GB2530843A

GB2530843A - Determining location of a receiver with a multi-subcarrier signal

Info

Publication number: GB2530843A
Application number: GB1511208.9A
Authority: GB
Inventors: Richard G Keegan
Original assignee: Deere and Co
Current assignee: Deere and Co
Priority date: 2014-07-09
Filing date: 2015-06-25
Publication date: 2016-04-06
Anticipated expiration: 2035-06-25
Also published as: GB201511208D0; GB2530843B

Abstract

Receiver 200 receives multiple multi-tone OFDM-like signals from respective transmitters 100 with known locations, each multi-tone signal having sub-carriers modulated with a pseudo-random noise code; a coarse estimate of a time of arrival of one or more of the sub-carriers is adjusted in accordance with a code correlation function (S302, Fig.11); a precise estimate of said time of arrival to align a slope at least two of the sub-carrier phases of the multi-tone signals is adjusted in accordance with a phase correlation function and the coarse estimate (S304, Fig.12); and receiver location or its range to a transmitter is estimated there-from (S306, Fig.11). Phase compensation data is determined for each ranging subcarrier in the multipath signal. It is based on a difference between an observed phase of the observed signal vector and a direct path phase of a direct path vector, where the direct path phase is estimated from prior measurements of a certain observed signal vector when an average amplitude of most ranges converge.

Description

DETERMINING LOCATION OF A RECEIVER WITH A MULTI-SUBCARRIER SIGNAL

Field

This disclosure relates to a method and system for determining location of a receiver with a multi-carrier or muiti-subcarrier signal, such as an OFDM-like signal.

Background

The accuracy with which time of arrival of a navigation signal can be measured drives the accuracy of a navigabon system Certain prior art uses orthogonal frequency division multiplex signals for local wireless area networks, such as those that are compliant with the Institute of Electrical and Electronic Engineers (IEEE) 602 11 standard Offer pnor art uses terrestrial navigation signals that mimic Global Positioning System GPS) signals, where navigation signals are transmitted from one or more terrestrial beacons or sites with known locations, However, the terrestrial navigation signals are susceptible to multipath distortion (e.g., from ground clutter, buildings or obstructions) that impacts the accuracy of a location-determining receiver that receives the navigation signals. Thus, there is need for a method and system for determining a location of a recever that provides sunerior inmunity or resistance to rnultipath distortion and accurate location determination.

Summary

In one embodiment, a method and system or determining a ocation of a receiver with a multi-carrier signal (e.g., muiti-subcarrier signal) comprises a receiver portion for receiving a plurahty of (orthogonal, frequency-division multiolex) OFDM like niultitone signals Each OFDM-like multitone signal has subcarriers (tones) and is modulated with a pseudo-random noise code and transmitted from a transmitter with a known location, A first estimator determines a course estimate of a time of arrival of a set of subcarriers (e.g., one or more subcarriers) of OFDM-like multitone sgnals by adjusting the course estimate of the time of arrival ir accordance with a code (first) correlation function A precise estimate is deten'nined for the time of amnval of tie set of subcarriers of the OFDM-like multitone signals by adjusting the precise time of arrival to align the phase slope of the subcarrier phases of the OFDM-like multitone signals in accordance with a phase (second) correlation function and consistent with the course estimate of the time of arrival associated with the code correlation function. A data processor estimates the location of the receiver or estimated range based on the precise estimate (e.g., and phase error compensation to drive the slope of the subearrier phases toward zero). The data processor determines phase compensatior data for each ranging subcarner compnsing an adjustment to the estimated range based on a difference between an observed phase of the observed signal vector and a dftect path phase of a direct path vector, where the direct path phase is estftnated based on one or more prior measurements of a certain observed signal vector when an average amplitude of all (or a majority of) ranging subcarriers converge to substantiafly the same value.

D es cr1 tion of the Draw in s FIG. I refers to FIG. 1A and FIG. 1 B collectively; FIG. 1 is a block diagram of a transmitter of a mull-carrier signal, such as an OFDM-like multi-carrier signal.

FIG. 2 refers to FIG. 2.A and FIG. 2B cdllectiv&y; FIG. 2 is a block diagram of a receiver of a multi-earner signal, such as an OFOM-like multi-carrier signal FIG. 3 is a block diagram of a group of transmitters and a receiver of a multi-carrier signal consistent with FIG. 1 and FIG. 2.

FIG. 4 5 a graph of variants of first correlation function (e.g., auto-correlation function).

FIG. 5 is a graph of variants of the first correlation function shown in greater detau.

FIG. 6 shows tone coherence intervals in a composite time series.

FIG 7 shows a first eariy minus late error function associated with different chip spacings.

FIG. 8 shows a phase slope with a second early minus late function (e.g., central early minus late error function).

FIG. 9A illustrates an exemplary vector length and phase angle difference for a multipath-impacted signal vector and direct path signal vector.

FIG 9B illustrates an exemplary vector length and phase angle difference for a multipath-impacted signal vector and direct path signal vector over a greater time period than FIG. QA.

FIG. 1 OA is a first illustrative graph of signal amplitude versus subcarrier identifier of a direct path signal and a multipath signal.

FIG. I OR is a second illustrative graph of signal amplitude versus subcarrier identifier of a direct path signal and a multipath signal FIG. I OC is a third illustrative graph of signal amplitude versus subcarrier identifier of a direct path signal and a multipath signal.

FIG 100 is a fourth illustrative graph of signal amplitude versus subcarrier dentiier of a direct path signal and a multipath signal. 2.

FIG. 11 is flow chart of one embodiment of determining the location or range of a receiver with a multi-carrier signal.

FIG. 12 is a flow chart of one embodiment of determining the location or range of a receiver with a multkcarrier signal with multipath compensation.

FIG. 13 refers to FIG. 13A and FIG. 13B collectively; FIG. 13 is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carrier signal with m ultipath compensation FIG. 14 refers to FIG. 14A and FIG. 14B collectively; FIG. 14 is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carrier signal with multipath compensation.

FIG. 15 refers to FIG. 15A and FIG. 15B collectively; FIG. 15 is a flow chart o another embodiment of determining the location or range of a receiver with a multi-carrier signal with multipath compensation.

FIG 16 refers to FIG. ThA and FIG. 166 collectively; FIG. 16 is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carrier signal with rnultipath compensation.

FIG. 17 is a flow chart of another embodiment of determining the location or range of a receiver with a niulti-carner signal to mnimlze or reduce the mpact Jf multipath distortion or error.

FIG. 18 refers to FIG. iSA and FIG. 186 collectively; FIG. 18 is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carder signal to minimize or reduce the impact of multipath distortion or error.

FIG. 19 refers to FIG. 1 9A and FIG. 196 collectively; FIG. 19 is a flow chart of another embodiment ci determining the location or range of a receiver with a multi-carner signal to minimize or reduce the impact of multipath distortion or error.

FIG. 20 refers to FIG. 20A and FIG. 208 collectively; FIG. 20 is a flow chart of another embodiment of deterniinirig the location or range of a receiver with a multi-carrier signal to minimize or reduce the impact of multipath distorton or erro' Detaile 4 De p t tftcF.:u.4 In one embodiment, a method and system for determining a location of a receiver with a multi-carner signal comprises a receiver portion for receiving a pluralty of OFDM-Iike multitone signals. Each OFDM-like multitone signal has subcarriers (tones) and is modulated with a pseudo random noise code and transmitted from a transmitter with a known location In a first trackhig loop, a first estimator determines a course estimate of a time of arrival of a set of subearners (e g one or more subcamers) of the OFDM-like multitone siqnals by adjusting the course estimate of the time of arrival to maximize the correlation of the received pseudo-random noise code of the at least one subearrier with the Local replica of the pseudo-random noise code in accordance with a code (first) correlation function. In a second tracking loop, a second estimator determines a precise estimate of the time of arrival of the set of subcarriers of the OFDM-hke multilone signals by adjusting the precise time of arrival to aflgn the slope of (at least to on the subcarrier hases of the OFDM-hke multitone signals in accordance with a phase (second) correlation function and consistent with the course estimate of the time of arrival associated with the code correlation function. A data processor estimates the location of the receiver or estimated range between the receiver and the transmitter based on the precise estimate and phase error compensation in a phase domain, a range domain, or both commensurate with driving the slope of the subcarrier phases toward zero (e.g., to compensate for phase or frequency error in the absence rnultipath.

In accordance with one embodiment, FIG. 1 shows a transmitter 100 for transmitting a multi-carrier signal or OFDM-Iike signal. FIG. 1 includes FIG. 1A and FIG. lB collectively. The transmitter 100 transmits an OFDM-Iike, multi-carrier signal that is further modulated by a ranging pseudorandom code.

The transmitter 100 is capable of transmitting an OFOM-like, multi-carrier signal with a known phase state or a known vector phase for each subcarrier (e.g., pilot tone), where the known vector phase of each ranging subcarrier is substantially fixed, individually adjustable, or generally varies in accordance with a known pattern (e.g., pilot sequence, a training sequence or randomized in accordance with a known sequence) over a time period. Here, known means that the phase state or vector phase (e g, an initial phase state) for a corresponding subcarrier of the transmitter 100 is known at the receiver 200 (e.g, a priori or prior to reception or demodulation of any subcarner received by the receiver 200 of FIG 2) A "anging subcarner means a subcarrier that is modulated by quasi-data or used for range estimation by a receiver 200, whereas a data-carrying subcarrier means a subcarrier that is used for data transmission and not for ranging, at least part of the time. A subearrier may refer to a ranging subcarrier, a data-carrying subcarrier, or both.

For exampie, for any ranging subcarrier the known vector phase can be defined in the in-phase and quadrature phase plane (e g 1-0 plane) that is associated with an in-phase vertical axis and a quadrature phase norizontal axis perpendicular to the n-phase vertical axis Sim larly, the known vector phase can be defined by equivalent complex number representations that correspond to the 1-0 plane representation. In one Hustrative embodiment, if each ranging subcarrier is modulated with quadrature phase shift keying (QPSK), each subcarrier may have four possible phase states for each code chip Qr symbol. In one embodiment, each ranging subearrier is encoded with the same phase state, which may change for each code chip or symbol. In another embodiment, each ranging subcarrier can be modulated with a fixed phase state (e g, distinct or not distinct from other subearriers) that is not changed over a tine penod to educe the complexity of the transmitter 100 by eliminating the inverse transform module 28 or at least same processing by the inverse transform module 28 during the time period; where -a time series that the inverse transform module 28 would otherwise create is stored in data storage device (e.g., memory) of an electronic data processing system that implements a digital portion of the transmitter 100.

In certain embodiments, the phase states or known vector phases of the ranging subcarriers can be randomized in accordance with a known pattern (e.g., pseudo-random pattern) to reduce the peak-to-average power ratio of the transmitted multi-carrier signal.

Further, each vector phase of each ranging subearner may be different from and random with respect to the vector phases of the other ranging subcarriers (e g that are not data-carrying subcanlers, but rather pilot subcarriers that carry quasi-data).

As illustrated in FIG. 1, quasi-data may be provided by a quasi-quasi-data source module 20 (e.g., signal generator), which is coupled to the transmitter 100 or integral with the transmitter 100. Quasi-data refers to dummy data, a training sequence, fixed data, or a pilot data pattern that is selected to modulate one or more ranging subcarner signas with target phase states or known vector phases that are known to the receiver 200 The quasi-data source module 20 may provide or output one or more quasi-data streams or logic level signals FIG. 1 illustrates m as a possible number of tones or subcarriers that are modulated by the quasi-data, such that up to ni data streams are possible inputs to an encoding bank 24. There is no requirement that all of the subcarriers are modulated with the same quasi-data or interleaved quasi-data. In certain configurations, each quasi-data stream may be unique.

Multiple quasi-data streams can directly feed he encoding bank 24 Because the quasi-data or the corresponding phase states or vector states on each ranging subcarrier are known at the receiver, the receiver 200 can use the ranging subcarriers for determining one or more ranges between the receiver 200 and one or more transmitters 100.

For example, because the quasi-data is known at the receiver 200! the transmitted pseudo-random noise code can be correlated eadilywth the local replica of the pseudo-random noise code at the receiver Tc the extent the phase state vectors are krowri at the receiver 200, the received phase state vectors can be compared to idea! phase state vectors (e.g., for QPSK states) in the IIQ plane to correct for phase noise or distortion A data-carrying subcarner is not used for any ranging (calculation) or phase quality detection because the receiver 200 does not know (a priori) what the observed phase vector should be. The data transmission of any data-carrying subcarrier is optional and does usually not support the ranging, range or position estimation of the receiver 200 in certain embodiments, unless the data transmission is used to transmit new data paerns or rew sequences of the known phase vectors for alignment at the transmitter 100 and the receiver 200.

In an aiternate embodiment, an optional data source 22 may comprise ranging data or navigation data that modulates one or more data-carrying suboarriers that are not modulated with the quasi-data on a dedicated basis or for some allocated time period. FIG. I illustrates n total subcarrier& that are set to zero and possibly avaiable for data-carrying as one or more data-carrying subcarriers. The ranging data or navigation data can be used by a location-determining receiver 200, such as the receiver 200 of FIG. 2. The ranging data or navigation data may modulate a subcarrier with any type of suitable modulation, such as phase shift keying (PSK or quadrature amplitude modulation (QAM), that is compatible with the OFDM-like multi-carrier signal. In FIG. 1, the optional data source 22 is illustrated in dashed lines because it is optional and may be deleted in certain configurations where data is not transmitted from the transmitter 100 to a receiver 200.

The digital modulator 27 may comprise, collectively, the quasi-data source module 20, the encoding bank 24, the inverse transform module 28, and the parallel-to-serial converter 30.

Although the optional cyclic prefix 962 and the optiona cyc'ic sutfix 961 are illustrated in FIG 1, the cyclic prefix 962 and the cyclic suffix 961 may be deleted in certain embodiments if circular convolution issues can be ignored. The data augmentation or appending, with the cyclic prefix 962 and cyclic suffix 961, provide for valid circular convolution results not degraded by adjacent pseudo noise code polarities. In one embodiment, the encoding bank 24 comprises one or more of the following components: (1) one or more signal generators for generating a set of subcarrier signals that are spaced apart by a target or desired frequency spacng (e g, bin spacing), (2) a set of subcarrier modulators 26 that modulates each subcarrier with a corresponding quasi-data output stream of the quasi-data source module 20 or a combination of the quasi-data output stream pf the quasi-data source module 20 and an optional data source 22, and (3) a constellation mapper.

1n one configuration, each subcarner modulator 28 may comprise a phase shift keying (PSK) modulator, a quadrature phase shift keying (QPSK) modulator, a quadrature amplitude modulator (QAM), or another suitable modulator. For example, for quasi-data, each ranging subcarrier may use a respective QPSK modulator, whereas for any data-carrying subcarrior, any suitable modulator may be used. For each inputted data stream or quasi-data stream, the encoding bank 24 or subcarrier modulators 26 can modulate each subcarrier at a corresponding data rate or quasi-data rate (e.g., generally uniform data rate).

In an alternate embodiment, the inverse transform module 28 performs phase modulation (e.g.. QPSK modulation) for the quasi-data such that one or more subcarder nodulators 28 can be omitted In one embodiment, the encoding bank 24 acts as a constellation mapper to map in-phase and quadrature data representations of each inputted quasi-data stream (e.g., among the streams that are inputted to the encoding bank 24) to corresponding subcarriers, In an alternate embodiment, the constellation mapper also maps the daa from the optional data saurce 22 o a data-carrying subcarrier or sacrificial subcarrier The subcarriers or tones of the encoding bank 24 are inputted to an inverse transform module 28 (e.g., inverse fast Fourier transform (LFFT) module) Each subcarrier signal may be modeled as a complex input that is individually settable, adjustable, dynamically changed, or fixed for input to each bin of the inverse transform module 28. For example, each subcarrier can e modeled as a complex input with a real and imaginary component, consistent with the I-Q vector plane. For example, if the encoding bank 26 uses a subcarrier modulator 26 such as a QPSK modulator then each subcarrier can have any of the following phase states or known phase vectors in the l-Q vector plane (1,1) (1-1), (-1,1) and (-1,-I) The inverse transform module 28 converts the subcarner signals (tones) and quasi-data (collectively called modulated subcarrier signals) from the frequency domain into time domain samples for processing.

Further, the inverse transform module 28 may derive cychc prefixes, cyclic suffixes or both for appending to the time domain samples to construct an appropriate baseband signal that is conveniently decoded at the receiver 200.

The inverse transform module 28 communicates to parallel-to-serial converter 30. The parallel-to serial converter 30 accepts the parallel input of multiple time domain samples (e.g.] appended with cyclic prefixes, suffixes or both) and outputs an aggregate time series 36 or aggregate baseband signal.

The parallel-to-serial converter 30 provides the aggregate baseband signal to a first input of a mixer 32. A code generator 34 generates a pseudo-random code or pseudo-random noise code. The content or sequence of pseudo-random code 34 or pseudo-random noise code is generafly known at the transmitter 100 and receiver 200. The code generator 34 provides a second input to the mixer 32. An output of the mixer 52 is fed to digital-to-analog converter 38.

From the digitako-analog converter 38, the analog baseband signal is fed to an upeonverter 42. The upconverter 42 converts the baseband signal to a radio frequency, microwave frequency signal, or another electromagnetic signal for transmission to one or more receivers 200 In one embodiment, the upconverter 42 compnses a mixer or multipier In an alternate embodiment, the upconverter 42 comprises multiple stages of mixers or multipliers in series. The upconverter 42. or its mixer, has an input for a transmitter local oscillator 40 to upconvert the signal to an electromagnetic signal at the desired frequency for wireless transmission. The electromagnetic signal includes subcarrier frequencies for tones of the transmitted OFDM-like signaL An amplifier 44 may comprise a radio frequency power amplifier, a microwave amplifier, or another amplifier for amplifying the electromagnetic signal for or prior to wireless transmission. The input of the amplifier 44 is coupled to the upconverter 42 and the output is coupled to an antenna 46 rnr transmitting the OFDM-like signal to one or more receivers 200 such as a the location-determining receiver 200 ol FIG 2 In practice, one or more filters (e g, passband filter) may be used in conjunction with the amplifier 44 or upconverter 42, for

example.

The transmitter 100 transmits pilot tones or ranging subcarriers that are modulated with quasi-data to produce an OFDM-Uke multi-carrier signal. In one embodiment, the multi-carrier sgnal or OFDM-like signal has pilot tones dispersed through its spectrun with known phase and amplitude at the time of transmission. For example, a pilot tone refers to one or more ranging subcarriers of the OFDM-like signal that carries one or more pilot signals (e.g., training sequences, preambles, known sequence of symbols) for measurement of the channel propagation conditions, such as measurement of phase shift (e.g., from multipath) for each ranging subcarrier or measurement of phase slope beeen two or more ranging subcarriers.

The phase and amplitude of the transmitted signal can be impacted by the propagation path or multipath reception between the transmitter 100 and the receiver 200, among other things.

In one embodiment, a receiver 200 is enabled to receive (e.g., simultaneously) multiple OFDM signals from different transmitters 100 (e.g., three or more transmitters) 100 to establish the position, location, attitude (e g roll, tilt, yaw) or geographc coordinates of the receiver 200 in two or more dimensions (e,g., three dimensions). To facilitate the receiver 200 discrimination between different received OFDM-like signals, the encoding bank 24 should not place the pilot tones adjacent to each other, but should separate the pilot tones by some minimum number (n)

B

of frequency bins, where a frequency bin is the fundamental spacing of the lEFT signal generation that is common to aD OFOM communicat'on systems The above frequency spacing facilitates reuse of the same demodulation hardre within the same frequency spectrum in each receiver 200. Accordingly, uniform spacing of the pilot tones across the band, such as by n frequency bins, provides simplified demodulation/measurement process, although other non-uniform spacing may be used for the pilot tones in any alternate embodiments.

In FIG. 1, the frequency domain representation of the signal is transformed into a time domain se ies using an inverse FF1 e g, by inverse FFT module 28) 1 he inverse FF1 is commonly used for OFDM transmitters 100, for example. At the output of the FF1 module 28, the resulting time series 36 is augmented by the parallel-ta-series converter 30. The parallel-to-series converter 30 can support augmentation of the time series 36 by a cyclic prefix 962, as is conmon in OFDM systems, by affixing a small segment of the erding time series 36 to the beginning of the time series 36. In addition, the parallel-to-series converter 30 can add a cyclic suffix 96i by affixing a small segment of the beginning of the un-augmented time series 36 to the end of the augmented time series 36. At the mixer 32, in one configuration the resulting time series 36 is then modulated by a ranging pseudorandom code that has a basic chip rate that is the same as the augmented symbol rate of the tones or tone modulation (e.g., by quasi-data).

As shown, the parallel-to-serial converter 30 may comprise N input parallel converter, where N is a whole number, where N= (m+ j)n+1) and where j = 2k1(n+1), where m is the number of ranging carriers or pilot tones, j is an integer, k is an integer number of time samples, which is proportional to a length of the optional cyclic prefix and cyclic suffix, n is the number of data-carrying subcarriers that can be reserved for data transmission or left empty without any modulated data to support co-frequency transmissions from other transmitters 100, and n + 1 is a number of subcarriers of which 1 is a tone or ranging subcarrier and the other n subcarriers are blank to facilitate the addition of other ranging beacons. Although the parallel-to-serial convertor 30 comprises an N input parallel-to-serial converter, any other suitable parallel-to-ser al converter 30 falls wthin the scope of this document The block diagram of FIG.1 merely illustrates one possible configuration for implementing a transmitter 100; others can be envisioned. For example, the transmitter 100 of FIG I can carry data using one or more data-carrying subcarriers (e g, n data-carry ng subcarriers) shown as being set to zero. However, in one embodiment the n data-carrying subcarners that are set to zero have no modulating data (e g payloac) and may support instead the transmission (e.g., co-frequency transmission) of ranging subcarriers of one or more other transmitters 100 for reception at a single receiver 200. Recah that the n data-carrying subcarriers can be offset by a fixed number of frequency bins from the ranging subcarrier frequency bins for reception at a receiver 200. Accordingly, if co-frequency transmission of other transmitters 100 is supported, the capacity of the data-carrying subcarriers may be Umited to some maximum number less than n, even if n subcarriers are set to zero.

In an alternate embodiment, if the system (a g, transmItters 100 and receiver 200 collectively) will not be required to transmit or carry any changing data or any changing quasi data (e.g., resulting in a pattern, training sequence, or sequence of target phase states), the time series 36 represented by the output of the parallel-to-serial converter 30 never changes; hence, could be stored in memory (e.g., nonvolatfie electronic memory or another data storage device) rather than reconstructed each time using an FF1 module 28. Accordingly, in such an alternate embodiment the combination of the lEFT module 28 and the paralel-to-serial converter 30 can be replaced with a data storage device (e.g., electronic memory or a series of shift registers) and a data processor that is capable of storing, retrieving, or generating a repetitive code or fixed data set.

FIG 2 discloses a receiver 200 for receiving a multi-carner sgnal for navigation or location determination. FIG, 2 includes FIG. 2A and FIG. 2B, collectively. The receiver 200 comprises a receiver front end 53 coupled to a data processor 104. In turn, the data processor 104 can communicate with a data storage device 98 (e.g., memory) via one or more data buses 103. In one embodiment, a digital baseband portion 82 comprises the data processor 104, one or more data buses 103, and the data storage device 98. The data processor 104 may comprise an electronic data processor, a programmable logic array (PLA), a microprocessor, a m crocontroiler, an application specific integrated circuit (ASIC), a diguta sgna processor (DSP), an arithmetic logic unit, a Boolean logic circuit, or any other data processing device. The data storage device 98 may comprise nonvolatile electronic memory, electronic memory, a magnetic storage device, an optical storage device, or any other suitable data storage device.

As used throughout this document, in certain embodiments the data processor 104 may compnse one or more of the blocks or modules within the dashed hnes of FIG 2, such as the tone signal detector 70, a tone phase measurement module 76, a multipath detector 94, a multipath compensator 96, a tracking module 86, a propagation time module 85, a range-domain, phase-slope compensator 92 and a range estimation module 84.

As illustrated in FIG. 1, and FIG. 2, the data paths that interconnect the blocks, modules and components may comprise physical data paths or virtual data paths, or both, where physical data paths means conductors, transmissions lines, cables or electromagnetic communication paths, and where virtual data paths means communication between software components, modules, libraries, data sets, or other electronic data. The modules, blocks and components may comprise hardware, software instructions or both. For example, dedicated hardware (e.g., an ASIC or DSP) can be substituted for or realized as software instructions (e.g., in the data storage device 98) executed by a general purpose computer or data processor (104).

To determine its locaflon, the rec&ver 200 needs to receive transmitted OFDM-like multi-tone signals from three or more transmitters 100, such as transmitters 100 similar to the transmitter 100 of FIG. 1. FIG. 3 provides ar illustrative exampie of a receiver 200 receiving signals from multiple transmitters 100. Each OFDM-like multitone signal has ranging subcarriers (tones) and is modulated with a pseudo-random noise code and transmitted from a transrrntter 100 with a known location or reference location In HG. 2, the receiver 200 comprises a receiver front end 53. In FIG. 2, the receiver front end 53 or another receiver portion receives a pluraflty of OFDM-Iike multitone signals. The receiver front end 53 comprises an antenna 146 coupled to an amplifier 52. An amplifier 52 is coupled to a downconverter mixer 54 The downconverter mixer 54 receives an input signal from a receiver local oscillator 56 and outputs an analog intermediate frequency signal or an analog baseband signal. The analog-to-digital converter 58 converts the analog intermediate frequency (IF) signal or the analog baseband signal to a digital IF or a digital baseband, where the baseband or IF signal can be described as a time series signal 38 or time series.

The baseband or IF signal and a pseudo-random noise (RN) code signal are inputted into a mixer 60 At the mixer 60, the received RN code on the baseband or IF signal is correlated with and temporally aligned with the locall) generated replica of the RN code from code generator 62 to remove or strip the PN code from the baseband or IF signal. The modulated IF or baseband signal has the RN code removed or stripped from the baseband or IF signal, such that any encoded information remains on the modulated IF or baseband signal.

The mixer 80 outputs a modutated IF or modulated baseband signal to a serial-to-parallel converter 64 As shown, the seriako-parallel conerter 64 may comprise N stage serial-to-parallel converter, where N is a whole number, where N m+ fln+1) and where j = 2k1(n÷1). where m is the number of ranging subcarriers or pilot tones, n is the number of data-carrying subcarriers or empty subcarriers, and k is an integer number of guard subcarriers or time samples outputted by transform module 68 of transmitter 100 or inputted to the parallel-to-serial converter 64 ot transmiter 100, although any other suitable senal-to-parallel convertor 64 falls within the scope of this document. 11.

The serial-to-parallel converter 64 communicates with a set of accumulators 66 or other data storage device for storing the paralel data outputted by the serial-to-parallel converter 64 In one embodiment, the number (N) of accumulators 68 is approximately (m+j) accumulators.

The accumulators 66 are capable of communication with a transfomi module 68 (e.g., Fast Fourier transform (FF1) module) that transforms the data from a time domain representation to a frequency domain representation for further processing. In one embodiment, the transform module 68 (e.g., FFT transform module) comprises multipoint or N point FFT module, where N = (m + fl(n+1) The output of the transform module 68 (e g FFT transform module) is coup'ed to an input of a tone signal detector 70.

The tone signal detector 70 may comprise a phase detector 72 and an amphtude detector 74 that detects the phase of each received subcarrier (e.g., ranging carrier, tone or piot tone) and the amphtude of each receved subcarrier or tone (e g, pilot tone), respectively for a given sampling interval. The tone signal detector 70 is adapted to communicate with one or more of the following modules or components: the tone phase measurement module 76, the tracking module 86, and the range-domain phase compensator 92.

A tone phase measurement module 76 is capable of communicating with one or more of the following components: the tone signal detector 70, the tracking module 86, and the range-domain phase compensator 92 In an alternate embodiment, the tone phase measurement module 76 is capable of communicating with one or more of the following components: the tone signal detector 70, the tracking module 86, the range-domain phase compensator 92, the multipath detector 94, and the multipath compensator 96 The multipath detector 94 and the multipath compensator 96 are indicated in dashed lines because they are optional and may be omitted in certain embodiments.

The lone phase measurement module 76 compnses a phase slope module 78 and an interpolator 80. The phase slope module 78 measures a first phase of a first pilot tone (or first ranging subcamner) and a second phase of a second oilot tone (or second ranging subcarrier) separated n frequency from tho first pilot tone (or first rangng suocamer) by a known frequency separation For example, the first pilot tone and a secord pilot tone may comprise outer or outermost frequency pilot tones (outermost ranging subcarriers) of the OFDfvl-i<e signal, or any other two tones or subcarriers (e.g., ranging subcarriers) within the OFDM-like signal that are separated by a known frequency separation, The phase slope module 78 determines a phase slope or the slope of the phases between the first pilot tone phase and the second pilot tone phase, or between any a: least two ranging subcarriers The phase slope module 78 does not use or filters out any data carrying subcarriers in the estimation of the phase slope because the phase of the data carrying subcarriers depends upon the data modulation and not any phase distortion from propagation or multipath.

In various embodiments of the method and receiver 200 set forth in this document, the prerequisite for the range measurement technique and the multipath mitigation techniques is that the vector phase of the transmitted tones (ranging suboarriers) are a priori known by the receiver 200, such as by storage in the data storage device 98 The fine range measurement s achieved by measuring an observed phase slope of the received tones (derived from phase of two or more ranging subcarriers) with respect to a reference phase slope derived from the a priori known phase of transmission of the OFDrvi-Iike signaL In one embodiment, the interpolator 80 or interpolation module can estimate the ooserved slope of the phase of the oTher pilot tones (eg. inner tones between the outer pilot tones in frequency) based on interpolation In an alternate embodiment, separately or cumulatively with the interpolation, linear regression or least squares analysis may be used to estimate the phase slope of the received OFDM-like signal. To measure the observed slope at least two data points (e.g., a set of at least two observed phases of ranging subearriers) are needed, the more data points available improves the measurement sgn-to-noise ratio (SNR) In the phase error function, the phase slope ambiguity is a functon of the tone frequency spacing between the observed phases of ranging subcarriers If the tone frequency spac ng is too large, the number of potential zero points increases (e.g., which leads to possible ambiguities in observed phase slope) Accordingly, there is a lower reasonable limit (probably much greater than two) for the number of tones (ranging subcarriers) used to measure the observed phase slope.

The phase slope may be caused by one or more of the following, among ottier factors: (1) a frequency or phase difference between the receiver local oscillator 56 (e g of FIG 2) and any transmitter local oscillator 40 (e g, of FIG 1), (2) uncompensated Dopole' frequency shift in the frequency of the received OFDM-like signal because of relative movement of the location-determining receiver 200 with respect to the transmitter 100 or transmitting beacon, (3) multipath signals assocated with received OFDM-hke signal or received by the receiver 200, or (4) other sources of phase or frequency error in the receiver 200 hardware (e.g., analog4o-digital conversion), In an alternate embodiment of the transmitter 100, the phase difference between received tones can be reduced or minimized if the transmitter 100 is configured to or adapted to omit the cyclic prefix and the cyclic suffix of the OFDM-like signal and if the transform module 68 (e.g., FFT transform module) of the receiver is synchronized with the coherence interval, as explained later.

In an alternate embodiment of the receiver 200, the optional muFtipath detector 94 and the optional (complex-plane) multipath compensator 96 and their respective communication paths, such as virtual or physical communication paths) are shown in dashed lines to indicate that the optional multipath detector 94 and the optional multipath compensator 96 are optional and can be deleted in alternate embodiments. In one embodiment, the optional multipath detector 94 detects the presence or absence of multipath signal in the received OFDM-like signal based on analysis of one or more ranging subcarriers, or the phase slope of two or more ranging subcarriers, in a phase domain. As used in this document, the phase domain shall mean vector space or the complex plane (e.g., 10 (in-phase and quadrature-phase plane signal representation) of one or more rang ng subcarners or plot tones, which conta ns information both on the amplitude and phase of each ranging subearrier. The optional muitipath detector 94 can detect the presence or absence of multipath distortion in the received OFDM-like signal based on one or more of the following: the phase aJope from the phase slope module 78, interpolated phase slope from the interpolator 80, the phase detector output data from the phase detector 72, or amplitude detector output data from the amplitude detector 74. If multrpath distornon!S detected the optional multipath detector 94 may trigger or control he multipath compensator 96 to compensate for the multipath distortion or may send a multipath-related data message to another module of the data processor 104 (e.g., second estimator 90).

In an alternate embodiment of the receiver 200, in the phase domain (e.g., complex plane or 10 plane) the multipath compensator 96 may optionally apply a phase compensation or adjustment to each subcarrier or pilot tone to approach a reference slope of the phases or a target slope of the phases (e g, flat slope or approximately zero slope consistent with the absence of a multipath signal); the reference slope or the target slope may be stored and retrieved from data storage, such as the data storage device 98 based on the a priori known phases of the ranging subcarriers that is transmitted from each transmitter 100, In one such alternate embodiment, the multipath compensator 96 may apply the phase compensation to each pilot tone or subcarrier at a rate equal to or greater than the symbol rate (e.g., of the quasi-data) in an attempt to drive the phase slope to the target slope (e.g., zero slope). However, if the multipath detector 94, the multipath compensator 96, or both are not used in a configuration, then the second estimator 90 in conjunction with the range estimation module 84 or the range domain phase slope compensator 92 can compensate for mulipah or phase slope devation from a reference phase slope or a target phase slope (e.g., approximately zero phase slope).

Further, in certain configurations, the range domain phase slope compensator 92 may be the first, preemptive, dominant or only process that is apphed to compensate for muitipath or phase slope deviation from the target phase slope.

The tone signal detector 70, the tone pnase measurement module 76, or the multipath compensator 96, or any combination of the above components, provide data to the tracking module 86. In one embodiment. the tracking module 86 comprises a first estimator 88 and a second estimator 90. In a first tracking 1oop, a first estimator 88 determines a course estimate of a time of arrival of a set of subcarriers (e.g., at least one subcarrier) of the OFDM-like multitone s'gnals by adjusting the course estimate of the time of arrival of a set of to maximize the correlation of the received pseudo-random noise code of the set (e.g., of at least one suboarrier) with the locally generated replica of the pseudo-random noise (PN) code by the code generator 62 in accordance with a code (first) correlation function. For example, the first estimator 88 and first oop control data 105 may adjust the timing of an early, prompt or late replica generated by the code generator 62. The first loop control data 105 can shift the time or phase of the rephca generated by the code generator 62, such as via shift registers, a clock sigral or another mechanism to adjust the time or phase of replica of the PN code with respect to the received signal or time series signal 36.

The first estimator 88 may communicate with the data storage device 98 or accumulators 66 to store, retneve or process accumulations or correlations 102 stored within the data storage device 98 or the accumulators 66. When the receiver 200 is properly receiving the received OFDM-Like signal (e.g., in a phase synchronized manner), certain accumulations or correlations 102 will conform to the first correlation function 99, which will be described later in greater detail in conjunction with FIG. 4. In one configuration, the data storage device 98 may be used to implement the accumulators 66, such that correlations 102 or accumulations are stored in the data storage device and are accessible to the transform module 68 (e g FF1 transform module).

In a second tracking 1oop, a second estimator 90 determines a precise estimate of the time of arrival of at east one subcarrier of the OFDM-Hke multitone signals by adjusting the precise time of arrival to align the slope of at least two of the ranging subcarrier phases (e.g., or the second of subcarriers) of the OFDM-Iike multitone signals in accordance with a phase (second) correlation unction and corsistent with the course estimate of the tine of arrival associated with the code correlation function. The phase (second) correlation function is described in more detail in conjunction with FIG. 8. In one configuration, second estimator 90 may communicate with the data storage device 98 (e g memory) to access retrieve or process accumulations or correlations that conform to a second correlation function 101 that is stored within the data storage device 98.

In one embodiment, the tracking modue 86 is capable of communicating with one or more of the following modules: a range estimation module 84, the code generator 62, the mnge-domain phase compensator 92, and the phase-domain) multipath compensator 96. For example, in accordance with the second tracking loop the second estimator 90 sends second loop control data 106 to the range estimaLion module 84, the range domain phase compensator 92, or both to comoensate for multipath distortion or pnase slope deviation from a referenco phase slope or target phase slope e.g., approximately zero phase slope) by adjusting the range estimates.

The tracking module 86 also provides output data to the propagation time module 85.

For exanple, the track ng module 86 outputs data that is consistent witn the first loop control data 105, the second loop control data 106, or both to the propagaton time module 85. In one corfiguration, the propagation time module 85 estimates one or more propagation times between the receiver 200 and one or more transmitters 100 based on reception of the ranging information in the received multi-carrier signal.

The range estimation module 84 converts the detected time of arrival of received signal into ranges (e g, diflncos) between receiver 200 and the transmitter 100 For example, the range estimation module 84 or a data processor 104 estimates the location of the receiver 200 or estimated range between the receiver 200 and the transmitter 100 (or between the receiver and multiple transmitters 100) based on the precise estimate and phase error compensation in a range domain commensurate with driving the slope of the subcarrier phases toward a target slope (e.g., zero slope in the absence of multipath). To estimate a precise position or location of the receiver 200, the receiver 200 determines ranges between the receiver 200 and at least three transmitters 100. The receiver 200 can estimate the precise position or location where the ranges (e.g., curves or spherical surfaces.) intersect or overlap.

The range domain phase compensator 92 communicates with the range estimation module 84 and the tone signal detector 70. The range-domain phase compensator 92 may use phase vector data amplitude vector data associated with the received OFDM-like signal from the phase detector 72 the amplitude detector 74, or both Further, the range-domain phase compensator 92 may communicate with the tone phase measurement module 76 to receive estimated phase sope, interpolated phase slope, or oTher phase slope measurements dunng a sampling period. In one embodiment, the range-domain phase compensator 92 can adjust the ranges in the a'ige domain to provide adjusted ranges that are consistent with dnving the slope of the subcarrier phases toward a reference slop or target slope (e.g., approximately zero phase slope) between two or more subcarrer phases, or otherwise provding a range adjustment to one or more range eslimates of the range estimation module 84 to compensate for multipath reception.

If present, for example, the optional multipath detector 94 carl determine whether rrultipath is present in a received signal dunng an evauation interval by analyzing the received amplitude of pilot tone signals or subcarriers anti the received pnase of the pilot tone signals or subcarriers. In an alternate embodiment, the optional multipath detector 94 or the optional multipath compensator 96 can cooperate with the range-domain phase compensator 92 to provide adjusted ranges or range adjustment data for corresponding range estimates that are consistent with driving the slope of the subcarrier phases toward a reference slope or target slope (e g, approximately zero phase slope) The receiver 200 can use various techniques to demodulate the multi-carrier received signal. Under the first technique, the unused suboarriers (or unused data-carrying subcarriers) are set to zero and employed by other transmitters 100 or beacons in the network. Within the transform module 68 (eq, FFT transform module) of the receiver 200 the FFT can be regarded as partial length because the unused subcarriers are set to zero. Here, in the illustrative example of the first technique, the time series 36 has been down-converted to a digital IF where resulting tone fwquencies are multiples of the FF1 execution rate and where the tone frequencies represented by the time series 36 fall directly at the output bins ci the FF1 A' the receiver 200, the received signal from each transmitter 100 would need a different down converson (e g filter), or equivalently separate dLgital lequency adjustment before demodulation. In this implementation the size of the FFT of the transform module 68 only has to be as large as the number of tones (ranging subcarriers) broadcast by a particuiar transmitter (plus any k guard tones). The reduction in size of the FF1 of the transform module 68 is a major advantage of having the broadcast tones spaced evenly through the spectrum. The accumulators 68 can do the orthogonalization (e.g., averaging) that would have been accomplished by the transform niodule 68 (a g FFT transform module) if the ± F were full length, as opposed to partial length.

Under a second technique for demodulating the multicarrier received signal, the serial-to-parallel converter 84 can be full length (e.g., N length, where N is defined above in this document) and use a full length FFT (e.g., N length) in the transform module 68. The tones are then seected off the appropriate laps of the FF1 output of erie transform module 68 (e g, FF1 transform module). The advantage to this full length approach is that data that had been modulated onto the unused or data-carrying subcarriers could be retrieved. It should be noted that the cyclic prefix and cyclic suffix, i a the expected time of each at the receiver 200, are eliminated from the FFT processing to provide enhanced multipath performance or immunity.

At a receiver 200, the received OFDM-like signal is subject to phase error or phase rotation with respect to the transmitted OFDM-Iike signal because of carrier frequency offset and clock offset between the transmitter 100 and the receiver 200, among other things. In the received OFDM-iike signal, the phase error tends to increase with increasing symbol index (e.g., of the quasi-data) n the time domain The aggregate phase error may be associated with two components: (1)first phase error component associated with carrier frequency offset between the OFDM transmitter 100 and receiver 200, where the first phase error component is generally modeled a constant errorwithin an OFDM symbol (e.g., quasi-data symbol), or one or more corstant errors that increase with a respectve increase in symbol index (a g $ of quasi-data), and (2) a second phase error associated with dock offset between the transmitter 100 and receiver 200, where the second phase error is a linear phase rotation with respect to the subcamer index (e g, of the ranging subearner) The aggregate phase error or the second phase error is associated with a generally linear phase slope between any two tones (e.g., pilot tones) in Lhe received OFDM signal (e.g., from the same OFDM transmitter 100 and transmitter site) Linear interoolatron may be used to estimate the aggregate phase error of each subcarrier (e.g., pilot tone) from an observed, measured phase slope associated with two or more tones (e g outer pilots) for corresponding OFDM symbols In addition to the above aggregate phase error, muttipath distortion of the OFOM-like signal may mimic the second phase error.

In one embodiment the range-domain phase compensator 92 may compnse a range-domain phase-slope compensator that compensates for any material deviation of the phase slope from a target phase slope (e.g., approximately zero phase slope). Under a flrst technique, the target phase slope may be approximately zero without an accounting, adjusting offset or compensation for any clock offset between the transmitter 100 and the receiver 200. Under a second technique, the target phase slope may ce approximately zero with an accounting, adjusting offset or compensation for any clock offset between the transmitter 100 and the receiver 200.

In an alternate embodiment, the optional multipath compensator 96 may determine a correction (e.g., phase compensation data or phase correction data) for each subcarrier (e.g., pilot tone) based on the aggregate estimated phase error for each subcarrier (a g, plot tone to yield a revised phase and a revised amplitude for each subcarrier (e.g., pilot tone). The muRipath compensator 96 may apply the correction in the phase domain or may support application in the range domain (e.g., after conversion from the phase domain). For example, the muitipath compensator 96 may determine a correction to produce the revised phase and the revised amplitude for two or pilots with a target slope (eg., an approximately zero phase slope or a phase slope that approaches zero).

If there is multipath between the OFDM transm'tter 100 and the receiver 200, the aggregate estinated phase error or the second phase error may depart fron a generally linear phase slope, or a target phase slope. However, the multipath correction module applies its multipath corrections after demodulation of the received symbols. In one embodiment, the multipath compensator 96 can adjust one or more estimated ranges between the OFDM receiver 200 and transmitter 100 to compensate for non-linearity in the phase slope or to compensate for deviation from the target phase slope The tracking module 86 facUitates estimation of the course time of arrival of one or more ranging subcarriers at the receiver 200. The tracking module 86 reads or accesses correlations, accumulations 102 that are consistent with the first correlation function 99, The correlations or accumulations 102 of the first correiation function 99 can be stored r the accumulators 66 or in the data storage device 98. In the receiver 200, the propagation time module 86 or data processor 104 measures the time of arrival of the pseudorandom code edge.

A first correlation function 99 may represent a code correlation function 404 of FIG. 5, for example, and an associated code error term are used to precisely estimate the code timing error. The time of arrival of the received signal may be measured based on the difference between a transrnssion time of a code edge and a reception time ol the code edge, where the code edges are constructed or demodulated from the received signal (e.g., time series signal 36) by alignment of the received pseudo random noise code with a replica pseudo random noise code signal (from the code generator 62 for a code epoch or sequence.

In a first tracking loop, the first estimator 88 or the tracking module 86 facilitates the propagation time module's (85) estimation of the course time of arrival of at least one subcarrier by minimizing an error associated with an early-minus late error fnCion (e g a first EML function 701, second EML function 702 or a third EML function 703 as illustrated in FIG. 7) The early-minus-late error function (e.g., 701,702,703) facilitates properallgnment of the locally generated replica and received code to realize the first correlation function 94 or code correlation function. As shown in HG. 2, the code correlation function can be performed prior to the transfor, module 68 (e g, FET transform modue) and occurs in the time domain, where al tones are equally affected. For example, the accumulators 68 or data storage device 98 store accumulations 102 or correlations that result from correlators ntegration and dump modules, or other signal processing within the data processor 104 In the frequency domain, the amplitude detector 74 detects a tone amphtude or subcarrier amplitude for providing data to the first estimator 88 or tracking module 86 for adjustment of a relative phase (e.g., in the phase domain or range domain) of the locally generated replica of the pseudo-random noise rode The first estimator 88 or the track ng module 86 supports an early-minus-late error function by providing a control signal or adjustment signal to adjust phase or relative timing of the local generated replica generated by the code generator 62 in the time domain.

In the second tracking loop, the tracking module 86 or the second estimator 90 can perform or track the error function (e.g., second error function or precise error function 805 of FIG. 8) based on one or more tones derived by the transform module 68 (ag., FF1 transform nodue), where the tones are in the frequency domain The second error function or Drecise error function of FIG. 8 facilitates proper alignment of the phase slope of the subcarrier phases of the received OFDM-fike signal with respect to the code error function, consistent with the second correlation function 101 (phase correlation function). The second error function or precise error functioi facilitates range estimatior based on phase information and code information of the received ranging subcarrier for greater accuracy of location or range than would otherwise be provided by the code information alone Because the tracking module 86 can operate on frequency domain representations, the tracking module 86 or second estimator can readily select one or more tones (e.g., two outer tones) or subcarriers for processing, while ignoring other tones to expedite processing and efficiently use processing resources.

As illustrated in FIG 3, a navigation system requires several transmitters 100 or beacons for a location-determinny recever or mobile receiver 200 to determine an accurate location or position based on the signals transmitted by the several transmitters 100. Each transmitter 100 may have its tones shifted with respect to other transmitters 100 (in a mutual reception or coverage area) by a number of frequency bins thus making the signals on each beacon orthogonal, or receivable without interference. However, the tones (e.g., co-frequency tones) could not be used for data transmission (e g, of ranging codes or other intormation) unless each transmitter 100 or beacon is assigned a different orthogonal pseudorandom noise (PN) code and unless each receiver 200 is programmed or configured to generate a replica of the PN code. In such a multi-transmitter system, where each transmitter 100 is assigned a unique set of tones, communications of data, voice signals, or both could stUl be supported by dedicat ng some of those assigned tones to be data carrying or payload carrying, rather than only modulated with ranging code or used for range measurements.

FIG. 4 shows a first correlation function (402, 404) (e.g., autocorrelation function) between the received code and the local gonerated replca of the received code In FIG 4 and FIG. 5, the horizontal axis 403 represents time, whereas the vertical axis 401 represents amplitude or magnitude of the cross-correlation or auto-correlation. The first correlation function (402, 404) is an illustrative example that is analogous to the first correlation function 99 of FIG. 2, where correlations can be stored in the data storage device 98.

To receive the OFDM-hke signal, the recever 200 needs to generate a restoring force consistent with an error function for a pseudorandom code or auto-correlation function (402, 404 or both), as illustrated in FiG. 4. For the OFDM-like signal transmitted by any transmitter 100, m is the total number of tones for the OFDM-like signal, n is the total number of subcarriers, and k is the number of guard suboarriers or K is an integer that represents the length of the cyclic pref x and suffix Consis'ent with FIG 1 and FIG 2, a simulation was constructed with m7 and n=7, with the cyclic prefix and suffix equal to 2k time samples (with k4) which resulted in a 64 point inverse Fast Fourier transform (IFFT) in the transmitter 100 to create the un-augmented time series 36. The simulation used a pseudorandom code of only length 31 to reduce time of a mulation, but the results are indicative of what would be expected with any maximal iength code appropriately scaled. A composite autocorrelation function was calculated as the receive signal strength (RSS, of the tone magnitudes that resulted from the FF1 and is reflected in FIG. 4.

The lull composite auto-correlation function 402 and a representative individual tone auto-correlation function 404 are shown in FIG. 4 with a zoom-in or enlarged central region shown in FIG 5 In one embodiment, the receiver 200 or data processor 104 performs the auto-correlation between the received signal and the locally generated replica of the received signal without the samples (e.g., time series 36) containing the corresponding the cyclic prefix and cyclic suffix, In one embodiment, the transmitLer 100 omits the cyclic prefix and the cyclic suffix such that the phase of each tone produced by the transform module 68 (e.g., FFT transform module) in the receiver 200 is coherent from symbol to adjacent symbol. At the transmitter 100, omitting the cyclic prefix and the cyclic suffice permits a phase-locked oop in the receiver 200 to estimate readfly any frequency error from transmitter 100 to receiver 200.

The frequency error estimation is further described in this document in conjunction with fine range estimation by the second estimator 90 in FIG. 8.

As illustrated in FIG. 5,the shape of the auto-correlation (502, 504, or both), while being simfiar, is not icentical to the shape that the urderlying maximal length pseucorandom code would produce. In F!G. 5, the enlarged, full composite auto-correlation function 502 corresponds to the full composite autocorre!atic'n function 402 of FIG. 4. Similarly, enlarged, the representative individual tone auto-correlation tinction 504 corresponds to the composte autocorrelafion function 404 of FIG & WhUe the envelope of each auto-correlation function (502, 504) is the classic triangular shape that is expected, each auto-correlation function (502, 504) has plateaus (506, 508, 510, 512) or steps that cascade downward from each side of the peak anpitude (503, 505, respectis..ely) of each auto-correlation funcUon (502, 504, respectively) As illustrated each plateau (506, 508) of the full composite auto-correlation function (502) has approximately the same duration; each each plateau (510, 512) of the full composite auto-correlation function (504) has approximately the same duration. The number of plateaus (506, 508, 510, 512) on each side of each corresponding auto-correlation function is equal to (n+1), the spectral spacing of the tones in IFFT bins of the Inverse transform nodule 68), where n is the number of subcarnors In the simulation (n+1) = 8 To understand the reason for these plateaus in the graph of FIG. 5, it is important to examine the time domain signal that results from the simulation described earlier. Each symbol, or code chip period, includes eight(S) cycles of tone coherence where all tones simultaneously return to zero phase.

FIG 6 isa tine domain signa graph that illustrates he tone coherence intervals (for a group of individual tones or suboarriers). The tone coherence interval means a duration or time (601) during which all tones successively or simultaneously return to zero phase-For example.

FIG. 6. is a graph that illustrates tone coherence intervals (as a composite tine series 36) for at least one code chip duration (602) or symbol period Each subcarrier is offset from other subcarriers by a frequency offset as previously indicated. The vertical axis 603 of FIG. 6 roorosents the amplitude of a series of tones, whereas the horizontal axis 605 represents the time, which includes a number of coherence intervals 601. In FIG. 6, the subcarriers or tones are aligned at zero phase for a certain number of tone coherence intervals during each symbol or code chip period, where at least one complete code ship period is shown. The cyclic prefix and suffix are shown to last for 0.5 chips of a coherence interval. The duration of each plateau coincides with a corresponding one of these coherence intervals. During one of these coherence intervals, the autocorrelation (502, 504) in FIG 5 remains constant with its amplitude e.g., tread of each step or plateau 506, 508, 510, 512 set to the amplitude of the interval mid-point of the code auto-correlation. The size of the peak plateau (503, 505) at the top of the auto-correlation function (502, 504) is a function of the cyclic prefix and cyclic suffix. The remaining plateaus (506, 508, 510, 512) begin as each coherence interval within the symbol is removed ftom the correlation The correlation remains relatively constant while a portion of that coherence interval is included in the correlation since the tone in the adjacent symbol has exactly the same phase fifty (50%) of the time since the correlation is performed over the expected duration of the tones without the cyclic prefix and suffix. This results in coherent phase across chip boundaries other than the phase reversal of the code polarity fifty (50%) of the time.

In the receiver 200 of FIG. 2, the code generator 62 may produce early, late and prompt outputs of the ocaly generated rephca for input to the mixer 60 In one embodiment, the receiver 200 creates a code error function by correlatng an earls, code and a ate code, separated by some fraction of a code chip and subtracting the two results. This code error function is commonly referred to as an early-minus-late EML) error function. One or more eady-minus-late (EML) functions can be reaUzed in the receiver 200 to decode the OFDM-like signal. The receiver 200 can calculate the code error function for various fractional code chip separation by using the composite autocorrelation function (e.g., 402 shown in FIG. 4), and subtracting an early version from a late version of the autocorrelation function (ag., 402 of FIG. 41 In one configuration, the EML is reahzed by the operation of the code generator 62 mixer 60, the data processor 200 (e.g., integration and dump module). and accumulators 66, where the data processor 200 may incude an ntegration and dump module between the mixer 60 and the accumulator 66 or between the mixer 60 and the serial to parallel converter 64.

FIG. 7 illustrates an EML function of the receiver 200 for various fractional code chip separations. In FIG. 7 and FIG. 8, the vertical axis 713 represents an amptude of the EML function, whereas the horizontal axis 711 represents time (e.g., in chips). The first EML function 701, indicated by the long dashed lines, has a chip spacing of S chLps, the second EML function 702, indicated by dot-dash lines, has a chip spacing of 25 chps, the thrd EML function 703, indicated by short dashed lines, has a chip spacing of.125 and a fourth EML function 704, indicated by dotted lines, has a chip spacing of.0625 chips.

The slope of this normalized EML error function (701, 702, 703) is constant near the zero crossing 705 for all separation greater than 0.125 chips. The dead zone 707 shown for fourth EML function 704 with the ± 0.0825 chip separation is the result of the duration of the plateaus, which are 0 125 chips long The plateau duration is a function of the number of coherence periods of the tones within a symbol, as discussed earlier.

FIG. 7 illustrates the EML functions over a range of approximately 3.5 chips and shows the peak amplitude 709 of the EML functions FIG 8 illustrates an EML function of the receiver for various fractional code chip separations in less detail than FIG. 7. In FIG. 8, the first EML function 801, indicated by the long dashed lines, has a chip spacing of 5 chip and corresponds to the first EML function 701 of FIG. 7; the second EML function 802, indicated by dot-dash fines, has a chip spacing of.25 chips and corresponds to the second EML function 702 of HG. 7; the third EML function 803, indicated by short dashed fines, has a chip spacing of chips, and corresponds to the third EML function 703 of FIG. 7; a fourth EML function 804, indicated by dotted lines has a chip spacing of.0625 chips and corresponds to the fourth EML function 704.

FIG 7 and FIG 8 provide illustrative graphs of earIyminus-late functions that can he used by the data processor 104, the first estimator 88, the tracking module 86, or the propagation time module 85 in the coarse error estimate. The course error estimate can be expressed in milliseconds error in the propagation time module 85, as opposed to a chip error.

For example, in the first control loop the first estimator 88 of the tracking module 86 may use the first [ML function (701, 801), the second EML function (702, 802), or the third EML function (703, 803) as the code error function, course error function or first error function for determining first ioop control data 105 to control the code generator 62. As illustrated in FIG. 7, the average slope near the zero crossing of the error function of FIG. 7 is the ideal early-minus-late slope of approximately one chip error per chip, or, as shown for the simulation, approximately 20 millisecond of error per cnip Because the symbol period (code chip) is expected to be relatively long, the tracking quality afforded by this type of code error function may not provide the degree of precision for position estimates, location estimates, or attitude estimates that are required for certain applications. However, the first control loop and the applicable EML function (e.g., code error function) can be used to achieve an initial estimate or course estimate of the correct time of arrival. Further, in one embodiment, the propagation time module 85 provides an initial estimate or course estimate that can be used as Lne seed for the precise erior function that will be described in greater detail in conjunction with FIG. 8. In one embodiment, the propagation time module 85 that rehes upon the appropriate [ML error function is capable of providing an initial accuracy of ±0.125 chips, or the equivalent time enor.

In FIG. 8, the tracking module 86 or the second estimator 90 applies precise error function 805 to make use of the phase slope (of the phases) of the received tones (e.g., ranging subearriers) of the OFDM-hke signal lithe FFT processing time, over which the FF1 is performed in the transform module 68 (e.g., FFT transform module), is synchronized to the coherence intervals, the phase slope of the phase of the tones will be zero or reduced from any propensity toward a greater than non-zero aggregate phase slope. Any error in the timing will result in the phase slope (of two or more ranging subcarriers) representing that time error. The &mple equation is: dcc radian.s seconds dw radians Tseconds In one embodiment, the precise error function 805, phase error function, or second error function is defined as the normalized slope of the phase of the subcarriers (or tones) with respect to correlation timing error and is shown in FG. 8. The precise error function 805 with its normalized siope of the phase is shown superimposed on the central area of the early-minus-late error functions (801, 802, 803, 804), previotsly described As illustrated, the precise error function SOS is a general linearization of the perfectly normalized error function without having to estimate a normalizing coefficient.

The precise errorfunction 805 of FIG. 8 is defined or characterized by a phase slope that intercepts a central EML error function about the zero crossing point As illustrated, the phase slope can provide unambiguous indication of the chip error in a region between approximately +/-05 chips about the zero crossing point of the EML functions The measurement noise from a conventional Delay-Locked Loop DLL.) typically used to track pseudorandom codes can be compared to the phase slope techniques describes above.

The performance of a conventional DLL has been analyzed thoroughly and the I a thermal noise tracking jitter for an Early-Minus-Late error function is given by: 2d2 4d 0 4 SNRL [2(1 -d) ± SNRJ where: d is the carrel ator spacing between early, prompt and late SNRL is the tracking loop SNR SNR0 is the predetection SNR For the simulation a code rate of 50Hz as used d isO 125 rn nirrum, a 4Hz loop bandw dth arid Signal-to-Noise ratio (SNRL) is chosen to be +40dB. Because the signals used for this illustrative analysis have high SNR's, the squaring loss represented by SNR0 can be ignored.

The resulting calculated tracking jitter using the above equation is.00234 chips or 14021 meters. Using the same parameters for Signal-to-Noise ratio SNR in the simulation, the standard deviation of the phase slope is calculated to be lA-2p seconds or 426 meters. The improvement in measurement noise, when using the phase slope method (e g, second tracking loop of second estrnator 90) as compared to the traditional code tracking loop (o g, first tracking loop of first estimator 88) with an appropriate EML function with the same SNR, is approximately 30dB for this set of parameters.

If the code rate is increased by a factor of 1000 but the 1 Hz loop bandwidth is retained, the SNR remains at +40dB the tracking error would be reduced by the same factor to approximately 14 meters. For example, in the second tracking loop of the second estimator 90 of the tracking module 86, the phase slope method with the tone frequencies can increase accuracy in a position estimate by the same factor of 1000 results in a measurement standard deviation of.426 meters or the same factor of 1000 decrease over the original value, with the same 30 dB advantage over the Early-Minus-Late approach of the first tracking loop aione.

FIG QA Hus'rates th9 impact of multipath distortion between the transmitter 100 and the receiver 200 over a shorter period than that illustrated in FIG. 9B. As illustrated in FIG. 9A, a direct path signal vector 901 may be aligned or coincident with a real axis 950 (e.g. in-phase axis), whereas the an imaginary axis 954 (e.g. quadrature axis) may be oriented substantially perpendicularly to the real axis, although other ahgnments of the axes (950, 954) with respect to the signal vectors (901, 902, 904) in the complex plane or vector space are possible. The axes are indicated in dashed lines in FIG. QA and FIG. 93.

In FIG. 9A, each multipath signal error vector 904 is a vector difference (in phase and magnitude) between the direct path signal vector 901 and each respective observed signal vector 902. From time to time, the observed signal vector 902 may also be referred to as the rrultipath-impacted signal vector Each observed signS vector 902 is the nieasred signal vector by the tone signal detector 70 of the receiver 200, for example. The observed signal vector 902 may contain multipath, phase distortion, or phase slope between two or more ranging subcamers The dtrect path signal vector 901 ard each observed signal vectors 902 intersect at the first vertex 900 at a phase error angle (0). Depending upon the degree of multipath component in the observed signal vector 902, the observed signal vector 902 forms a corresponding phase error angle (e g Oo Qi 02 03 0, 05 and Os as illustrated in FIG 9A) The direct signal vector 9W and each multipath signal vector 904 intersect at the second vertex 906 at instantaneous angle (B). Depending upon the degree of multipath component in the observed signal vector 902, the direct signal vector 904 and each multipath signal error vector 904 forms a corresponding instantaneous angle (e.g. 00, °i'°2'3' 94,95,9, as illustrated in FIG. 9A).

The observed signal vectors 902 and respective multipath signal error vectors 904 intersect a series of corresponding third vertices 952. As illustrated to simply FIG. 9A and FIG. 93, the magnitude or amplitude of each multipath signal vector 904 is fixed or constant, although in practice the amplitude of each multipath signal vector would va'y For nlustrative purposes in FIG, QA and FIG. 96, circumference 905 is spaced by a radius about the second vertex 906 to reflect that the observed signal signal (902, 908) is assumed to have a uniform signal strength below the direct path signal strength (a g 6 dB lower) of direct path signals 901 although the method and system disclosed in this disclosure is in no way limited by this iUustrative assumption.

In FIG. 9A, the phase of each tone observed signal vector 902 can be associated with a reflerbon of a coriesponding indirect path signal transmitted from the transmitter location Fach tone observed signal vector 902 can vary in phase and amphtude with respect to the drect path signal vector 901 Ror example, the phase of the tone observed signal vector 902 or resultant multipath-influenced phase slope of the tones depends upon a first propagation dis'arice of the direct path signal, a second propagation distance between the transmitter and reflecting object, a third propagation distance between the reflecting object and the receiver, and the phase shift (e g, phase reversal) associated with the reflecton from the surface of the reflecting object FIG. 9A may represent an observed signal vector 902 of a ranging subcarrier or pilot tone over multiple sampling intervals within a time period, or observed signal vectors 902 of corresponding ranging subcarriers or pilot tones during a single sampling interval. If the latter is ilustrated in FIG 9A, each tone cbserved signal vector 902 has a phase shift with respect to the other tones or subcarriers of an OFDM-Iike signal, where in the aggregate the tones may be characterized by a multipath phase slope ln practice, with respect to an OFDM-like signal that is impacted by multipath, each received subcarrier or tone of the observed signal vector 902 tends to have a slightly different phase, even if the direct path signal 901 compnses tones that are generally aligned in phase e g, or have a lesser phase slope) Accordingly in FIG 9A each ray or arrow of the tone obse'ved signal vectors 902 can represent a corresponding subcarrier phase that is offset from the direct path signal vector 901 by a corresponding multipath signal error vector 904; each of the multipath signal error vectors 904 can refer to a different subcarrier or tone of the transmitted OFDM-like signal that are transmitted synchronously or simLltareously (e g from diferent transmitters 100) Accordingly, the muitipath-impacted signal or observed signal vector 902 can reduce the Qrecision of the position estimate of the receiver 200, unless the range-domain phase Gompens&tor 92 or a phase slope compensator (e.g., multipath compensator 96) applies appropriate compensation.

In FIG. 93, the observed signal vector 908 of FIG. 93 is analogous to the observed signal vector 902 (or the multipath-impacted signal vector) of FIG. 9A; and the resultant multipath signal vector 944 of FIG 9B is analogous to the resultant multipath signal vector 904 Like reference numbers indicate like elements in FIG. 9A and FIG. 93.

FIG. 9B illustrates the rnpact of canceled or reduced multipath distortion between the transmit location of the transmitter 100 and the mobile receiver location r the receiver 200 over a minimum threshold time period or long-term time period (e.g., greater than the time period represenled in FIG. 9A). In FIG. 98, the receiver location is mobile. In FIG. 98, as the tone observed signal vector 908 changes by motion of the mobile receiver 200 over a greater time period e.g., and possibly a greater spatial area) than illustrated in FIG. 9A, as depicted, the resultant observed signal vectors 908 change and the distribution of resultant observed signal vectors 908 change.

FIG. 93 illustrates the contribution of each observed signal vector 908 of a pilot tone to a phase of the average e.g., mean,) observed signal vector 907 of the observed signal vector of the pilot tone of the OFDM-iike signal over a long time period or after the expiration of a minimum threshold time period Similarly, FIG 913 illustrates the nontr,buton of each multipath signal error vector 944 of a pilot tone to a phase of the average (e.g., mean) multipath signal error vector 903. For example, the tone signal detector 70 and the rnultipath detector 94 may use the average phase (e.g., mean phase) of the multipath signal error vector 944 of a single pilot tone or the average (e.g., mean) phase slope of the rnultipath error vectors 944 of multiple pilot tones (relative to vertex 906) to provide one or more performance indicators (e.g., figures of merit) for judging the mult'path compensation performance of the receiver 200 or the degree of multipath received in the uncompensated OFDM-like signal.

If one compares FIG. 9A and FIG. 9B, it becomes obvious that eventually, as the mobile receiver 200 moves, each third vertex 956, ts corresponding observed signal vector 908 (e.g., multipath impacted signal vector) and its corresponding multipath signal error vector 944 will traverse the entire circumference 905, hence, each multipath signal error vector 944 for each pilot tone experiences corresponding phases that are associated with the circumference 905. In one embodiment, if the data processor 104 (e g, range estimation module 84 of the range-domain phase compensator 92) a'erages the observed signal vectors 908 to obtain an average resultant observed signal vector 907 or an average multipath-impacted signal for each tone or for a group of tones of the OFDM-like signal, sucn that the average (e g, mean) resultant observed signal vector (909) represents nearly or approximately the direct path phase or aligned phase (e.g., with reduced or eliminated multipath phase component). Similarly, the average (e g, mean) resultant multipath error signal vector 903 is cons'stent with nearly or approximately the direct path phase or aligned phase of the direct path signal 909. Depending upon the multipath-impacted signa vector 908, the direct pain signal may compnse generally a normal direct path vector 909, an attenuating direct path vector 907, or an oversized direct path signal vector 958.

jfl an alternate embodiment, the data processor 104 (e.g., range estimation module 84, or the range domain phase compensator 92, or multipath compensator 96) may determine an average (e.g., mean) observed signal vector during minimum threshold time period to yield the average resultant observed signal vector 909 Accordingly in the receiver 200 any disagreement of or nonconformity of the average or mean of the indMdual instantaneously observed resultant signal vector with respect to the average resultant signal vector 909 would be a rnultipath metric on the quality of the observed measurement or the degree of multipath distortion the observed measurement for each ranging subcarrier. In one embodiment, the average resultant signal vector 909 of one or more subcarriers is used to define a reference phase slope or target phase slope for the OFDM4ike signal in the absence of material multipath components.

FIG. 1OA through FIG. 100, inclusive, show the amplitude versus subcarrier identifier (e.g. tone identifier) for a number of different subcarriers (e.g., approximately 32 different carners) Although 32 subcarriers are illustrated, tfts disclosure applies to virtually any number of subcarriers greater than two. As illustrated, the vertical axis 960 shows amplitude of each subcarrier, whereas the horizontal axis 962 shows the subcarrier identifier of each subcarrier.

In FIG. WA through FIG. 1OD, the multipath impacts the amplitude of each suboarrier (tone) differently because the frequencies of each subcarrier and; hence, wavelength of each subcarrier are sightly d iferent Accordingly the amplitude curve (969, 966, 961, and 968, as depicted by the rectangular symbols) varies with the corresponding subcarner identifier The generally linear plot or curve 964 depicts an average amplitude of each subcarrier, where the generally linear plot or curve 964 is indicated by the diamond symbols. In one configuration, the generally linear plot or curve 964 is consistent with the average length of vector (909 or 907) in FIG. 9B, for example.

In HG. 10k the amplitude curve 969 decreases (e.g., somewhat linearly) with increasing frequency of the subcarrier or subcarrier identifier of the OFDM-Iike signal. In FIG. lOB, the amplitude curve 966 has amplitude peaks at the lowest subcarrier frequency and the highest suboarrier frequency, with a somewhat concave curve between the lowest suboarrier frequency and the highest subcarrier frequency In FIG IOC, tne arnpltude curve 967 increases (e.g., somewhat linearly) with increasing frequency of the subcarrier or the subcarrier ident'fior In IG IOD, the amplitude curve 968 has amplitude minimums at the lowest subcarrier frequency and the highest subearrier frequency, with a somewhat convex curve between the lowest subcarrier frequency and highest subcarrier frequency. In FIG. IOA through 1OD, ,nclusive, the above amplitudes of the subcarners (tones) change as the multipath changes due to path length changes such as vehicle movement to produce a series of plots.

In one configuration, the generally linear plot or curve 964 is consistent with the average length or amplitude of vector signal 909 in FIG. 9B. The average ength of the direct path vector may be based on the average of a lowest amplitude of the direct oath vector signal 909 (as indicated by reference numeral 907) and a greatest amplitude of the direct path vector signal 909 (as indicated by reference number 909 plus reference number 958, illustrated as a dotted line). In FIG. 9B, when the observed signal vector 908 (or multipath-impacted signal vector) and the multipath signal error vector 944 align in opposite phase, the amplitude of the shortest vector is at 6 o'clock of circle 905. This shortest vector or lowest amplitude has the amplitude of subcariie' identifier 16 in curve 966 Likewise, whcn the observed signal vector 908 (or multipathmpacted signal vector) and direct path signal vectors 909 have the same phase, they add together for a maximum which is shown in FIG. 9B as the oversized direct path vector 958 (that includes direct path signal vector 909) at 12 o'clock (of circle 905), which is the equivalent of subcamer identifier 16 in curve 968 of FIG 1OD The average of the amplitudes of subcarrier identifier 16 from FIG. lOB and FIG. IOD, in curves (966, 968,) can be used to determine the A. ±A average amplitude (A) of the drrect path sgnal, or A = rn' 2 max whereA,,, is the minimum amplitude of a given subcarrier and where Amax is the maximum amplitude the given subcarrier.

Similarly the multipath amplitude}.) of the given subcarrir is approximately one-half of the difference between the amplitudes ot subcarrier identifier 16 from FIG. lOB and FIG. 100, in curves (966, 968), which is Amflftoth = AmcA -. With the average amplitude of the direct path signal 909, the corresponding multipath amplitude of the muitipath signal error vector 944, and the multipathimpacted signal vector 908 (observed from the received signal), the receiver or data processor can determine the phase error angle or correction oats to remove its infiuence on the calculated phase. Alternately, the receiver or data processor can determine the instantaneous angle) and the length and the corresponding length of the multipath signal error vector (904 944) to remove its influence on the multipath-imoacted signal vector (902, 908). The receiver can apply the above correction data to any given subcarrier from FIG. 1OA through FIG. IOD, inclusive.

In an alternate embodiment, another multipath compensation technique makes use of the fact that the average of the length of the tones impacted by multipath (indicated by squares in curves 969, 966, 967, 968) is equal to the length of the direct path tone Qndicated by diamonds in curves 964). In HG. 1 OA through FIG. 100, as the amphtudes of subcarriers (tones) are averaged, the length and phase of the direct path results. Accordingly, averaging the amplitudes of the subcarriers (eg,associated with subcarner identifiers I through 32, inclusive) until all of amplitudes of the subcarriers (tones) have the same average vector length, provides a set of tones free from rnultipath; where the corresponding phase s the phase of the direct path signal.

FIG. 11 is flow chart of one embodiment of determining the location or range of a receiver 200 with a multi-carrier signal. The method of FIG. 11 begins in step S300.

In step S300, a reGeiver 200 portion receives or is adapted to receive a plutahty of OFOM-like multitone signals. Each OFDM-like multitone signal has subcarriers (tones) and is modulated with a pseudo-random noise code and transmitted from one or more transmitters 100 with corresponding known locations.

In step 3302, in a first tracking loop a first estimator 88 deterinires a course estimate of a time of arrival of a set (e.g.. at least one subcarrier) of the OFDM-Iike multitone signals by adjusting the course estimate of each time of arrival of the set in accordance with a code (first) correlation function. For example, in a first tracking loop, a first estimator 88 determines course estimates of respective times of arrival of each ranging subcarrier within the set of the OFDM-like multitone signals by adjusting the course estimates of each time of arrival of the set in accordance wth a code (first) correlation function Each set of, one or more, ranging subcarriers from a transmitter lOUis generally associated with a different or unique time of arrival for each course estimate, unless all of the transmitters happen to be located equidistantly from the receiver 200. The first estimator 88 or data processor 104 is adapted to maximize the correlation of the received pseudo-random noise code of the at east one subcarrier with the local replica of the pseudo-random noise code in accordance with a code (first) correlation function In step 3302, the determining of the course estimate may be executed in accordance with various techniques, which may be applied separately or cumulatively. Under a first technique, in the first tracking loop, the determining of the course estimate further comprises minimizing an error associated with an early-minus-late error function to determine the course estimate of the time of arrival Urder a second technique, the early minus late spacing is selected based on a ratio of the pseudo random noise code period and frequency spacing between subcarriers, Under a third technique, the early-minus-late error function has an early mnus late spacing between approximately 0625 and 5 chips Under a rourth technique, the code correlation of the ffrst tracking roop occurs in the time domain. Under a fifth technique, in a frequency domain the tone signal detector 70 or amplitude detector 74 detects a tone amplitude associated with the at least one subcarrier for providing an adjustment signal to adjust a relative phase of the locally generated replica of the pseudo-random code to the recer.ed code Under a sixth technique, the code correlation function comprises an autocorrelation function for the received signal and a locally generated pseudo random noise code signal that is expressed in magnitude versus time, wherein the autocorrelation function has a central peak associated with a maximum magnitude of a correlator output signal and a series of step-like transitions on each side of the central peak.

In step 3304, in a second tracking loop, a second estimator 90 determ nes a precise estimate of the time of arrival of the set (e.g., the at least one subcarrier; typically at least two ranging subcarriers per transmitter 100) of subcarriers of the OFDM-like multitone signals by adjusting the precise time of arrival to align the slope of (at least two of the) subcarrier phases of the OFDM-like multitone signals in accordance with a phase (second) correlation function and consistent with the course estimate of the time of arrival associated with the code correlation fLnction For example, in step S304 in a second tracking boo, a second estimator 90 determines precise estimates of the respective times of arrival of the set of subcarriers of the OFDM-like multitone signals by adjusting the precise times of arrival to align the slope of (at least two of the) subcarrier phases of the OFDM-like multitone signals in accordance with a phase (second) correlation function and consistent with the course estimates of the time of arrivals associated with the code correlation function. Each set of, two or more, ranging subcarriers, from a transmitter 100 is generally associated with a different or unique time of arrival for each precise estimate, unless all of the transmitters happen to be located equidistantly from the receiver 200. Two or more ranging subcarriers can establish a phase slope for single reference transmitter 100, for example.

In step 3304, the determining of the precise estimate is carried out n accordance with one or more procedures, which may be carried out alternatively or cumulatively. Under a first procedure, in the second tracking loop, the data processor 104 or the receiver 200 the determines the precise estimate further comprises an error function with a target or goal of driving an average observed tone phase slope to a zero phase slope, or another target phase slope (e g associated with the abserce of mulipath signals) Undor a second procedure, the data processor 104 or the receiver 200 adjusts the precise time of arrival to align the slope of the subcarrier phases of the OFDM-Uke rnultitone signals in accordance with a phase (second) correlation function comprises adjusting a central region of the phase slope, between adjacent ambiguity regions eg., with generally vertical slope) in the phase slope, to be aligned with a zero crossing point of the first early minus late function.

Under a third procedure, the receiver 200 or data processor 104 processes the received OFDM-like signal encoded without any cyclic prefix or cyclic suffix such that the phase o each tone produced by a transform module 68 (e.g., FFT transform nodule) in the receiver 200 is coherent from symbo to adjacent symbol. Accordingly, under the third procedure, the target phase slope may approach approximately a zero phase slope.

Under a fourth procedure, the receiver 200 or data processor 104 synchronizes the processing time, over which a time domain-to-frequency domain transform (e.g., FF1 transform) is performed by a transform module 68 (e.g., FF1 module) to a coherence interval (e.g., described in coniunction with FIG. 6) such that phase slope of the subcarriers wifl approach zero, or be reduced from any propensity toward a greater non-zero aggregate phase slope, for the target phase slope.

In step 3306, a data processor 104 or range est mation module 84 estimates the location of the receiver 200 or estimated range between the receiver 200 and the transmitter based on the precise estimate and phase error compensation in a phase domain, a range domain, or both commensurate with driving the slope of the subcarrier phases toward zero (e.g., to compensate for phase or frequency error in the absence multipath). For example, the data processor 104 or the range estimation mocule 84 estimates the location of the receiver 200 or estimated range between the receiver 200 and he transmitter 100 based on the precise estimate after any compensation for multipath and phase error compensation in a phase doma n, a range domain, or both commensurate with dnving the slope of the subcarrier phases toward zero.

The range-domain phase compensator 92, the multipath detector 94cr data processor 104 can compensate for certain multipath induced distortion by applying one or more procedures as described in any of the embodiments are examples set forth in FIG. 12 through FIG. 16, inclusive. Forexample, theembodimentsof FIG. l2through FIG. 16, inclusive, may measure vector lengths of a multipath-impacted signal and a direct path signal, among other tnings, to denve a phase error correction (e g, phase error angle between the direct path signal and the multipath-impacted signal). Accordingly, if the frequency or wavelength of a reference ranqing subcarner or reference tone s known, then the data processor 104 or range domain phase compensator 92 can determine a phase correction, or its equivalent range correction, for each subcarrier vector based on the length of the multipath-impacted vector, the length of the dftect path vector, and the frequency or wavelength of the subcarrier.

HG 12 is flow chart of one embodiment of determining the location of a receiver 200 with a multi-earner signal The method of FIG 12 is sinidar to the method of FIG 11 except the method of FIG. 12 further comprises steps S350 and S352. Like reference numbers in FIG. 11 and FIG. 12 indicate like elements.

In step S350, a data processor 104 or multipath detector 94 identifies a multipath-impacted signal (e g, multipatn signal component) ansing from the transmitted OFOM-like signal, wherein the identifying is accomplished by evaluating a received amplitude of the set (e.g., the at east one subcarrier) of subcarrlers of the OFDM-like signal. The identWying of the multipath-rnpacted signal may be accomplished by the application of various techniques which may be applied individually or cumulatively. Under a first technique, during a group of adjacent sampling intervals, the data processor 104 or multipath detector 94 determines one or more vector lengths of the direct path signal during peak ampltudes and the data processor 104 or multipath detector 94 determines one or more vector lengths of the multipath-impacted signal during an intervening lower amplitude lower than the peak amplitudes. Under a second technique, during a group of sampling intervals, the data processor 104 or multipath detector 94 determines whether the amplitude of the tones o g, pilot tones) dur ng different sampling periods have a dispersion level exceeding a minimum threshold indicative of the existence of a multipath-impacted signal The second technique triggers mLltipath compensation b, range adjustment the estimated range based on the dispersion level exceeding the minimum threshold.

Under a third technique, the data processor 104 or multipath detector 94 averages a plurality of observed vector lengths for each ranging subcarrier to determine corresponding averaged subearrier vector lengths, and the rrultipath detector 94 identifies the direct signal vector length when the corresponding averaged subcarrier vector lengths converge to substantially the same value that is designated as the direct signal vector length Under a fourth technique, the data processor 104 or tone signal detector 70 detects a greatest amplitude of one of the subcarriers of the OF UM-lice sigral and a lowest amphtude of one the subcarriers of the OFDM-like signal (e g, by recording through a generally continuous evaluation period that contains the greatest amplitude and the lowest amplitude), the data processor 104, the tone signal detector 70, or the muitipath detector 94 determines the direct path amplitude based on the greatest amplitude and the multipath signal based on a lowest amplitude; and estimates a multipath phase error for a particular ranging subcarrier based on the corresponding direct path arnphtude and multipath amplitude for any subcarrier with an observed amplitude that is not equal to or substantially less than the direct path amplitude of the direct path signal.

In step S352, a data processor 104 or multipath compensator 96 compensates with a multipath phase compensation for a multipath phase error in the at least one subcarrier phase assocated with the dentified mutipath-impacted signar arising from the transmitted OFDM-like signal; wherein the phase compensation data results in or comprises an adjustment to the estimated range (between a receiver and a corresponding transmrter position) based a direct signal vector length of a direct path signal and a multipath-impacted signal vector length of multipath-impacted signal that are determined through observations (e g observe signals or signal vectors) of the receiver 200. As previously indicated, the multipath-impacted signal is also referred to as an observed signal vector. The data processor 104, muthpath compensator 96 or the range-domain, phase-slope compensator 92 can use compensation data for a first, representative or reference ranging subcarrier to estimate additional compensation data for the other ranging subcarriers based on one or more of the following d iferent frequency or waveiengths of the ranging subcarners, an observed phase slope of the ranging subcarriers, and a target phase slope of the ranging subcarriers.

Step 5352 may be executed in accordance with various techniques that may be applied separately or cumulative. Under a first technique for step S352, a data processor 104 or multipath compensator 98 compensate produces phase compensation data that results in or comprises an adjustment to the estimated range based on tie direct signal vector length, the multipath-impacted signal vector length (e.g., observed signal vector length), and a multipath error rector signal. Under a second technique for step S352, the data processor 104 or multipath compensator 9$ comprised a phase error angle between the direct signal vector and the muitipath-impacted signal vector determined based on the direct signal vector length, the rnultipath-impacted signal vector length (or an observed signal vector), arid an rnultipath error vector length Under a third technique for step S352, the data processor 134 or multipath compensator 96 estimates the multipath error vector length as a substantially fixed amplitude below the amplitude of the direct path s gnal Under a fourth technique for step 5352, the data processor 104 or multipath compensator 96 estimates a multipath error vector length derived from an instantaneous angle between the direct path signa and the multipath error vector and a ratio of the direct signal vector length to the muitipath signal vector length (or the reciprocal of the ratio) Under a fifth technique for step 5352, the data processor 104 or multipath compensator 96 estimates a phase error comprising an angle between the direct signal vector and the niultipath signal vector (or observed signal vector), where the data processor 104 or the range domain phase slope compensator 92 separately converts the phase error for each ranging subcarrier of the received OFD.M-Iike signal to the phase compensation data, and where the propagation time module 85, the range estimation module 84, or both convert the phase error to a corresponding time error to a distance error, and where the distance error is based on the propagation at the speed of light.

FIG. 13 refers to FIG. 1 3A and FIG, 1 3B couectively. FIG. 13 is a flow chart of another embodiment of determining the location or range of a receiver 200 with a multi-carriei signal with multipath compensation. The method of FIG. 13 is similar to the method of FIG. 12, except the method of FIG. 13 further comprises step S354; ike reference numbers in FIG. 12 and FIG. 13 indicate like elements, steps or procedures.

n step S354, a data processor 104 or multipath cornpeisator 96 determines the direct signal vector length based on (e.g., an average of an observed (approximately) shortest vector length and) an observed (approximately) longest vector length for at least one subcarrier within the set of subcarriers during a sampling period, wherein the direct signal vector length comprises an amplitude of the direct path signal. For example, during a group of adjacent sampling intervals, a data processor 104 or multipath compensator 96 determines one or more direct signal vector lengths of the direct path signal during peak amphtudes and determining one or more multipath-impacted signal vector lengths of the multipathimpacted signal during an intervening lower amplitude lower than the peak amplitudes, wherein the amplitude measurements comprise the one or more direct signal vector lengths and the one or more mutipath-impacted signal vector Ienqths Further, the data processor 104 estimates the mutipath error vector length as a substantially ixed amphtude below the amphtLde of the direct path signal. Alternately, the data processor 104 multipath error derives the vector length from an instantaneous angle between the direct path signal and the multipath error vector and a ratio of the direct signal vector Length to me multipath signal vector length FIG. 14 refers to FIG. 1 4A and FIG. 14B collectively. FIG. 14 is a flow chart of another embodiment of determining the location or range of a recever w th d muli-car ner s gnal with multipath compensation. The method of FIG. 14 is similar to the method of FIG. 12, except step 5356 replaces step S352 and steps 5354 and S358 are is added. Like reference numbers in FIG. 12, FIG. 13 and FIG. 14 indicate like elements, steps or procedures.

In step 5358, a data processor 104 or multipath compensator 96 compensates with a multwath phase compensation or a multipath phase error in the at least one subcarner phase associated with the identified multipath-impacted signal, wherein the phase compensation data results in or comprises an adjustment to the estimated range based on the direct signal vector lengTh, the multipath-impacted signal vector length, and a multipath error vector signal In step 5354, a data processor 104cr niultipath compensator 96 determines the direct signal vector length based on e..g, an average of an observed (approximately) shortest vector length and) an observed approximately) longest vector length for at least one subcarder within the set of subcarriers during a sampling period, wherein the direct signal vector length comprises an amplitude of the direct path signal. Variations of step 5354 discussed in conjunction with FIG 13 apply equally to FIG 14 In step S358, a data processor 104 or multipath compensator 96 determines the multipath error vector signal based on approximately one-half of a difference between an observed (approximately) shortest vector length and an observed (approximately) longest vector length fe-a subcarrier within the set of subcarners during a samping penod FIG. 15 refe to FIG. 1 GA and FIG. 1 SB collectively and is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carrier signal with multipath compensation. The method of FIG, 15 is similar to the method of FIG. 12, except step 8360 is added. Like reference numbers in FIG. 12 and FIG. 14 indicate like &ements, steps or procedures.

In step 5360, dunng a group of adjacent sampling intervals, a data processor 104 or nuitipath compensator 96 determines one or more direct signal vector lengths of the direct path signal during peak amplitudes and determining one or more multipath-impacted signal vector lengths of the multipath-irnpacted signal dunng an intervening lower amplitude lower than tne peak amplitudes, wherein the amplitude measurements comprise the one or more direct signal vector lengths and the one or more multipath-impacted signal vector lengths.

FIG 16 refers to FIG 1 GA and FIG 1GB collectively and is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carrier signal with multipath compensation. The method of FIG. iSis similar to the method of FIG. 12, except step S360 is added. Like reference numbers in FIG. 12 and FIG. 16 indicate ike elements, steps or procedures.

In step S362, a data processor 104 or multipath compensator 96 determines the direct signal vector length of the direct path signal of a ranging subcarrier by averaging observed vector lengths over a period of time in which the observed vector lengths of different ranging subcarriers converge to the direct signal vector length within a certain tolerance or deviation.

In certain embodiments of the methods set forth in FIG 17 through FIG 20, rnclusi'ìe, the procedure for multipath phase mitigation can depend less upon accurate measurement of the received vector lengths and can require much less computation than the methods set forth in FIG 12 through FIG 16, inclusive In the methods set forth n FIG 17 through FIG 20, inclusive, the procedure for mitigating the effects of multipath described above may require the multipath vectors be averaged for an extended period of time (e.g., because of ow vehicte dynamics). When the observed vector length of the tone is consistent with a long term average vector length representative of the direct path vector length, the muitipath distortion is minimal and; hence, the associated reference phase slope of the subcarriers can be observed in the absence of multipath. The resultant reference phase slope measurement would serve as the target phase slope, which could differ from a zero target phase slope, for the second estimator in the second tracking loop.

FIG. 17 is a flow chart of another embodiment of determining the location or range of a receiver with a muiti-carrier sigral to minimize or reduce the impact of muftipath distortion or error.

In step S300, a receiver 200 portion receives or is adapted to receive a plurality of OFDM-like multitone signals. Each OFDM-iike multitone signal has subcarriers (tones) and is modulated with a pseudo-random noise code and transmitted from one or more transmitters 100 with corresponding known locations.

In step S302, in a flrst tracking loop, a first estimator 88 determines a course estimate of a time of arrival of a set (e.g., at least one subcarrier) of the OFDM-like multitone signals by adjusting the course estimate of the time of arrival of the set in accordance with a code (first) correlation function. For example, the phase error angle, and with angle and length remove its influence on the calculated phase to maximize the correlation of the received pseudo-random noise code of the at least one subcarner with the local replica of the pseudo-random noise code in accordance with a code (first) correlation function.

In step S302, the determining of the course estimate may be executed in accordance with various techniques, which may be applied separately or cumulatively. Under a first technique, in the first tracking loop, the determining of the course estimate further comprises minimizing an error assocated with an earls.-mrnus-late error function to determine the course estimate of the time of arrival. Under a second technique, the early minus late spacing is selected based on a ratio of the pseudo random noise code period and frequency spacing between subcarriers. Under a third technique, the early-minus-late error function has an early minus late spacing between approximately.0525 and.5 chips. Under a fourth technique, the code correlation of the first tracking loop occurs in the time comain Unde a fifth tecbniquo, in a frequency domain the tone signal detector 70 or amplitude detector 74 detects a tone amplitude associated with the at least one subcarrier for providing an adjustment signal to adjust a relative phase of the locaUy generated replica of the pseudo random code to the received code Under a sixth technique, the code correlation function comprises an autocorrelation function for the received signal and a locafly generated pseudo random noise code signal that is expressed in magnitude versus time! wherein the autocorrelation function has a centrai peak associated with a maxmum magnitude of a correlator output signal and a series of step-hke transitions on each side of the central peak.

In step 6304, in a second tracking loop, a second estimator 90 determines a precise estimate of the time of arrival of the set (e.g., the at east one subcarrier) of subcarriers of the OFDM-like multitone signals by adjusting the precise time of arrival to align the slope of the subearher phases of the OFDM-like multitone signals in accordance with a phase (second) correlation funct'on and consistent with the course estimate of the time of arrival associated with the code correlation function.

In step 6304, the determining of the precise estimate is carried out in accordance with one or more procedures, which may be carried out alternatively or cumulatively. Under a first pro:edure, in the second tracking loop, the data processor 104 or the receiver 200 the determines the precise estimate further comprises an error function with a target or goal of driving an avetage observed tone phase slooe to a zero phase slope, or another target phase slope (e.g. associated with the absence of multipath signals). Under a second procedure, the data processor 104 or the receiver 200 adjusts the precise time of arrival to align the slope of the subcarrier phases of the OFDM-like multitone signals in accordance with a phase (second) correlation function comprises adjusting a centra region of the phase sope, between adjacent ambiguity regons (e g with generally vertical slope) in the phase slope, to be aligned with a zero crossing point of the first early minus late function.

Under a third procedure, the receiver 200 or data processor 104 processes the received OFDM-like signal encoded without any cyclic prefix or cyclic suffix such that the phase of each tone produced by a transform module 68 (e.g., FFT transform module) in the receiver 200 is coheent from symbol to adjacent symbol Accordngly, under the third procedure, the target phase slope may approach approximately a zero phase slope.

Under a fourth procedure, the receiver 200 or data processor 104 synchronizes the processing time, over which a time domain-to-frequency domain transform (e.g., FF1 transform) is performed by a transform module 68 (e.g., FFT module) to a coherence nterval (eg., descr bed in conjunction with FIG 6) such that phase slope of the suocarriers will approach zero, or be reduced from any propensity toward a greater non-zero aggregate phase slope, for the target phase slope.

In step S306, a data processor 104 or range estimation module 84 estimates the location of the receiver 200 or estimated range between the receiver 200 and the transmitter based on the precise estimate and phase error compensation in a phase domain, a range domain, or both commensurate with driving the slope of the subcarrier phases toward zero (e.g..

to compensate for phase or frequency error in the absence rnultipath). For example, the data processor 104 or the range estimation module 84 estimates the location of the receiver 200 or estimated range between the receiver 200 and the transmitter 100 based on the precise estimate after any compensation for multipath and phase error compensation in a phase domain, a range domain, or both commensurate with driving the slope of the subcarrier phases toward zero.

In step S375, a data processor 104 or multipath compensator 96 determines phase compensation data for each ranging subcarrier comprising an adjustment to the estimated range based on difference between an observed phase of the observed signal vector and a direct path phase of a direct path vector, where the direct path phase is estimated based on one or more prior measurements of a certain observed signal vector when an average amplitude of all (or a majority of) rargng subcarriers converge to substantially the sane value The data processor 104, multipath compensator 96 or the range-domain, phase-slope compensator 92 can use compensation data for a first, representative or reference ranging subcarrier to estimate additional compensation data for the other ranging subcarriers based on one or more of the following: different frequency or wavelengths of the ranging subcarriers, an observed phase slope of the ranging subcarriers, and a target phase slope of the ranging subcarriers.

Step S3Th may be carried out in accordance with various techniques that may be used separately or cumulatively. In accordance with a first technique, the data processor 104, multipath detector 94, or tone signal detector /0 determines that the certain observed signal vector comprises a direct path signal vector for which no phase compensation is required for a respective time period when an average amplitude of all ranging subcarriers converge to a same value. In accordance with a second technique, the data processor 104, niuftipath compensator 96, or tone signal detector 70 determines phase compensation data during a multipath compensation mode, where the direct phase is estimated during an uncompensated mode and where the multipath compensation mode and the uncompensated mode or nonsynchronous or temporally mutually exclusive In accordance with a third tethnique, the data processor 104, multipath compensator 96, or tone signal detector 70 derive phase measurements from the certain observed sgnal vector where the derived phase measurements are generally free of muitipath contamination and require no multipath phase compensation for the phase measurements.

In accordance with a fourth technique, if the respective received subcarrier phases of muRiple subcarriers of the OFDM-Uke signal have short-term amplitude that thffers from a long-term ampitude during an evaluation time period by equal to or greater than a deviation threshold, the data processor 104 or multipath detector 94 determines that the dentifled rnultipath-inipacted signal is present and that the compensating for multipath phase error is or becomes in an active mode (multipath phase compensation mode) for at least the evaluation time period. In accordance with a fifth technique, if the respective received subcarrier phases of multiple subcarriers of the OFDM-like signal have a short-term amplitude that is within less than a deviation threshold of the long-term amphtude during an evaluation time period, the data processor 104 or the multipath detector 94 determines that the identified multipath-impacted signal is not present and that the compensating for phase error is or becomes in an inactive mode for the evaluation time period.

In accordarce with a sixth technique, the data processor 104 oror esses the received OFDM-like signal encoded without any cyclic prefix or cyclic suffix such that the phase of each tone produced by a transform module in the receiver is coherent from symbol to adjacent symbol.

In accordance with a seventh technique, the data processor 104 synchronizes the processing time over which a transform performed by a transform moaule, to a coherence interval, such that the phase slope of the subrarriers will approach zero or be reduced from any propensity toward a greater non-zero aggregate phase slope.

FIG. 18 refers to FIG. IBA and FIG. 18B coRectively and is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carrier signal to minimize or reduce the impact of muitipath distortion or error. The method of FIG 18 is similar to the method of FIG. 17 except the method of FIG. 18 further comprises steps S376, 8377, and S378 Like reference numbers in FIG 17 and FIG 18 indicate like elements procedures or steps.

In step S376, dunng a group of sampling intervals, the data processor 104 determines whether the amplitudes of the subcarriers or pilot tones during different sampling periods have a dispersion level exceeding a minimum threshold indicative of the existence ala multipath signal.

If the amplitudes of the subcarriers have dispersion level that exceeds the minimum threshold, 4' the method continues with step S378. However1 if the amplitudes do not have a dispersion level that exceeds the dispersion level, the method continues with step 5377 In step 5377, the receiver or data processor 104 waits an interval, return to 5300 or both.

In step 5378, a data processor 104cr multipath compensator 96 triggers multipath compensation by the range adjustment of the estimated range if the dispersion level exceeds the minimum dispersion threshold.

FIG 19 refers to FIG I 9A and FIG 195 collectively and is a flow chart of another embodiment of determining the location or range of a receiver with a multi-carrier signal to minimize or reduce the impact of multipath distortion or error. The method of FIG 19 is similar to the method of FIG. 17 except the method of FIG. 19 further comprises steps S379, S3TI, and 5380. Like reference numbers in FIG. 17 and FIG. 19 indicate like elements, procedures or steps.

In step 5379, during a group of sampling intervals, a data processor 104 determines whether the amplitudes of the subcarriers or pilot tones during different sampling periods deviates from a ong4erm average (e g mean) of the amolitude of the sbcarriers f the amplitudes of the subcarriers deviate from the long-term average, the method continues with step 5380 However, if the amplitudes do not have a dispersion level that ex'eeds the dispersion level, the method continues with step 5377 In step S377, the receiver or data processor 104 waits an interval, return to 5300 or both.

In step 5380, a data processor 104 or multipath compensator 96 triggers multipath compensation by the range adjustment if an amplitude of an observed signal vector differs from a long-term average amplitude of the ranging subcarriers by more than a deviation threshold.

FIG, 20 refers to FIG. 20A and FIG. 2DB collectively and is a flow chart of another embodiment of determining the location or range of a receiver with a muiticarner signal to minimize or reduce the impact of multipath distortion or error. The method of FIG. 20 is similar to the method of FIG. 17 except the method of FIG. 20 further comprises steps 5381, S382, 5383 and S384. Like reference numbers in FIG. 17 and FIG. 20 indicate like elements, procedures or steps.

In step 5381, a data processor 10401 multipath compensator 98 averages observed amplitude of each pilot tone or each subcarrier having multipath content over an extended time period to produce a reference mean ampUtude or long-term amplitude of the subcarriers.

In step 3382, the data processor 104 or rnultipath compensator 96 associates the reference mean amplitude with a reference direct path signal component.

Instep 5333, the data processor 104 or multipath compensator 96 measures an observed amplitude of the pilot tone over a sampling period less than the extended time In step 3384, the data processor 104 or muftipath compensator compares the observed amplitude to the reference mean amplitude to determine an estimated level of multipath in Vie observed phases of at least one subcarrier.

The system and method of this disclosure defines a navigation signal that has two major components that facilitate an extremely precise measure of time of arrival. The signal structure supports the ability to make extremely precise time-of-arrival measurements that are significantly immune from muitipath distortion. The receiver 200 uses a scheme of demodulation and range measurement that produces extremely precise range measurements.

Multipath interference can be easily detected and measurements can be produced that are virtually free of multipath contamination that varies with time. If multipath signal is highly time correlated, a metric of the multipath contamination or multipath figure of merit can be developed This metnc can then be used to estimate the quality of the range measurement Having described the preferred embodiment, it will become apparent that various modifications can be made without departing from the scope of the invenUon as defined in the accompanying claims.