CN1225207A

CN1225207A - Pre-acquisition frequency offset removal in a GPS receiver

Info

Publication number: CN1225207A
Application number: CN97196332A
Authority: CN
Inventors: D·D·哈里森; J·J·蒂曼
Original assignee: General Electric Co
Current assignee: General Electric Co
Priority date: 1996-07-12
Filing date: 1997-07-03
Publication date: 1999-08-04
Anticipated expiration: 2017-07-03
Also published as: TW344801B; CN1319283C; TW358269B; TW351033B

Abstract

A direct sequence spread spectrum receiver samples an incoming signal and stores the sample in memory. Prior to sampling and storage, the incoming signal is translated to an IF signal. Also prior to storage, the IF signal is corrected for a frequency offset signal. The frequency offset may be caused by many sources, Doppler shift or local oscillator error, for example. Once the signal is corrected for the frequency offset, the signal sample is stored in memory. The signal sample is read from memory as necessary to process the signal. Such a receiver is useful in global positioning satellite (GPS) signal processing where the incoming signal contains several satellite transmissions encoded with CDMA encoding.

Description

The method and apparatus of the pre-acquisition frequency offset of elimination in the GPS receiver

The relevant application of cross-reference

The application requires in the rights and interests of the tentative Application No. 60/021618 of submission on July 12nd, 1996.

It is relevant that the application number of submitting on the application's invention disclosed theme and May 31 nineteen ninety-five is that 08/456229 the application of awaiting the reply " is used for from the center system based on GPS for the reduction power consumption of multiple target tracking ", the applicant of this application is Daniel D.Harrison, Anantha K.Pradeep, Glen W.Brooksby and Stephen M.Hladik, and assigns to the application's assignee.The disclosure of this application is incorporated by reference in this application.

The series number of the application's related application is [a reel number: RD-25187; RD25671; RD-25730; RD-25731; RD-25732; RD-25733; RD-25758; ], these applications are submitted to simultaneously and are assigned to the applicant's assignee.

Background of invention

Invention field

Generally, the application relates to spread spectrum communication system, particularly relates to the structure and the method for the power down signal treatment circuit that is used for band spread receiver.

Background note

Has advantage in the application that requires high reliability communication of spread spectrum communication under noise circumstance.Main noise is normally had a mind to or artificial disturbance unintentionally.In concrete application, communication environment may comprise a plurality of potential reverberations, produces serious multipath and disturbs.This multipath disturbs normally selects the form of decline to cause the communication disruption of the degree of depth with frequency.Spread spectrum communication especially is good at aspect difficult overcoming these.

The spread spectrum system that several types is arranged comprises Resistant DS Spread Spectrum System, frequency-hopping system, time-hopping systems, pulse frequency modulated system (being the chirp system) and various hybrid system.Certainly, this direct sequence spread spectrum (DSSS) system and frequency-hopping system may be to implement system the most widely.Following discussion also is to concentrate on binary system DSSS system.

In binary system DSSS communication, the bandwidth carrier signal is modulated by the message signale of arrowband.This bandwidth carrier signal is normally by adopting a binary system pseudo noise (P/N) sign indicating number sequence the single-frequency carrier wave to be carried out the quarter-phase modulation produces.The P/N sign indicating number normally produces with one or more high speed shift registers, and according to basic multinomial, each register has the feedback of mould two.The high speed P/N sign indicating number that produces is added to a balanced modulator (multiplier) subsequently, and another input of this modulator is a narrowband carrier.The output signal of this balanced modulator is a broadband signal that often is known as " bandwidth carrier ".In order to transmit data, this bandwidth carrier is carried out the quarter-phase modulation by a binary message data stream.This message data speed is normally much lower than this P/N code sign or " chip " speed, and the edge of these data and chip is normally synchronous.This DSSS technology directly depends on the ratio of this spreading rate and data rate for the supression ability of disturbing.In many application, in each message bit, there are several thousand chips.

Reproducing signals by the original narrowband carrier that produces with this machine multiply each other by (for example correctly tuning local oscillator) and earlier the DSSS signal down to base band, just can receive this DSSS signal.If the frequency of the carrier wave that duplicates (with phase place) is identical with the frequency of the original narrowband carrier that receives, then the signal of this multiplier output will be bipolar " wideband data " data flow, and this data flow is the product of bipolar P/N sign indicating number and information code sequence.Subsequently by the data flow of this wideband data and this machine duplicate that produce and that received this P/N sign indicating number of P/N sign indicating number time alignment being multiplied each other and this P/N sign indicating number being removed.The despreading process of data that Here it is, and produce the data flow of origination message data at the output of multiplier.

In the process of the despreading of data, this wideband data power spectrum is pooled original narrower data bandwidth again, in this bandwidth data power is risen on the background noise.The amount that this power level is raised is called processing gain, and directly is proportional to the ratio of bit rate and data rate.And received narrow band interference is duplicated modulation by sign indicating number and is expanded, thereby will be reduced in the interference power level in the frequency ranges of data widely.

Receive the relevant task of difficulty the most with the DSSS signal and be to produce frequency and all correct carrier wave duplicate of phase place, and produce the P/N sign indicating number with correct speed and reasonable time adjustment (skew).In many DSSS communication systems, in advance and do not know, and these parameters must can be determined till observing big signal in the output at digital filter by attempting different values on receiver in the skew of necessary carrier frequency, carrier phase and P/N sign indicating number.This is referred to as one and searches or acquisition process, and the DSSS signal will can obtain when all being determined when correct frequency, phase place and sign indicating number skew.

In many DSSS used, the level of this DSSS signal was more much lower than background noise and/or interference level, and can be detected up to despreading correctly with after being low pass filtering.When the SNR (SNR) of received signal when being very low, this filter must be very narrow bandwidth, so that realize for input and required processing gain for obtaining.Because narrow filter needs a long comprehensive cycle, so before can detecting judgement, must add up with the sample value multiplied result of the corresponding P/N sign indicating number that duplicates for a plurality of P/N sign indicating number sample values that received.This to multiply each other and be accumulated between the P/N sign indicating number sequence that has received and duplicate be crosscorrelation, and in order to realize low SNR, this sequence must not be not very long.

The use of this DSSS method has realized that a plurality of users adopt code division multiple access (CDMA) technology, side by side shares same broad band channel.Utilize this technology, each transmitter adopts a different P/N sign indicating number, so that make the crosscorrelation between different sign indicating numbers be actually zero.Specific transmission signal is selected and detected to receiver by selecting suitable substance P/N sign indicating number and carrying out and obtain search.In some cases, and do not know which transmitter may send, and this obtains search and must comprise the different P/N sign indicating number of check from a known tabulation.As many different sign indicating numbers, sign indicating number skew with carrier frequency must be verified and this SNR when being low value, then this obtaining of task will be a thing consuming time and power consumption.An important aspect of the present invention is to reduce the time in this DSSS signal acquisition and the consumption of energy.

The description of the spread spectrum communication system of direct sequence and other type can be referring to for example by RobertC.Dixon, John Wiley﹠amp; Sons is at " spread spectrum system " (SpreadSpectram Systems) of the third edition in 1994 with by the article " spread spectrum communication " of people such as M.K.Simon on PC Magazine (ComputerSciena Press) (1985) second volume.For example can be referring to by Andrew J.Viterbi (" the CDMA principle of spread spectrum communication " (in CDMA Principles of SpreadSpectram Communication) that A/Ddison-Wesley publishing house publishes nineteen ninety-five about the description of CDMA technology.

Knowing with the global positioning system signal of popular employing is a kind of important use that DSSS communicates by letter.In the last few years, in six elliptic planes during Navstar global positioning system (GPS) satellite has been transmitted in the height Earth's orbit, each tilts with respect to the equator was 55 degree.This complete satellite layout comprises 21 satellites and several backup satellites.Can make receiver in the position of coming accurately to determine the time and it self near ground from the signal of these satellite transmissions.Each satellite all sends the data that the exact position of its satellite situation is provided, and can realize the measurement of the distance from this satellite to the receiver user antenna.Be used to these information from least four gps satellites, the user can calculate its oneself position and speed and time parameter by known triangle location technology (being navigational solution).Usually, be reduced to if user's receiver has the expedite spacious visual field very near horizontal words, then normally have seven (being four at least) satellites will by on ground or near the user on ground observed to.Each satellite sends signal according to two frequencies that are called L1 (1575.42MHz) and L2 (1227.6MHz), and all satellites all are to adopt the CDMA/DSSS technology of formerly describing to share these frequencies.

More particularly, each satellite sends a single DSSS signal on frequency L2, and sends the DSSS signal that this identical signal adds another low resolution on frequency L1.The DSSS signal of this low resolution comprises a P/N sign indicating number, and this P/N sign indicating number has the repetition period that the 1.023MHz chip is cut apart speed and 1.0ms, and has 50 bps speed message data sequence (NAV data).What high-resolution DSSS signal used is that the 10.23MHz chip is cut apart speed and the repetition period of being longer than a week.Identical NAV data flow is used in all DSSS signals from a given satellite.These NAV data from this given satellite comprise the gps signal transmitting time, be used for astronomy (position) data of this satellite, be used at all satellites of this star battle array astronomical yearbook data (reducing the chronometer data of accuracy) with the transmission code word of excessive relevant use from low resolution to the high-resolution code tracking.Should be low and the high-resolution sign indicating number be referred to as process/obtain (C/A) sign indicating number and accurate (P) sign indicating number respectively.

After obtaining, the skew of each yard makes receiver determine in the satellite of correspondence and the scope between the user together with the signal sending time from these NAV data.By the C/A sign indicating number of P sign indicating number and repetition is included in the signal of transmission, makes and to realize that the classification of faster P sign indicating number obtains, and provide that worldwide navigation is professional to divide two-layer setting.This P sign indicating number can provide accurately approximately 3 meters position, and the C/A sign indicating number can provide the order of magnitude of 30 meters accuracy.In general, the use of the business of low resolution is unrestricted, and high-resolution business then is limited in military use by the control technology of encryption or other high-resolution P/N sign indicating number.

In common military receiver, the C/A sign indicating number is at first obtained.From this NAV data flow, read the transmission code word subsequently.This transmission code word has been stipulated the side-play amount roughly of the P sign indicating number relevant with gps time (sending) in time mark, and its use reduce significantly in P sign indicating number acquisition process the number of skew of the different sign indicating number that must search.Obtaining of C/A sign indicating number is in fact easy than obtaining of P sign indicating number, because the every 1ms of this C/A sign indicating number repeats once, and therefore has only the skew of 1023 different sign indicating numbers to be searched (if carry out this search in half common chip, this number will double).

The gps signal that receives normally will be offset from the frequency of normal L1 and L2 carrier frequency, because gps satellite moves several thousand kilometers each second on its track, produces tangible Doppler frequency shift.Therefore satellite orbit is known in advance usually, if the position of this GPS receiver is known words then the carrier frequency of this Doppler frequency shift will be foreseeable.Regrettably, the position of receiver do not learn in advance, and often has the error of using the local oscillator that cheap receiver causes.This will cause uncertainty (promptly duplicating the uncertainty in the carrier frequency) in reception carrier frequency strengthen (for example ± 7.5KHz), and this frequency range may must search in the process of obtaining at gps signal.Duplicate by the crossing dependency that repeats to receive sample value and this machine that is used for different local oscillators normally that P/N sequence (carrier wave duplicates) frequency is carried out frequency or Doppler searches.Interval between the frequency step is enough little by what do, so that avoid lossing signal when using the long crosscorrelation time of integration (narrow filter bandwidht).The long detection that has improved low SNR signal the time of integration.In common civilian GPS used, the crosscorrelation integration (single C/A sign indicating number cycle) of use 1ms drew the bandwidth of the equivalent Doppler filter that is roughly the 500HZ step-length.Utilize the step-length of 30 500Hz to search ± frequency range of 7.5KHz.The acquisition of this GPS has then realized the search in satellite code, sign indicating number skew and Doppler frequency.

A main control station (MCS) and several monitoring stations have constituted the control section of this GPS.This monitoring station is followed the tracks of all gps satellites in the visual field passively, is collected from the range finding collection and the satellite clock data of each satellite.These information are sent to MCS, and wherein the drift of the chronometer data of satellite and clock is predicted.The astronomy and the clock data that upgrade are input to each satellite, are used for the transmission again in the NAV of each satellite information.

In operation, common GPS receiver is carried out each the following operation that is used at least four satellite-signals:

1) obtain the DSSS signal,

2) synchronous NAV data flow and read time mark, clock correction, ionosphere delay and the chronometer data of satellite,

3) go out the position of satellite from these astronomical data computation,

4) read the clock of self receiver, so as to determine the receiver relevant with the reception in time mark epoch time and

5) by deducting the value of time mark the time from relevant receiver, the walking the time of estimating signal.

This time difference is multiplied each other by the light velocity, so that obtain the estimated range corresponding to this satellite.If the clock that this GPS receiver has and the clock synchronization of this satellite be synchronous (or error is known) in good condition, then as long as the satellite of three this scopes can be realized the accurate localization of receiver.But because common GPS receiver uses is the clock of crystal oscillator at a low price, thus exist the skew (becoming error slowly) of clock, but satellite but is being equipped with atomic clock.The skew of this clock is found and uses this measured value to eliminate this deviation effect by measuring apart from four gps satellite distances (propagation time) and in the system of four equatioies with four unknown numbers (receiver x, y, and z, and time).For the general information that is relevant to GPS, the reader can be with reference to " NAVSTAR Global Positioning System " book of being shown by Tom Logsdon of being published by Van Nostrand Reinhold publishing house in 1992.

Best use of the present invention is location and the tracking to the facility that uses gps system, for example railcar, steamer or freight house, truck etc.In these are used, be used for that special power supply power supply normally can not be arranged, all these GPS receivers are normally battery powered.Therefore the consumption of the energy by reducing receiver will help to increase the life-span of battery.

In general band spread receiver, the front end of receiver (being RF and IF electronic installation) will consume a large amount of power in it is connected.If the acquisition of signal and the time that takies synchronously are long, this will cause high energy consumption.GPS receiver in most prior arts does not have the storage device (memory) of signal but must handle the signal of reception in real-time mode.And, these receivers or the search of use order, or side by side search a spot of satellite/sign indicating number-skew/Doppler (SCD) receiver (bin) and realize the acquisition of signal.This receiver must receive and handle each satellite-signal continuously, up to the SCD of signal receiver identified and necessary NAV data decoded till.Owing to will be passed through the actual time before identifying at the SCD receiver that is relevant to each gps signal in the sequential hunting, so the energy consumption of this sequential hunting is high.In addition, a plurality of SCD receivers can be searched concurrently so that reduce the used time, but because existing processing method is not the processing method of very low energy consumption, so the energy consumption of this and line search is still very high.And, because need a large amount of circuit arrangements, be very limited so adopt the degree of the parallelization of existing processing method.

In a system of the present invention, the necessary a plurality of equipment (for example railcar) of following the tracks of of central facilities or station.Each tracked target has for the GPS receiver of handling from the signal data of several visible gps satellites; But do not carry out accurate position determines at the receiver place.On the contrary, just carry out the processing of part and the result of centre is returned to central station at receiver.Do not need the to navigate decoding or of these intermediate object programs from other data of gps signal.Therefore this system time span of allowing GPS receiver and signal processor only to be energized to be enough to obtain satellite-signal gets final product (determining this SCD receiver).Utilize this system, main energy consumption is acquisition process, and if the acquisition time of signal and energy are considerably reduced, the energy that this GPS receiver uses on each device will greatly reduce.

The United States Patent (USP) 5420593 of authorizing Nilles uses the time interval that memory is stored the received signal that comprises a plurality of gps satellite signals.The signal that has received is sampled and be written in the memory with a speed, and reads with faster rate subsequently.When reading, signal is digitized processing, so that the gps satellite signal that acquisition receives and synchronous with it.This just realizes the short time spent for the acquisition of gps signal.But after signal storage, receiver does not cut out immediately, thereby can not use the signal of low power supply to obtain.And, can not realize the reduction of substantial energy.

Authorize the United States Patent (USP) 5225842 of Brown and described a kind of centralized infrastructure tracking system, by avoiding on this facility navigational solution being calculated the cost that reduces the GPS receiver on each tracked facility based on GPS.Each described facility all is loaded with the GPS receiver, handles the signal from several visible gps satellites, and this central station accurately calculates the navigation position of tracked target according to the result who handles.In fact this system does not reduce the energy consumption of the GPS receiver on facility, and life-span of extension fixture battery or reduce to change the number of times of battery not in fact.And, do not adopt the parallel correlation technique of low-power consumption.

Brief summary of the invention

An object of the present invention is to provide a kind of direct sequence spread spectrum (DSSS) signal processing structure, it can make receiver be closed in the most time of obtaining in the stage that receives, thereby is reduced significantly the turn-on time of the device that realization links to each other with receiver front end.

Another object of the present invention provides a kind of signal processing structure, can realize low-power consumption in the stage of obtaining that the DSSS signal receives.

Another object of the present invention provides a kind of parallel correlation technique of low-power consumption, is convenient to be realized and used by available integrated circuit the method acquisition DSSS signal of low power consuming.

Another object of the present invention provides a kind of gps signal and handles structure, utilize the sequence processor (promptly slow relatively and have little information source) of standard to carry out the P/N sign indicating number and Doppler search (with obtain handle relevant), and do not require that during searching the front end of receiver is switched on.

A further object of the present invention provides a kind of gps signal and handles structure, requires to follow the tracks of the facility that uses GPS with very little energy.

Another purpose of the present invention provides a kind of signal processing structure, can be according to the relevant and incoherent generalized time of exchange at the needs of the signal to noise ratio (snr) of concrete received signal and dynamically.

A further object of the present invention is to adopt a kind of low receiver to export sampling rate and obtain obtaining regularly of accurate sub-chip DSSS signal.

A further object of the present invention provides a kind of gps signal and handles structure, shows obtaining of gps satellite signal fast.

Another purpose of the present invention provides the method that a kind of gps signal obtains, though when GPS use a cheapness, also can realize obtaining of low energy signal fast in the time of may being coarse local oscillator.

According to an aspect of the present invention, during the DSSS signal obtains, use the high-speed parallel correlator of a low-power consumption, so that reduce the energy consumption of acquisition process.Owing to the time that the speed of parallel correlator makes receiver be in turn-on condition is lowered, so the energy consumption of receiver is further reduced.

According to another aspect of the present invention, the time interval of receiver dateout is sampled and be stored in the memory, and subsequently receiver is turn-offed.The length in the time interval is enough to make to realize to be included in obtaining of any CDMA/DSSS signal in the receiver dateout of storage.The receiver dateout of storage is repeated to read from storage and can requires to obtain each required CDMA signal so repeatedly.In order to keep the low energy consumption of processing procedure, parallel (partial simulation) correlator of a low energy consumption of available use in this acquisition process.The energy that this scheme is used than existing be much smaller at the use energy of cross correlation.

Concentrated GPS tracking system according to the present invention makes tracked facility consume the receiver that low-down energy consumption and use have cheap local oscillator.Coarse local oscillator is used in the permission of parallel correlator, because the many frequencies in the acquisition process that uses low-power consumption are promptly searched.And the use of parallel correlator can be avoided the reception and the demodulation of GPS NAV data, has further reduced the average turn-on time of receiver.No longer need almanac data to support the satellite search time of an attenuating, because should can on all satellite code, search apace by parallel correlator.Owing to do not need navigation procedure at tracking cell, so only need the very low limited processing procedure of energy consumption, this navigational solution produces at central station.

Brief Description Of Drawings

Believe that novel these features of the present invention are described in the appended claims.But further purpose of the present invention and advantage can be with reference in conjunction with the drawings following descriptions and are understood more thoroughly.

Fig. 1 is the block diagram according to a remote tracking system of the first embodiment of the present invention;

Fig. 2 is the block diagram according to a tracking cell on tracked target of the present invention;

Fig. 3 is the block diagram that conventional sequence is obtained structure;

Fig. 4 is the block diagram of the structure of sequential signal processing according to an embodiment of the invention;

Fig. 5 is the block diagram of parallel signal processing structure according to another embodiment of the invention;

Fig. 6 is a schematic diagram of the data and the parallel correlation procedure between the waveform reproduction of slip of storage;

Fig. 7 is the block diagram that expression produces the structure of this replication sequence;

Fig. 8 is that the block diagram in the embodiment simplification of duplicating data storage, reproducing signals register and parallel correlator in the production method of Fig. 7 is used in expression;

Fig. 9 is the block diagram of an embodiment of the one-level of the 2-sequence parallel correlator among Fig. 8;

Figure 10 is the block diagram that expression is used to produce a kind of method of separate code and Doppler's replication sequence;

Figure 11 is the block diagram of an embodiment of the expression data register that duplicates production method, sign indicating number copy register, Doppler's copy register and the parallel correlator that use Figure 10;

Figure 12 is the block diagram of embodiment of one-level of parallel correlator of the 3-sequence of Figure 11;

Figure 13 be represent differential digital-analog converter according to an aspect of the present invention and get and the schematic diagram of principle,

Figure 14 illustrates the schematic diagram of the noiseless waveform automatic relevant with noise, is used for an of short duration part near the 1.0ms C/A sign indicating number P/N sequence of correlation peak;

Figure 15 is the block diagram of expression low power consumption parallel correlator for the system that obtains on basis, has complete IQ and handles, and RF/IF wherein partly produces the output of two quadratures;

Figure 16 is the block diagram of an embodiment of square A/D converter;

Figure 17 is data storage, duplicate generation, the block diagram of parallel correlator combination, and this combination has reduced the number of register shift of the sign indicating number of each yard/Doppler's combination;

Figure 18 is the schematic diagram of the digital-analog convertor unit of one 3 sequence multiplier;

Figure 19 is the schematic diagram that a ping-pong data storage constitutes;

Figure 20 is the schematic diagram with parallel correlator that data-Doppler multiplies each other in advance;

Figure 21 is the block diagram with whole I-Q processors that data-Doppler multiplies each other in advance;

Figure 22 is the block diagram with whole I-Q processors that data-Doppler's plural number multiplies each other in advance;

The block diagram of the low-yield register wiring method that is to use a scan shift register shown in Figure 23.

The detailed description of the embodiment of the invention

A plurality of gps satellites 12 have been shown among Fig. 1 have followed the tracks of a target (facility), for example carried motor vehicle railcar and central station 16 of a tracking cell 14.As previously mentioned, each satellite 12 sends a signal, and the GPS receiver in tracking cell 14 uses the propagation delay (if wish learn the speed of facility, then measure delay speed) of this signal measurement from this satellite to receiver antenna.This satellite-signal comprises also and repeats the NAV data periodically that this is for determining that from measured time delay navigational solution is necessary.Because the low rate (50 bps) of NAV data in this gps signal, so if collect this NAV data, receiver will be opened a quite macrocyclic length time (from 1 to a few minutes).And the NAV data of a particular satellite change in time, and this GPS MCS monitors that these change, and to provide almost be the NAV data of upgrading by the hour.In order to guarantee accurate navigational solution, any GPS is that the navigation system on basis all must be used not old in about four hours NAV data.If the supervision of the position of facility is more frequent than per four hours, then these new NAV data must be to collect once in per four hours.The maintenance of NAV data then require the operating period of receiver each hour be average per 15 seconds once, if all carry out, will cause considerable energy requirement at each tracked facility.

According to an aspect of the present invention, navigation separate be calculate at central station rather than on facility, calculate.On tracked facility without any need for the NAV data.Have only the relevant number of propagating with the gps signal between each satellite and this facility of delay on this facility, to measure, and these data are sent to central station subsequently.These NAV data can be with communicating by letter with in central station 16 and determine by standard GPS receiver in central station or by what carry out with the standard GPS receiver of suitable location.As required, these NAV data or navigational solution can by needs still less the communication link of the higher rate of energy be sent to receiver on the facility of being followed the tracks of.Owing on this facility, do not need the NAV data decode, thus the gps signal on this facility obtain that to become mainly be this GPS Processing tasks, and the acquisition methods by low energy consumption of the present invention makes and concentrates the flexibility of tracking system greatly to strengthen.

As shown in Figure 2, the tracking cell 14 of a railcar comprises the receiver 2 of response from the signal that receives at antenna 5 of gps satellite; A processor 3 and a reflector 4.The signal that receives is handled in processor 3, so that find out and utilize from the difference in transmission time between the signal of these gps satellite receptions.By utilizing these time differences, alleviated the needs of learning this gps signal time mark at this facility, and thereby need not carry out the decoding of data flow at this facility.Owing to do not need the data decode of gps data stream, the processing procedure of receiver is reduced to obtaining of this gps signal and calculates the relevant time difference (if determine the speed of facility, it is poor then also to calculate Doppler frequency) from the result who obtains.The time difference that calculates sends to central station by transmitter 4 from antenna 6 with the time that shows the satellite relevant with this difference.This related application No.08/456229 that the centralization tracking system of obtaining the major part of paying as the GPS energy of gps signal is found in above referenced Harrison, Pradeep, Brooksby and Hladik.

Fig. 3 illustrates and uses a conventional serial of serial correlator to obtain structure.In the GPS of routine receiver, the demodulation of obtaining the synchronous and NAV time that is afterwards carrier wave and P/N sign indicating number of signal, but the module of these processing does not illustrate in Fig. 3.The system configuration that sort signal obtains comprises RF/IF (radio frequency/intermediate frequency) part 21, and this part comprises antenna 211, RF amplifier 212, frequency mixer 213, local oscillator 214 and low pass filter 215.This low pass filter 215 is provided to an analog to digital (A/D) transducer 22.A/D transducer 22 to that received and signal down converted and normally takes a sample with the integral multiple of the speed of duplicating the C/A chip and change, and a Serial No. is provided to a serial digital correlator 23.This correlator 23 calculates serially from the subsequence of the digitlization receiver of A/D converter 22 output with from the inner product of the C/A replica code subsequence of sign indicating number/Doppler's (or duplicating) generator 24.The generation of inner product realizes by carrying out following process: at first first in two subsequences is multiplied each other in multiplier 231, and the result is stored in the relevant accumulator 232, subsequently second in two subsequences is multiplied each other and they long-pending is added in the relevant accumulator 232, by that analogy.According to the order of the subsequence that can get from A/D converter 22, this inner product is to carry out in real-time mode.As in usual manner, subsequence is with crossing over the single cycle that repeats the C/A sign indicating number.As calculated after this inner product, this relevant accumulator comprises the sampling of the crosscorrelation that receives subsequence in a C/A sign indicating number cycle with reproducing signals, is used for the skew of specific C/A sign indicating number, sign indicating number and by duplicating the Doppler frequency that generator produces.After from the subsequence of A/D converter 22, several times are carried out in the operation of asking inner product, used the identical subsequence of duplicating the C/A sign indicating number simultaneously.The result of the interior product of this subsequence is subsequently by 29 squares of squarers, and in irrelevant accumulator 30, get and, draw relevant treatment result's uncorrelated integration.The output signal of this irrelevant accumulator 30 is detected by threshold detector 31, and if the level of the signal in irrelevant accumulator 30 is sufficiently high words, then produces the order of " signal acquisition ".When obtaining a signal, controller 27 monitors the skew and the Doppler frequency of relevant C/A sign indicating number index (satellite index), sign indicating number, and command code/Doppler's generator 24 or be transformed into different C/A sign indicating number (being used for different gps satellites) and begin another and search is if or received enough satellite-signals then stop to search.If do not have picked up signal after for the plurality of sub series processing from A/D converter 22, then controller 27 instruction this yard/Doppler's generators 24 change to different C/A sign indicating numbers, sign indicating number skew or Doppler frequency.Along with the acquisition of each satellite-signal, controller 27 is provided to the synchronous and NAV time processing unit (not illustrating) of gps signal to relevant sign indicating number index, sign indicating number side-play amount or Doppler frequency.

Must be provided for searching the scheme of all possible C/A sign indicating number, sign indicating number skew and Doppler frequency.Controller 27 is by desirable C/A sign indicating number of Instruction Selection and sign indicating number skew for satellite code generator 243 and skew generator 244.With respect to the bit data flow from A/D converter 22, skew generator 244 provides from the time offset of the replica code of satellite code generator 243 generations.The sinusoidal numeral by the combination of the Doppler frequency shift of controller 27 realizations and local oscillatory frequency error is represented in 242 generations of Doppler I/Q generator.The product that the reproducing signals that is produced by multiplier 241 duplicates as this sinusoidal signal and C/A sign indicating number.For guarantee gps signal detection, the output signal that this RF/IF part 21 must both produce homophase (I) also produces quadrature (Q) output signal (not illustrating among Fig. 3), and these two signals are processed so that realize obtaining of signal.And each of this I and QRF/IF output signal must utilize I and Q Doppler sinusoidal component to handle, and is known as the technical staff in the GPS field.

The conventional scheme of exporting in Fig. 3 requires to handle as quickly as possible after obtaining from the receiver dateout of RF/IF part 21 outputs, and the processing of carrying out is subjected to the restriction of the bit rate in the received signal.The formation that is used for the processor that improved sequence signal obtains according to the present invention is shown in Figure 4.The formation of this system is similar with the formation of Fig. 3, but difference is to have added a signal storage memory 33, and is controlled by controller 35 for the power supply of RF/IF part 21.The gps signal that RF/IF part 21 provides converts digital form to by RF/IF part 21, but the A/D sampling rate of this moment can be set at the non-integral multiple of C/A bit rate.Memory 33 storage is enough to the input signal length that realizes that signal obtains, and makes this RF/IF part shutoff after storage.This acquisition process is read and may is to read the data that are stored in the memory again subsequently.Because this RF/IF part 21 consumes most energy, energy consumption in this way be considerably reduced.And the process of obtaining no longer is subjected to resembling the restriction of the bit rate of the received signal of pointing out in prior art.This non-integral multiple input A/D sampling rate makes this obtain processor to determine accurate gps signal propagation data poor (is necessary for definite this position result), use much lower sampling rate simultaneously.And, as SNR when not being very low, or when the requirement of accuracy when not being very high, so this low non-integral multiple sampling rate makes the time difference of determining have enough accuracy, thereby make it possible to avoid required usually traditional carrier wave and P/N sign indicating number Synchronous Processing (for example be used for the Costas phase-locked loop of carrier track and be used for sign indicating number tracking leading-delay phase-locked loop lags).

The data set of received signal is stored in the memory and is that according to the advantage that the needs of handling different SCD receivers are read this memory again relevant treatment can be carried out and not have to obtain losing of accuracy owing to the unsteadiness or the caused signal of inexactness of local oscillator on a time cycle.And, if the signal of storage also is used to be extracted as the delay that obtains the needed time of navigational solution, then just need not remain on GPS reception obtain and Phase Tracking between accurate timing.For not requiring the occasion of using this navigational solution in measurement (for example tracking of facility) afterwards immediately, can use one very the integrated circuit of low-power consumption constitute a very GPS receiver of low-power consumption, sequence correlator (and other circuit) and low power consumption data memory are combined.The unusual large scale integrated circuit of low-power consumption (for example 1.5V, 0.35 micron complementary metal oxide semiconductors (CMOS), i.e. CMOS technology) aborning in industry.

Utilize this gps system, as in most DSSS system, SNR before signal processing is very low, and the cycle of reality that must be by received signal is the peak value of necessary high SNR correlation so that obtain detecting for the reliable signal of being carried out by threshold detector 31.Use for general civilian GPS, must processed and storage from the signal of about 20ms of RF/IF part 21.In order to make this memory remain on low capacity, from the signal of RF/IF part 21 by with the low rate sampling and be quantized into and have only several few level.For civil applications, common GPS receiver realizes that generally 30 meters GPS fixes (location) precision.Location with this precision GPS can be calculated from the side-play amount of signal code, and this side-play amount is with less than the error metrics of 1/10th C/A chip.By the side-play amount of indicating that the reproducing signals sign indicating number side-play amount that is associated with the crosscorrelation peak value is come the measuring-signal sign indicating number.In one embodiment of the invention, produce sequence as a result, keep the invariable of sign indicating number exponential sum Doppler frequency simultaneously so that increase the side-play amount of replica code from irrelevant accumulator 30.If observe big correlated results, sequence as a result and the estimation sign indicating number side-play amount related that a kind of interpolation algorithm is used to produce then with correlation peak.Realize 1/10th sign indicating number side-play amount accuracy of chip.And simultaneously with this signal of sample at rates of the twice that is similar to this C/A spreading rate.Some traditional GPS receiver can not be realized the accuracy of desirable 1/10th chips in the acquisition process of signal; On the contrary, some GPS receiver is at this signal of taking a sample with the C/A spreading rate of twice, and can realize this accuracy between the sign indicating number sync period, and tuning this sampling phase is as the part on the leading-delay phase-locked loop road that lags.A kind of in addition scheme, other conventional GPS receiver is by taking a sample this signal and indicate that this yard skew realizes the accuracy of 1/10th chips with ten times of C/A spreading rates, and this will obtain maximum correlation peak.Reduce sampling rate and require less memory and lower processing speed.

At the SNR of the input of A/D converter 22 far below zero, so the bad change of signal when this A/D converter only uses three correct expression level of selecting, just seldom occurs.Each sampling can only adopt two data bits to be encoded easily according to the amplitude of symbol.In order to guarantee the detection of signal, the RF/IF input signal of I and Q all must be produced, be stored and be handled (only showing a RF/IF channel in Fig. 3 and Fig. 4).Known as GPS technical professional, use three or more expression level that substantial antijamming capability for non-gps signal can be provided.But,, then just require less data storage, and this has the advantage on the price in some applications if just adopt two expression level (bit) for I and Q signal.In one embodiment of the invention, the RF/IF output signal of I and Q is by digitlization side by side and storage, and the length of this data storage is enough to remain the whole series of data that realizes required time (for example 20ms).The RF/IF signal that utilizes every millisecond of 1023 C/A chips and I and Q all is with two sample values of each chip, and each sample value has two bits, and then having about 170000 bits for the signal data section of 20ms needs storage.For convenience's sake, this I and Q data can be considered to be stored in other I of branch and the Q memory.In, conversion sampled when desired data and when storage,, this receiver just can be closed, and the data that write down can be processed.

In the method for order of the present invention, the data of storage are reset (reading) once in skew that is used for each combination of sign indicating number, sign indicating number and Doppler frequency shift processing procedure.In the system of Fig. 4, the sequence of the digital data samples of storage is read from memory 33, reads a sample value at every turn.Multiply each other from each sample value of the sequence of memory 33 sample value, and the result is accumulated in the relevant accumulator 232 by the sequence of the sign indicating number/Doppler's generator 24 in multiplier 231.From this yard/Doppler or duplicate the sequence of generator or data segment is produced at concrete sign indicating number, sign indicating number skew and Doppler frequency under test condition.So the length of the memory sequence of handling is exactly the comprehensive length of being correlated with, and is chosen as a total length of this C/A sign indicating number usually, promptly is 1.0 (ms).Handle the data segment of several (for example 20) adjacent 1.0ms data in this way and do not change duplication sequence.After the data segment of each 1.0ms is processed, is stored in data value table in the relevant accumulator 232 and is shown in correlation between duplicating of this 1ms and the data segment (sequence).This value is by 29 squares of squarers and be added to irrelevant accumulator 30.Before the data segment of first 1.0ms that is directed to given reproducing signals is processed, irrelevant accumulator 30 is reset to zero, so that make the total relative recording (score) that is used for special code, sign indicating number skew and Doppler frequency that final accumulation result representative is limited by this reproducing signals.Similarly, this relevant accumulator was reset before the data segment of each 1.0ms is processed.Threshold detector 31 monitors record that this is relevant, and if this record when being thresholding greater than regulation, then produce " signal acquisition " signal.When receiving this " signal acquisition " signal, controller 35 is carried out simple peak value-search and interpolative operation (following will the description), so that find the best estimate of the sign indicating number skew relevant with given sign indicating number exponential sum Doppler frequency under testing conditions.Subsequently, sign indicating number, sign indicating number skew and Doppler frequency combination that controller 35 selections are other, and instruct and duplicate generator change reproducing signals, so that reflect this change.For a plurality of reproducing signals corresponding to sign indicating number, sign indicating number skew and the Doppler frequency that will search, the acquisition process of this signal is repeated repeatedly, and when obtained, stops this processing procedure in the presence of the number (different C/A sign indicating numbers) of desirable gps signal.Controller 35 produces and obtains the designator of the sign indicating number of signal correction, the deviant and the Doppler frequency of estimation subsequently, as output signal.

Use a digital correlator 23 as shown in Figure 4, the memory data of I and Q can sequentially be handled (for example handle all I data, and then handle all Q data).In addition, the memory data of I and Q can side by side be handled by using other digital correlator of branch.Under any condition, in order to guarantee obtaining of signal, I and Q Doppler handle and must both carry out for the I memory data, also the Q memory data are carried out.Therefore have four kinds of I-Q combinations, and they can sequentially be handled by a digital correlator, perhaps handle simultaneously by a plurality of correlators.In a kind of method of sequential processes, calculate according to following mode with the correlated results of a given replica code index, sign indicating number skew and Doppler frequency for whole memory data orders: at first relevant and uncorrelated accumulator is reset to zero.Subsequently, duplicate the processing of I memory data, and relevant result is added up in relevant accumulator with I Doppler.Duplicate with Q Doppler subsequently the Q memory data is handled, and relevant result is further added up in relevant accumulator.Whole relevant accumulation results is subsequently by 29 squares of squarers and be added in the irrelevant accumulator 30.Should reset subsequently by relevant accumulator.Subsequently, the Q memory data is by I Doppler replication processes, and a relevant result is accumulated in the relevant accumulator, subsequently, the I memory data is by Q Doppler replication processes, and relevant result's anti-phase (multiply by-1) and further be accumulated in the relevant accumulator.Total relevant result who adds up is subsequently by 29 squares of squarers and be added to irrelevant accumulator 30.The memory data section that this process is directed to each 1.0ms repeats, and does not reset for irrelevant accumulator between data segment, and requires completely that readout interval is used for I and Q memory data.Divide other Doppler IQ generator 242, multiplier 241 and digital correlator 23 by using, can reduce the time and the energy that obtain, so that side by side handle four all I-Q combinations.

In a traditional mode, sign indicating number/Doppler's generator 24 comprises Doppler I/Q generator 242, satellite code generator 243, skew generator 244 and multiplier 241 as shown in the figure.Doppler I/Q generator 242 produces sine with the desirable frequency that is relevant to sampling rate or the cosine wave numerical order is represented, and for example can be realized by the ROM that address counter drove.Order be can be read out by the sine sampling of control store and sinusoidal wave frequency and phase place (I or Q) (for example the selection by address discriminating and start address realizes respectively) selected.Other digital sine baud generator also is known for the professional and technical personnel.For GPS C/A sign indicating number, the state machine of knowing can be used to produce C/A sign indicating number bit with correct order.Multiply each other for Doppler's sine wave and C/A code value by multiplier 241, so that produce reproducing signals.Being used to search all available yard schemes that are offset is by using skew generator 244 to provide for the time migration of duplicating from the memory code of satellite code generator 243.This state machine that utilization is realized by satellite code generator 243 by before the beginning relevant treatment state machine being preset to relevant state, is realized the special code skew.Initial condition corresponding to different yards skews can be stored among the ROM, and by a simple binary address counter retrieval.

In practice, by at first selecting a sign indicating number, obtain retrieval and carried out easily with a Doppler frequency and subsequently by the different yard retrievals that skew is carried out.Subsequently, Doppler frequency is changed and different sign indicating number skew is checked again.Irrelevant accumulator 30 can be used as the array of an accumulator and realize.Each accumulator is used for a Candidate key skew in the order of adjacent skew.The formation of this accumulator makes that the mark that is used for adjacent skew is side by side checked, and finds the subdata field offset value relevant with the peak value of this real marking so that make it possible to use an interpolation algorithm.If employing order interpolation algorithm, this peakdeviation value can be interpolated, although only use the signal accumulator element.

Availablely reduce total acquisition time by side by side handling a plurality of yards skews.For example, a plurality of digital correlators 23 can be used, and each different delay version by reproducing signals is driven.The delay line of a tap that the available usefulness of different delays is coupled to the output of sign indicating number/Doppler's generator 24 is realized.Each of this different tap can drive other serial digital correlator 23 of branch subsequently, and the result of the output of each correlator can by the relevant parts in irrelevant accumulator array respectively quadratic sum add up.

The scheme that is used for the GPS receiver that low-power consumption and fast signal obtain according to another aspect of the present invention is shown in Figure 5.Structural similarity shown in structure shown in this figure and Fig. 4, but wherein serial digital correlator, serial Doppler and P/N sign indicating number generator and series read-out memory are replaced by Parallel Digital correlator 36, parallel doppler and P/N sign indicating number generator 37 and parallel read-out memory 33 respectively now.One aspect of the present invention be carry out in parallel correlator that a large amount of Parallel Simulation are got and method.This simulation get and, form, duplicate generator and correlator unit together with memory a large amount of parallel, making that the time takies with energy consumption in correlation procedure reduces widely.Get and the result of this simulation are converted to digital form by A/D converter 38, and according to described such subsequently, this transducer 38 can be used and the squarer combination.This parallel organization has also utilized the advantage of traditional low-power consumption complementary metal oxide semiconductors (CMOS) (CMOS) integrated circuit, so that realize the use of low energy consumption.Power consumption in cmos circuit is determined by the charging of the node capacitor of circuit and discharge; Those voltages static (not changing) and or the energy consumption at the very little node place of electric capacity very little.Utilize the present invention, Doppler and P/N sign indicating number duplicate generator and register, data storage and parallel correlator are configured the CMOS node that reduces to be recharged and discharged in the process of relevant treatment number.

Fig. 6 shows and the notion of line correlation and the waveform that the signal of having received He duplicate does not all have Doppler's shift frequency is shown.Digital signal data is written sequentially to data storage, and making can be used for RF/IF part 21 and A/D converter 22.Data storage 33 is constructed for large batch of and line output, so that make that the long sequence of data is side by side obtained in output.And the selectable reproducing signals of shift register 1004 input, and its formation is to be used to have and data from the identical in enormous quantities and line output of the data length of data storage.The parallel data sequence and the sampling crosscorrelation between the parallel replication sequence (being inner product) that are used for given Doppler frequency, the skew of sign indicating number exponential sum sign indicating number are all produced immediately by parallel correlator 1000, the multiplier of the correspondence of use in multiplier array, each composition of memory data sequence is multiplied each other by the tie element of replication sequence.The output signal of multiplier side by side got and, so that be formed on correlator output relevant treatment result.Keep the stable of this memory data by being shifted step-length of this copy register simultaneously, and produce the relevant result that is used for the adjacent code skew.In addition, although this memory data is shifted, it is constant that this reproducing signals can be held.

In an attractive embodiment of the present invention, this parallel output data and replication sequence are the length of 1.0ms, and it has covered the one-period of this C/A sign indicating number.By taking a sample with the speed that is similar to two sample values of each C/A chip and being digitized into three level, from the RF/IF output signal, produce data sequence with the A/D converter that uses a dibit symbol-amplitude form.If this A/D threshold level is correctly selected, the sampling rate of level and number will be avoided the bad change with substantial SNR obscured of data, and be created on the sequence length of 2100 number of samples magnitudes.This symbol-amplitude form makes and use simple multiplier in above-mentioned multiplier array.The length of the data representation of other available consideration, sampling rate and parallel formation sequence is possible, and is obvious for the professional and technical personnel.

Fig. 7 shows and is used to produce an also method of line output replication sequence.In sign indicating number/Doppler's generator 1008, with the phase place and the Doppler frequency of hope, a C/A sign indicating number generator 1001 produces desirable C/A sign indicating number sequence, and a digital sinusoidal signal generator 1002 produces digital sinusoidal wave sequence.Multiplier 1003 sequentially multiplies each other this yard with by the sinusoidal wave sequence order ground that sign indicating number/Doppler's generator 1008 provides so that produce the sequence of duplicating, and this sequence by bit shift in parallel sign indicating number/Doppler's (duplicating) register 1004 of importing.

Replication sequence preferably shows as the symbol-amplitude form of three level (two bits), and perhaps two level (bit) signal formats are because these forms for the form that uses more bits, have reduced the complexity and the power consumption of parallel correlator.Use three or level still less, the Doppler's composition in replication sequence will have higher harmonic content, and these harmonic contents can with the input the signal spurious correlation.The avoiding of this problem realized from zero frequency offset by the RF/IF local oscillator frequency being chosen to this RF/IF output signal reality.By side-play amount being selected enough big, all harmonic waves that produce necessary Doppler's frequency reproduction all will be on the Doppler frequency reproduction the highest far above this.This harmonic problem is known for the superhet professional and technical personnel of design.If it is wish, available by in Doppler and replication sequence sign, adopting the more bits of more bits and each sampling in copy register to reduce Doppler's replica harmonic level.But this will increase power consumption and complexity (size) that signal obtains implement device.

Fig. 8 illustrate the reproducing signals production method that uses Fig. 7 data storage 33, duplicate the local pie graph of shift register 1004 and parallel correlator 1000.Data sequence and replication sequence all are to use the symbol-amplitude of dibit to represent, and because the row of the unit in each of memory 33, shift register 1004 and correlator 1000 is to aim at according to the row mode each other, so from the symbol (S) of the corresponding sample of two sequences and amplitude (M) can be fed into correspondence easily with the value of each input multiplier 1200.Be used to the symbol-amplitude input value from these two list entries, for example-1 or 1, each multiplier 1200 all produces-1,0 or 1 as output signal.Independent digital-analog convertor 1300 converts the output signal of each digital multiplier to analog form.By output signal is provided to the expression this simulation correlated results common output, the output signal of all D/A converters got with.Adopt electric charge get and just can realize easily this simulation get with, but other simulation is got and form also be available.This get and method effective especially, need very little power, do not require any analog memory, and fast especially.Another benefit of the structure of this parallel correlator is that this sign indicating number and Doppler's sequence are programmable, allows single correlator to search apace on a plurality of yards indication, skew and Doppler frequency in acquisition process.

Shown in Fig. 9 is the embodiment of the one-level of the two sequence parallel correlators in Fig. 8.Multiplier 1200 produces symbol and amplitude output bit, is controlled at the switch 1400 and 1500 in the D/A converter 1300.Switch is got a end with capacitor 1100 to electric charge and is connected to that positive voltage reference connects end or the negative voltage benchmark connects end, perhaps is connected to output reference voltage and connects end (for example ground connection).Get and be to have the process of two steps to produce.At first, charging-reset line is set to low level, and charge closing reset switch 1600 and the electric capacity that force switch 1400 handles in each parallel correlator to be correlated with are received output reference and connect end (ground among Fig. 9).This will be to all capacitor discharges.Subsequently, the charging reset line is set to high level, operation charging reset switch 1600, and make the data in each parallel correlator circuit stages control relevant symbol and magnitude switch 1500 and 1400 respectively with copy data.The processing procedure of this two step has guaranteed excessive charging not occur on time period with electric capacity work is got in charging.Its advantage is can for example metal-metal cross capacitor and binary system electronic switch be realized this parallel correlator by using.Finally, be can be to get simply that extension with circuit realizes that electric charge is got and expansion on a plurality of integrated circuits for its advantage.

Figure 10 shows the another kind of method that produces reproducing signals.Utilize this method, sign indicating number and Doppler's replication sequence are stored in the parallel output register.The C/A sign indicating number generator of sign indicating number/Doppler's generator 1010 produces the C/A sign indicating number sequence of wishing, and this sequence is moved in yard copy register 1005.Similarly, the digital sine baud generator 1002 of sign indicating number/Doppler's generator 1010 produces digitized sinusoidal wave sequence with desirable phase place and Doppler frequency, and this sequence is moved in Doppler's copy register 1006.

Figure 11 shows data storage 33, sign indicating number-copy register 1006 and the parallel correlator 1000 of the copy data production method that uses Figure 10.Use and divide other yard register 1005 and Doppler's register 1006, draw other memory data of branch, sign indicating number and Doppler's sequence.Be fed to corresponding multiplier 1201 corresponding to symbol (S) and amplitude (M) bit from the sampling of these three sequences.Utilize signal-amplitude input value, for example from-1,0 or 1 of data and Doppler's sequence, and from-1,0 or 1 of sign indicating number sequence, each multiplier 1201 produces one-1,0 or 1 as output signal.Utilize the front at the simulation of the description of two sequence parallel correlators get and, different multiplier output signals is taken a sample fully simultaneously.

Figure 12 shows the one-level embodiment of circuit of three sequence parallel correlators among Figure 11.The content of the two sequence parallel correlators among this correlator circuit level and Fig. 9 is similar, but this multiplier has an other XOR gate 1205, makes the bit of the register that divides other yard act on the symbol of the structure that multiplies each other.

Fig. 8 and Figure 11 two and three sequence parallel correlators can be summarized as a plurality of sequence parallel correlators.With regard to two sequence correlator, be that the correlator of these three sequences removes one of crosscorrelation sequence (copy data) and become be the correlator of two sequences.In general, this decomposition both can be implemented on data, also can be implemented on replication sequence, so that a multisequencing parallel correlator, i.e. inner product device are provided.

" difference " simulation in parallel correlator is got and can be had lower noise susceptibility and other advantage.A common difference has been shown among Figure 13 gets and configuration.In each circuit stages in parallel correlator, the identical symbol that two D/A converters are origin auto-correlation multipliers and the output signal of amplitude drive.It is positive that transducer is marked as, and its capacitor receives and officially enroll and circuit, and that another transducer is marked as is negative, and its capacitor is received negative getting and circuit.As scheme this specialty known, realize with electronic switch at the switch shown in each transducer.The operation of two transducers is identical, is reversal connections but the symbol switch 1500 in negative transducer is symbol switches with respect to this positive transducer.Utilize this difference method, positive and negative get and the result between difference must be used to determine final relevant result.As the professional and technical personnel was known, this was available by using differential amplifier for example high-speed linear or switching capacity to realize.In addition, positive and negative get with the result can be carried out respectively A/D conversion and carry out quadratic sum irrelevant get and before carry out numerical calculation for their difference.

The parallel correlation technique of three sequences of Figure 11 is than two sequence method uses power still less of Fig. 8, because have only 1 bit-depth sign indicating number register to need displacement when generating the correlated results of the sign indicating number skew that is used for subsequently; Sequence in dividing other Doppler's register can be held and immobilize.An energy that register consumed of the displacement dibit degree of depth approximately is the twice of the energy that consumes of a register of displacement one bit-depth.Because the displacement of register is to adopt the main energy consumption of this parallel relevant programme, so the method for this three sequences has advantage.On the other hand, the method for two sequences is used less shift register parts (bit) and less XOR gate in multiplier, therefore has less side circuit.Difference on this realization size reduces along with the increase of the bit number of Doppler's replication sequence.

D/A converter 1300 shown in Fig. 9 and Figure 12 utilizes the switch of serial to control charging and gets charging with electric capacity.Another embodiment that is used for multiplier and analog converter (D/A) is illustrated, and is used for the situation of three sequences of Figure 18.The advantage of present embodiment is that its digital logic functions drives directly that electric charge is got and capacitor, and does not need the connected in series of switch.The combination of this multiplier-D/A is to be convenient to realize in ordinary numbers CMOS handles.In fact capacitor 1101 and 1102 has identical value and three level D/A is provided conversion jointly, for example-1,0 and+level of 1.-1 is used for two electric capacity are driven and is low level (digital grounding), + 1 level is used for two electric capacity are driven and is high level (digital Vdd), and 0 level to be used to drive an electric capacity be high level and another electric capacity is low level.In multiplier 2001, the fellow disciple 1210 and data of this dibit and doppler values are multiplied each other with door 1211 is so that form symbol-amplitude product.Fellow disciple 1212 converts the product of this dibit to A and B signal subsequently, drives this two capacitors in identical or different direction.Finally, fellow disciple 1213 is with binary system code value and A and B signal multiplication, and not gate 1214 is used for for capacitor grounding, so that discharge for electric capacity in reseting stage.

For the given length of the gps signal of storing, be that having one between relevant and the irrelevant processing trades off.Along with coherent correlation length increases, each relevant treatment result's SNR is increased at that time can be for the relevant treatment result's who carries out that quadratic sum is irrelevant and add up (integration) decreased number.The professional and technical personnel is known, 10 times of the every increases of relevant treatment length, and this SNR just improves 10dB, but just improves 5dB for 10 times of increases of irrelevant treated length.So,, make the SNR that detects in advance be maximized by carrying out a length coherent correlation process for a given storage data length., but the receiver of nonanticipating move or the shake of local oscillator will limit the line length of relevant treatment.And along with the increase of coherent correlation length, the bandwidth of doppler receiver is reduced, and produces the doppler receiver that more must be searched.The system configuration of Fig. 4 and Fig. 5 is convenient to resetting and for the bypass of square function, and realize the dynamic exchange of relevant and irrelevant treated length by the control of duplicating generator, relevant and irrelevant accumulator.

General situation is that the SNR that detects in advance that needs in order to realize reliable detection is lower than being used for the required SNR in accurate correlation peak location (being interpolation).So, by using short coherent correlation process to make acquisition time and energy be reduced (and therefore in the enterprising line search of less doppler receiver, these receivers are the same wide with the reliability restriction of the detection that will allow), till observing a coherent detection, and use subsequently to handle to have the more data of long coherence correlation, so that increase SNR and carry out more senior yard skew interpolation near the sign indicating number skew that draws this testing result and Doppler displacement.The structure of Fig. 4 and Fig. 5 easily realizes the processing procedure in these two steps.

Figure 14 shows the part of the automatic waveform correlation of noiseless of the C/A-sign indicating number P/N waveform (not having sampling) that is used for 1.0ms.For all side-play amount of amplitude greater than a chip width, this autocorrelation value is to approach very much zero, and be used for-1 and+trigonometric function of skew between the 1 chip width.Another aspect of the present invention is, is not to use in fact the sampling rate than the high twice of C/A spreading rate, and is to use a non-integral multiple speed of this C/A spreading rate to realize the estimation of accurate sub-chip correlation peak time.Utilize each the Integer N of sampling of C/A chip, when the bandwidth of receiver was in fact wide than gps signal bandwidth, the automatic coherent sampling value of discrete time on the input signal time migration variable quantity of the 1/N of C/A chip period was to keep in fact constant.This is a kind of form of quantification, and if N draw actual quantization error under the little situation.For example, the offset measurement precision of a C/A chip 1/10th will need N=10 doubly to a sampling rate of C/A spreading rate.Adopt the bandwidth of further limited receiver, this automatic coherent sampling value changes along with the time migration of input signal, still there is no need the skew retention wire sexual intercourse that is and imports.If adopt non-integral times to take advantage of relation, then according to the order of signal C/A chip, the instantaneous process or the progress of sampling are sampled.In fact the automatic correlation that this process make to be calculated changes linearly along with the moving of time of input signal, and has the much smaller time and move.Although above-mentioned discussion is the relevant automatically problem that concentrates on binary system C/A sign indicating number, the method that this sampling is carried out also goes for the signal of general multiple level.

Utilize the replica code exponential sum to be matched with the Doppler frequency of noise input signal, the correlator of this serial or parallel reveals computational chart the noise samples of auto-correlation function.By selecting the speed of sample of signal, make the integer sampling was arranged in C/A sign indicating number cycle of every 1ms (for example 2183), correlated results (using identical replica code skew) from the data storage sequence of for example 20 continuous 1ms will draw 20 noise sample values of same point on the automatic waveform correlation of this C/A sign indicating number.In the acquisition process of signal, Doppler frequency of duplicating and phase place are seldom strictly mated with the Doppler frequency and the phase place of signal, make relevant correlated results is carried out square necessitating, so that guarantee positive correlated results.Also illustrate among Figure 14 typically by square 1ms coherent correlation sampling point (x), be aimed near several yards calculations of offset of received signal, and have sign indicating number exponential sum Doppler frequency and phase place with the input signal coupling.By averaging for the sample value from the quilt of identical sign indicating number skew square, one that produces corresponding correlation automatically than the low noise estimated value.Typically also be shown among the figure by average coherent sampling automatically.This on average is so-called irrelevant adding up.Phase place with respect to the sampling grid of trigonometric function is the phase place of receiving waveform that depends on respect to input A/D sampling clock.

In the waveform of Figure 14, automatically the time of correlation peak is the parameter of institute's perception interest, and this time to peak is not on this takes a sample grid.But, using the average automatic correlation of quilt around this time to peak, this time to peak can be with not estimating by interpolation.For the professional and technical personnel, the method for various interpolations is known, carries out classification but a kind of method is the skew of estimating the increase sign indicating number for average automatic coherent sampling, and searches the adjacent item (entries) of two maximums subsequently.Left side option (one of morning) and its adjacent option (in the drawings some B and A) in left side have defined a line jointly, and the selection (in the drawings some C and D) adjacent with its right side of the option on right side defined another line jointly.As SNR when being enough high, the somewhere that the two lines intersect between two originating endpoints (A in the drawings and D point), and to intersect related time with this be exactly the correlation peak time of this estimation.This method requires two finding the solution of equation simultaneously, available use serial or parallel method, and by 35 calculating of the controller shown in Fig. 4 and Fig. 5.

Figure 15 shows according to having whole IQ processing and obtains a low power consumption parallel correlator of system, and RF/IF part (not illustrating) wherein produces the output signal of two quadratures, is sent to A/D converter 22 and 56.Number 35 is added to this RF/IF part and A/D converter 700 and 701.A/D transducer 56 to power supply and produces the I data of sampling, be stored in the data I storage, and number 22 produces the Q data of sampling, is stored in the data Q memory.

In an embodiment of this system, input a/d converter is with 1,000,000 sampling period ground sampling I of per second 2.183 and Q input signal, and each C/A sign indicating number cycle produces 2183 sample values, and is that each C/A chip has about 2.1 sample values.These sample values are quantified as 3 level, and use dibit symbol-amplitude to represent, as previously mentioned.Each of I and Q data storage memory is read to realize that even as big as storage signal obtains required whole data segment.For the occasion of the utilization of demarcating, normally 20ms.After I and Q data have been stored 20ms (2183 * 20 sample value), partly remove power supply and remove power supply from this RF/IF, and begin to obtain processing procedure from input a/d converter.Each is 20 row that are constituted as 2183 dibit samplings for this I and Q data storage, has the output (full line) of 2183 whiles.Use other three sequences parallel correlator of four branches, be designated as II, IQ, QI and QQ.These memories comprise compound parallel correlator 70.Each of correlator that should be parallel is used the sequence length of 2183 samplings.The I data storage drives II and IQ correlator, and the Q data storage drives QI and QQ correlator.And I Doppler register drives II and QI correlator, and Q Doppler register drives IQ and QQ correlator.The order of data processing is set up, so that before this yard index and Doppler frequency are changed, and the replica codes skew of all wishing at given replica code exponential sum Doppler frequency check.And, before this row is changed, at I that is stored and the whole sign indicating number skews of wishing of Q data processing.This will reduce the use of the energy of reading of the memory that makes up and sign indicating number register shift most possibly.The order that other memory lines, sign indicating number exponential sum Doppler frequency are handled also is possible, and this order can be selected to the consumption that reduces energy.The result of this II and QQ relevant treatment got and, so that form II+QQ by getting of join dependency with line, and by connect get accordingly with line make the result of QI and IQ relevant treatment got and.For correct I-Q processing procedure, QI or IQ result must be by anti-phase, and availablely realize by for example sensing paraphase to the symbol switch in the analog converter of the parallel correlator of all selections.Form IQ-QI in the present embodiment.

Utilize this gps signal, the SNR of each II+QQ and IQ-QI signal is 0-4dB normally.These signals are expressed as several level (for example-1,0,1) fully, and convert digital form to by A/

D converter

700 and 701, and are undertaken square by squarer 702 and 703 subsequently.Square signal by digital adder 46 get and and be stored in the accumulator 44.Accumulator 44 has the position of other memory of branch that is associated with the skew of each standby sign indicating number, and the relevant result with each yard of the different data segments that receive data adds up in the memory location that is associated.

Figure 16 illustrates an embodiment of the high-speed a/d converter that carries out square, and it combines the function of A/D and square operation.Two comparators 61 wherein and 62 determine the analogues value whether with simulation characterize level-1,0 and+1 two threshold values that are associated on, under or between.Logical block 63 transforms to suitable square digital value to comparator in conjunction with output state subsequently.If desired, by adding more comparator and threshold voltage, this transducer is used to characterize level more than three easily.

The length of sampling rate, storage sequence and while memory output length can be selected to be suitable for best GPS and use.For example, be necessary data segment by storing whole for obtaining of signal, this RF/IF part can be turn-offed afterwards and not the shortest possible time (relative with the reception of required data segment).The data set of storage is used to the processing of the whole SCD receivers in obtaining search subsequently.In addition, the memory that available use is shorter, its cost is to keep RF/IF one longer period of part.For example, used " ping-pong " memory construction shown in Figure 19 among another noticeable embodiment.Each of I wherein and Q data storage memory 192 and 194 has only the length of 2.0ms respectively and is constituted as two also row of the 1.0ms of line output (for example as in the past, in each row 2183 samplings being arranged).In I and Q data channel, a data line is extracted by parallel, so that realize and line correlation, other data line is to write with the digitalized data from this RF/IF part simultaneously.Each of whole I and Q data sequence length is occupied the data segment of 1.0ms.Whole 20ms data sets then approximately is being treated to single S CD receiver among the 20ms.Because whole data sequence is not stored, so if another SCD receiver is not tested, then this RF/IF part must and produce another complete sequence by energy supply.On average, in the acquisition process of signal, this will increase the RF/IF part must be held supplying time.But reducing of memory-size is very considerable.(for example military receiver under serious congestion condition) in some applications, GPS SNR is low-down, and the length that is used to realize the data sequence obtained can be long be unpractiaca degree to storing whole sequence.Under this condition, the structure of obtaining of ping-pong has advantage.

In some applications, will be useful if can further lower the size of data storage memory.Utilize the storage organization of two data segments (ping-pong), can handle the continuous data stream of the length of receiver dateout.With longer acquisition time is cost, can be reduced for the memory of single data segment according to section I and Q data storage 192 and 194 (Figure 19) from one or two number of the 1.0ms of each memory by elimination.By this reduction, the receiver dateout section of the 1.0ms that time is adjacent is no longer processed, have only separately 1.0ms data segment to be caught in the memory and processed, and the processing time of the signal data section of this reception of being used to handle given number of will having extended.

In order further to reduce the size of memory and the complexity of parallel correlator, the length of parallel correlator and the RS data segment that is associated can be made as less than desirable relevant treatment length.Desirable relevant treatment length can be used by handling a plurality of the combination relatively than the data segment of short data and their result and realize.For example, the relevant treatment length of a 1.0ms can be passed through at the parallel correlator that uses a single 0.5ms length on the data segment of two adjacent 0.5ms data and after the A/D conversion its result digitally be realized the phase Calais.The various combinations that width is searched in the number of correlator length, relevant treatment length, memory data section and sign indicating number skew all are possible.

When the cost size of common serial input-serial output (SISO) memory and power consumption when being fully low, in such memory memory all need I and Q sequence, to close the RF/IF receiver section and sort memory is used in combination with ping-pong or forms data segment memory structure to be useful, so that detect the SCD receiver that is hopeful.For each tested SCD receiver, all I and Q data sequence are once read from the SISO memory, and this readout will consumed energy.But, owing to RF/IF partly is closed, so if the power consumption of SISO memory is very low for the power consumption of RF/IF receiver section, then the energy consumption of system will be very low.The length and the sampling rate of the length of the sequence of other storage, memory output simultaneously also all are attractive.

In the embodiment of two sequences shown in Fig. 8 and Figure 11 or three sequence parallel correlators, the product of related data and copy data sampling is to carry out in this correlator in parallel mode.This will make that different doppler receivers is detected so that realize obtaining, and needn't collect new data accepted.Figure 20 shows an other embodiment of parallel correlator and related parallel storage, and data wherein and Doppler's sample value are multiplied each other by multiplier 2001 before in being stored in parallel storage 2002.Advantage is to make the multiplier in parallel correlator simplify.Like this, each correlator multiplier multiplies each other the only same single associated code bit of the sampling of associated storage.The multiplier of combination that is used for this correlator embodiment is similar with the situation shown in Figure 18 with D/A converter, but comprises fellow disciple 1210 and 1212 and the front that is removed to the data storage outside the correlator with the data-Doppler's multiplier and the A/D converter of door 1211.The input for fellow disciple 1213 that disconnects is coupled to the output of data storage.Utilize this embodiment,, then must store new data, but this is not a shortcoming in some applications if carry out the test of different doppler receivers in order to realize obtaining.

The parallel correlator of the combination shown in Figure 20 and parallel storage can be used for forming the complete I-Q processing GPS deriving means of a simplification.In the I-Q deriving means shown in Figure 19, four other parallel correlators of branch are arranged, be coupled to I and Q and the Doppler I and the Q register (parallel storage) of data, and be coupled to parallel sign indicating number register.The embodiment that the another kind of I-Q fully of the structure of more simple correlator shown in use Figure 20 and memory handles the GPS deriving means is shown in Figure 21.In the present embodiment, 4 parallel data-Doppler's memories are keeping II, QQ, IQ and QI data Doppler product sequence respectively.As previously mentioned, each parallel storage can both constitute ping-pong or single data segment memory.There are not independent data and Doppler's memory.Each parallel memory is received related parallel correlator, and each parallel correlator is also received single parallel sign indicating number register.This interconnectivity that is lowered between memory portion and correlator part will make the embodiment of the I-Q processor that this is complete have the integrated circuit facility scheme of simple (volume is little and cost is low).

Addition by carrying out II+QQ before storage and QI-IQ subtract each other the complexity that will further reduce the embodiment of Figure 21.Before storing, the addition and the output signal of subtracting each other often can or be punctured into two bits and almost not make the bad change of SNR by abbreviation.So as shown in figure 22, only need two data-Doppler's memories (is used for the II+QQ data, and another is used for the QI-IQ data), two parallel correlators and a sign indicating number register.As previously mentioned, each data Doppler memory can both be constituted as ping-pong or single data segment memory.Parallel correlator, parallel storage and sequence doubly take advantage of other structure of mode may and to have advantage equally.

Figure 17 shows a kind of data storage of combination, the structure (only being used for the II relevant treatment) of copy data generator parallel correlator, this structure is reduced to 2183 displacements to the number of the sign indicating number register shift of each yard/Doppler's combination, and has eliminated the needs for any irrelevant RAM of adding up.Cost is that input store must be by segmentation, so that make all storage data by access side by side.These data are to be added to other parallel correlator of branch in the data set of 1.0ms.These parallel correlators are driven by identical sign indicating number and Doppler's copy data, and their (in a complete yard cycle) 20 other 1.0ms correlated serieses of branch of generation side by side, and these sequences are the time sequencings that before added up.Divide other square A/D converter that the simulation correlated series is converted to digital form, adder tree structure 80 wherein forms the correlated series that adds up.Can handle the use of fully avoiding the RAM that adds up by being carried out the search of peak value and interpolation for this sequence when producing at the correlated series that adds up.In this structure, sign indicating number and Doppler's register are to drive 20 loads rather than originally drove a load like that, reduce certain energy consumption like this.

When parallel correlator and data and copy data register one are used from the situation of two sequences, when perhaps being used from the situation of three sequences with data, doppler data and sign indicating number register one, it is useful reducing import the energy that uses in (writing) for various registers.Doppler data and sign indicating number register circuit level are write at any time.Figure 23 illustrates a scanner shift register 304, is the situation that is used for three sequences with data register 300, doppler data register 301 and 302 combinations of sign indicating number register.In a method of operation, the sampling that is used for the sequence of numerical data, doppler data and sign indicating number sequentially arrives, and the speed of sampling is identical.Along with being used for each not arrival of homotactic sampling, they are written at them and respectively organize in the position of the correspondence in the register.Synchronous with the sampling that arrives, single logical one makes the sampling that arrives be written to the corresponding ordinal position in other register of branch along the displacement of binary system scan register.Owing to have only two adjacent scanner register position to change their storing values in displacement each time, and other register do not change in write operation, so write operation is low-down energy consumption.

Though invention disclosed is to be used for obtaining of GPS C/A signal, also can be used for reducing significantly in order directly to obtain military gps signal P (Y) rather than at first to obtain the C/A signal and the time and the energy of needs.The content of the one yard generator of satellite by changing the sign indicating number/Doppler generator in Fig. 4 or Fig. 5, so that produce P (Y) yard rather than C/A sign indicating number, Method and circuits structure then disclosed herein just goes for the situation of P (Y).

Although just described definite preferred feature of the present invention at this, but can carry out multiple correction and improvement for the professional and technical personnel, therefore, should understand that appended claim attempts to cover all these corrections and the improvement in the spirit scope of the present invention.

Claims

1. can receive at one and a plurality ofly comprise from the device in the Receiver of Direct-sequence Spread Spectrum of a plurality of signals that divide other transmitter:

Tuner is used to receive said at least one signal and produces an intermediate-freuqncy signal, and said tuner comprises a transducer, and said intermediate-freuqncy signal is used to take a sample;

Be used for removing the device of frequency shift (FS), thereby produce a correction signal from the output signal of said transducer;

Memory is used to store the signal of said correction; With

Processing unit is used to handle said correction signal.

2. according to the receiver of claim 1, wherein:

Said receiver is the GPS receiver, and said transmitter is the gps satellite transmitter.

3. according to the receiver of claim 1, wherein:

Said frequency shift (FS) comprises Doppler frequency shift.

4. according to the receiver of claim 3, wherein:

Said Doppler frequency shift is known.

5. according to the receiver of claim 1, wherein:

Said frequency shift (FS) comprises the error of local oscillator.

6. according to the receiver of claim 5, wherein:

The error of said local oscillator is known.

7. according to the receiver of claim 1, wherein:

Be used for comprising a multiplier from the device of said received signal removal frequency shift (FS).

8. according to the receiver of claim 1, wherein:

Said processing unit comprises a correlator.

9. receiver according to Claim 8, wherein:

Said correlator comprises a parallel correlator.

10. according to the receiver of claim 1, wherein:

Said processing unit is applied to read according to the needs of handling said sample value the sample value of said correction signal.

11. comprise a method that can receive a plurality of Receiver of Direct-sequence Spread Spectrum from a plurality of signals that divide other transmitter:

Reception has an input signal of signal(-) carrier frequency;

Produce a local oscillator signal, this signal has the frequency from said carrier frequency shift;

Said input signal and said local oscillator signal mixing, so that produce an intermediate-freuqncy signal;

Said intermediate-freuqncy signal is converted to multiple sampling numeral signal;

The signal mixing of a said multiple sampling numeral signal and a frequency shift (FS), so that produce the signal of an emending frequency;

The signal storage of frequency correction in a memory; With

Handle the signal of said emending frequency.

12. according to the method for claim 11, wherein:

13. according to the method for claim 11, wherein:

Said frequency shift (FS) comprises Doppler frequency shift.

14. according to the method for claim 13, wherein:

Said Doppler frequency shift is known.

15. according to the method for claim 11, wherein:

Said frequency offset signals comprises the error of local oscillator.

16. according to the method for claim 15, wherein:

The error of said local oscillator is known.

17. according to the method for claim 11, wherein:

The step that the signal of a said multiple sampling numeral signal and a frequency shift (FS) is carried out mixing comprises carries out digital multiply.

18. according to the method for claim 11, wherein:

The signal that said processing comprises said frequency correction is relevant with a reproducing signals.

19. according to the method for claim 18, wherein:

Said relevant comprise one and line correlation.

20. according to the method for claim 11, wherein:

Said processing is used to the step of reading said memory according in order to realize said processing.