RADIO FREQUENCY DISTANCE DETERMINATION

Abstract

A method of determining a distance between a radio frequency device and a target is disclosed in which the radio frequency device receives a radio frequency signal from the target. The method comprises determining a time domain channel response from the received radio frequency signal, determining an amplitude of a largest peak in the time domain channel response, determining an amplitude of a second, earlier, peak in the time domain channel response, comparing the second peak amplitude to a threshold based on the largest peak amplitude, identifying the largest peak as a shortest path peak if the second peak amplitude is less than the threshold, identifying the second peak as a shortest path peak if the second peak amplitude is greater than the threshold, and calculating the distance between the radio frequency device and the target based on a time corresponding to the shortest path peak.

Claims

1. A method of determining a distance between a radio frequency device and a target, the method comprising: the radio frequency device receiving a radio frequency signal from the target; determining a time domain channel response from the received radio frequency signal; determining an amplitude of a largest peak in the time domain channel response; determining an amplitude of a second, earlier, peak in the time domain channel response; comparing the second peak amplitude to a threshold based on the largest peak amplitude; identifying the largest peak as a shortest path peak if the second peak amplitude is less than the threshold; identifying the second peak as a shortest path peak if the second peak amplitude is greater than the threshold; and calculating the distance between the radio frequency device and the target based on a time corresponding to the shortest path peak.

2. The method of claim 1, wherein the radio frequency signal received from the target comprises a second radio frequency signal based on a first radio frequency signal transmitted previously to the target by the radio frequency device.

3. The method of claim 1, wherein the radio frequency signal comprises a plurality of frequencies.

4. The method of claim 3, wherein the radio frequency signal comprises a sequence of radio frequency signals having different carrier frequencies.

5. The method of claim 3, wherein the radio frequency signal comprises a bandwidth of at least 10 MHz.

6. The method of claim 1, wherein the target comprises a second radio frequency device that generates and transmits the radio frequency signal.

7. The method of claim 1, wherein the time domain channel response is determined using a frequency domain channel response of the received radio frequency signal.

8. The method of claim 7, wherein the frequency domain channel response comprises the output of a Multi-Carrier Phase Distancing process applied to the radio frequency signal.

9. The method of claim 1, wherein the threshold corresponds to an amplitude larger than a side-lobe amplitude of the largest peak.

10. The method of claim 1, comprising calculating the distance between the radio frequency device and the target by determining the distance travelled by an RF signal in a time corresponding to the shortest path peak.

11. The method of claim 1, further comprising: identifying a closest local minimum in the time domain channel response earlier than the shortest path peak; determining a time separation between the closest local minimum and the shortest path peak; comparing the time separation to an expected time separation based on a bandwidth of the received radio frequency signal; and if the time separation is greater than the expected time separation, calculating said time corresponding to the shortest path peak as a time corresponding to the closest local minimum plus the expected time separation.

12. The method of claim 11, wherein the expected time separation comprises a Rayleigh criterion for the time domain channel response.

13. A radio frequency transceiver device arranged to: receive a radio frequency signal from a target; determine a time domain channel response from the radio frequency signal; determine an amplitude of a largest peak in the time domain channel response; determine an amplitude of a second, earlier, peak in the time domain channel response; compare the second peak amplitude to a threshold based on the largest peak amplitude; identify the largest peak as a shortest path peak if the second peak amplitude is less than the threshold; identify the second peak as a shortest path peak if the second peak amplitude is greater than the threshold; and calculate the distance between the radio frequency device and the target based on a time corresponding to the shortest path peak.

14. A method of determining a distance between a radio frequency device and a target, the method comprising: the radio frequency device transmitting a first radio frequency signal; the radio frequency device receiving a second radio frequency signal from the target based on the first radio frequency signal; determining a time domain channel response from the second radio frequency signal; determining an amplitude of a largest peak in the time domain channel response; determining an amplitude of a second, earlier, peak in the time domain channel response; comparing the second peak amplitude to a threshold based on the largest peak amplitude; identifying the largest peak as a shortest path peak if the second peak amplitude is less than the threshold; identifying the second peak as a shortest path peak if the second peak amplitude is greater than the threshold; and calculating the distance between the radio frequency device and the target based on a time corresponding to the shortest path peak.

15. A radio frequency transceiver device arranged to: transmit a first radio signal; receive a second radio frequency signal from a target based on the first radio frequency signal; determine a time domain channel response from the second radio frequency signal; determine an amplitude of a largest peak in the time domain channel response; determine an amplitude of a second, earlier, peak in the time domain channel response; compare the second peak amplitude to a threshold based on the largest peak amplitude; identify the largest peak as a shortest path peak if the second peak amplitude is less than the threshold; identify the second peak as a shortest path peak if the second peak amplitude is greater than the threshold; and calculate the distance between the radio frequency device and the target based on a time corresponding to the shortest path peak.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0070] One or more non-limiting examples will now be described, by way of example only, and with reference to the accompanying figures in which:

[0071] FIG. 1 is a schematic diagram of an arrangement for use in embodiments of the present invention;

[0072] FIG. 2 is a schematic diagram of another arrangement for use in embodiments of the present invention;

[0073] FIG. 3 is a flow diagram illustrating a method according to an embodiment of the present invention;

[0074] FIG. 4 shows a simulated time domain channel response for signals sent between an RF device and a target without multipath effects;

[0075] FIG. 5 illustrates a time domain channel response for signals sent between an RF device and a target with multipath effects;

[0076] FIGS. 6-8 illustrate time domain channel responses containing two incoming paths of equal strength with a various different separations;

[0077] FIG. 9 is a flow diagram illustrating steps of a method according to an embodiment of the present invention;

[0078] FIG. 10 is a magnified view of the time domain channel response shown in

[0079] FIG. 11 is a simulated time domain channel response used in another embodiment of the invention;

[0080] FIGS. 12 and 13 are simulated results illustrating the improved distance estimation performance of embodiments of the invention; and

[0081] FIG. 14 illustrates phase differences for an example of two-way ranging.

DETAILED DESCRIPTION OF EMBODIMENTS

[0082] FIG. 1 shows a radio frequency (RF) device 2 comprising a transmitter 4, a receiver 8, and a processor 14. The RF device 2 is located a distance r from a passive reflector 18. In use, the transmitter 4 transmits a sequence of radio frequency signals 16 with different carrier frequencies (e.g. following a predetermined frequency-hopping pattern). The radio frequency signals 16 are reflected by the reflector 18 and received by the receiver 8. The received radio frequency signals 16 may be used to estimate the distance r between the RF device 2 and the reflector 1, as described in more detail below with reference to FIGS. 3-8.

[0083] FIG. 2 shows another arrangement featuring a first RF device 102 and a second RF device 122, separated by a distance, r. The first RF device 102 comprises a transmitter 104, a receiver 108 and a processor 112. The second RF device 122 also comprises a transmitter 124, a receiver 128 and a processor 132.

[0084] In use, the transmitter 104 of the first RF device 102 transmits a sequence of radio frequency signals 146 with different carrier frequencies (e.g. following a predetermined frequency-hopping pattern). The radio frequency signals 146 are received by the receiver 128 of the second RF device 122, and then re-transmitted by the transmitter 124 (i.e. acting as an active reflector). For instance, the second RF device 122 may comprise a local oscillator that is locked to the frequency of the received radio frequency signals 146 (thus being locked to the same frequency as a local oscillator of the first RF device 102). The re-transmitted (i.e. reflected) signals 146 are then received by the receiver 108 of the first RF device 102.

[0085] As with the arrangement shown in FIG. 1, the received radio frequency signals 146 may be used to estimate the distance r between the first and second RF devices 102, 122, as described in more detail below with reference to FIGS. 3-8.

[0086] A method of determining the distance, r, between an RF device (e.g. the RF device 2 in FIG. 1, or the first RF device 102 in FIG. 2) and a target (e.g. the passive reflector 18 in FIG. 1, or the second RF device 122 acting as an active reflector in FIG. 2) will now be described with reference to FIGS. 3-5. A further improvement on this method is then described with reference to FIGS. 6-10.

[0087] In this example, in step 150, the RF device transmits a sequence of RF signals (i.e. signals 16, 146 in FIGS. 1 and 2) in different frequency channels spaced at 2 MHz intervals between 2404-2478 MHz. This means the RF signals have a total bandwidth BW of 74 MHz.

[0088] These signals are returned by the target and received back at the RF device in step 152. The received signals have a frequency-dependent amplitude and phase that depends (amongst other things) upon the distance, r, between the RF device and the target.

[0089] The received signals are analysed (e.g. by the processors 14, 112) to produce a frequency domain channel response (FDCR), H(jω), which is a representation of the relative amplitude and phase of the received signals across the different frequency channels. The FDCR may be produced by Fourier-transforming a time-series of the received signals (e.g. using a Fast Fourier Transform (FFT) technique). The FDCR may be produced by performing a Multi-Carrier Phase Distancing (MCPD) procedure on the received signals. This may enable both one-way and two-way ranging.

Two-Way Ranging

[0090] In one example of two-way ranging (2WR), the local oscillators (LOs) of two RF devices, an initiator and a reflector, are locked. The LOs are not assumed to be phase-locked so there exists a time offset ΔT which results in a channel dependent phase offset θ.sub.LO=ω.sub.LOΔT between the two LOs. We can describe the initiator's LO instantaneous carrier phase as ω.sub.LOt and the reflector's as ω.sub.LO(t+Δ.sub.T).

[0091] The initiator first sends a tone ω.sub.LOt, and the receiver gets ω.sub.LOt−ω.sub.LOr/c, which it then down mixes to baseband and measures the phase (Step 1). In Step 2, the reflector then sends ω.sub.LO(t+Δ.sub.T) and the initiator measures the received baseband phase. Note that the LOs are kept on during Step 1 and Step 2, but each side changes roles in between. The devices can then exchange the measured phase values (ψ.sub.I, ψ.sub.R) by a data channel to perform a distance measurement.

TABLE-US-00001 TABLE 1 2WR phase measurement on a single tone Step 1 Initiator TX RF Reflector RX RF Reflector RX BB ω.sub.LOt ω.sub.LOt − ω.sub.LOr/c ψ.sub.R = −ω.sub.LOΔ.sub.T − ω.sub.LOr/c Step 2 Reflector TX RF Initiator RX RF Initiator RX BB ω.sub.LO(t + Δ.sub.T) ω.sub.LO(t + Δ.sub.T) − ω.sub.LOr/c ψ.sub.I = ω.sub.LOΔ.sub.T − ω.sub.LOr/c Step 3 Reflector sends Initiator calcs ψ.sub.R ψ = ψ.sub.R + ψ.sub.I −ω.sub.LOΔ.sub.T − 2πr/c ψ = −2ω.sub.LOr/c

[0092] This is shown in Table 1 and also FIG. 14. We can see that if we take the sum of the received phases, we get something which is a function of the distance, and the time offset between the devices is cancelled out. Conversely, if we take the difference of the received phases, we get something which is a function of the time offset between the devices.

[0093] We see that the initiator and receiver's phase measurement (ψ.sub.I, ψ.sub.R) are equally offset by the distance, but are offset in opposite directions because of the LO phase offset between devices

[00013]${\psi }_I={\omega }_{LO}{\Delta }_T-{\omega }_{LO}r/c{\psi }_R=-{\omega }_{LO}{\Delta }_T-{\omega }_{LO}r/c$

[0094] So we have

[00014]${\varphi }_{\mathrm{dist}}={\psi }_I+{\psi }_R=-2{\omega }_{LO}r/c$${\varphi }_{{\Delta }_T}={\psi }_I-{\psi }_R=-2{\omega }_{LO}{\Delta }_T$

[0095] From which we can measure the distance and the time offset. In reality ϕ.sub.dist will be disturbed by an additional phase offset due to each device's internal delays. As long as these are constant during a sweep over the band they have no effect.

[0096] To generate the FDCR, i.e. the channel response {tilde over (H)}(f.sub.k), we assume the I and Q values of the reflector and initiator are given by

[00015]$PC{T}_{REFL}\left(f_k\right)=A_{REFL}\left(f_k\right)e^{i{\theta }_{\mathrm{REFL}}\left(f_k\right)}\mathrm{and}{\mathrm{PCT}}_{\mathrm{INIT}}\left(f_k\right)=A_{\mathrm{INIT}}\left(f_k\right)e^{i{\theta }_{\mathrm{INIT}}\left(f_k\right)}$

[0097] Where f.sub.k is the kth channel.

[0098] Assuming the physical channel is symmetrical between initiator and reflector then the measured phases are dependent on both the physical communication channel as well as the relative difference in phase of the RF carrier between the devices i.e. that θ.sub.REFL(f.sub.k)=θ.sub.CH(f.sub.k)+Δθ.sub.LO(f.sub.k) and θ.sub.INIT(f.sub.k)=θ.sub.CH(f.sub.k)−Δθ.sub.LO(f.sub.k) where

[0099] θ.sub.CH(f.sub.k) is the phase delay of the channel, and Δθ.sub.LO(f.sub.k) is the relative difference in phase of the RF carrier between the devices

[0100] An estimate of the square of the actual channel response {tilde over (H)}(f.sub.k)is made e.g.

[00016]${\tilde{H}}^2\left(f_k\right)=A_{\mathrm{REFL}}\left(f_k\right)e^{i{\theta }_{\mathrm{REFL}}\left(f\right)}\times A_{\mathrm{INIT}}\left(f\right)e^{i{\theta }_{\mathrm{INIT}}\left(f\right)}=A_{\mathrm{CH}}^2\left(f\right)e^{i2{\theta }_{\mathrm{CH}}\left(f\right)}$

[0101] From {tilde over (H)}.sup.2(f.sub.k), can construct an estimate of the actual channel transfer function H(f.sub.k). Unfortunately both {tilde over (H)}(f.sub.k)=√{square root over (H.sup.2(f.sub.k))} and −√{square root over (H.sup.2(f.sub.k))} are valid solutions. To determine the correct solution one may for example assume to minimize the Euclidean distance between successive values in {tilde over (H)}(f.sub.k), or use other means based on assumption of phase coherence of the PLL between different frequencies.

One-Way Ranging (1WR)

[0102] In 1WR all events happen on a timing grid, such that Δ.sub.T is kept constant over the full channel sweep.

TABLE-US-00002 TABLE 2 1WR in multipath Step n Initiator TX RF Reflector RX RF Reflector RX BB e.sup.jω.sup.LO.sup.t H(jω.sub.LO)e.sup.j(ω.sup.LO.sup.t) γ.sub.R = H(jω.sub.LO)e.sup.−j(ω.sup.LO.sup.Δ.sup.T)

[0103] So we need to estimate Δ.sub.T. This can be done using a set of 2WR measurements as explained above.

[00017]$y_R=H\left(j{\omega }_{\mathrm{LO}}\right)e^{-j{\omega }_{\mathrm{LO}}{\Delta }_T}$$y_1=H\left(j{\omega }_{\mathrm{LO}}\right)e^{j{\omega }_{\mathrm{LO}}{\Delta }_T}$

[0104] And therefore

[00018]$\angle \left(y_R^{*}{y}_I\right)=2{\omega }_{\mathrm{LO}}{\Delta }_T$

[0105] So we can measure Δ.sub.T by looking at the slope of ∠(y.sub.R*y.sub.I) over frequency. This slope is still a straight line in the presence of multipath. One way ranging may require highly accurate frequency correction, with the LO and CLK frequencies are tied together.

[0106] This fine frequency measurement can be done by measuring on the same frequency on many tones spread over a large amount of time

[0107] Returning now to the two-way method illustrated in FIGS. 3-5, once produced, the FDCR is then inverse-Fourier transformed using a 2048-point Inverse Fast Fourier Transform (IFFT) technique to produce a time domain channel response (TDCR), |h(t))| in step 154.

[0108] FIG. 4 shows the TDCR 202 for a simulated situation in which the RF device and the target are separated by 25 m and where there are no multipath effects. The largest peak 204 in the TDCR is found at time index n.sub.peak=342. Because there are no multi-path effects, the distance between the RF device and the target can therefore be calculated using the time index of the largest peak 204 according to:

[00019]$d_{est}=\frac{n_{\mathrm{peak}}}{N_{\mathrm{IFFT}}}.Math.\frac{c}{\Delta f}=25.05m,$

which is only 0.05 m away from the true distance (this error is only due to the resolution of the IFFT).

[0109] However, in real-world situations multi-path effects can cause frequency-dependent increases and decreases in received signal strength. In some such scenarios the largest peak in the TDCR may thus not actually represent the shortest path between the two radio nodes (i.e. the largest peak may not be the “shortest path peak”).

[0110] A more realistic TDCR 302 is shown in FIG. 5, for a simulated situation in which the RF device and the target are separated by 11.5 m and there are realistic multi-path effects. In step 156, the amplitude of the largest peak 304 is determined to be approximately 0.014 and at a time index of n.sub.largestpeak˜400 which corresponds to a distance of approximately 29 m, i.e. 17.5 m away from the true distance.

[0111] Thus, simply using the largest peak 304 may not be particularly accurate in a situation where multi-path effects are significant. To improve the accuracy of the distance estimation, in step 158 the next highest peak 306 which has an earlier time index in the TDCR is selected and analysed to determine if it corresponds to a shorter path between the RF device and the target.

[0112] Because the TDCR 302 is derived from a bandlimited FDCR (i.e. limited by the bandwidth of the transmitted radio frequency signals), peaks in the TDCR 302 corresponding to paths between the RF device and the target appear as sinc functions with a first sidelobe peak having a amplitude of approximately −13.3 dB compared to the main peak (i.e. approximately 4.7%. of the amplitude of the main peak). Thus, the first side lobe peaks in the TDCR 302 for the largest peak 304 would be expected to have a amplitude of approximately 0.0007.

[0113] Therefore, to determine whether the earlier peak 306 in the TDCR 302 actually corresponds to a shorter path between the RF device and the target than that corresponding to the largest peak 304, or if it is just a side lobe of the largest peak 304, in step 160 a threshold test is used where the earlier peak 306 is compared to a threshold 308 based on the amplitude of the largest peak 304. The threshold 308 has a amplitude of −10.0 dB compared to the largest peak 304 (i.e. 10% or approximately 0.0014), i.e. greater than the expected amplitude of the first side lobe of −13.3 dB (two times greater). If the amplitude of the earlier peak 306 is lower than the threshold, the largest peak is identified as the shortest path peak in step 162. However, as can be seen in FIG. 5, the amplitude of the earlier peak 306 is greater than the threshold 308 and it is thus identified as the shortest path peak in step 164.

[0114] Once the shortest path peak 306 has been identified, it can be used to determine the distance, r, between the RF device and the target with greater accuracy than simply using the largest peak 304. The distance between the RF device and the target can be determined in step 166 by simply using the time index of the shortest path peak in conjunction with equation (4). For the example shown in the FIG. 5, the shortest path peak 306 lies at a time index n.sub.sp of approximately 170, which corresponds to a distance of approximately 12.5 m, i.e. an error of only 1 m.

[0115] However, the accuracy of the distance estimation may still be improved. Because the TDCR 302 is bandlimited, individual peaks corresponding to similar length paths may overlap, such that a single peak in the TDCR 302 may not actually correspond to a single path. In this embodiment, where the bandwidth, BW, of the transmitted signals is 74 MHz, the Rayleigh criterion (which denotes the minimum time separation of two paths that can both be accurately individually resolved in the TDCR) is equal to

[00020]$\frac{1}{BW}=13.5\mathrm{ns}.,$

corresponaing to a aistance of approximately 4.05 m.

[0116] The uncertainty introduced when two paths are closely separated is illustrated in FIGS. 6-8. FIG. 6 shows a TDCR 600 (in which the x-axis has been converted to distance for clarity) for a situation in which two paths 602, 604 of equal strength are separated by 2 m, i.e. roughly half the Rayleigh distance of 4.05 m (all the other parameters of the TDCR are the same as that described above). Only one peak 606, located at 3 m, is visible in the TDCR. FIG. 7 shows a TDCR 700 for a situation in which two paths 702, 704 are separated by 3 m, i.e. roughly 0.75 of the Rayleigh distance. Two separate peaks are visible in this TDCR, but they are still located roughly 1 m away from the true path distances. FIG. 8 shows a TDCR 800 for a situation in which two paths 802, 804 are separated by 4 m, i.e. roughly equal to the Rayleigh distance. In this example the peaks 802, 804 are still roughly 70 cm away from the true path distances.

[0117] Therefore, in order to further improve the accuracy of the distance determination, the shortest path peak 306 is further analysed to ascertain whether it is the actually the produce of two different paths. This process will now be explained with reference to the flowchart of FIG. 9 and with reference to FIG. 10, which is a magnified view of the TDCR shown in FIG. 5

[0118] In step 350 a closest local minimum 310 in the TDCR of FIG. 10 earlier than the shortest path peak 306 (i.e. at a lower value of n) is identified. In step 352 a time separation Δn between the shortest path peak 306 and the closest local minimum 310 is then determined. In step 354 this time separation is compared to the separation Δn.sub.BW which would be expected to be observed if the closest local minimum 310 was simply the first null of the shortest path peak 306 (i.e. the first null of the sinc function that describes peaks in the TDCR).

[0119] In this example the expected separation Δn.sub.BW (calculated using equation 5) is 55 time index units. If the time separation is less than the expected separation, the time of the shortest path peak 306 would be used to calculate the distance to between RF device and the target in step 356. However, in this is example, the local minimum 310 is located at n=100, such that the time separation Δn measured between the shortest path peak 306 and the closest local minimum 310 is Δn=70. The time separation Δn is greater than the expected time separation Δn.sub.BW which suggests that the shortest path peak 306 is actually made up of two overlapping peaks. To determine the time index of the earlier of these two peaks (i.e. corresponding to the shortest path), the expected time separation Δn.sub.BW is simply added on to the time index of the local minimum 310 (as this is where the corresponding peak to that minimum should lie) in step 358 to produce a corrected time index n.sub.corrected=155, which is used to calculate a distance (using equation 7) of approximately 11.4 m, which is within 0.1 m of the actual distance.

[0120] FIG. 11 illustrates another simulated TDCR 900 in which the above described methods are implemented, produced from a situation where an RF device and a target are separated by 8.828 m. The TDCR 900 comprises a largest peak 902 located at an index corresponding to approximately 13.184 m (i.e. an error of 4.356 m). The amplitude of an earlier peak 904 is compared to a threshold based on the largest peak 902, but falls below this threshold, so the largest peak 902 is identified as the shortest path peak.

[0121] The closest local minimum 906 earlier than the shortest path peak 902 is then identified, and a time separation between the local minimum 906 and the shortest path peak 902 is compared to an expected time separation based on the bandwidth of signals transmitted between the RF device and the target. The time separation exceeds the expected time separation, and so a corrected time is determined by adding the expected time separation to the time of the local minimum. This corrected time is used to produce a final distance estimate of 8.496 m (i.e. an error of only 0.332 m).

[0122] FIG. 12 shows simulation results in terms of a cumulative distribution function (CDF) of absolute distance estimation errors using a prior art approach and approaches according to embodiments of the invention. The simulation was performed using generated MCPD data from a Matlab model.

[0123] The lower line 1002 shows distance estimation error using a prior art phase slope (MCPD) approach. The middle line 1004 shows distance estimation error when a TDCR-based approach using the shortest path peak threshold test (i.e. as described above with reference to FIGS. 3-5) is used. The upper line 1006 shows distance estimation error when further local minimum analysis is used (i.e. as described above with reference to FIGS. 6-10). It can be seen that embodiments of the invention produce more accurate distances in the vast majority of simulations, with only roughly 20% of simulations resulting in an error above 1 m compared to roughly 65% for the prior art approach.

[0124] FIG. 13 shows the same simulation results as FIG. 12, but for only those simulations in which the local minimum analysis 1006 resulted in a correction to the time used for distance estimation (i.e. where the closest local minimum in the TDCR was greater than the expected time separation). In these cases, roughly 85% of simulations using the further local minimum analysis 1006 improvement result in an error of less than 2 m, compared to roughly 50% for simulations using just the shortest path peak threshold test 1004, and just 25% for the prior art MCPC approach 1002.

[0125] While the invention has been described in detail in connection with only a limited number of embodiments, it should be readily understood that the invention is not limited to such disclosed embodiments. Rather, the invention can be modified to incorporate any number of variations, alterations, substitutions or equivalent arrangements not heretofore described, but which are commensurate with the scope of the invention. Additionally, while various embodiments of the invention have been described, it is to be understood that aspects of the invention may include only some of the described embodiments. Accordingly, the invention is not to be seen as limited by the foregoing description, but is only limited by the scope of the appended claims.