UWB RADAR MEASUREMENT EVALUATION METHOD AND ARRANGEMENT

Abstract

It is described a method of evaluating radar measurement data, the method comprising: forming plural subsequent channel impulse response profiles (14) from radar signals (6) across subsets of plural subsequent sample time intervals (15), the radar signals (6) being received due to reflection and/or interaction of plural UWB radar radiation pulses (3) transmitted towards a target (4); identifying, in each of the plural channel impulse response profiles (14), one of the plural sample time intervals (15) as a target sample time interval (16) in which received radar signals (6) are comprised originating due to reflection from the target (4); evaluating plural total radar signals of interest (17), each defined as respective magnitude and phase of the plural channel impulse response profiles (14) at the target sample time interval (16), in order to determine for at least one total radar signal of interest a phase (?) of a target contribution (21) to the total radar signal of interest (17); determining location and/or movement information of the target (4) based at least on the at least one phase (?) of the target contribution (21).

Claims

1. A method of evaluating radar measurement data, the method comprising: forming plural subsequent channel impulse response profiles from radar signals across subsets of plural subsequent sample time intervals, the radar signals being received due to reflection and/or interaction of plural UWB radar radiation pulses transmitted towards a target; identifying, in each of the plural channel impulse response profiles, one of the plural sample time intervals as a target sample time interval in which received radar signals are comprised originating due to reflection from the target; evaluating plural total radar signals of interest, each defined as respective magnitude and phase of the plural channel impulse response profiles at the target sample time interval, in order to determine for at least one total radar signal of interest a phase of a target contribution to the total radar signal of interest; determining location and/or movement information of the target based at least on the at least one phase of the target contribution.

2. The method according to claim 1, wherein evaluating the total radar signals of interest comprises: determining, for each of the total radar signals of interest, the associated phase of the target contribution to the respective total radar signal of interest, wherein determining the location and/or movement information of the target is based on the phases of the target contribution to all of the respective total radar signals of interest, wherein each channel impulse response profile in particular defines magnitude and phase of the received radar signals for some of the plural sample time intervals.

3. The method according to claim 1, wherein determining the location and/or movement information of the target comprises: using the target sample time interval to derive a target location interval and using at least one phase of the at least one target contribution to estimate movement and/or position of the target within the target location interval.

4. The method according to claim 1, wherein each of the total radar signals of interest comprises: a crosstalk contribution, being a contribution due to back coupling and/or crosstalk between a transmitter, transmitting the pulses, and a receiver, receiving the radar signals; the target contribution, being a contribution due to interaction and/or reflection of the transmitted pulses at the target.

5. The method according to claim 1, wherein evaluating total radar signals of interest comprises: determining the crosstalk contribution by processing the plural total radar signals; subtracting the crosstalk contribution from each of the total radar signals of interest, in order to derive the target contributions, each comprising at least a target phase and in particular a target magnitude.

6. The method according to claim 1, wherein the crosstalk contribution is determined in an approximate manner by averaging the plural total radar signals of interest over an averaging time range, wherein the approximate crosstalk contribution is subtracted from the total radar signals of interest within the averaging time range, in order to derive approximate target contributions within the averaging time range, each comprising at least an approximate target phase and in particular an approximate target magnitude.

7. The method according to claim 1, wherein determining the crosstalk contribution by processing the plural total radar signals comprises: forming cumulative phases of the approximate target phases over different cumulative phase time ranges; determine a cumulative phase time range of interest, where the cumulative phase of the approximate target phases satisfies a criterium, in particular being substantially or at least 2 Pi or 4 Pi.

8. The method according to claim 1, comprising, if the cumulative phase of the approximate target phases does not satisfy the criterium: acquiring further radar measurement data and determining further total radar signals; determining further cumulative phase time ranges, until a further cumulative phase time range of interest satisfies the criterium.

9. The method according to claim 7, wherein determining the crosstalk contribution further comprises: fitting the plural total radar signals of interest across the cumulative phase time interval of interest onto a predetermined curve shape, in particular arc or circle or spiral, the curve shape being associated with a curve center, in particular center of a circle or spiral; setting the crosstalk contribution to the curve center.

10. The method according to claim 1, wherein if the target phase indicates that target has moved such that its radar signals are received in another target time interval and/or another target sample time interval is identified in the identifying step, the method comprises: evaluating other total radar signals of interest, associated with the other target time interval; determining another crosstalk contribution in the other total radar signals of interest by approximation and/or curve shape fitting; deriving other target phases; deriving another target location/movement based on the other target time interval and/or the other target phases.

11. The method according to claim 1, wherein identifying the target sample time interval comprises: forming at least one difference between channel impulse response profiles, in particular relative to a reference impulse response profile, and/or forming an average over plural channel impulse response profiles; performing peak finding in the at least one difference and/or in the average.

12. The method according to claim 1, wherein a channel impulse response profile is formed as an average of received signals in accordance with a code according to which the pulses are transmitted; wherein forming plural subsequent channel impulse response profiles from the received radar signals comprises at least one of: removing a carrier radiation contribution; demodulating the received radar signals, in order to extract in-phase (I) components and quadrature components.

13. The method according to claim 1, wherein at least one of the following holds: the target is moving in one direction or is moving back and forth; the radar signals are received at a rate between 0.5 and 2 GHz, in particular at substantially 1 GHz corresponding to a sample interval width of 30 cm; a pulse width of the transmitted radar pulses is between 0.5 ns and 2 ns, in particular substantially 1 ns; the method adhering to IEEE 802.15.4, at least the version corresponding to the priority date of this application.

14. A method of performing a radar measurement, the method comprising: transmitting plural UWB radar radiation pulses towards a target; receiving, at plural subsequent sample time intervals, radar signals due to reflection and/or interaction of the transmitted pulses at the target; performing a method of evaluating radar measurement data according to claim 1.

15. An arrangement for evaluating a radar measurement data and in particular performing a radar measurement, the arrangement comprising: in particular: a transmitter adapted to transmit plural UWB radar radiation pulses towards a target; in particular: a receiver adapted to receive, at plural subsequent sample time intervals, radar signals due to reflection and/or interaction of the transmitted pulses at the target; the arrangement comprising: a processor adapted: to form plural subsequent channel impulse response profiles, across subsets of the sample time intervals, from the received radar signals; to identify, in each of the plural channel impulse response profiles, one of the plural sample time intervals as a target sample time interval in which received radar signals are comprised originating due to reflection from the target; to evaluate total radar signals of interest, each defined as respective magnitude and phase of the plural channel impulse response profiles at the target sample time interval, in order to determine for at least one total radar signal of interest a phase of a target contribution to the total radar signal of interest; to determine location and/or movement information of the target based at least on the at least one phase of the target contribution.

16. The arrangement according to claim 15, configured to perform a radar measurement, the arrangement further comprising: a transmitter adapted to transmit plural UWB radar radiation pulses towards the target; a receiver adapted to receive, at the plural subsequent sample time intervals, radar signals due to interaction of the transmitted pulses at the target.

17. The arrangement according to claim 15, wherein each of the total radar signals of interest comprises: a crosstalk contribution, being a contribution due to back coupling and/or crosstalk between the transmitter, transmitting the pulses, and the receiver, receiving the radar signals; and the target contribution, being a contribution due to interaction and/or reflection of the transmitted pulses at the target.

18. The arrangement according to claim 17, wherein the crosstalk contribution is determined in an approximate manner by averaging the plural total radar signals of interest over an averaging time range, and wherein the approximate crosstalk contribution is subtracted from the total radar signals of interest within the averaging time range, in order to derive approximate target contributions within the averaging time range, each comprising at least an approximate target phase and in particular an approximate target magnitude.

19. The arrangement according to claim 17, wherein the processor is further adapted to: determine the crosstalk contribution by processing the plural total radar signals; and subtract the crosstalk contribution from each of the total radar signals of interest, in order to derive the target contributions, each comprising at least a target phase and in particular a target magnitude.

20. The arrangement according to claim 17, wherein the processor is further adapted to: form cumulative phases of the approximate target phases over different cumulative phase time ranges; and determine a cumulative phase time range of interest, where the cumulative phase of the approximate target phases satisfies a criterium, in particular being substantially or at least 2 Pi or 4 Pi.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0048] FIG. 1 schematically illustrates an arrangement for performing a radar measurement and also for evaluating a radar measurement according to an embodiment of the present invention;

[0049] FIG. 2 schematically illustrates pulses transmitted according to an embodiment of the present invention towards a target;

[0050] FIG. 3 illustrates an example of a channel impulse response profile as evaluated according to embodiments of the present invention;

[0051] FIG. 4 illustrates a total radar signal of interest as considered in embodiments of the present invention;

[0052] FIG. 5 illustrates an evolution of a total radar signal of interest as evaluated according to embodiments of the present invention;

[0053] FIGS. 6A, B, C, D illustrate finding a center of a curve according to embodiments of the present invention;

[0054] FIG. 7 illustrates cumulative phase difference within a measurement window plotted across time which shows that when target is moving in one direction, there are no zero crossings in this signal, but when target moves back and forth, there are zero crossings;

[0055] FIG. 8 illustrates sum of cumulative phases as determined according to embodiments of the present invention;

[0056] FIG. 9 illustrates a time evolution of a channel impulse response during target movement as considered according to an example;

[0057] FIG. 10 illustrates a method scheme according to an embodiment of the present invention; and

[0058] FIG. 11 illustrates evaluation results compared to conventional methods.

DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

[0059] The arrangement 1 for performing a radar measurement according to an embodiment of the present invention schematically illustrated in FIG. 1 comprises in particular a transmitter 2 (Tx) which is adapted to transmit plural UWB radar radiation pulses 3 towards a target 4. The arrangement 1 further comprises a receiver 5 (Rx) which is adapted to receive, at plural subsequent sample time intervals, radar signals 6 which are due to reflection and/or interaction of the transmitted pulses 3 at the target. The arrangement 1 further comprises a processor 7 which is adapted to perform a method of evaluating radar measurement data according to an embodiment of the present invention.

[0060] In the embodiment illustrated in FIG. 1, the receiver 5 comprises also an I/Q demodulator 8 which is adapted to demodulate the received radar signals in order to determine an in-phase component I and a quadrature component Q. Thereby, multiplication elements 9 are provided which multiply the received signals by a sine 10 of the carrier frequency (times the time) and a cosine 11 of the carrier frequency (times the time) in order to obtain the in-phase component I and the quadrature component Q, respectively. Both components I and Q are received by the processor 7.

[0061] FIG. 2 illustrates in a coordinate system having an abscissa 12 indicating time and having an ordinate 13 indicating an intensity or magnitude the pulses 3 as transmitted by the transmitter 2 illustrated in FIG. 1. The pulses 3 may for example be transmitted at regular time intervals such that the time spacing ?t is for example constant between subsequent pulses 3.

[0062] FIG. 3 illustrates in a coordinate system having an abscissa 12 indicating the time delay and having an ordinate 13 indicating the intensity, an example of a channel impulse response profile 14 which is formed by the processor 7 illustrated in FIG. 1 from plural radar signals across subsets of plural subsequent sample time intervals, the plural sample time intervals being labelled with reference sign 15. In particular, the width of the sample interval 15 may be much smaller than the time interval ?t at which the pulses are subsequently transmitted. The channel impulse response profile 14 an example of which is illustrated in FIG. 3 may be for example obtained by averaging multiple reflected pulses 3 which are reflected at the target 4.

[0063] The processor 7 is adapted to perform a method of evaluating radar measurement data, first forming plural subsequent channel impulse response profiles (an example of which is illustrated in FIG. 3) from the received radar signals 6. The processor further is configured to identify in each of the plural channel impulse response profiles, one of the plural sample time intervals 15 as a target sample time interval 16 in which received radar signals 6 are comprised originating due to reflection from the target 4. As can be appreciated from FIG. 3, the exemplary channel impulse response profile 14 exhibits relatively high intensity at the first sample time intervals which slowly decreases towards higher sample time intervals. In the target sample time interval 16, however, the intensity slightly increases which can be detected for example by evaluating differences between subsequent channel impulse response profiles or forming averages of such channel impulse response profiles. The slightly increased intensity in the target sample time interval 16 is due to reflection of radiation from the target 4.

[0064] FIG. 3 illustrates thereby only a magnitude of the plural channel impulse response. However, as has been explained with respect to FIG. 1, the processor 7 not only receives the in-phase signal I, but also the quadrature signal Q, thus magnitude and also phase of the radar signals 6 are obtained by the processor 7. The processor 7 is then configured to determine for at least one total radar signal of interest a phase of a target contribution to the total radar signal of interest. Based on the at least one phase of the target contribution, the processor 7 is then further configured to determine location and/or movement information of the target.

[0065] FIG. 4 illustrates a total radar signal of interest as considered in embodiments of the present invention and being processed by the processor 7. The total radar signal of interest 17 (comprising magnitude and phase of the CIR at the target sample time interval 16) is illustrated in a complex coordinate system having a real axis 18 and an imaginary axis 19. The total radar signal of interest 17 comprises a disturbance contribution or crosstalk contribution 20 being a contribution due to disturbance or back coupling and/or crosstalk between a transmitter 2 and a receiver 5 for example or which may be due to other measurement errors. The total radar signal of interest further comprises a target contribution 21 being a contribution due to interaction and/or reflection of the transmitted pulses 3 at the target 4.

[0066] The target sample time interval 16 can then be used to determine a target location interval 22 being one of plural possible location intervals of the target 4 (see FIG. 1, 3). The target location interval 22 together with the phase (1) of at least one target contribution 21 is then utilized by the processor 7 to estimate movement and/or position of the target within the target location interval 22. When the disturbance contribution 20 of the total radar signal of interest 17 is assumed to be (at least approximately) constant and the target is moving in a same direction in a linear manner for a particular distance, the target contribution 21 would describe a circle 23 in the coordinate system illustrated in FIG. 4. In order to estimate or determine the disturbance contribution 20 according to embodiments of the present invention the center 24 of the circle 23 is determined. The crosstalk contribution 20 is then subtracted from the total radar signal of interest 17 in order to derive the target contribution 21. The center 24 of the circle 23 may also be determined by averaging (and/or fitting) the total radar signal of interest 17 across a particular averaging time range which is selected or determined such that in fact the target contribution describes a full circle around the center 24.

[0067] Below, particular details according to embodiments of the present invention are provided. However, the invention is not restricted to the explained and disclosed details:

Phase Tracking

[0068] The doppler motion is encoded in the CIR phase. Transmitting pulses according to cos(2?f.sub.ct), we would receive cos(2?(f.sub.c+f.sub.d)t)=cos(2?f.sub.ct+?(t)) at the receiver, where the phase ?(t)=2?f.sub.dt depends on the doppler frequency f.sub.d

[0069] After removing the carrier, this phase will modulate the CIR. We look at the phase change over a given time,

[00001]$?\left(?t\right)=2?f_d?t=2?f_c\left(\frac{v}{c}\right)?t=2?\left(\frac{f_c}{c}\right)\left(2?d\right)$

where ?d=v?t/2 is the distance the target 4 has moved in the given time.

[0070] We then get

[00002]$?d=\left(\frac{?}{4?}\right)??\left(?t\right),$

so every 2? angle change corresponds to ?/2 movement of the target

[0071] A key problem is finding ?(t), since what we have is a phasor representing the reflected signal modulating the phasor from TX-RX leakage.

[0072] The angle of the resultant phasor is no longer ?(t), but is <(?e.sup.j?(t)+re.sup.j?)

[0073] The complex phasor

[00003]$\left(?e^{j?\left(t\right)}+{\mathrm{re}}^{j?}\right)={?}_s\left(e^{j?}+\frac{?}{r}{e}^{j?\left(t\right)}\right)$

indicates that if the reflected signal power is much smaller than the leakage power at any given tap, it gets harder to extract ?(t)

[0074] To extract ?(t), we then need to separate the two phasors

[0075] Taking one time instance t o as a reference does not help, since that gives us ?e.sup.j?(t.sup.0.sup.)+re.sup.j? and subtracting that from the time varying phasor, we have ?(t)e.sup.j?(t)??(t.sub.0)e.sup.j?(t.sup.0.sup.) and there is no way to get to ?(t) from this either since we don't know ?(t.sub.0)e.sup.j?(t.sup.0.sup.)

Circle Fitting and Finding the Center

[0076] One way to estimate the static phasor c.sub.l+jc.sub.Q, is to build a system of equations and solve for the unknowns

[0077] With (x.sub.l(t)+jx.sub.Q(t))=(c.sub.l+jc.sub.Q)+?(t)?(cos ?(t)+j sin ?(t)), we have no way to do that since we have 4 unknowns and cannot get a rank-4 matrix that we can invert regardless of how many observations we collect

[0078] We can however use the fact that the static phasor forms the center of the arc/circle/spiral the signal phasor traces out as it changes over time

[0079] We have (x.sub.l.sup.(t)?c.sub.l).sup.2+(x.sub.Q(t)?c.sub.Q).sup.2=?.sup.2(t), and if we have 3 points where ?(t)=?, which is still unknown, but remains constant, we can form the following system of equations to estimate c.sub.l+jc.sub.Q,

[00004]$({x_I\left(t_1-c_I\right)}^2+{\left(x_Q\left(t_1\right)-c_Q\right)}^2-\left({\left(x_I\left(t_0\right)-c_I\right)}^2+{\left(x_Q\left(t_0\right)-c_Q\right)}^2\right)={\left(x_I\left(t_2\right)-c_I\right)}^2+{\left(x_Q\left(t_2\right)-c_Q\right)}^2-\left({\left(x_I\left(t_0\right)-c_I\right)}^2+{\left(x_Q\left(t_0\right)-c_Q\right)}^2\right)=0$$\frac{\left(x_I^2\left(t\right)-x_I^2\left(t_0\right)\right)+\left(x_Q^2\left(t\right)-x_Q^2\left(t_0\right)\right)}{2}=\left(x_I\left(t\right)-x_I\left(t_0\right)\right)c_I+\left(x_Q\left(t\right)-x_Q\left(t_0\right)\right)c_Q$

Which simplifies to:

[0080] For t=t.sub.1 and t=t.sub.2, this gives us two equations in two unknowns that we can solve for

[0081] Note however that the choice of t.sub.0, t.sub.1 and t.sub.2 are critical for this approach. If we pick then too close, then we get an ill-conditioned matrix that we have to invert. If we pick them too far, the assumption that ?(t)=? is not strictly true.

[0082] By the circle fitting, the center 24 can be found. The accuracy of the center determination and the distance errors depend on the accuracy of the disturbance contribution determination 20. This in turn influences the accuracy of phase tracking of the target phase. Due to the particular movement manner of the target, the target may move only in an arc segment and not in a full circle. According to embodiments of the present invention, it is determined when the target contribution 21 runs through at least a full circle. Below, some particular details are given in order to identify the target sample time interval 16 also referred to as target tap and to determined cumulative phases:

[0083] Use the clutter removed CIR to identify the target tap

h.sub.ref(?)=h.sub.m.Math.(L.sub.win.sub.?L.sub.ovlp.sub.)(?) where m=0,1,2, . . .

[0084] {hacek over (h)}.sub.n(?)=h.sub.n(?)?h.sub.ref(?) with n=m .Math.(L.sub.win-L.sub.ovlp)+ being the absolute time index and =1,2, . . . , L.sub.win being the index within each window

[00005]${?}_{\mathrm{tgt}}=\underset{?,m}{\max }.Math.{\stackrel{?}{h}}_n\left(?\right).Math.$

or use a target peak finding algorithm such as one used for vital signs detection

[0085] With h.sub.n(?)=h.sub.s(?)+h(t, ?), we have:

h.sub.ref(?)=h.sub.s(?)+h(t.sub.0, ?) and {hacek over (h)}.sub.n(?)=h(t, ?)?h(t.sub.0, ?)=?(t)e.sup.j?(t)??(t.sub.0)e.sup.j?(t.sup.0.sup.)

[0086] Estimate phase change {tilde over (?)}(?t) within each window m by using the clutter removed CIR {hacek over (h)}.sub.n(?), and using mean tap within that window to estimate the center:

{tilde over (c)}.sub.m(?.sub.tgt)=E{{hacek over (h)}.sub.n(?)} where E{} is the expectation operator

{tilde over (?)}(t)e.sup.j{tilde over (?)}(t)={tilde over (h)}(t, ?.sub.tgt)={hacek over (h)}.sub.n(?.sub.tgt)?{tilde over (c)}.sub.m(?.sub.tgt), and

[00006]$\widetilde{?}\left(?t\right)=?\left(\frac{\widetilde{?}\left(t\right)e^{j\widetilde{?}\left(t\right)}}{\widetilde{?}\left(t+?t\right)e^{j\widetilde{?}\left(t+?t\right)}}\right),$

and the cumulative sum {tilde over (?)}.sub.m()=?.sub.1.sup.l {tilde over (?)}(?t) where t=n.Math.T.sub.RIFR and the coarse accumulated phase at the end of every window is {tilde over (?)}.sub.m=?.sub.1.sup.m {tilde over (?)}.sub.m(L.sub.win)

[0087] Though this will cause incorrect angle estimation, the direction in which the accumulated phase moves will be maintained

[0088] Sum the cumulative phase changes in each window: ?.sub.m=?.sub.l{tilde over (?)}.sub.m()

[0089] FIG. 5 illustrates in a three-dimensional coordinate system having a real axis 25 and imaginary axis 26 and an axis 27 indicating the index of the channel impulse response profile the target CIR tap evolution across time. Between the different layers 28, the target moves farther by 10 cm.

[0090] FIGS. 6A, B, C, D illustrate in coordinate systems having a real axis 25 and an imaginary axis 26 evolution of total signals of interest 17 and an iterative determination of approximate or accurate centers 24 during different phases of movement of the target 4. The curves 17 illustrate the evolvement of the total radar signal of interest during the different movement phases.

[0091] FIG. 7 illustrates the cumulative phase 17P of the total radar signal of interest across plural channel impulse response profiles whose index is indicated in the abscissa 29. The cumulated phase of the total radar signal of interest is indicated along the ordinate 30.

[0092] The sum of cumulative phase change within each window, ?.sub.m, will show multiple zero crossing across windows if the movement is periodic (short to and from movement such as breathing) and will be always positive or negative for several consecutive windows if the target moves closer or farther away.

[0093] Use the zero crossing to detect time instances when the target is moving in one direction

[0094] Check the coarse accumulated phase {circumflex over (?)}.sub.m to see if it has crossed 4? rotation.

[0095] Even though the accumulated phase across windows, {tilde over (?)}.sub.m, is under or over estimated due to incorrect center, we would at least have either one full revolution of the signal phasor, or more than 2 revolutions

[0096] Use the mean of h.sub.n(?.sub.tgt) across this movement window to estimate the center. Since we are likely to have a full circle, center estimation is accurate.

[0097] Use the center to remove the static phasor and accumulate angles to map to distance movement

[0098] Update center as long as we have movement, but update only if the accuracy doesn't drop.

[0099] The multiple subsequent channel impulse responses at the target sample time interval (in particular) is evaluated and the zero crossings or crossing is used to detect time window when the target is moving in one direction. Between the circled regions 31, the target moves farther by 10 cm.

[0100] FIG. 8 shows in a coordinate system having an abscissa 32 indicating the time window index m and having an ordinate 33 indicating the sum of the cumulative phase per window, the sum of cumulative target phase represented by a curve 34. As seen here, during the windows where there is a target movement on one direction (see the circles traced in FIG. 5), we see that there are no zero crossings. In all the other windows indices, the cumulative phase per window crosses zero several times.

[0101] FIG. 9 illustrates in a three-dimensional coordinate system having a real axis 25, an imaginary axis 26 and an axis 27 indicating the CIR index, the time evolution of the CIR tap (e.g. corresponsing to a total signal of interest 17) during the target movement. The portion 17a, 17b, 17c, 17d illustrate the total radar signal of interest, wherein different target time sample intervals are occupied or used such that the target moves during the entire evolution from one target sample time interval to an adjacent interval and so forth. For each of the portions 17a, 17b, 17c, 17d, a centering algorithm may be performed.

Recentering Across Taps

[0102] As the accumulated phase grows beyond 15 cm, looking at the time evolution of the target tap is suboptimal, since the peak due to target is now captured in the adjacent tap.

[0103] But switching to the adjacent tap means computing the center again

[0104] Recentering is done by computing the center again during the movement window

[0105] FIG. 10 illustrates a method scheme 35 as applied or performed according to embodiments of the present invention. In a method step 36, clutter removal within each processing window of the received radar signals is performed.

[0106] In a method step 37, the target tap is centered around the local mean (computed within the window).

[0107] In a method step 38, the phase change across adjacent CIRs is computed, accumulated and the phase change within the window is accumulated.

[0108] In a method step 40, the coarse phase accumulation across windows is computed.

[0109] In a next method step 41, the sum of the cumulative phase for each window is computed, wherein it is looked for set of windows with no zero crossings.

[0110] In a method step 42, the target tap during movement windows (no zero crossings) is used to compute the accurate center (of the disturbance contribution).

[0111] In a next method step 43, the target tap is centered within movement windows using the new center estimate.

[0112] In a next method step 44, the phase change across adjacent CIRs is computed, and the phase change is accumulated within window.

[0113] In a next method step 45, the fine phase accumulation across windows is computed.

[0114] In a next method step 46, it is checked if the distance from the fine phase accumulation crosses distance per tap crosses tap. If this is the case, it is switched to an adjacent target tap and it is recentered.

[0115] The method step 46 then refers back to method step 42 which again leads to subsequent method steps 43, 44, 45 and again to step 46.

[0116] If method step 46 evaluates to false, then in a method step 47 (if the target moves <15 cm in one direction and then stops to perform a periodic motion, then step 47 is performed), the fine phase accumulation using known center even after movement stops is continued. The method step 47 leads back to method step 46.

[0117] According to embodiments of the present invention, one or more method steps illustrated in FIG. 10 may be omitted.

[0118] FIG. 11 illustrates, in a coordinate system having an abscissa 50 indicating time and having an ordinate 51 indicating the estimated distance, a conventional peak finding methodology 52 (on 1 ns tap), a peak finding 53 with a time interpolation and a phase tracking for distance estimation 54 as obtained according to embodiments of the present invention. Thereby, a robot was breathing from 5 cm to 45 cm in 10 cm steps. As can be appreciated from FIG. 11, the curve 54 representing the results according to embodiments of the present invention accurately determines the movement of the robot and also the locations of the robot being 10 cm spaced apart.

[0119] Thus, embodiments of the present invention provide a sub-centimeter accuracy for movement tracking at close distances while even eight times interpolation with peak finding is not accurate enough. According to embodiments of the present invention, in particular linear motion is tracked very well by estimating the distance using phase much better than conventional peak finding algorithms.

[0120] Conventional target detection may use the CIR magnitude; phase information, however, is ignored. In particular, a phase is not usable directly in systems with practical limitations such as high TX leakage and CIR drift across temperature.

[0121] According to embodiments of the present invention, an algorithm is disclosed that makes use of the CIR phase to estimate target motion in sub-centimeter accuracy. The method may even work in case of radar devices with practical limitations, such as high TX to RX leakage where separating the CIR phase changes due to target movement from the self-interference may become a challenging task.

[0122] Embodiments of the present invention allow for the UWB modems radar functionality to be used for applications where high distance estimation accuracy is required, such as under the wall imaging, high resolution range imaging, high accuracy time-of-flight for camera autofocus, etc.