjustment.
21~117~
g The amplitude and delay characteristics provided by logic circuitry 520 along with signals P, and PQ are coupled to complex equalizer 510 wherein they are used to adjust the operation thereof. Specifically, in well-known fashion, the amplitude and delay characteristics are used to determine which equalizer coefficients arenon-zero and signals P~ and PQ are used to adjust the values of these non-zero coefficients.
The receiver embodiment shown in FIG. 6 is identical in function and operation to that already described in reference to FIG. 5 except that the serial connection of the correlators and shift-and-add Cii~;uiLly is reversed. As a result, the training sequences are replicated in FIG. 6 as opposed to replicating the correlator output in FIG. 5. This reversal is equivalent.
An illustrative output of the inphase shift-and-add circuitry 517 of FIG. 5, de~ign~ted as P" is shown in FIG. 7. The output of the quadrature shift-and-add circuitry 518 and dçsign~ted as PQ is related to P~ by a function of the phase shift of the co~ llunications medium through which the received signal has propagated.As is well known, a correlator is designed to provide a large amplitude output pulse upon detectin~ a training sequence. At other times, the correlator output may beambiguous and the correlator output amplitude is subst~nti~lly less than when the training sequence is detecte-l Signal P~ includes four signal peaks 701 through 704 wherein each peak represents a training sequence or its replicas generated by shift-and-add circuitry of FIG. 5. These peaks are spaced apart by 35.6 ,~Ls time intervals. It should be noted that the intervals between peaks 701 and 702 and between peaks 703 and 704 are respectively corrupted by incompletely correlated postcursor and precursor ghosts of the training sequence. These incompletely correlated postcursor and precursor ghosts are respectively represented by waveforms 705 and 706. The presence of such incompletely correlated ghosts lessens the ability to clearly discern ghost characteristics in those portions of intervals 707 and 708 cont~ining waveforms 705 and 706. However, either one of these intervals could be used to determine signal dispersion characteristics and, consequently, the shift-and-addcircuitry in the receiver need only provide at least one replica of its input. It is however, advantageous, for such ~ ;uilly to provide three replicas of its input signal so as to provide the four illustrated peaks 701 through 704 and thereby 21~117~
provide buffer intervals 707 and 708 for an innermost interval 709, dc~ign~ted as a "quiet zone". Such buffer intervals sufficiently separate quite zone 709 in timefrom incompletely correlated ghosts so that it is devoid of all ghost signals except those of the transmitted training sequence that undergo full correlation. Such ghosts can be either postcursor ghosts, i.e., they are received after the training sequence, or precursor ghosts, i.e., they arrive at the receiver prior to the c~"illed training sequence. F~IG. 7 depicts one such postcursor ghost as signal peak 710 which is located at a time interval 711 following peak 702 and one precursor ghost by peak 712 which is located at a time interval 713 preceding peak 703. Accordingly, the methodology used by the present invention and provided within logic cil~uiLly 519 is to determine the amplitude of any ghost by using amplitude comparators which determine the presence of any signals in the quiet zone above a predetermined level and the magnitude of such signals. The delay ofany such ghost is then determined by the location of any detected signal above this threshold relative to peaks 702 or 703. Note that there is virtually no ambiguity in discerning between postcursor and precursor ghosts as the duration of quiet zone709 extends for at least the range extending from the m~ "u", delayed precursor ghost, typically dçsign~ted by a negative time value, to the maximum delayed postcursor ghost, which is typically de~ign~te-l by a positive time value.
FIGs. 8 and 9 show alternate embodiments of a passband receiver which incorporates the present invention. In these receivers the received QAM signal on lead 801 is coupled through AGC cil~;uiLly 802 to A/D converter 803 and thence to passband Hilbert filter pair 804 and 805. Filters 804 and 805 provide outputs which are coupled to complex equalizer 810 and to store-and-subtraction circuits812 and 813. The outputs of complex equalizer 810 are connected to rotator 825 to produce the tr~ncmitte-l information signals on outputs 826. The operation ofstore-and-subtraction circuit 812 and 813, correlators 815 and 816, shift-and-add ;uiLly 817 and 818, combiner 819, logic cil~;uiLly 820 and phase estim~tor 821 are functional equivalents to their counterparts in the baseband receiver. FIG. 9 is similar in operation to FIG. 8 but for the reversal in the series connection of shift-and-add circuitry 817 and 818 and correlators 815 and 816.
In FIGs 5, 6, 8 and 9, it is preferable that the correlator filter coefficients be binary numbers corresponding to those in the training sequence since such coefficients can have only one of two values, such as +1 and -1, and implementation is thereby simplified. However, the use of such binary coefficients in both the correlator and the l~ liLLer training sequence generator can producedc bias in the quiet zones. Such dc bias, which is inherent due to the autocorrelation nature of a pseudorandom sequence, can be çlimin~te~ using well-known techniques. For example, in one such technique the dc bias can be measured and then subtracted from P, and PQ or from the signal combiner output signal P. In another technique, the dc bias can be elimin~tcd by passing Pl and PQ
or P through a p,op~ chosen dc blocking filter (not shown), such as those which subtract a delayed version of the signal from the signal itself.
A modified pseudorandom sequence can be tr~n~mitte(l in lieu of a pseudorandom sequence. In a modified pseudorandom sequence, a dc constant is added to each symbol of a pseudorandom sequence so that the transmitted symbols are either zero or a positive value rather than the equal positive or negative values of the original pseudorandom sequence. The advantage of using a modified pseudorandom sequence is that, unlike the original pseudorandom sequence, there is no dc bias in any quiet zone at the output of a correlator. Accordingly, it is preferable to utilize both binary correlators and modified pseudorandom trainingsequences in the preferred embodiment of the present invention. For another arrangement, a pseudorandom sequence having +1 and -1 symbol values is transmitted and the coefficients of the binary correlator take values of either +1 or 0 instead of the +1 and -1 values.
In FIGs. 5, 6, 8 and 9, it is advantageous for A/D converter 508, store-and-subtract 512, correlator 515, and shift-and-add ~ Ui~ly 517 to operate at a ratewhich is a multiple, n, of the symbol rate, where n is an integer > 1, in order to provide higher resolution of the signal dispersion delay. With such an arrangement, it is preferable to implement the correlator as shown in FIG. 10. As illustrated in FIG. 10, the correlator, e.g., 515, is implemented by ~ltili~ing a tapped delay line 1001, clocked at twice the symbol rate, 2FC, multipliers 1002 and summer 1003. Each multiplier utilizes a fLl~ed correlation coefficient which multiplies the symbol value stored in one of the taps of delay line 1001. The delay line taps 21111 ~ O
which are combined by summer 1003 are spaced apart by the symbol rate. The intermediate taps are not associated with a multiplier and therefore are not inputted to summer 1003. In the preferred embodiments, where a binary correlator is lltili7.ori the correlation coefficients are either +1, -1 or +1, 0. With 5 such coefficients, the multipliers 1002 can be realised by pass through elements for +1 multiplication, sign inversion elements for a -1 multiplication, and signal blocking elements for a 0 multiplication.
In many applications, the sampling rate at the receiver is generally chosen to be an integer multiple, I, of the training sequence symbol rate. Therefore, each 10 active correlator tap is separated by I-1 zero or null taps. Each active tap multiplies its associated symbol value by a non-zero value while each null tap, in effect, multiplies its associated symbol value by zero. One computationally-simple implementation of a correlator has been shown in FIG. 10. Another alternative isto use the sub-correlator structure 1100 shown in FIG. 11. Such a structure 15 elimin~tes the I-1 null taps separating each active tap. In FIG. 11, a demultiplexer 1101 sequentially distributes s~lcc~ssive received samples on lead 1102 to a different one of I sub-correlators 1103-1 through 1103-I. Each sub-correlator has an ordered sequence of tap settings which is equal to the ordered sequence of symbol values in the training sequence. The output of each sub-correlator is then 20 sequentially multiplexed via multiplexer 1104 to provide a correlator output on lead 1105.
In the foregoing embodiments of the present invention, a training sequence extending for a predetermined time interval, i.e., interval 102 of FIG. 1, was utili7e~1 This time interval is a function of the symbol rate and the number of 25 symbols in the sequence. Of course, either of these two parameters can be varied in different applications so long as time interval 102 is at least as long as the sum of the m~Ailllulll postcursor and precursor ghost delays. Each of these delays will vary with different system applications and is dependent upon the geographical terrain through which the ll~"~",iLIecl signal propagates. FIG. 12 shows an 30 alternate embodiment of a LlallsllliLLer 1200 which may be used in lieu of the one shown in FIG. 4 to provide any of a plurality of training sequences wherein eachsuch training sequence includes a different number of symbols or a different symbol rate so that each sequence has a different associated time interval 102.
7 n For a television application, the symbol rate is generally fixed and the number of symbols in the training sequence provided by any training sequence generator is different from that provided by any other generator. As shown in FIG.12, to provide any such training sequence, transmitter 1200 includes a plurality of training S sequence generators 1201-1 through 1201-K, where K is a predetermined integer, and single-pole, multi-throw switch 1202. Switch 1202, whose switch position is set for a particular system application, permits any of these training sequence generators to be selected and couples the output of the selected generator to transmit raised cosine filter 404. Each of the training sequence generators is clocked by timing generator 403. Filter 404, information source 401, timing generator 403, signal gateway 405 and modulator 406 function in an identical fashion as described in reference to FIG. 4. Advantageously, the time intervals of the training sequences provided by generators 1201-1 through 1201-K should be selected to provide an appropriate range of different training sequence durations to meet the range of expected ghost delay characteristics for a number of dirreLellt expected system applications.
FIG. 13 shows an embodiment of a receiver 1300 for use with the Ll~lsllli~Ler of FIG. 12. Receiver 1300 operates in similar fashion to receiver 500 of FIG. 5 and shares a number of components with receiver 500. Such components in FIG. 13 bear the same reference numerals as their counterparts in F~G. 5. Receiver 1300 also includes a number of components which have been added to provide operation co.llpa~ible with that provided by transmitter 1200. As shown, receiver 1300 includes a plurality of correlators 1301-1 through 1301-K
coupled to the output of store and subtraction circuit 512 in the "I" or inphasechannel and a plurality of correlators 1301-1 through 1301-K coupled to the output of store and subtraction circuit 513 in the "Q" or quadrature channel. Each of these correlators is associated with a different one of the training sequences provided by generators 1201-1 through 1201-K in FIG. 12. The function of each correlator is to provide an output pulse upon detectin~ the associated training sequence. A pair of decision circuits in the "Q" channel, examines such outputs and determines which of the K training sequences has been transmitted. The determination is then used to control the delay provided within shift-and-add circuitry 1305 and 1306 via a signal on leads 1307 and 1308. In this regard it should be noted that Cil- uilly 1305 and 1306 is identical to shift-and-add circuitry 517 and 518 in FIG. 5 except that programmable delay devices are used in lieu offixed delay devices. Such fixed delay devices provide a delay of N symbols or the duration of the training sequence. Now, however, since this duration can be any on of K different values, the control signal on leads 1207 and 1208 sets the delay of each programmable delay element to equal the duration of the detected training sequence.
FIG. 14 shows an alternate embodiment of a receiver 1400 which can be used with the LlallslllilLer 1200 shown in FIG. 12. Receiver 1400 is analogous to receiver 600 of FIG. 6 and shares a number of components which bear the same reference numerals. To be compatible with tr~n~mitter 1200, receiver 1400 includes a plurality of shift-and-add circuits 1401-1 through 1401-K in each of the I
and Q channels. Each such circuit incorporates the same number of delay elements and all of the delay elements in any one shift-and-add circuit provide the same delay. In addition, each of the delay elements in different ones of the K
shift-and-add circuits 1301-1 through 1301-K provides a delay corresponding to the duration of a different one of the K training sequences provided by ~lal~ er 1200 of FIG. 12. That is, the delay provided by each delay element in shift-and-add circuit 1301-1 is equal to the duration of the training sequence provided bytraining sequence generator 1201-1, and the delay provided by each delay elementin shift-and-add circuit 1301-K is equal to the duration of the training sequence provided by training sequence generator 1201-K. The outputs of the shift-and-addcircuits are coupled to correlator 1301-1 through 1301-K in the I channel and correlators 1302-1 through 1302-K in the Q channel and thence to decision circuits 1303 and 1304. The correlators and decision circuits operate in the same manner as described in reference to FIG. 13.
Receivers of the passband variety for use with tr~n~mitter 1200 can also be devised. Such passband receivers can be provided by modifying the receivers of FIGs. 8 and 9 in a manner analogous to that provided to the circuitry of FIGs. Sand 6.
It should, of course, be understood that while the present invention has been disclosed in reference to specific embodiments, numerous other arrangementsshould be apparent to those of ordinary skill in the art. For example, while thepresent invention has been disclosed in reference to a VSB or QAM signal, the 5 invention is applicable to any analog or digital communication system regardless of the presence, absence or type of modulation. In addition, the present invention is not limited to pseudorandom or modified pseudorandom training sequences and, indeed, can employ any type of training sequence. Furthermore, the complex equalizer shown in FIGs. 5, 6, 8 or 9 can be positioned so that its outputs are 10 coupled to each correlator input. In such an arrangement, the equalizer serves as a unit impulse fixed filter during system initi~ tion and serves as an adaptive equalizer thereafter.
Claims (18)
1. Apparatus for use in a communications system wherein a received signal is sampled, said apparatus comprising:
processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a predetermined symbol rate, forprocessing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the received signal samples onto a plurality of output terminals in a cyclic fashion;
a plurality of sub-correlators each connected to a different one of said output terminals and providing an output;
multiplexing means for multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion; and means responsive to the multiplexed outputs of said plurality of said sub-correlators for determining at least one signal dispersion characteristic.
processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a predetermined symbol rate, forprocessing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the received signal samples onto a plurality of output terminals in a cyclic fashion;
a plurality of sub-correlators each connected to a different one of said output terminals and providing an output;
multiplexing means for multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion; and means responsive to the multiplexed outputs of said plurality of said sub-correlators for determining at least one signal dispersion characteristic.
2. The apparatus of claim 1 wherein said received signal includes information transmitted at a predetermined symbol rate and wherein said multiplexing and demultiplexing means each utilize a clock signal which is n times said predetermined symbol rate, where n is an integer ~ 1.
3. The apparatus of claim 1 wherein said received signal is a training sequence.
4. The apparatus of claim 3 wherein said training sequence is a pseudorandom sequence.
5. The apparatus of claim 1 wherein said received signal is a training sequence and said apparatus further includes means, serially connected to said multiplexing means for replicating said correlated training sequence.
6. The apparatus of claim 1 wherein each of said sub-correlators utilizes a predetermined set of binary numbers as coefficients.
7. The apparatus of claim 6 wherein the binary numbers have equal positive and negative values.
8. The apparatus of claim 6 wherein the binary numbers have values which are zero and non-zero.
9. The apparatus of claim 1 further including means, serially connected to said demultiplexing means, for replicating said received signal one time and said plurality of sub-correlators correlates said corresponding received signal and correlates the replica of said received signal.
10. The apparatus of claim 1 further including a source of at least one carrier signal and said determining means determines the phase characteristic of said signal dispersion and controls said carrier signal source.
11. The apparatus of claim 1 further including an equalizer and said determining means adjusts the operation of said equalizer.
12. The apparatus of claim 1 wherein said determining means determines the amplitude and delay characteristics of said signal dispersion.
13. The apparatus of claim 1 further including an automatic gain control circuitthrough which said received signal is coupled and said determining means includes means for controlling the gain of said automatic gain control circuit.
14. In a communication system having a signal dispersion with an expected range of delay, a method of correlating comprising the steps of:
generating a training sequence, said training sequence having a duration at least equal to said range of delay;
demultiplexing samples of a signal onto a plurality of output terminals in a cyclic fashion;
coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a different output terminal and providing a correlated output signal;
multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion; and determining at least one signal dispersion characteristic to respond to the multiplexed outputs of said plurality of sub-correlators.
generating a training sequence, said training sequence having a duration at least equal to said range of delay;
demultiplexing samples of a signal onto a plurality of output terminals in a cyclic fashion;
coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a different output terminal and providing a correlated output signal;
multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion; and determining at least one signal dispersion characteristic to respond to the multiplexed outputs of said plurality of sub-correlators.
15. Apparatus for use in a communications system wherein an information signal is transmitted at a predetermined rate, said information signal being received after propagation through a communications link and sampled at an integer multiple of said predetermined rate, said apparatus comprising:
processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a predetermined symbol rate, forprocessing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the samples of the received information signal onto a plurality of terminals in a cyclic fashion;
a plurality of sub-correlators each connected to a different one of said terminals and providing a correlated output; and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a predetermined symbol rate, forprocessing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the samples of the received information signal onto a plurality of terminals in a cyclic fashion;
a plurality of sub-correlators each connected to a different one of said terminals and providing a correlated output; and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
16. A method of correlating in a communications system wherein an information signal is transmitted at a predetermined rate, said information signal being received after propagation through a communications link and sampled at aninteger multiple of said predetermined rate, said method comprising the steps of:
generating a training sequence, said training sequence having a duration at least equal to an expected range of delay of signal dispersion;
demultiplexing the samples of the received information signal onto a plurality of output terminals in a cyclic fashion;
coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a different output terminal and providing a correlated output signal; and multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion.
generating a training sequence, said training sequence having a duration at least equal to an expected range of delay of signal dispersion;
demultiplexing the samples of the received information signal onto a plurality of output terminals in a cyclic fashion;
coupling each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator receiving the samples from a different output terminal and providing a correlated output signal; and multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion.
17. Apparatus for use in a communication system wherein a received signal is sampled, said apparatus comprising:
processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a predetermined symbol rate, forprocessing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the received signal samples onto a plurality of terminals in a cyclic fashion;
a plurality of sub-correlators, each receiving a common sequence and received signal samples from a different terminal, each sub-correlator correlating its received sample using said common sequence to provide a correlated output;
and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
processing means responsive to a received training sequence including a plurality of predetermined symbols occurring at a predetermined symbol rate, forprocessing said received training sequence using correlation and signal replication, the processing means including;
means for demultiplexing the received signal samples onto a plurality of terminals in a cyclic fashion;
a plurality of sub-correlators, each receiving a common sequence and received signal samples from a different terminal, each sub-correlator correlating its received sample using said common sequence to provide a correlated output;
and means for multiplexing the correlated outputs of said plurality of sub-correlators in said cyclic fashion.
18. A method of correlating in a communication system wherein a received signal is sampled, said method comprising the steps of:
generating a training sequence, said training sequence having a duration at least equal to an expected range of delay of signal dispersion;
demultiplexing said received signal samples onto a plurality of output terminals in a cyclic fashion;
coupling the received signal samples on each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator being supplied with a common sequence and correlating its received signal samples with a said common sequence to provide a correlated output; and multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion.
generating a training sequence, said training sequence having a duration at least equal to an expected range of delay of signal dispersion;
demultiplexing said received signal samples onto a plurality of output terminals in a cyclic fashion;
coupling the received signal samples on each output terminal to an associated sub-correlator in a plurality of sub-correlators, each sub-correlator being supplied with a common sequence and correlating its received signal samples with a said common sequence to provide a correlated output; and multiplexing the outputs of said plurality of sub-correlators in said cyclic fashion.
Priority Applications (8)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| DE69022715T DE69022715T2 (en) | 1990-03-13 | 1990-12-07 | Determination of dispersion properties of a telecommunication transmission channel from a received test sequence after correlation and duplication. |
| EP90313332A EP0446523B1 (en) | 1990-03-13 | 1990-12-07 | Determination of dispersion characteristics of a communication channel from a received training sequence processed by correlation and replication |
| CA002033931A CA2033931C (en) | 1990-03-13 | 1991-01-10 | Technique for determining signal dispersion characteristics in communications systems |
| KR1019910003643A KR940008123B1 (en) | 1990-03-13 | 1991-03-07 | Technique for determining signal dispersion characteristics in communication systems |
| JP3069500A JPH04220892A (en) | 1990-03-13 | 1991-03-11 | Transmitter corresponding to signal dispersion, television signal transmitter apparatus, receiver, communication system and method of deciding signal dispersing characteristic |
| US07/728,824 US5260972A (en) | 1990-03-13 | 1991-07-09 | Technique for determining signal dispersion characteristics in communications systems |
| US08/089,153 US5406586A (en) | 1991-07-09 | 1993-07-08 | Signal correlation technique |
| CA 2111170 CA2111170C (en) | 1990-03-13 | 1993-12-10 | Technique for determining signal dispersion characteristics in communications systems |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US49301790A | 1990-03-13 | 1990-03-13 | |
| US07/728,827 US5090600A (en) | 1990-06-25 | 1991-07-09 | Liquid pressure opened pouring spout |
| CA 2111170 CA2111170C (en) | 1990-03-13 | 1993-12-10 | Technique for determining signal dispersion characteristics in communications systems |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CA2111170A1 CA2111170A1 (en) | 1995-06-11 |
| CA2111170C true CA2111170C (en) | 1999-06-29 |
Family
ID=27169655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA 2111170 Expired - Fee Related CA2111170C (en) | 1990-03-13 | 1993-12-10 | Technique for determining signal dispersion characteristics in communications systems |
Country Status (1)
| Country | Link |
|---|---|
| CA (1) | CA2111170C (en) |
-
1993
- 1993-12-10 CA CA 2111170 patent/CA2111170C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CA2111170A1 (en) | 1995-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5260972A (en) | Technique for determining signal dispersion characteristics in communications systems | |
| US5406586A (en) | Signal correlation technique | |
| EP0430531B1 (en) | Technique for determining signal dispersion characteristics in communications systems | |
| US6094464A (en) | Burst mode receiver | |
| KR100339110B1 (en) | Carrier recovery system for a vestigial sideband signal | |
| CA1338153C (en) | Interference canceller | |
| US4290139A (en) | Synchronization of a data communication receiver with a received signal | |
| CA2079292C (en) | Fast response matched filter receiver with decision feedback equalizer | |
| US4262360A (en) | Method and device for detecting a pseudo-random sequence of carrier phase changes of 0° and 180° in a data receiver | |
| CA1210877A (en) | Method of synchronizing parallel channels of orthogonally multiplexed parallel data transmission system and improved automatic equalizer for use in such a transmission system | |
| US3508172A (en) | Adaptive mean-square equalizer for data transmission | |
| US3794921A (en) | Differentially adaptive communication system | |
| US4376308A (en) | Control of coefficient drift for fractionally spaced equalizers | |
| KR0163729B1 (en) | Phase detecting method and ptl of vsb modulation system | |
| EP0573696B1 (en) | Timing recovery method and system | |
| US6148037A (en) | Sampling timing phase error detector for VSB modulation signal | |
| KR100273763B1 (en) | Tv reception apparatus using same ghost-cancellarion circuitry for receiving different types of tv signals | |
| CA2111170C (en) | Technique for determining signal dispersion characteristics in communications systems | |
| CA2033931C (en) | Technique for determining signal dispersion characteristics in communications systems | |
| US3581207A (en) | Joint setting of demodulating carrier phase, sampling time and equalizer gain parameters in synchronous data transmission systems | |
| EP0265131A2 (en) | Simplified recovery of data signals from quadrature-related carrier signals | |
| Wang et al. | Training signal and receiver design for multipath channel characterization for TV broadcasting | |
| Kim et al. | A symbol timing recovery using the segment sync data for the digital HDTV GA VSB system | |
| KR20010074497A (en) | Symbol sign directed phase detector | |
| US4207528A (en) | Timing-phase control system for duobinary transmission |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request | ||
| MKLA | Lapsed |