WO2006042810A1

WO2006042810A1 - Positioning device

Info

Publication number: WO2006042810A1
Application number: PCT/EP2005/055147
Authority: WO
Inventors: Nicolas Martin
Original assignee: Thales
Priority date: 2004-10-22
Filing date: 2005-10-11
Publication date: 2006-04-27
Also published as: FR2877437A1; FR2877437B1

Abstract

The invention relates to a positioning device comprising a local clock (15) which has a drift and a first receiver (10) of a radio-navigation signal. The inventive device also comprises a second receiver (20) of a radio-electric signal originating from a fixed beacon (30) with a stable time reference. According to the invention, the local clock (15) is common to both the first and second receivers (10, 20). The invention further comprises an element for calculating the drift of the local clock from the radio-electric signal originating from the fixed beacon, said element being connected to the first receiver.

Description

POSITIONING DEVICE

The field of the invention is that of receiving a radionavigation signal from a satellite positioning system such as the GPS system (acronym for the English expression "Global Positioning System").

In the following, we denote by the generic term GNSS receiver any type of GPS receiver, GLONASS, GALILEO, etc.

To function well, current receivers usually require direct satellite reception. Positioning deteriorates rapidly, particularly in terms of accuracy and acquisition time, when the reception is disturbed as is the case inside a building or more generally in a degraded environment (obstacles, multipaths, interference , interior of a building, etc.).

During degraded reception conditions the recovery of the navigation signal with low signal-to-noise ratio is possible but requires a longer integration time to filter the noise.

The difficulty lies in the fact that to integrate the received signal over a long time and retrieve information it is necessary that the apparent dynamics at the receiver is very low during this time. When the receiver (or its antenna) is fixed there is more dynamic due to movement. By cons it remains the apparent dynamics due to the drift of the clock of the receiver.

We will clarify what this drift is by recalling the operating principle of a GNSS receiver. Shows Figure 1a a code transmitted by a satellite at t _em ission and 1b, this code arriving at the receiver to tréception or .DELTA.t _pr op = tréception - témission seconds later .DELTA.t _pr op being the propagation time of the signal between the satellite and the receiver. This propagation time multiplied by the speed of light in the vacuum gives a measure of distance; it is measured as follows. A replica of the code transmitted by the satellite and represented in FIG. 1c can be generated locally by the receiver, in phase. The difference between the received code (FIG. 1b) and the generated local code (that is to say the replica represented in FIG. 1c) corresponds to the desired time difference Δtprop. This shift represented in FIG. 1d is measured by setting phase the received code and the local code; the phasing criterion corresponds to the maximization of the correlation function of the two codes. If the satellite and receiver clocks are perfectly synchronized, the offset that maximizes the correlation function of the two codes provides the time Δtprop-

In practice this is not the case. FIG. 2 shows the time actually measured; is the true time at which the signal is emitted by the satellite, receiving the true time at which the signal reaches the receiver, t _h this time indicated by the clock and Δt _h through the receiver clock. Finally the measured propagation time is: .DELTA.t _pr op measured .DELTA.t = _pr op + .DELTA.t _h. In the ideal case of a perfect clock, Δt _h = 0.

Moreover, when the integration time of the correlation between the local signal and the received signal is long, the bias Δt _h of the receiver clock may vary. In fact, in the case of consumer applications, the low-cost clocks used in receivers have very high _h (t) drifts and are very unstable. It is recalled that the drift of the clock is by definition the derivative with respect to the time of the clock bias: dh (t) = d (Δt h (t)) / dt.

For signal reception inside a building, this clock drift can prevent useful signal recovery and positioning. One solution is to reduce the coherent integration time of the correlation but this increases the losses due to the noise and therefore also the time required to obtain the positioning; another solution possibly combined with the previous one is to test several hypotheses of clock drifts in parallel, but this increases the computational load. Moreover, when the drift is very unstable, the task becomes very difficult, if not impossible.

An important object of the invention is therefore to compensate for the apparent dynamic due to the drift of the receiver clock, so as to allow a sufficiently long correlation integration time to be used to distinguish the signal from the noise, despite degraded reception conditions.

To achieve this object, the invention proposes a positioning device, which comprises a local clock having a drift, and a first receiver of a radionavigation signal, mainly characterized in that it further comprises a second receiver of a radio signal from a fixed beacon which has a stable clock, the local clock being common to the first and the second second receiver, and in that it comprises connected to the first receiver, an element for calculating the drift of the local clock, from the radio signal from the fixed beacon.

This device thus makes it possible to compensate the clock drift of a GNSS receiver in order to integrate the received signal over a long time. We can then determine the position of the receiver in a difficult environment.

The second receiver is for example a mobile telecommunication device.

The radionavigation signal comes from a satellite positioning system and / or pseudolites.

The invention also relates to a method of positioning from a radionavigation signal and a local clock signal, characterized in that it comprises a step of receiving a signal from a fixed beacon which presents a stable clock, and calculating the drift of the local clock signal from the signal of the fixed beacon.

Other features and advantages of the invention will become apparent on reading the detailed description which follows, given by way of nonlimiting example and with reference to the appended drawings in which: FIG. 1 already described schematically illustrates the operating principle of a GNSS receiver, FIG. 2 already described, schematically illustrates the time actually measured, FIG. 3 schematically represents examples of curves of time t true and time t _h (t) indicated by the clock at time t true, FIG. 4 schematically illustrates examples of results of the correlation of the signal received with the local signal as a function of several time assumptions, τ being the delay of the local code represented in FIG. 1d with respect to the received code represented in FIG. 1 b, 5 schematically shows examples of change in the integration time, on the one hand the time t _received (t) which is transmitted by the satellite code C (t _re Cu (t)) received by the receiver to time t and for different assumptions true time t _re ceived estimated (t) estimated received in the case where it does not compensate for clock drift, Figure 6 schematically shows a GNSS receiver according to the prior art, 7 schematically shows examples of change in the integration time, on the one hand the time t _re Cu (t) received true and different assumptions time t _re ceived estimated (t) estimated received in the case where the the clock drift is compensated, FIG. 8 is a schematic representation of a positioning device according to the invention, FIG. 9 schematically represents a non-GNSS receiver for determining the local clock drift, and FIG. 'a say positive positioning according to the invention.

It will express in more detail the times indicated in Figures 1 and 2. a: _re ceived t (t) = t - .DELTA.t _prop (t), t _re Cu (t) and t is the real time.

The GNSS signal emitted by the satellite is of the form:

Semis (t) = Code (tem (t)). sin (φ _é put (t)) = Code (t). sin (ω. t) with temis (t) = t φemise (t) = ω. t

The GNSS signal received by the receiver is of the form: Sreçu (t) = Code (treçu (t)). sin (φ _get (t)) with set (t) = t - Δt _prop (t) φ get (t) = ω (t - Δt _prop (t))

Furthermore, since the clock has a bias Δt _h (t), we have: th (t) = t + Δth (t), th being the time indicated by the receiver clock at time t. We typically have drifts Δth (t) of 10 ^-5 seconds per second for low-end clocks, a variation of the clock bias of 10 microseconds over an interval of 1 second, that is to say 10 times the duration of a code chip, which is unacceptable .. Figure 3 shows the curves illustrating t and th (t).

But we do not know t therefore Δt _h (t). We actually have:

Testified (t) = t _h (t) - estimated Δt _h (t) (1)

Similarly, we do not know t _re ceived (t) or .DELTA.t _pr op (t). We have in fact: estimated cost (t) ⁼ t _e stimé (t) - Δtprop estimated (t) (2)

We have (1) and (2), estimated cost (t) = th (t) - Estimated Δth (t) - Estimated Δtprop (t) (3)

During signal acquisition by the receiver, and in particular during the correlation calculation of the received signal with the locally generated signal, the receiver can make a rough estimate of time: this estimate t _re ceived estimated (t) n is not precise enough to sufficiently match the local signal with the received signal.

For this, the receiver tries several correlations with different bias assumptions clock estimated .DELTA.t estimated _h (t) and thus time T _Rec estimated u (t) offset from it by an amount T such that at least one of the hypotheses is sufficiently close to the time t _re Cu (t) true for the signals are correlated, as shown in Figure 4. the result of the correlation depends on τ with t ⁼ treçu (t) - treçu estimated (t)

T = (t - evidenced (t)) "(Δtprop (t) - Estimated Δtprop (t))

The error due to the estimate of the propagation time is considered as zero because it is supposed that one knows the position of the satellite with a sufficient precision. We can write: τ ≡ Δth (t) - Δt estimated _h (t)

In Figure 5, the variations are shown on the integration time, on the one hand the time t _re Cu (t) and real different assumptions estimated cost (t) in the case where the clock drift is not compensated, that is to say where it is assumed that Δt _h (t) and therefore Δt _he stimulated (t) are constant .

The curves in Figure 5 represent t _re ceived estimated (t) differ from each other by taking as different hypotheses for .DELTA.t _h estimated (t), nT, (n + 1) T, (n + 2) T, etc. .

Because of the drift of the clock, no estimated received time hypothesis coincides with the time t _received (t) true over the integration time. As a result, the receiver is unable to determine which hypothesis coincides with the received signal, even by increasing the integration time. FIG. 6 shows an example of such a receiver 10. It comprises, in a conventional manner, a generator 11 of a local carrier of the form e ' ^ωt , a carrier demodulator 18, a correlator capable of multiplying the received signal demodulated by the local code coming from a local code generator 12 and integrating it by means of an interval integrator 13, a servocontrol device 14 for acquiring and tracking the phases of the received signals producing corrections in speed of the code loop and the carrier loop, a local clock generating a periodic pulsed signal of frequency F _h , an NCO (Numerically Controlled Oscillator) code 16 and a carrier NCO 17 integrating over an interval T _in tegration these speed adjustments and corrections that take into account the Doppler effect, to provide phase estimates respectively (or time) of code and carrier. This figure shows the elements for a reception channel corresponding to a satellite. There is usually in a receiver several reception channels in parallel.

It is recalled that an NCO code carries out the operation: estimated sum (t) = th (t) - Δth estimated (t) + I [th (to), th (t)] V _CO of (x) dτ with Vcode = - Radial velocity of the satellite / c + Correction of code in speed (or in frequency) c: speed of light.

The term "radial velocity of the satellite / c" corresponds to the taking into account of the Doppler effect. We therefore have (3):

J [th (to), th (t)] Estimated Vcode (τ) dτ = - Δtprop (t) According to a variant, the NCO performs the following operation: estimated cost (t) = th (t) - Δth estimated (t) + J [th (to), th (t)] V _CO of (x) dτ

We have: Δth estimated (t) = Δth estimated (to) + I [th (to), m)] dh estimated (τ) dτ by definition

And: th (t) = th (to) + I [th (to), th (t)] 1 dτ

We finally obtain: estimated cost (t) = th (to) - estimated Δth (tθ) + I [th (to), th (t)] (1 "estimated dh + V _CO of) (x) dτ

The implementation of this variant is shown in FIG. 10. A carrier NCO performs the same operation by further multiplying the result by the pulsation ω.

According to the invention, the GNSS receiver 10 is associated with a receiver of a non-GNSS communication signal (designated a non-GNSS receiver) which has a high signal-to-noise ratio and which is used to identify the drift of the clock. local 15 common to both receivers and compensate, as shown in Figure 8.

The non-GNSS signal is transmitted by a fixed beacon 30 having a clock 31 or more generally a stable time reference, typically with a lower drift in absolute value at 10 ^-10 seconds per second.

FIG. 9 shows an example of a non-GNSS receiver 20 for determining this clock drift. It comprises in particular a phase locked loop ("PLL") which conventionally comprises a carrier demodulator 22, a servo device 24 including a loop filter, an integrating NCO 27 connected to a clock 15 and a generator 21 of local carrier of the form e ' ^ωt . The output of the phase integrator NCO provides the local phase φ'iocai-

The recovery of the signal at the non-GNSS receiver makes it possible to estimate the drift of the local clock in the following manner. The non-GNSS signal emitted demodulated is of the form: Semis (t) = sin (φ'emis (t)) with φ'emiss (t) = ω '. t

Failure GNSS received signal: Sreçu (t) = sin (φ _'re Cu (t)) with φ' _re Cu (t) = ω '. (t - Δt ' _P ro _P (t)) (4)

Δt'prop (t) being the propagation time between the fixed beacon and the non-GNSS receiver, which is constant because the receiver and the beacon are supposed to be immobile.

The local non-GNSS signal:

Siocal I (t) = Sin (φ'iocale (t)) Siocal û (t) = COS (φ'bcale (t)) with φ'iocale (t) = φ'reçue estimated (t) = ω '. (T _'re çu estimated (t)) (5)

The non-GNSS receiver may for example implement a phase-locked loop ("PLL") to recover the non-GNSS communication signal.

When the PLL converges the received and local phases coincide, and we obtain: φ'iocai = φ 'received + nx 2π + ε ε' π / 2 (phase error) either under (4) and (5): estimated receipt (t) = t - Δt'prop + n X 2π / ⁽ θ '+ ε (t) / ω' (6)

In addition we have:

Δt _h (t) = t _h (t) - t under (6):

Δth (t) = t _h (t) - the estimated receipt (t) + Δt'prop - n X 2π / (θ '- ε (t) / (θ'

Neglecting the error ε phase (t) and when .delta.t _pr op (t) is constant because the non-GNSS receiver does not move, we finally obtain:

Δth (t) - Δth (t) ≡ (t _h (t) - th (to)) - (t _'re çu estimated (t) - t'reçu estimated (tθ))

Δth (t) - Δth (to) ≡ (t _h (t) - th (to)) - (φ'iocai (t) / ω '- φ'iocai (to) / ω').

We have :

Δth (t) - Δth (to) = I [to, t] d _h (τ) dτ by definition, th (t) - th (to) = I _[th (to), th (t)] 1 dτ by definition, φ'ι∞aie (t) / ω '- φ'iocaie (to) / ω' = Wo ), m)] (1 + Vp _L ι_ (τ)) dτ at the output of the NCO of the non-GNSS receiver.

We can write:

I [to, t] dh (τ) dτ ≡ J [th (to), th <t)] 1 dτ - Imto), th (t)] (1 + Vp _L ι_ (τ)) dτ

Where: d _h (t) ≡ - VPLLW

Thus one is able to determine the instantaneous clock drift d _h (t), from the speed carrier correction of the PLL which is sent to the carrier NCO of the PLL.

According to the invention, this local clock is common to the two receivers, that is to say the GNSS receiver and the non-GNSS receiver. The estimated clock drift in seconds / second is added to the speed corrections (from servocontrol and Doppler compensation) of the local GNSS signals for the correlation, namely the local code and the local carrier. In Figure 7, there is shown the results thus obtained that is to say the variations in the integration time on the one hand the time t _re Cu (t) and different assumptions true in the case where the we compensate the clock drift. In this case when the received time hypotheses are sufficiently close, there is one which coincides with the time received true, which allows an energy detection at the output of the correlation and therefore the estimation of a received time.

FIG. 10 shows an exemplary positioning device according to the invention. It comprises a local clock 15, a GNSS receiver 10, a receiver 20 of a non-GNSS radio signal from a fixed beacon having a stable time reference, the local clock being common to both receivers.

It also includes, connected to the GNSS receiver, an element for calculating the clock drift from the non-GNSS radio signal from the fixed beacon, the calculation being for example as described above. This element is connected to the NCOs (the NCOs 16 and 17 in FIG. 10) of the local GNSS signals, namely the local code and the carrier local, so as to add this instantaneous clock drift to the existing corrections. This calculation element is for example a microprocessor or a circuit including an NCO to perform the integration. In the example of FIG. 10, the drift equal to -Vp _L i (t) is calculated by the servocontrol device 24 of the PLL of the non-GNSS receiver.

The GNSS receiver 10 comprises, as in the case of FIG. 6, a local clock 15, a generator 11 of a local carrier of the form e ' ^ωt , a carrier demodulator 18, a correlator able to multiply the signal received by the local code from a local code generator 12 and integrating it by means of an interval integrator 13, a servocontrol device 14 for acquiring and tracking the phases of the received signals producing corrections in speed of the code loop and the carrier loop, and an NCO code 16 and a carrier NCO 17 incorporating these speed corrections and corrections to take into account the Doppler effect and the clock drift, to provide estimates of phase (or time) of code and carrier.

The non-GNSS receiver 20 includes, as in the case of FIG. 9, a PLL which conventionally comprises a carrier demodulator 22, a servocontrol device 24 including a loop filter, an integrating NCO 27 and a generator 21. local carrier of the form e ' ^{ω t} . The output of the NCO phase integrator 27 provides the measurement of the received signal phase φ '= _re ceived estimated φ'i∞aie-

This calculation element of the clock drift (the device 24 in the figure) is connected to the corrector (the NCO 27 in the figure) of the carrier of the non-GNSS receiver, in the case where it is interesting to correct the local carrier. with this drift, as shown in the figures. It can be integrated in the non-GNSS receiver as in the example of Figure 10 or in the GNSS receiver. The non-GNSS receiver is for example a mobile telecommunication device possibly comprising a transmitter.

The so-called GNSS signal can come from a satellite positioning system and / or pseudolites.

Claims

Positioning device, which comprises a local clock (15) having a drift, and a first receiver (10) of a radionavigation signal, characterized in that it further comprises a second receiver (20) of a radio signal from a stationary beacon (31) having a stable time reference, the local clock (15) being common to the first and second receivers (10, 20), and including the first one being connected to the first one receiver (10), an element for calculating the drift of the local clock, from the radio signal from the fixed beacon (31).

2. Positioning device according to the preceding claim, characterized in that the calculating element of the local clock drift is integrated with the second receiver.

Positioning device according to the preceding claim, characterized in that the second receiver (20) comprises a phase-locked loop provided with a servo-control device (24) and in that the servo-control device (24) is able to calculate the local clock drift.

4. Positioning device according to one of the preceding claims, characterized in that the second receiver (20) is a mobile telecommunication device.

5. Positioning device according to one of the preceding claims, characterized in that the radionavigation signal comes from a satellite positioning system and / or pseudolites

6. Positioning method from a radionavigation signal and a local clock signal, characterized in that it comprises a step of receiving a signal from a fixed beacon which has a time reference. stable, and calculating the drift of the local clock signal from the signal of the fixed beacon.