SYSTEM AND METHOD FOR CHARACTERIZING PROPERTIES OF EM SIGNALS

Abstract

A method and system are presented for determining properties of an electromagnetic waveform. The method comprises: providing measured parametric EM field data indicative of measured vector components of electric and magnetic fields of an EM waveform measured in at least one instance of time; providing reference data indicative of a plurality of reference data sets, each data set comprising: a reference steering vector parameters indicative of a certain respective direction of arrival (DOA), and a corresponding parametric EM field reference data including reference vector components of an electric and magnetic field pertaining to a wavefront propagating with the DOA of the corresponding reference steering vector parameters; determining a matching score between the measured parametric EM field data and the parametric EM field reference data of one or more of the reference data sets; and in case the matching score of a certain reference data set complies with a certain threshold condition, determining that said measured parametric EM field data corresponds to said EM waveform having a single EM wavefront thereby enabling to discriminate between measured EM waveforms having a single wavefront and measured EM waveforms having multiple wavefronts.

Claims

1. A method for determining properties of an electromagnetic waveform, the method comprising: providing measured parametric EM field data indicative of measured vector components of electric and magnetic fields of an EM waveform measured in at least one instance of time; providing reference data indicative of a plurality of reference data sets, each data set comprising: a reference steering vector parameters indicative of a certain respective direction of arrival (DOA), and a corresponding parametric EM field reference data including reference vector components of an electric and magnetic field pertaining to a wavefront propagating with the DOA of the corresponding reference steering vector parameters; determining a matching score between the measured parametric EM field data and the parametric EM field reference data of one or more of the reference data sets; and in case the matching score of a certain reference data set complies with a certain threshold condition, determining that said measured parametric EM field data corresponds to said EM waveform having a single EM wavefront thereby enabling to discriminate between measured EM waveforms having a single wavefront and measured EM waveforms having multiple wavefronts.

2. The method of claim 1, comprising, upon determining that said EM waveform has a single EM wavefront, estimating a DOA of said single EM wavefront based on the reference steering vector parameters of said certain reference data set whose matching score complies with said certain threshold condition.

3. The method of claim 1, comprising determining a plurality of matching scores between the measured parametric EM field data and the parametric EM field reference data of each reference data set of said plurality of reference data sets respectively; and at least one of the following: in case none of said plurality of matching scores complies with said certain threshold condition, determining that said measured parametric EM field data corresponds to the EM field measured under interference conditions, by which DOA cannot be accurately estimated; at least in case one or more than one matching scores associated with a certain respective one or more candidate reference data sets comply with said certain threshold condition, applying further processing to said certain respective candidate reference data set, based on predetermined data on properties of a source of said electromagnetic waveform, to thereby filter out from said one or more candidate reference data sets, candidate reference data sets whose respective reference steering vectors' parameters do not comply with said predetermined data.

4. (canceled)

4. The method of claim 4, wherein said predetermined data pertains to a state of polarization of said source, thereby enabling to discriminate between single path waveforms and multipath waveforms which are associated with reflection/scattering from objects in relative vicinity to said source, as compared to their distance from the vector sensor.

6. (canceled)

7. The method of claim 1, wherein the measured parametric EM field data is indicative of nominal vector components of said electric and magnetic fields obtained from a plurality of measurements of said electric and magnetic fields taken in a plurality of time instances.

8. The method of claim 7, wherein said nominal vector components of said electric and magnetic fields are obtained by processing said plurality of measurements of said electric and magnetic fields which are taken in a plurality of time instances and wherein at least one of the following: said processing comprises averaging of said plurality of measurements; said processing comprises coherent integration of said plurality of measurements; said measured vector components are obtained from a sensor being under movement conditions.

9-13. (canceled)

14. The method of claim 7, wherein said measured parametric EM field data pertains to a certain frequency and wherein the method comprises processing said plurality of measurements of said electric and magnetic fields taken in the plurality of time instances, to determine frequency contents of the measured electric and magnetic fields in at least said certain frequency such that said nominal vector components pertain to amplitudes of components of said electric and magnetic fields associated with said certain frequency.

15. (canceled)

16. The method of claim 1, wherein the matching score between the measured parametric EM field data and the parametric EM field reference data of a certain reference data set is determined by computing an inner product between the measured parametric EM field data and the parametric EM field reference data of the certain reference data set; and wherein said inner product is computed by at least one of the following: said inner product is computed between normalized values of the measured parametric EM field data and the parametric EM field reference data of the certain reference data set; and said inner product is computed as a generalized (coherent) inner product computed based on a covariance of said measured parametric EM field data and said the parametric EM field reference data of the certain reference data set.

17. (canceled)

18. The method of claim 7, wherein said nominal vector components of the electric and magnetic fields pertain to electric and magnetic fields of a certain temporal form; and wherein said plurality of reference data sets are associated with one or more predetermined temporal forms.

18. The method of claim 18, wherein at least one of the following: said certain temporal form and said one or more predetermined temporal forms are pulse forms of predetermined widths; the matching score between the measured parametric EM field data and the parametric EM field reference data of a certain reference data set is determined by computing a correlation between normalized values of the measured parametric EM field data and the parametric EM field reference data of the certain reference data set.

20-21. (canceled)

22. The method of claim 1, wherein the reference steering vector parameters of at least one reference data set of said reference data, is indicative of at least one of the following: a two dimensional direction of arrival (DOA) of the corresponding parametric EM field reference data characterized by (Θ, φ); and wherein determining compliance of the matching score with said certain threshold condition thereby provides an un-ambiguous estimate of two dimensional direction of arrival (DOA) of a wavefront associated with said measured parametric EM field; polarization parameters of the corresponding parametric EM field reference data; and wherein determining compliance of the matching score with said certain threshold condition thereby provides an un-ambiguous estimate of the direction of arrival (DOA) of a wavefront associated with said measured parametric EM field and polarized with said polarization parameters.

23. (canceled)

24. The method of claim 1, wherein the measured parametric EM field is obtained by at least one parametric EM field sensing device (vector sensor) comprising three or more sensing modules (antenna elements) arranged to span sensing of at least one of an electric field or a magnetic field in 3D coordinates.

25-29. (canceled)

30. The method of claim 1, wherein said reference data is adjusted to a certain location at which the vector sensor is located so as to compensate over warping effect of objects in the vicinity of said location on the measured parametric EM field sensed at said location.

31. A system for determining properties of an electromagnetic waveform, the system comprising: a measurements' preprocessor module connectable to a vector sensor and adapted to provide measured parametric EM field data indicative of measured vector components of electric and magnetic fields of an EM waveform measured in at least one instance of time; a reference data provider module connectable to a reference data repository comprising reference data and adapted to provide a plurality of reference data sets, each data set comprising: a reference steering vector parameters indicative of a certain respective direction of arrival (DOA), and a corresponding parametric EM field reference data including reference vector components of an electric and magnetic field pertaining to a wavefront propagating with the DOA of the corresponding reference steering vector parameters; a comparison module configured and operable for determining matching scores between the measured parametric EM field data and the parametric EM field reference data of one or more of the reference data sets; and an interference estimator module configured to determine whether said EM waveform having a single EM wavefront based on compliance of at least one of said matching scores with a certain threshold condition, thereby enabling to discriminate between measured EM waveforms having a single wavefront and measured EM waveforms having multiple wavefronts.

32. The system claim 31, wherein the comparison module is adapted to determine as candidate reference data sets, one or more of the reference data sets whose respective matching scores comply with said certain threshold condition; and wherein at least one of the following: said interference estimator module further comprises a filtration module configured and operable to filter out, form said one or more candidate reference data sets, certain candidate reference data sets whose respective reference steering vectors' parameters do not comply with said predetermined data indicative of properties of a source of said EM waveform; the system comprises a DOA/SOP estimator module configured and operable for estimating a DOA of said EM wavefront based on the reference steering vector parameters of said certain reference data set whose matching score complies with said certain threshold condition said interference estimator module is adapted to determine that in case none of said matching scores complies with said certain threshold condition, said measured parametric EM field data corresponds to the EM field measured under interference conditions, by which DOA cannot be accurately estimated.

33-34. (canceled)

31. The system of claim 31, wherein said predetermined data pertains to a-priory data indicative of a state of polarization of said source.

36. The system of claim 31, wherein the measurements preprocessor module is adapted to obtain a plurality of measurements of said electric and magnetic fields taken in a plurality of time instances and for processing said plurality of measurements to obtain said measured parametric EM field data as indicative of nominal vector components of said electric and magnetic fields.

37-38. (canceled)

39. The system of claim 36, wherein said measured vector components are obtained from a sensor being under movement conditions.

40. The system of claim 36, wherein said nominal vector components of the electric and magnetic fields pertain to electric and magnetic fields of a certain frequency; and wherein said plurality of reference data sets are frequency specific reference data sets whose parametric EM field reference data pertain to said certain frequency.

41-42. (canceled)

43. The system of claim 31, wherein said comparison module is configured and operable for determining said matching scores between the measured parametric EM field data and the reference data sets of the by respectively computing inner products between the measured parametric EM field data and the corresponding parametric EM field reference data of the respective reference data sets; and wherein at least one of the following: said inner products are computed between normalized values of the measured parametric EM field data and the corresponding parametric EM field reference data of the respective reference data sets; and said inner products are computed as generalized (coherent) inner products computed based on covariance of said measured parametric EM field data and the corresponding parametric EM field reference data of the respective reference data sets.

44. (canceled)

45. The system of claim 36, wherein said nominal vector components of the electric and magnetic fields pertain to electric and magnetic fields of a certain temporal form; and wherein said plurality of reference data sets are associated with one or more predetermined temporal forms; and wherein said comparison module is configured and operable for determining the respective matching scores between the measured parametric EM field data and the parametric EM field reference data of respective reference data sets by computing correlations between normalized values of the measured parametric EM field data and the corresponding parametric EM field reference data of the respective reference data sets.

46-47. (canceled)

48. The system of claim 31, wherein the reference steering vector parameters of at least one reference data set of said reference data, is indicative of one or more of the following: direction of arrival (DOA) of the corresponding parametric EM field reference data characterized by (Θ, φ); polarization parameters of the corresponding parametric EM field reference data.

49. The system of claim 31, comprising or being connected to a parametric EM field sensing device (vector sensor) and wherein at least one of the following: said reference data is specific to a type of said parametric EM field sensing device, and includes compensation over warping/distortion effect of said EM field sensing device on the measured parametric EM field sensed by the EM field sensing device; said parametric EM field sensing device being a single parametric EM field sensing device located at a certain location or platform and wherein said reference data is adjusted according to said certain location or platform so as to compensate over warping effect of objects in the vicinity of said certain location or platform location. cm 50. (canceled)

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0087] In order to better understand the subject matter that is disclosed herein and to exemplify how it may be carried out in practice, embodiments will now be described, by way of non-limiting example only, with reference to the accompanying drawings, in which:

[0088] FIG. 1A is a schematic illustration exemplifying a configuration of an EM vector sensor;

[0089] FIG. 1B illustrates, in a self-explanatory manner, an arbitrary EM waveform having a single planar wavefront traveling in the −{circumflex over (r)} direction (e.g. towards a vector sensor);

[0090] FIG. 2A shows a block diagram of a system 100 for discrimination between waveforms sensed/measured by a vector sensor having one or multiple wavefronts and optionally subsequent determination of direction of arrival (DOA) and possibly state of polarization (SOP) of the electromagnetic wavefront, according to an embodiment of the present invention;

[0091] FIG. 2B to 2D together present a flow chart of a method 200 for performing such discrimination between measured waveforms having one or multiple wavefronts and optionally subsequently determining/estimating the DOA and possibly SOP of the electromagnetic wavefront, according to an embodiment of the present invention; whereby: [0092] FIG. 2B specifically presents the operation/sub-method carried out in order to obtain measured parametric EM field data custom-character {right arrow over (X)} indicative of vector components of electric and magnetic fields measured during one or more time instances/frames; [0093] FIG. 2C specifically presents the operation/sub-method carried out in order to provide, compute and/or measure reference data, including reference data sets, which are to be used according to the technique of the present invention; and [0094] FIG. 2D specifically presents the operation/sub-method carried out in order to determine matching score(s) between the measured parametric EM field data custom-character {right arrow over (X)} and one or more of the reference data sets {DS.sub.i}, and thereby determining/estimating whether the measured parametric EM field data {right arrow over (X)} is characterized by one wavefront or by multiple wavefronts, and optionally subsequently determining/estimating the DOA and possibly SOP parameters of the measured parametric EM field data custom-character {right arrow over (X)};

[0095] FIGS. 3A and 3B are schematic illustrations of two examples of multipath scenarios that can be identified according to various embodiments of the present invention, whereby in both the examples a signal/waveform is shown to propagate from a signal source 1 to a vector sensor 10 directly WF1 as well as indirectly WF2 via a scattering/reflecting object 2 (whereby in the example of FIG. 3A the scattering/reflecting object 2 resides relatively close to the vector sensor, and in the example of FIG. 3B the scattering/reflecting object 2 resides relatively close to the signal source); and

[0096] FIG. 4 is a graphical illustration showing the statistics (commutative distribution function) of highest matching scores obtained according to the technique of the present invention for multipath (interfered) waveforms (e.g. yielding a plurality of wavefronts), and single path (non-interfered waveforms (e.g. having a single wavefront).

DETAILED DESCRIPTION OF EMBODIMENTS

[0097] Reference is now made together to FIGS. 2A to 2D, in which FIG. 2A is a block diagram of a system 100 for discrimination between waveforms sensed/measured by a vector sensor having one or multiple wavefronts (e.g. waveforms not subjected to interference/multipath conditions, or waveforms that are subjected to such conditions) and optionally subsequent determination of direction of arrival (DOA) and possibly state of polarization (SOP) of an electromagnetic wavefront according to an embodiment of the present invention; and FIG. 2B to 2D together presenting a flow chart of a method 200 for performing such discrimination between measured waveforms having one or multiple wavefronts and optionally subsequently determining/estimating the DOA and possibly SOP of the electromagnetic wavefront according to an embodiment of the present invention. The system 100 may be configured and operable for implementing various embodiments of the method 200. Accordingly, the configuration and operation of the system 100 and method 200 are described herein below together with reference to FIGS. 2A to 2D.

[0098] It should be noted that system 100 may generally include analogue and or digital modules adapted for connecting to the vector sensor (e.g. to the antenna elements thereof, their feeding points, and/or RF front-ends associated therewith, and/or digital processing modules of such vector sensors, to obtain therefrom signals/data indicative of the components of a waveform sensed/measured by the antenna elements of the vector sensor. The system may generally include any suitable analog/digital signal processing modules which may be adapted for performing preliminary processing on the received signals whereby such preliminary processing may include filtration of the signals (e.g. frequency filtration e.g. utilizing analog and/or digital pass/notch filters), down conversion of the signal, demodulation (e.g. utilizing signal mixers and/or digital demodulators) of the signals (e.g. to obtain the modulation and/or carrier components separately, down convection, phase shifting and/or analogue to digital conversion by suitable A/D convertors, or vice-versa digital to analogue conversion by suitable D/A convertors, all as may be required in particular embodiments of the present invention for the preliminary processing of the specific type of the signals which are to be analyzed by such embodiments. Moreover, system 100 may also include signal processors, typically digital processors, which may be implemented by proper CPU(s) and/or DSPs and memory/storage modules included in the system 100, or by other means (e.g. analog processing means as known in the art (e.g. signal delay lines, phase shifters, mixers, filters etc.) in order to apply the processing described below with reference to the system 100 and method 200 below. To this end, the modules of system 100 described in detail below may be implemented by such analog or digital means, and a person of ordinary skill in the art will readily appreciate that the present invention as described below may be implemented utilizing digital means and/or analog means and/or a combination of digital and analog means as indicated above for carrying out the invention according to the various embodiments thereof as described in detail below.

[0099] System 100 includes a wavefront measurements preprocessor module 110 connectable to a vector sensor 10 and adapted to receive/sample the parametric EM field components x.sup.t=[E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.t measured by the vector sensor 10 at a certain one or more time instances, t, (e.g. at a certain sampling rate). Here, t may be the time/sample index of the obtained measurements/samples from the vector sensor and [E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.t present the measured electromagnetic field and may correspond to the voltages measured at respective feeding ports of 6 antenna elements of the vector sensors, each designed to measure a different component of the magnetic and electric fields sensed by the sensor (e.g. in case the vector sensor comprise three orthogonal pairs of magnetic and electric antennas). The wavefront measurements preprocessor module 110 is adapted to obtain the samples x.sup.t of the measured electromagnetic field at one or more (typically a plurality) of time-instances/samples and preprocess those measurements (e.g. in the manner discussed in detail below with respect to FIG. 2B), so as to obtain measured parametric EM field data (e.g. a measured steering vector) custom-character {right arrow over (X)} for further use for accurately determining/estimating the DOA and/or SOP of the sensed EM field.

[0100] In this regard, for understanding the below description, it may be noted that the measured samples may be assumed to comply with the following general sensing model of the vector sensor. The measured EM field/waveform {right arrow over (x)}.sub.t sample at a sampling instant t may be linked to the signal Sig.sub.t that is transmitted by the signal source and received/sensed by the vector sensor as follows:

{right arrow over (x)}.sub.t=h.sub.ESig.sub.t+{right arrow over (q)}.sub.t

where t is the sampling index, Sig.sub.t is the raw signal (e.g. a complex number/scalar) that was transmitted from the source at the t-th sample, h.sub.E is the steering filter (e.g. being an operator in the temporal domain or frequency domain, which is sometimes generally referred to as vector effective height of the receiving antenna-system/vector-sensor) presenting how the waveform of the specific signal Sig.sub.t is received and measured as {right arrow over (x)}.sub.t by the antenna elements of the vector sensor. Typically the operator h.sub.E is independent of t and, while the t dependency of the measurements {right arrow over (x)}.sub.t is associated with the temporal form of the specific signal/modulation Sig.sub.t. In this connection it should be noted that the general description of the signal is not limited to its presentation in the frequency or time domain, and h.sub.E is the response operator presenting the operation/response function of the vector sensor to the signal Sig.sub.t. The operator h.sub.E may be a vector function of the signal Sig.sub.t (e.g. convolution of the signal with vectorial impulse response function of the vector sensor, which may be the case when the processing is performed in the time domain) or in some cases the operatorh.sub.E may be presented as a vector multiplying the signal Sig.sub.t and presenting the how the vector components of the EM field would be measured by the vector sensor in response to the signal Sig.sub.t (e.g. when considering the frequency domain, the operator h.sub.E may be a function of the frequency).

[0101] System 100 includes a reference data repository (e.g. local or remote memory module) storing reference data indicative of the response of the vector sensor 10 to being impinged by various single warfronts arriving from different directions (having different DOAs), and/or arriving with different states of polarization (SOPs). To this end the reference data REF is indicative of a plurality of reference data sets REF={DS.sub.i}, each data set DS.sub.i comprising: a reference steering vector parameters {right arrow over (S)}.sub.i and a corresponding reference parametric EM field data {right arrow over (A)}.sub.i (e.g. a corresponding reference steering vector). The reference steering vector parameters {right arrow over (S)}.sub.i is indicative of at least a certain respective direction of arrival, {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i], and typically, according to some embodiments of the present invention is also indicative of certain SOP parameters [γ.sub.i, τ.sub.i] so that {right arrow over (S)}.sub.i is given as {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i]. θ.sub.i, ϕ.sub.i are azimuth and elevation angles of the direction of propagation of the single reference wavefront, and γ.sub.i, τ.sub.i are tilt and phase lag of cardinal polarization components. The corresponding reference steering vector (referred to herein as reference parametric EM field data) {right arrow over (A)}.sub.i is a generalized EM field vector {right arrow over (A)}.sub.i=[Ê.sub.x.sup.i Ê.sub.y.sup.i Ê.sub.z.sup.i Ĥ.sub.x.sup.i Ĥ.sub.y.sup.i Ĥ.sub.z.sup.i] presenting the response of the vector sensor 10 to an electromagnetic field arriving with the DOA and SOP parameters of its respective steering vector parameters {right arrow over (S)}.sub.i. To this end the reference parametric EM field data {right arrow over (A)}.sub.i=[Ê.sub.x.sup.i Ê.sub.y.sup.i Ê.sub.z.sup.i Ĥ.sub.x.sup.i Ĥ.sub.y.sup.i Ĥ.sub.z.sup.i] is indicative of the vector components of electric and magnetic fields that are expected to be measured by the vector sensor 10 in response to an EM wavefront propagating with DOA and/or SOP corresponding to the reference steering vector parameters {right arrow over (S)}.sub.i. Generally, the reference data includes a plurality of such data sets {DS.sub.i}. Thus, the reference data, REF, may be presented as REF={DS.sub.i}={[{right arrow over (A)}.sub.i, {right arrow over (S)}.sub.i]}. For example the reference data may be stored in the reference data repository 120 in the form of a lookup table or a model (e.g. empirical/analytical model in which each line is a reference dataset DS.sub.i=[{right arrow over (A)}.sub.i, {right arrow over (S)}.sub.i]=[[Ê.sub.x.sup.i Ê.sub.y.sup.i Ê.sub.z.sup.i Ĥ.sub.x.sup.i Ĥ.sub.y.sup.i Ĥ.sub.z.sup.i], [θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i]] (this is e.g. a four dimensional table with the dimensions corresponding to the parameters of the steering vector parameters: θ, ϕ, γ, τ). Alternatively or additionally in some embodiments, as indicated below, the reference data is split to partial reference data (the function F) depending on the DOA parameters θ, ϕ and partial reference data (the function g) depending on the SOP parameters γ, τ. In this case, the two dimensional lookup-tables/modes and may be provided in the reference data. In this regard, it should be noted that in some embodiments the part of the reference data (the function G) which depends on the SOP parameters γ, τ, may be considered to be less affected by distortions (e.g. the matrix D below may be considered to be independent from the SOP parameters γ, τ, or having only negligible dependence on these parameters), and thus this part may be modelled by the analytic function g as provided below or a somewhat modified one). It should be noted that the part of the reference data (the function F) which depends on the DOA parameters θ, ϕ, is generally more dependent on the distortions affecting the waveform by objects surrounding the vector sensor (e.g. the matrix D below may be considered to be relatively strongly dependent from the DOA parameters θ, ϕ). Accordingly, this part of the reference data is often provided in a lookup table presenting the response of the vector sensor (e.g. inclusive of the platform at which it is located/fixed and/or the surroundings at the vicinity of the vector sensor) to waveforms from different DOAs. However, nonetheless, still in some embodiments/cases, a model (e.g. empirical/analytical) may also be constructed to represent F relatively accurately and in such cases there may be no need for representing F in a lookup table.

[0102] To this end the reference parametric EM field data {right arrow over (A)}.sub.i may be represented as follows:

[00008] $\overset{.fwdarw.}{A} = [\begin{matrix} {\hat{E}}_{x} \\ {\hat{E}}_{y} \\ {\hat{E}}_{z} \\ {\hat{H}}_{x} \\ {\hat{H}}_{y} \\ {\hat{H}}_{s} \end{matrix}] = \overset{\overset{? (θ, ϕ)}{︷}}{D [\begin{matrix} \cos θcos ϕ & - \sin ϕ \\ \cos θsin ϕ & \cos ϕ \\ - \sin θ & 0 \\ \sin ϕ & \cos θ \cos ϕ \\ - \cos ϕ & \cos θ \sin ϕ \\ 0 & - \sin θ \end{matrix}]} \overset{\overset{\overline{g} (γ, τ)}{︷}}{[\begin{matrix} \sin τ e^{j γ} \\ \cos τ \end{matrix}]}$ $? indicates text missing or illegible when filed$

[0103] where D is an empirical/measured/calculated matrix containing parameters like η (the wave impedance) and other potential normalization factors or other objects affecting warping of the EM field in the vicinity of the vector sensor 100.

[0104] Thus the matrix F(θ, ϕ) in the reference parametric EM field vector {right arrow over (A)} indicates the part describing (θ, ϕ) dependency, and the vector {right arrow over (g)}(γ, τ) presents the polarization parameters related to (γ, τ), and generally spans to orthogonal polarizations. In the specific non-limiting example above the {right arrow over (g)} vector indicates vertical and horizontal polarization, (where (γ, τ) are the phase lag between the components and tilt angle). Accordingly, the two columns matrix F respectively may be configured as follows: the first column relates to the vertical polarization {circumflex over (θ)} components of the reference parametric field {right arrow over (A)} and the second column relates to the horizontal polarization {circumflex over (ϕ)} components of the reference parametric field {right arrow over (A)}.

[0105] F may include the theoretical projection matrix of the EM fields on the axes (e.g. Cartesian) of the vector sensor as well as the warping/distortion matrix D which presents the empirical/measured/calculated warping parameters/factors affecting the EM field/waveform due to objects/surroundings of the vector sensor, the medium's wave impedance, and other optionally also other normalization factors. Thus, D may be determined by calibration measurements (e.g. e.g. via antenna measurement system) and/or analytical computations. D may be dependent as described below on the frequency ω).

[0106] As will be appreciated from the below detailed description and accompanying drawings, in some implementations the reference data REF is spanned in the frequency domain {ω}. Namely REF={DS.sub.i, ω} In this case the reference parametric EM field data {right arrow over (A)}.sub.i in each data set is also specific to a certain to wavefront's frequency ω. In this case the reference parametric EM field data is {right arrow over (A)}.sub.i.fwdarw.{right arrow over (A)}.sub.i, ω and is indicative of the frequency response of the vector sensor to an EM wavefront of that specific frequency propagating with the DOA and/or SOP of the corresponding reference steering vector parameters {right arrow over (S)}.sub.i of the respective data set. Thus, in this case the reference data REF is generally given by REF={DS.sub.i, ω}={[{right arrow over (A)}.sub.i, ω, {right arrow over (S)}.sub.i]}. It should be understood that in cases where the reference data REF is spanned in the frequency domain {ω}, the preprocessing of the measured parametric EM field data custom-character {right arrow over (X)} by preprocessor 110 further includes applying frequency filtering to the measured parametric EM field data {right arrow over (X)} according to an expected frequency of a signal of interest ω.sub.0 to obtain only the measured parametric EM field data {right arrow over (X)} custom-character .sub.ω.sub.0 oscillating at that frequency of interest. In this case the measured parametric EM field data {right arrow over (X)}.sub.ω.sub.0 may be a generalized vector of complex scalars presenting the amplitudes and phases of the vectorial components [E.sub.x, E.sub.y, E.sub.y, H.sub.x, H.sub.y, H.sub.z,].sup.ω.sup.0 of the measured electric and magnetic fields which oscillate with the frequency of interest ω.sub.0: custom-character {right arrow over (X)}.sub.ω.sub.0=[E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.ω.sup.0. For instance, the samples of the parametric EM field components x.sup.t=[E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.t measured by the vector sensor 10 at certain one or more time instances t may be processed/transformed (e.g. by Fourier Transform; or by digital down converter (DDC)) to the frequency domain and only the complex components in this domain pertaining to the frequency of interest ω.sub.0 may be extracted to obtain custom-character {right arrow over (X)}.sub.ω.sub.0.

[0107] For example the measured parametric EM field custom-character {right arrow over (X)}.sub.ω.sub.0 per frequency of interest ω.sub.0 may be obtained by filtering the measured samples {right arrow over (x)}.sub.t=h.sub.ESig.sub.t+{right arrow over (q)}.sub.t according to the frequency of interest ω.sub.0 to obtain {right arrow over (x)}.sub.t(ω.sub.0) and taking the Eigen vector of the covariance matrix Q=Σ.sub.t {right arrow over (x(ω.sub.0))}.sub.t{right arrow over (x(ω.sub.0))}.sub.t.sup.H whose corresponding to the maximal Eigen value λ

[0108] ({right arrow over (X)}).sub.ω.sub.0=EigenVector(Q)|.sub.λ max In this regard it should be noted that generally the time dependency of the measurement is cancelled out in the covariance matrix Q (the covariance matrix is time independent).

[0109] Alternatively, or additionally, as will also be appreciated from the below detailed description and accompanying drawings, in some implementations the reference data REF is spanned in the time domain {δ}. Namely REF={DS.sub.i, δ}. In this case the reference parametric EM field data {right arrow over (A)}.sub.i in each data set is also specific to a certain temporal wavefront's form, characterized here by generalized parameter δ, which presents a certain temporal response function (e.g. impulse response function) of the vector sensor to an EM wavefront of that specific certain temporal wavefront's form δ propagating with the DOA and/or SOP of the corresponding reference steering vector parameters {right arrow over (S)}.sub.i of the respective data set. For example, the reference data REF may include a plurality of data sets in which the reference parametric EM field data {right arrow over (A)}.sub.i presents the impulse responses of the vector sensor 10 to different pulse shaped EM waveforms where the parameter δ associated with each data set characterizes the temporal widths of the pules. E.g., in a dataset DS.sub.i, δ may be the Full Width at Half Maximum (FWHM) parameter of the corresponding EM waveform, and the corresponding reference parametric EM field data {right arrow over (A)}.sub.i is the impulse response of the vector sensor 10 to such a pulsed waveform arriving with the respective properties (DOA and/or SOP) of the steering vector parameters {right arrow over (S)}.sub.i of the dataset DS.sub.i. In this case the reference parametric EM field data indicative of the temporal response of vector sensor 10 to waveform with temporal parameter δ is a function of time {right arrow over (A)}.sub.i.fwdarw.{right arrow over (A)}.sub.i, δ(t). Thus, in this case the reference data REF is generally given by REF={DS.sub.i, δ}={[{right arrow over (A)}.sub.i, δ(t), {right arrow over (S)}.sub.i]}. It should be understood that in cases where the reference data REF is spanned in the time domain {δ}, the measured parametric EM field data custom-character {right arrow over (X)} processed by the preprocessor 110 is maintained as a function of time {right arrow over (X)}(t). Thus in this case, for instance, the samples of the parametric EM field components x.sup.t=[E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.t measured by the vector sensor 10 at certain one or more time instances t may be used to present the developments/functions of the parametric EM field components in time: ({right arrow over (X)})(t)=[E.sub.x(t), E.sub.y(t), E.sub.z(t), H.sub.x(t), H.sub.y(t), H.sub.z(t)]. To this end, the preprocessing may be carried out for one time frame, or more generally in case the received waveform is repetitive (e.g. self-correlative in time) it may be carried out for a plurality of time frames, whereby per each time frame, a detection process is carried out to identify whether the measured parametric EM field component x.sup.t includes a measured signal of interest. Then the relevant time frames (e.g. including the signals of interest), in case of a plurality of time frames, may be respectively filtered (e.g. time aligned) and possibly summed to obtain a nominal measured parametric EM field data custom-character {right arrow over (X)}(t). Otherwise, in case of using a signal time frame, the nominal measured parametric EM field data {right arrow over (X)}(t) may be similar to the parametric EM field components x.sup.t obtained during the time frames. Nonetheless it should be noted that such time alignment may also be performed implicitly during the comparison/matching between the reference data and the measured parametric electric field—e.g. by calculating the correlation function between the reference parametric EM field data {right arrow over (A)}.sub.i, δ(t+Δt) (being the temporal response of vector sensor 10) of one or more data sets, and the measured parametric EM field data custom-character {right arrow over (X)}(t) thereby finding the maximal correlation between them for any plausible time lag Δt . The reference data REF stored in the repository 120 according to various embodiments of the present invention is described in further detail with reference to FIG. 2C below and with respect to the operation 220 of method 200.

[0110] System 100 additionally includes a comparison module 130 configured and operable for comparing/processing the measured parametric EM field data custom-character {right arrow over (X)} and the parametric EM field reference data {right arrow over (A)}.sub.i of one or more of the reference data sets {DS.sub.i} and determining a matching score between the measured parametric EM field data {right arrow over (X)} and each of the one or more of the reference data sets {DS.sub.i} being compared. The comparison module 130 methodology implemented here for finding one matching data set DS.sub.i or a plurality of matching data sets {DS.sub.i} (if any one or more matching data sets exist), may be different from embodiment to embodiment. The comparison module 130 may include a reference data provider 131 (also referred to herein interchangeably as data set estimator/looper module 131) configured and operable for connecting to the reference data repository 120 and estimating/looping to sequentially retrieve one or more relevant data set for which matching score is to be determined; a matching score calculator 132 connectable to the preprocessor 110 for receiving therefrom the measured parametric EM field data custom-character {right arrow over (X)} and to the reference data provider 131 (data set looper module 131) for receiving the parametric EM field reference data {right arrow over (A)}.sub.i of a relevant data set, and adapted for determining a matching score BF.sub.i between them; and a match analyzer 136 connectable to a threshold provider 134 for receiving therefrom a certain threshold condition TH, and determining, based on the respective matching scores, one or more data sets whose steering vectors' parameters are candidates “Candidates {right arrow over (S)}.sub.i” for being representatives of the steering parameters (DOA and/or SOP) of the measured parametric EM field data custom-character {right arrow over (X)}.

[0111] The relevant data sets to be possibly considered/looped by the dataset estimator/looper 131, may include all the data sets, or only the subset of data sets pertaining to the frequency of interest or to a certain temporal waveform shape(s)), and possibly also to predetermined/a priori known parameters of the measured EM field/waveform.

[0112] As indicated above, the reference parametric EM field data {right arrow over (A)}.sub.i in each data set may be represented as follows, where F(θ.sub.i, ϕ.sub.i) represents the DOA dependency of the data sets and {right arrow over (g)}(γ.sub.i, τ.sub.i) represents the polarization dependency:

[00009] $\overset{.fwdarw.}{A_{i}} = [\begin{matrix} {\hat{E}}_{x} \\ {\hat{E}}_{y} \\ {\hat{E}}_{z} \\ {\hat{H}}_{x} \\ {\hat{H}}_{y} \\ {\hat{H}}_{s} \end{matrix}] = F (θ_{i}, ϕ_{i}) \overset{.fwdarw.}{g} (γ_{i}, τ_{i})$

[0113] Therefore, as the dependency on the DOA and on the polarization is separated, the reference data may not necessarily include 4 dimensional data spanning the entire DOA and SOP space (θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i), but may instead include separate data, separately spanning the DOA space and the polarization space. For example in the reference data, the DOA space (θ.sub.i, ϕ.sub.i), may be represented by a 2D lookup table of F or an analytic representation/model of F, as described above, per two orthogonal polarizations. As for the polarization space (γ.sub.i, τ.sub.i) this may be represented by the function {right arrow over (g)}(γ.sub.i, τ.sub.i) described above which presents the linear combination of the two polarizations represented by F. In this way, the amount of reference data required may be reduced as compared to cases where the reference data includes a table presenting the full 4D space (θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i), Consequently, in case the reference data is presented/stored in the reduced way above, the dataset estimator/looper 131 may be configured and operable for determining the data sets per each steering vector parameters (θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i) based on the reference data given/stored for F(θ.sub.i, ϕ.sub.i) and the reference data/analytic function {right arrow over (g)}(γ.sub.i, τ.sub.i): {right arrow over (A)}.sub.i=F(θ.sub.i, ϕ.sub.i){right arrow over (g)}(γ.sub.i, τ.sub.i).

[0114] In this connection, while looping or estimating the datasets {DS.sub.i}, the estimator/looper 131 may not necessarily loop/recur over all the possibilities in the 4D space (θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i) in brute force. In order to provide efficient processing/searching over the reference data, an initial starting point in the 4D space (θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i) can be estimated (e.g. analytically for example by obtaining the rough DOA based on Poynting vector calculations from the measurements custom-character {right arrow over (X)} and/or optionally also estimating the polarization, and/or based on prior knowledge (e.g. from other sources) regarding the properties of the expected signal, such as rough DOA or type of polarization (linear, circular)).

[0115] In this connection, it should be noted that according to some embodiments of the present invention, the estimator/looper 131 actually carries out only a two dimensional search (loop) over the relevant data sets (e.g. those being stored in the reference data memory or constructed based on the reference data F(θ.sub.i, ϕ.sub.i), {right arrow over (g)}(γ.sub.i, τ.sub.i) as described above). The 2D search/loop may be carried out only over the DOA space in case where the parametric EM field data {right arrow over (A)}.sub.i presented/indicated in the reference datasets are all normalized to a certain similar value {right arrow over (A)}.sub.(θ, ϕ, γ, τ).sup.H{right arrow over (A)}.sub.(θ, ϕ, γ, τ)=const. In such cases one can first estimate θ, ϕ by finding the {right arrow over (F)}.sub.(θ, ϕ) that maximizes the largest Eigenvalue of the matrix {right arrow over (F)}.sub.(θ, ϕ).sup.HQ{right arrow over (F)}.sub.(θ, ϕ), and the matching Eigenvector is {right arrow over (g)}(γ, τ), where Q is the covariance matrix of the measurements custom-character {right arrow over (X)}.

[0116] Thus, in some embodiments, the estimator/looper 131 is configured and operable for estimating and looping over all of the 4D space parameters (θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i) of the steering vectors' parameters {right arrow over (S)}.sub.i in the datasets until the best/highest matching score(s) is found by the Matching Score Calculator. Alternatively or additionally, in some implementations the estimator/looper 131 is configured and operable for estimating and looping over a subset of the datasets which are in the vicinity of a certain a priori determined starting point in the 4D space parameters (θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i) of the steering vectors' parameters {right arrow over (S)}.sub.i, thus saving processing time. Yet alternatively or additionally, in some implementations where parametric EM field data {right arrow over (A)}.sub.i in the reference data are all normalized to a certain similar value, the estimator/looper 131 may be adapted to carry out only a 2D search over the reference data sets.

[0117] As said above, methodology implemented by the comparison module 130 for finding one matching data set DS.sub.i (candidate) or a plurality of matching data sets {DS.sub.i} (candidates), may be different from embodiment to embodiment. For instance, in some embodiments the looper module 131 may loop over all of the relevant data sets in the reference data, retrieve each relevant data set from the reference data repository 120, such that comparison module 130 determines the matching score between measured parametric EM field data custom-character {right arrow over (X)} and the parametric EM field reference data {right arrow over (A)}.sub.i of each of the relevant data sets, and the match analyzer 136 selects the one or more datasets for which the matching score is maximal and satisfies a certain threshold condition that is given by the threshold provider 134, as candidates for representing the measured EM field custom-character {right arrow over (X)}. In some embodiments the match analyzer 136 selects only one candidate data set whose matching score is the highest and satisfies the threshold condition. In some embodiments several candidate data sets satisfying the threshold condition may be selected. In the latter case resolving which of these candidates actually represents the parametric EM field data custom-character {right arrow over (X)} may be resolved by other means (e.g. by location systems and/or spectral analysis systems and/or beam-forming/phased array systems receiving the DOAs and/or SOPs of the plurality of candidate data sets and independently determining their match to signals independently sensed by such systems). In another methodology, the looper module 131 loops over the relevant data sets (e.g. one by one) until the match analyzer 136 finds a candidate data set for which the matching score between the parametric EM field reference data {right arrow over (A)}.sub.i and the measured parametric EM field data custom-character {right arrow over (X)} satisfies the threshold condition. Thus, the match analyzer 136 outputs candidate data indicative of one or more candidate datasets, or one or more steering vectors' parameters associated with such one or more candidate datasets, for which the threshold condition is satisfied, or the candidate data indicates that no candidate datasets are found. In this regard it should be understood that in some implementations in case a priori knowledge about the expected steering parameters of the measured parametric EM field data custom-character {right arrow over (X)} is known (e.g. rough DOA and/or SOP properties), then the match analyzer 136 may file out from the candidate datasets, those candidates whose steering parameters mismatch prior knowledge about the expected steering parameters.

[0118] To this end, system 100 may optionally also include an interference/steering estimator 140 connectable to the comparison module 130 (e.g. to the match analyzer 136) for receiving the candidate data. Interference/Steering Estimator 140 may include an Interference Estimator module 142 that processes the candidate data “Candidates {right arrow over (S)}.sub.i” and in case the candidate data does not include any candidate steering vectors' parameters/data sets, the Interference Estimator module 142 determines that the measured parametric EM field data custom-character {right arrow over (X)}, corresponds to the sensing of the interfered electromagnetic field by the vector sensors (e.g. an electromagnetic field suffering from multipath artifacts of other harsh interference conditions).

[0119] In this regard, it is noted that often interference and/or multipath effects are sporadic (namely, they are not temporally and/or spatially stable) and may exist at one instant/period of time in the EM field at the location of the vector sensor as a result of multiple interfering wavefronts, while at the next moment in time, or in a close by location, interference/multipath effects may not exist. Therefore in some embodiments the system 100 (e.g. Interference/Steering Estimator 140) includes a measurement repeater module 143 that is adapted to receive data indicative of the detected interference conditions from the interference estimator module 142, and optionally in case such conditions are detected, operate the system 100 for repeating the operation for an additional one or more time periods in an attempt to determine the direction of arrival (DOA) and possibly state of polarization (SOP) of the electromagnetic wavefront at these additional time periods (in case no interference conditions will be present at least one of these periods).

[0120] In this regard, referring to FIG. 3A, a signal/waveform from a signal source 1 propagating to the vector sensor 10 directly WF1, and indirectly WF2 via a scattering/reflecting object 2, is shown. In cases where the scattering object 2 is relative proximate to the vector sensor 10 and far from the signal source 1 (as shown in FIG. 3A), the direct signal/waveform WF1 and the reflected/scattered signal/waveform WF2 arrive at the vector sensor at substantially different DOAs, whereby the angular difference between their DOAs, ΔΘ; Δφ, may be larger than the angular resolution (beam lobe widths) of the vector sensor 10. In this case the matching scores of the waveform sensed by the vector sensor may be low, and no candidate data set may be found to match the sensed/measured EM waveform, as its multipath represents sufficiently distinct DOAs. Accordingly, in this case no candidate data sets may be found and the interference estimator module 142 may conclude/estimate that interference exists.

[0121] In case the candidate data includes one or more candidates, the interference estimator 140 determines that the measured parametric EM field data custom-character {right arrow over (X)} is highly likely to represent a sensing/measurement of a single wavefront by the vector sensor 10 and thus the candidate data may be passed for estimation of the DOA and/or SOP parameters of the single waveform by the DOA/SOP estimator module 144 of the interference/steering estimator 140.

[0122] Optionally, in some embodiments of the present invention, particularly in cases where the angular resolution of the vector sensor 10 is relatively low (namely its lobe width ΔΘ; Δφ is relatively wide), there may exist one or more candidate datasets whose matching score is relatively high (e.g. above the threshold) even in case of multipath conditions. This may be for instance the case in the example illustrated in FIG. 3B, in which the signal/waveform from the signal source 1 propagates to the vector sensor 10 directly WF1, and indirectly WF2 via a scattering/reflecting object 2 that resides in relative proximity to the signal source 1 and relatively far from the vector sensor 10 so that the direct signal/waveform WF1 and the reflected/scattered signal/waveform WF2 arrive at the vector sensor at relatively similar DOAs, whereby the angular difference between their DOAs, is smaller than the angular resolution ΔΘ; Δφ (beam lobe widths) of the vector sensor 10. In this case, candidate datasets, whose matching score is relatively high, may be found even in multipath conditions, simply because the angular resolution of the vector sensor 10 is low. To this end, the measured parametric EM field <X> resulting from the sum of the wavefronts of the multipath waveforms, WF1 and WF2, as sensed by the vector sensor 10, can be estimated/matched to candidate data set having polarization different than those of the interfering waveforms WF1 and WF2 (for example whereby WF1 and WF2 may have different linear polarizations), the measured parametric EM field <X> may be indicative of circular/elliptical polarization. Thus in some embodiments of the present invention the system (e.g. the interference estimator 140) also includes an optional filtration module 141, also referred to herein below as state of polarization (SOP) filter module 141, configured and operable for applying further filtration to the candidate datasets based on prior data on the polarization properties of the signal source 1. The optional SOP filter module 141, may be configured and operable carry out operation 241 of method 200, so as to obtain data about the properties of the signal source, particular indicative of its polarization properties—whether it is linear or circular etc., and in case the steering vector parameters of one or more of the candidate data sets are associated with different polarization properties (different from the source), these datasets are removed from the “list” of candidate datasets. Also here, in case no candidate data-sets remain, the SOP filter module 141 determines that the measured parametric EM field <X> pertains to multipath conditions. In this case, optionally in some embodiments (as shown in operation 243), the operations of method 200 may be repeated for an additional one or more times (e.g. until a measured parametric EM field data custom-character {right arrow over (X)} not associated with interference conditions, is obtained).

[0123] Thus in case after the optional operation of module 141, the candidate data includes one or more candidates, it may be determined that the measured parametric EM field data custom-character {right arrow over (X)} is highly likely to represent a sensing/measurement of a single wavefront by the vector sensor 10. Optionally, the candidate data may be passed for estimation of the DOA and/or SOP parameters of the single waveform by the DOA/SOP estimator module 144 of the interference/steering estimator 140.

[0124] In turn, in case/embodiment(s) where the candidate data is indicative of only a single candidate data set, or its respective steering vector parameters, the interference/steering estimator 140 outputs that steering vector parameters as representing the DOA/SOP of the measured parametric EM field data custom-character {right arrow over (X)}. Otherwise, in case/embodiment(s) where the candidate data may be indicative of a plurality of candidate data sets, or their respective plurality of steering vectors' parameters, the interference/steering estimator 140 may either output the respective plurality of steering vectors' parameters as possible representations of the DOA/SOP of the measured parametric EM field data custom-character {right arrow over (X)}, or it may connect to additional signal processing systems (e.g. external beam former resolver 150) which may utilize additional data/measurements of the electromagnetic field that had been measured by the vector sensor to further resolve, from the plurality of candidate data sets, a single one that is most likely representative of the parametric EM field data custom-character {right arrow over (X)} DOA/SOP of the measured parametric EM field data {right arrow over (X)}. This resolving may be based on the spatial discrimination of the electromagnetic field propagation direction that may be obtained by a phased array system 150 under the assumption (verified by the interference estimator 142) that the EM field is not interfered and represents a single wavefront. In this connection it is noted that generally in cases where multiple candidates exist, the filtering performed by the system 100, in high probability, filters out candidates whose DOAs are associated with similar angular orientations. Accordingly, conventional phased array-based beam forming systems may generally have the sufficient angular resolution to resolve between the different candidate steering vectors' parameters that are obtained by system 100.

[0125] Referring now together to FIGS. 2B to 2D, a method 200, which may be implemented by the system 100, for determining whether an EM field sensed by the vector sensor is the product of interfering wavefronts (e.g. resulting from multipath or other interference conditions), and in case not, estimating the DOA/SOP of the EM field by the vector sensor, will now be explained in more detail. Method 200 includes the following sub-methods/operations:

[0126] Operation 210, which is illustrated specifically in FIG. 2B includes the provision/measurement of the measured parametric EM field data custom-character {right arrow over (X)} indicative of measured vector components of electric and magnetic fields as those are measured at a certain location and at/during one or more time instances/frames;

[0127] Sub-method/Operation 220, which is illustrated in FIG. 2C, includes the provision/computation/measurement of the reference data REF. The reference data REF includes a plurality of reference data sets REF={DS.sub.i}, each data set DS.sub.i includes a reference steering vector parameters {right arrow over (S)}.sub.i indicative of at least a certain respective direction of arrival, {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i], and the corresponding reference parametric EM field data {right arrow over (A)}.sub.i=[Ê.sub.x.sup.i Ê.sub.y.sup.i Ê.sub.z.sup.i Ĥ.sub.x.sup.i Ĥ.sub.y.sup.i Ĥ.sub.z.sup.i] including components of electric and magnetic fields of an EM wavefront (indexed i) propagating with DOA and possibly SOP corresponding to the reference steering vector parameters {right arrow over (S)}.sub.i. As indicated above, the reference parametric EM field data {right arrow over (A)}.sub.i may be considered as representing the response (e.g. output signals) of the vector sensor 10 outputted in response to an EM wavefront propagating with DOA and/or SOP corresponding to the reference steering vector parameters {right arrow over (S)}.sub.i. It is understood that the reference parametric EM field data {right arrow over (A)}.sub.i may actually represent the expected output signals of the vector sensor 10, after the latter, in case needed, are possibly aligned and/or transformed and/or scaled to a certain six dimensional space of the vector space of representing the three orthogonal components of the electric field and the three orthogonal components of the magnetic field (e.g. such alignment and/or transformation and/or scaling may be needed for instance in case the electric and/or magnetic antenna elements of the vector sensor are not aligned with respect to Cartesian coordinates, or the output signals (e.g. from their feeding ports) are not scaled). Thus, REF={DS.sub.i}={[{right arrow over (A)}.sub.i, {right arrow over (S)}.sub.i]}.

[0128] Sub-method/operation 230, which is illustrated in FIG. 2D, includes determining matching score(s) BF.sub.i between the measured parametric EM field data custom-character {right arrow over (X)} and the reference parametric EM field data {right arrow over (A)}.sub.i of one or more of the reference data sets {DS.sub.i}, and, based on the matching scores, determining whether the steering vectors' parameters {right arrow over (S)}.sub.i of one or more data sets DS.sub.i are candidates for being indicative of the DOA and/or SOP parameters of the measured parametric EM field data custom-character {right arrow over (X)}. The matching score may be obtained by computing the inner product (e.g. generalized inner product or simple inner product) and/or by computing the correlation between the measured parametric EM field data {right arrow over (X)} and the reference parametric EM field data.

[0129] Then in sub-method/operation 240, which is also illustrated specifically in FIG. 2D, the candidate data sets DS.sub.i are processed to determine whether the measured parametric EM field data custom-character {right arrow over (X)} represents interference conditions, such as multipath, or otherwise determine/estimate the DOA and/or SOP parameters of the measured parametric EM field data {right arrow over (X)}.

[0130] Turning now more specifically to operation 210, the following should be noted:

[0131] Operation 212 includes obtain the parametric EM field data custom-character {right arrow over (X)} and may optionally be obtained by measuring/sampling the vectorial components {right arrow over (x)}.sup.t=[E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.t of the electric and magnetic fields at one or more time instances t. As clarified in 214, in some embodiments the parametric EM field data custom-character {right arrow over (X)} to be measured/obtained in 210 should represent the vector components of electric and magnetic fields as those are measured by a vector sensor (e.g. of a certain type) that is positioned at a certain location. Accordingly in 212 the lower case {right arrow over (x)}.sup.t represents a single set of samples of the electric and magnetic field components obtained at particular time point t: {right arrow over (x)}.sup.t=[E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.t. Generally/typically a plurality of such sets of samples are acquired/obtained from the vector sensor during one or more time frames (indexed frm) including several sampling time points (e.g. t.sub.1, t.sub.2, t.sub.3, . . . ) at which a set of samples {right arrow over (x)}.sup.t of the electric and magnetic field is measured/sampled by the vector sensor.

[0132] Thus, for each time frame frm, the electric and magnetic fields are measured/obtained as a function of time (e.g. in digital/analog representation) as follows: {right arrow over (x)}.sub.frm(t)=[E.sub.x(t), E.sub.y(t), E.sub.z(t), H.sub.x(t), H.sub.y(t), H.sub.z(t)]. This can be performed for one or more time instances (e.g. for one or more time frames of certain durations extending about the one or more time instances.

[0133] Then, optionally, preprocessing operation 216 may be performed on the measured/sampled vector components of the electric and magnetic fields. The type of preprocessing, carried out at this stage, depends on which embodiment/implementation of the present invention is performed, namely whether the technique of the present invention is implemented in the frequency domain, or in the time domain.

[0134] Indeed, in the frequency domain, implementations/embodiments of the present invention, optional operation 216.2 may be carried out in order to extract the parts of the measured electric and magnetic fields pertaining to some particular frequency of interest ω.sub.0, from the electric and magnetic field components {right arrow over (X)}.sub.frm(t) that are measured as a function of time during the one or more time frames frm. As will be appreciated by those versed in the art considering the technique disclosed herein, operating in the frequency domain may be conveniently used in particular when the DOA and/or SOP of a certain relatively narrow band wavefront are to be determined. In this case, applying operation 216.2 provides for filtering out the components of the sensed/measured EM field which are not in the frequency of interest ω.sub.0, and thus determining whether harsh interference/multipath conditions exist only for that part of the EM field which oscillates at the frequency of interest ω.sub.0.

[0135] Thus, in this case operation 216.2, which is carried out, includes providing frequency of interest ω.sub.0 for which it is to be determined whether there exist harsh interference conditions exist or not, and in the latter case (where no or not harsh interference conditions exist), the DOA and/or SOP of the wavefront in that frequency may be estimated. 216.2 includes determining, per each time frame frm measurements {right arrow over (X)}.sub.frm(t), the amplitudes and phases of the vector components of the measured electric and magnetic fields oscillating at the frequency of interest ω.sub.0—namely: {right arrow over (X)}.sub.frm(ω.sub.0). This may be achieved for instance by converting the parametric EM field data {right arrow over (X)}.sub.frm(t) of the time frame to frequency representation {right arrow over (X)}.sub.frm(ω) (e.g. using Fourier Transform or similar transform or otherwise any other suitable narrow band filtration about ω.sub.0) and extracting therefrom the amplitude and phase {right arrow over (X)}.sub.frm={right arrow over (X)}.sub.frm(ω.sub.0) of the frequency of interest ω.sub.0 (whereby here {right arrow over (X)}.sub.frm(ω.sub.0) is a generalized vector of complex numbers representing both the amplitude and the phase for each of the EM field components [E.sub.x, E.sub.y, E.sub.z, H.sub.x, H.sub.y, H.sub.z,].sup.ω0.

[0136] Then, optionally, in case the parametric EM field data {right arrow over (X)}.sub.frm(t) is obtained for a plurality of time frames frm, an accurate estimation of the parametric EM field in the frequency of interest may be obtained by carrying out operation 218 to average over {right arrow over (X)}.sub.frm(ω.sub.0) that is obtained for the plurality of time frames: custom-character {right arrow over (X)}=Average({right arrow over (X)}.sub.frm). This may be, for example performed in the frequency domain implementation/embodiment by implementing operation 218.2 as follows: {right arrow over (X)}≡{right arrow over (X)}|.sub.ω.sub.0={right arrow over (X)}.sub.frm(ω.sub.0) custom-character .sub.frm.

[0137] Alternatively, or additionally, in time domain implementations/embodiments of the present invention, optional operation 216.6 may optionally be carried out. It should be noted that time domain implementations of the present invention allow analysis of the interference conditions at which a broad band signal (such as a narrow temporal pulse) is measured/sensed by the vector sensor, and in case no harsh interference conditions are found, estimating the DOA and/or SOP of such broadband signals. Actually, in this case, typically there may be no particular advantage for measuring the plurality of time frames frm of the EM field, but a single time frame long enough to include the temporal signal/waveform transmitted by the emitter may suffice in order to determine accurately whether a temporal signal/waveform sensed by the vector sensor (i.e. the measured parametric EM field) was subjected to interference/multipath conditions. Thus, in case of only a single time frame, operation 216.6 may be skipped/obviated, as well as the operation 218 (e.g. operation 218.4 may be implemented according to which the measured parametric EM field custom-character {right arrow over (X)}≡{right arrow over (X)}.sub.frm(t) is similar to the measurements obtained during the time frame).

[0138] However, for example in case the emitter of the temporal signal/waveform repeatedly transmits the similar waveform shape, in such cases sensing/measuring the parametric EM field, {right arrow over (X)}.sub.frm(t) of the signal received by the vector sensor during a plurality of time frames, at which this similar waveform's shape is received, may further be used to improve the accuracy of the determination of whether or not the received signal suffers from interference/multipath effects or not, and/or for estimation of the DOA/SOP of the received signal. However, in such cases (namely when operating in the time domain and also measuring the parametric EM field {right arrow over (X)}.sub.frm(t) during a plurality of time frames), operation 216.6 may be carried out prior to operation 218, in order to respectively filter/temporarily align the parametric EM fields {right arrow over (X)}.sub.frm(t) measured in the plurality of time frames, so that the waveform shapes of interest included therein would be aligned. To this end, in case of more than one time frame, operation 216.6 may include applying temporal alignment to the parametric EM fields {right arrow over (X)}.sub.frm(t) obtained during the different time frames with respect to one another. Presuming the signal of interest (e.g. being the signal from the emitter) is predominant in the received parametric EM fields {right arrow over (X)}.sub.frm(t), such alignment may be determined by calculating the correlation between parametric EM fields {right arrow over (X)}.sub.frm(t) of two or more frames to determine a relative time delay(s) between them at which the correlation is maximal, and then shifting one or more of the parametric EM fields {right arrow over (X)}.sub.frm(t) of the different frames according to the found delays so as to maximize the overlap/correlation between them. In a particular example this may be achieved by selecting (e.g. arbitrarily) one of the time frames to serve as a reference frame rfrm and applying time delay/shift to each of the other time frames, such that its parametric EM field data {right arrow over (X)}.sub.frm(t) has maximal correlation with the parametric EM field data {right arrow over (X)}.sub.rfrm(t) of the selected reference frame rfrm.

[0139] Thus, once aligned, operation 218 may be carried out in order to average the aligned parametric EM fields {right arrow over (X)}.sub.frm(t) and obtain the average parametric EM field as a function of time. This may be for example implemented by carrying out the operation 218.6 as follows: custom-character {right arrow over (X)}≡{right arrow over (X)}(t)={right arrow over (X)}.sub.frm(t).sub.frm.

[0140] In optional operation 219 electric field components and the magnetic field components of the parametric EM field custom-character {right arrow over (X)} obtained from the one or more time frame measurements may be further normalized (e.g. such that: {right arrow over (X)}=[Ê.sub.x Ê.sub.y Ê.sub.z Ĥ.sub.x Ĥ.sub.y Ĥ.sub.z]). It should be noted that this operation is indeed optional as the reference parametric EM field data {right arrow over (A)}.sub.i in the different data sets are generally also normalized, and therefore, in this case, accurate results would be obtained, even if the measured parametric EM field custom-character {right arrow over (X)} is not normalized.

[0141] Turning now to FIG. 2C, operation 220 of providing the reference data REF, is now described in more detail. As indicated above, the reference data REF includes a plurality of reference data sets REF={DS.sub.i} whereby each data set DS.sub.i includes a reference steering vector parameters {right arrow over (S)}.sub.i indicative of at least a certain respective direction of arrival, {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i], and the corresponding reference parametric EM field data {right arrow over (A)}.sub.i=[Ê.sub.x.sup.i Ê.sub.y.sup.i Ê.sub.z.sup.i Ĥ.sub.x.sup.i Ĥ.sub.y.sup.i Ĥ.sub.z.sup.i].

[0142] Optional operation 221 indicates that the reference parametric EM field data {right arrow over (A)}.sub.i of each reference data set DS.sub.i being provided in the reference data REF, is preferably normalized. More specifically, in some implementations of the present invention, the reference parametric EM field data {right arrow over (A)} presents the normalized electric field vector components and the normalized magnetic field vector components as would be measured by a vector sensor (e.g. of a certain particular type) in case an EM wavefront (indexed i) propagating with parameters of the reference steering vector parameters {right arrow over (S)}.sub.i impinges the vector sensor. This provides that the results of matching between the measured parametric EM field custom-character {right arrow over (X)} and different reference parametric EM field data {right arrow over (A)} are co-scaled (are similarly scaled) and thus can be compared to accurately determine the reference parametric EM field data {right arrow over (A)} (and corresponding data set DS.sub.i or steering vector parameters {right arrow over (S)}.sub.i) which best matches the measured parametric EM field custom-character {right arrow over (X)}.

[0143] As indicated in optional operation 222 the reference steering vector parameters {right arrow over (S)}.sub.i in each data set DS.sub.i being provided, may also include reference State of Polarization (SOP) parameters, γ.sub.i, τ.sub.i indicative of a certain respective SOP of the corresponding reference parametric EM field data {right arrow over (A)}.sub.i in the respective data set. Thus the complete steering vector parameters per each reference parametric EM field data {right arrow over (A)}.sub.i may be given as {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i] (where are tilt and phase lag of cardinal polarization components and i spans a plurality of pairs of DOAs & SOPs of interest).

[0144] In this regard, it should be noted that embodiments of the present invention in which the steering vector parameters {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i] of the data set DS.sub.i includes the State of Polarization (SOP) parameters, γ.sub.i, τ.sub.i, (in addition to the DOA parameters θ.sub.i, ϕ.sub.i) were found by the inventors of the present invention to provide better and more accurate/reliable results for estimating/determining whether an EM field sensed/measured by the vector sensor is subjected to (or is the product of) harsh interference/multipath conditions or not. This is because in case the data sets pertain to complete steering vectors' parameters {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i], the respective reference parametric EM field data {right arrow over (A)}.sub.i in the data sets are also specific to the complete steering vectors' parameters of the respective data set DS.sub.i. Accordingly, in case a match is found between the measured parametric EM field data custom-character {right arrow over (X)} and a parametric EM field data {right arrow over (A)}.sub.i that is specific to the complete steering vector parameters including both DOA and SOP parameters, such a match provides much more reliable indication (higher probability) that the measured parametric EM field custom-character {right arrow over (X)} has the steering vector parameters properties {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i], as compared to the case where the parametric EM field {right arrow over (A)}.sub.i is not specific to the complete steering vector parameters, but is actually indicative of the nominal parametric EM field {right arrow over (A)}.sub.i components for signal of any polarization arriving from certain DOA indicated by the steering vector parameters.

[0145] Optional operation 223 includes providing the reference parametric EM field data {right arrow over (A)}.sub.i of the respective reference data sets DS.sub.i. As indicated above, the reference parametric EM field data {right arrow over (A)}.sub.i in each reference data set DS.sub.i, represents a generalized vector of the sensing of an EM field waveform having the corresponding steering parameters {right arrow over (S)}.sub.i by the vector sensor. It should be noted that in some embodiments, the reference parametric EM field data {right arrow over (A)}.sub.i of the respective reference data sets {DS.sub.i} is a priori determined (e.g. measured/calculated) and stored in the reference data repository 120. Alternatively or additionally, in some embodiments the reference parametric EM field data {right arrow over (A)}.sub.i for the data sets {DS.sub.i}, or some of them, may be computed in real time, during operation of the system 100 (e.g. based on an analytical model of the vector sensor 10 and optionally of objects in its vicinity), and the respective steering vectors' parameters of the data sets {DS.sub.i}.

[0146] In this connection, it should be understood that the reference data REF may be stored in the form of a lookup table and/or a model by which including all the datasets of interest can be determined/obtained (e.g. based on their steering parameters θ, ϕ, γ, τ). In this case the data set looper module 131 actually retrieves the relevant data sets (e.g. those pertaining to steering vectors' parameters of interest) from the lookup table in the repository 120.

[0147] Additionally or alternatively, the stored reference data may include a stored model (e.g. indicative of the response/distortion matrix Di characterizing the response of the vector sensor 10 to the EM field of different steering vector properties (of different DOAs and/or different SOPs), and possibly also characterizing the distortions applied to such EM fields by the presence of objects (e.g. metallic objects) in the vicinity of the vector sensor 10 (in case accurate processing should be obtained for vector sensor 10 that is positioned at a predetermined location or on a predetermined platform introducing various distortions/warping to electric fields propagating nearby). In this case, the data set looper module 131 may actually be configured and operable to compute the reference parametric EM field data {right arrow over (A)}.sub.i of the relevant data sets (e.g. those pertaining to steering vectors' parameters of interest) based on an analytic model including the response/distortion matrix D.sub.i of the vector sensor (e.g. matrix of size 6×6).

[0148] The analytic model may be for example given as:

[00010] $\overset{.fwdarw.}{A_{i}} = [{\hat{E}}_{x}^{i} {\hat{E}}_{y}^{i} {\hat{E}}_{z}^{i} {\hat{H}}_{x}^{i} {\hat{H}}_{y}^{i} {\hat{H}}_{z}^{i}] = [D_{i}] [{DOA}_{i}] [{SOP}_{i}] = \overset{\overset{? (θ, ϕ)}{︷}}{[D_{i}] [\begin{matrix} \cos θcos ϕ & - \sin ϕ \\ \cos θsin ϕ & \cos ϕ \\ - \sin θ & 0 \\ \sin ϕ & \cos θ \cos ϕ \\ - \cos ϕ & \cos θ \sin ϕ \\ 0 & - \sin θ \end{matrix}]} \overset{\overset{\overline{g} (γ, τ)}{︷}}{[\begin{matrix} \sin τ e^{j γ} \\ \cos τ \end{matrix}]}$ $? indicates text missing or illegible when filed$

where [D.sub.i] is the response/distortion matrix (e.g. which may be determined empirically or estimated analytically), θ, ϕ, γ, τ are the DOA and SOP parameters of the respective steering vector parameters {right arrow over (S)}.sub.i=[θ.sub.i, ϕ.sub.i, γ.sub.i, τ.sub.i] for which the respective reference parametric EM field data {right arrow over (A)}.sub.i is computed, [DOA.sub.i] is a generalized rotation matrix associated with the DOA parameters of the reference parametric EM field θ.sub.i, ϕ.sub.i; and [SOP.sub.i] is a matrix representing the SOP (γ.sub.i, τ.sub.i) of the reference parametric EM field {right arrow over (A)}.sub.i.

[0149] To this end, as indicated in optional operation 223.1, the parametric EM field data {right arrow over (A)}.sub.i may be computed/provided, in real time (e.g. during the operation of system 100) or a priori (e.g. during the manufacturing/calibration of the system 100), based on the above indicated analytic model {right arrow over (A)}.sub.i=[D.sub.i][DOA.sub.i][SOP.sub.i]. The response/distortion matrix D.sub.i characterizing the response and optionally EM distortions that would be applied to the sensing of EM field waveform with steering vector parameters {right arrow over (S)}.sub.i may be empirically measured or analytically estimated, and may be particular to a certain type/model of a vector sensor and/or optionally to a certain location/platform at which it is positioned, or in some cases may suit various vector sensors. In an ideal case, where the response function of the vector sensor 10 is substantially flat (in the spectral/temporal regime of interest and there are no significant distortions that are applied to the EM field by the vector sensor or objects in its vicinity, the response/distortion matrix D.sub.i is unity. The [DOA.sub.i] and [SOP.sub.i] matrices are generally mathematical entities typically independent of the vector sensor's type or location.

[0150] Alternatively, or additionally, as indicated in optional operation 223.3, the parametric EM field data {right arrow over (A)}.sub.i may be provided/computed based on results of empirical measurements performed (e.g. a priori) on the vector sensor 10 (e.g. when the vector sensor is placed at a designated location/platform of interest). For instance the response of the vector sensor to EM fields characterized by various steering vectors' parameters {{right arrow over (S)}.sub.i} and optionally the EM distortions that would be applied to the sensing of such EM fields by the vector sensor itself and/or by objects in its vicinity, may be determined/measured empirically by “bombarding” the vector sensor (e.g. sequentially) by the different EM fields characterized by steering vector parameters {right arrow over (S)}.sub.i and measuring its response (its output signals—e.g. from the feeding ports of its antenna elements) to the different EM fields. In this connection, the output signals of the vector sensor may represent the parametric EM field data {right arrow over (A)}.sub.i. Moreover the response/distortion matrices D.sub.i of the vector sensor 10 to the different EM fields may also be determined empirically noting that a bombarding EM field characterized by steering vector parameters {right arrow over (S)}.sub.i may be represented by the above described matrices [DOA.sub.i][SOP.sub.i]. Accordingly the response/distortion matrices D.sub.i for a given bombarding EM field with steering vector parameters {right arrow over (S)}.sub.i and given resulting/measured output {right arrow over (A)}.sub.i of the vector sensor may be determined empirically by solving the equation: {right arrow over (A)}.sub.i=[D.sub.i][DOA.sub.i][SOP.sub.i].

[0151] It should be noted that according to optional operation 224.2, the reference parametric EM field data {right arrow over (A)}.sub.i in the reference data sets is provided/computed such that the reference data sets {DS.sub.i} are spanned over the frequency domain. In other words, the reference data is such that one or more reference data sets {DS.sub.i} characterize the response of the vector sensor to a certain respective frequency {ω}. In this case it should be noted that the dependency of the vector sensor's response to EM fields of different frequencies may be actually represented as the dependence of the response/distortion matrices [D.sub.i].fwdarw.[D.sub.i, ω] on the frequency ω. Accordingly, by determining/providing (e.g. empirically in the manner indicated in 223.3, or analytically in the manner indicated in 223.1) the response/distortion matrices [D.sub.i, ω] of the vector sensor for EM fields of different frequencies and different steering vectors' parameters, the corresponding reference parametric EM field data {right arrow over (A)}.sub.i.fwdarw.{right arrow over (A)}.sub.i, ω for different frequencies may be obtained (computed/provided). Analytic determination of the reference data in this case, as in 223.1, may be achieved by computing a model of the vector sensor and/or objects in its vicinity, which affect the propagation of EM fields of different frequencies and different steering vectors' parameters. Empiric determination of the reference data, as in 223.3, may be achieved by bombarding the vector sensor with EM fields of different frequencies and measuring its output parametric EM field data {right arrow over (A)}.sub.i, ω in response to each frequency.

[0152] Alternatively, or additionally, it should be noted that according to optional operation 224.6, the reference parametric EM field data {right arrow over (A)}.sub.i in the reference data sets may be provided/computed such that the reference data sets {DS.sub.i} are spanned over the time domain. In other words, the reference data is such that one or more reference data sets {DS.sub.i} characterize the response of the vector sensor to wavefronts of certain respective temporal shapes (the different temporal shapes are indicated here by the generalized parameter(s) δ characterizing a certain temporal waveform shape. For instance, the waveform shapes considered may be different pulses characterized by different widths (where in this particular example δ may be just a single parameter indicative of the FWHM characteristic of the pulse). In this case the dependency of the vector sensor's response to EM fields of different temporal shapes may be actually represented as the dependence of the response/distortion matrices [D.sub.i].fwdarw.[D.sub.i, δ] on the temporal shape parameter(s) δ. Accordingly, by determining/providing (e.g. empirically in the manner indicated in 223.3, or analytically in the manner indicated in 223.1) the response/distortion matrices [D.sub.i, δ] of the vector sensor to EM fields of different temporal shapes δ and different steering vectors' parameters, the corresponding reference parametric EM field data {right arrow over (A)}.sub.i.fwdarw.{right arrow over (A)}.sub.i, δ(t) for different temporal shapes may be obtained (computed/provided). In analytic determination, as in 223.1, this may be achieved by computing a model of the vector sensor and/or objects in its vicinity and how those affect the propagation of EM fields with different temporal shapes arriving with different steering vector parameters properties). In empiric determination, as in 223.3, this may be achieved by bombarding the vector sensor with EM fields of different temporal shapes and different directions and/or polarization and measuring its output parametric EM field data {right arrow over (A)}.sub.i, δ in response to each temporal shape and steering vector properties (indicating the direction and/or polarization). Moreover, it should be noted that in case where spanning the reference data in the time domain as described above, the reference parametric EM field that is acquired, should generally include data indicative of the development of the reference parametric EM field in time {right arrow over (A)}.sub.i, δ(t) (namely it should be provided/computed/stored as a function of time or as a time series indicative of the development of vector sensor's output in time in response to an applied EM field of corresponding temporal shape δ). In this connection it should be noted that analytic determination of the reference data, as in 223.1, may be achieved by computing a model of the vector sensor and/or objects in its vicinity and how those affect the propagation of EM fields with different temporal shapes δ and different steering vectors' parameters. Empiric determination of the reference data, as in 223.3, may be achieved by bombarding the vector sensor with EM fields of different temporal shapes and measuring its output parametric EM field data {right arrow over (A)}.sub.i, δ in response to temporal shape as a function of time.

[0153] Turning now to FIG. 2D, this illustrates a flow chart of the operations 230 and 240 which are further carried out in method 200 in order to determine/estimate whether the EM field sensed by the vector sensor suffers from interference effects, and, in case it does not, determine/estimate the DOA and possibly SOP of the sensed EM field.

[0154] In operation 230 the matching score BF.sub.i between the measured parametric EM field data custom-character {right arrow over (X)} and the reference parametric EM field data {right arrow over (A)}.sub.i of one or more of the reference data sets {DS.sub.i} are determined and one or more data sets whose matching scores BF.sub.i satisfy a certain threshold condition TH are selected as candidates for representing the sensed/measured EM field custom-character {right arrow over (X)}. In various embodiments of the present invention operation 230 may include the following:

[0155] Carrying out operation 231 for providing one or more reference datasets for calculating their matching scores with the measured parametric EM field data custom-character {right arrow over (X)}. This operation may be performed by the dataset looper/estimator module 131 in the manner described above.

[0156] Carrying out operation 232 for computing the matching scores between the one or more reference datasets and the measured parametric EM field data custom-character {right arrow over (X)}. This operation may be performed by the matching score calculator module 132 in the manner described above.

[0157] As indicated above, in some embodiments the matching score BF.sub.i may be obtained by computing the inner product (generalized or simple inner product) and/or the correlation (in case A and X are functions of time) between the measured parametric EM field data custom-character {right arrow over (X)} and the reference parametric EM field data {right arrow over (A)}.sub.i.

[0158] For instance, generally the matching score a simple inner product, such as that exemplified in optional operation 232.2 may be computed as follows:

[0159] BF.sub.i(θ, ϕ, γ, τ)={right arrow over (A.sub.t)}.sup.H.Math. custom-character {right arrow over (X)} where here {right arrow over (A.sub.t)}.sup.H is the Hermitian (i.e. transpose conjugate) of the reference parametric EM field data vector {right arrow over (A)}.sub.i (e.g. for real values, the Hermitian conjugate being just the transposed).

[0160] Alternatively or additionally, in some embodiments a generalized inner product is computed, as shown in 232.4 by taking computing the covariance matrix custom-character {right arrow over (X)}{right arrow over (X)}.sup.H of the measured parametric EM field data {right arrow over (X)} and multiplying the reference parametric EM field data vector {right arrow over (A)}.sub.i from one side and by its Hermitian conjugate {right arrow over (A.sub.t)}.sup.H from the other side. For example, as follows:

[00011] ${BF}_{i} (θ, ϕ, γ, τ) = \frac{\overset{.fwdarw. H}{A_{i}} (\overset{.fwdarw.}{X}) {(\overset{.fwdarw.}{X})}^{H} \overset{.fwdarw.}{A_{i}}}{\overset{.fwdarw.}{A} \overset{H .fwdarw.}{A}}$

[0161] In this example also the normalization coefficient is

[00012] $\frac{1}{\overset{.fwdarw.}{A} \overset{H .fwdarw.}{A}}$

multiplied, however this may not be needed in case the reference parametric EM field data vector {right arrow over (A)}.sub.i in the reference data are already normalized.

[0162] To this end, for example an estimate of the steering parameters θ, ϕ, γ, τ of the measured parametric EM field custom-character {right arrow over (X)} based on the reference data may be obtained by the following computation:

[00013] $\hat{θ}, \hat{ϕ}, \hat{γ}, \hat{τ} = \underset{θ, ϕ, γ, τ}{argmax} \frac{{\overset{.fwdarw.}{A}}_{(θ, ϕ, γ, τ)}^{H} Q {\overset{.fwdarw.}{A}}_{(θ, ϕ, γ, τ)}}{{\overset{.fwdarw.}{A}}_{(θ, ϕ, γ, τ)}^{H} {\overset{.fwdarw.}{A}}_{(θ, ϕ, γ, τ)}}$

where Q is the covariance matrix of the measurements (indicated above) and {right arrow over (A)}.sub.(θ, ϕ, γ, τ) is the reference parametric EM field data vector {right arrow over (A)}.sub.i whereby instead of the index i, it is presented as spanned over the four dimensions of the steering parameters θ, ϕ, γ, τ, and {right arrow over (A)}.sub.(θ, ϕ, γ, τ).sup.H is its Hermitian.

[0163] Yet, alternatively or additionally, in some embodiments, particularly when operating in the temporal domain in which both the measured parametric EM field data custom-character {right arrow over (X)} and multiplying the reference parametric EM field data vector {right arrow over (A)}.sub.i are functions of time, matching score BF.sub.i may be computed as the maximal correlation (for any Δt) between those temporal vector functions, as shown in 232.6, and as follows:

BF.sub.i(θ, ϕ, γ, τ)=Corr({right arrow over (A.sub.i)}.sup.H(t), custom-character {right arrow over (X)}(t+Δt))

[0164] It should be understood that the present invention is not limited by the specific technique by which the matching score is computed and that generally, as would be readily appreciated by those versed in the art, other computations/equations may be used to determine a match/matching score between the measured parametric EM field data custom-character {right arrow over (X)} and the reference parametric EM field data vector {right arrow over (A)}.sub.i.

[0165] It should be noted that in cases/embodiments where both the measured parametric EM field data custom-character {right arrow over (X)} and the reference parametric EM field data vector {right arrow over (A)}.sub.i are normalized (e.g. according to operations 219 and 221 respectively), the maximal possible value of the matching score BF.sub.i is unity. The matching score BF.sub.i reaches its maximum (e.g. unity) when the steering vector parameters {right arrow over (S)}.sub.i corresponding to the reference parametric EM field data vector {right arrow over (A)}.sub.i is of the same as the direction and polarization of the measured parametric EM field data custom-character {right arrow over (X)}.

[0166] Operation 234 may be carried out by determining a threshold condition TH for determining candidate datasets based on the matching scores computed in 232. The threshold condition TH may be a predetermined (a priori set condition), or dynamically determined condition (e.g. which may be set to allow only up to a limited number of candidate datasets). Generally, the threshold condition TH may be set in order to optimize a ratio between false positive candidates (false alarms—these are candidate data sets whose matching scores are above the threshold, but yet whose steering vector parameters {right arrow over (S)}.sub.i does not actually represent the DOA and/or SOP of the measured EM field) and false negatives (that is miss-identification as candidate of data set whose steering vector parameters {right arrow over (S)}.sub.i does actually represent the DOA and/or SOP of the measured EM field). The considerations of selection of the threshold condition TH are discussed in more detail below with reference to FIG. 4.

[0167] In operation 236 (e.g. which may be performed by the match analyzer module 136) the candidate data-sets (or steering vectors' parameters) are determined. In case the matching score BF.sub.i between the measured parametric EM field data custom-character {right arrow over (X)} and the reference parametric EM field data {right arrow over (A)}.sub.i of a certain reference data set DS.sub.i complies with the threshold condition TH, then it is estimated that the measured parametric EM field data {right arrow over (X)} may correspond to an EM wavefront propagating with the properties (DOA and/or SOP) of the steering vector parameters {right arrow over (S)}.sub.i of the certain reference data set. Accordingly, such a certain reference data set DS.sub.i is selected as candidate.

[0168] In operation 240, which may be performed for example by the—interference/steering estimator module 140 of system 100 includes determining whether the measured parametric EM field data custom-character {right arrow over (X)} represents the existence/sensing of interference conditions, and in case it does not, determining/estimating the steering parameters (DOA and/or SOP) of the measured parametric EM field {right arrow over (X)}.

[0169] As shown in 242, in case that no candidate data-sets are identified in 236, the measured parametric EM field data custom-character {right arrow over (X)} may be considered not to be matching to any one single reference parametric EM field {right arrow over (A)}.sub.i and thus its DOA and/or SOP cannot be associated with any particular steering vector parameters. This is the case where none of the matching scores BF.sub.i between the measured parametric EM field data custom-character {right arrow over (X)} and any of the reference parametric EM field data {right arrow over (A)}.sub.i of reference data sets { DS.sub.i} comply with the certain threshold condition. In this case, it is determined (e.g. under the assumption that the reference data includes all the relevant reference parametric EM fields {{right arrow over (A)}.sub.i} of interest) that the measured parametric EM field data custom-character {right arrow over (X)} cannot be associated with a certain single wavefront (at least none which are included/approximated in the reference data), and therefore it is determined that the measured parametric EM field data {right arrow over (X)} was obtained due to the sensing/measurement of harsh interference/multipath conditions by the vector sensor 10. Therefore, in this case, the result of method 200 might be that the measured parametric EM field data custom-character {right arrow over (X)} was sensed under the interfering conditions (e.g. it represents interfering EM fields). However, optionally in some embodiments (as shown in operation 243) in such cases the operations of method 200 are repeated for an additional one or more times (e.g. until a measured parametric EM field data custom-character {right arrow over (X)} not associated with interference conditions, is obtained).

[0170] As shown in 242, in case that no candidate data-sets are identified in 236 (or the candidate(s) identified in 236 were subsequently further filtered out in operation 241 described below—e.g. by the SOP filter 141), the measured parametric EM field data custom-character {right arrow over (X)} may be considered not to be matching to any one single reference parametric EM field for which particular steering vector parameters exists. This is the case where none of the matching scores BF.sub.i between the measured parametric EM field data {right arrow over (X)} custom-character and any of the reference parametric EM field data {right arrow over (A)}.sub.i of reference data sets {DS.sub.i} comply with the certain threshold condition, determining that the measured parametric EM field data {right arrow over (X)} cannot be associated with a certain single wavefront due to the presence of harsh interference/multipath conditions. Therefore, in this case, the result of method 200 is that the measured parametric EM field data custom-character {right arrow over (X)} was sensed under the interfering conditions (e.g. it represents interfering EM fields). In this case, optionally operation 243 is further carried out to repeat the method 200 for an additional one or more times, during which interference conditions might not exist.

[0171] Optionally, in some embodiments, operation 242 includes carrying out the optional operation 241 to further filter the candidate data sets obtained in 236, based on prior knowledge about properties of the expected waveform. This may be performed for instance as explained above with reference to module 141.

[0172] Optional operation 244, may be carried out in case single candidate datasets DS.sub.i is found/obtained in operation 236 or after the optional SOP filtration of operation 241. In this case the single candidate dataset DS.sub.i is that for which {right arrow over (A)}.sub.i matches custom-character {right arrow over (X)}). Accordingly, the steering vector parameters {right arrow over (S)}.sub.i of the single candidate datasets DS.sub.i dataset is determined to provide an accurate estimate of the DOA and/or SOP of the measured parametric EM field data {right arrow over (X)}.

[0173] Optional operation 246, may be carried out in case multiple candidate datasets are found/obtained in operation 236 or after the optional SOP filtration of operation 241. In this case, the reference parametric EM field data {right arrow over (A)}.sub.i of all candidates seem to match the measured parametric EM field data custom-character {right arrow over (X)}, and thus the steering vector parameters {right arrow over (S)}.sub.i of any of those candidate datasets may be representative of the DOA/SOP properties of the measured parametric EM field {right arrow over (X)}. Accordingly, there may be several options for determining the DOA and/or SOP of the measured parametric EM field data custom-character {right arrow over (X)}, as follows:

[0174] In some embodiments/implementations, the method 200 may be repeated for an additional one or more times (e.g. as in operation 243) until only a single candidate data set is found.

[0175] In some embodiments/implementations, the threshold condition/bar TH is raised (e.g. operations 234 and 236 described above are repeated for the candidate data sets, but only with stricter threshold conditions). In this way, one or more of the candidate data sets may be eliminated to remain with a single candidate dataset, which is then handled according to operation 244.

[0176] In some embodiments/implementations, the correspondence between one of the candidate datasets and the measured parametric EM field data custom-character {right arrow over (X)} is resolved by additional systems such as a phased array antenna system and/or beamforming system having high spatial resolution. Indeed, in cases where there are several candidate datasets, often their DOAs are substantially different. Therefore, beam forming arrays/phased-array systems having high spatial resolution may be used to resolve which of the candidates actually match the measured EM field.

[0177] Turning back to FIG. 2C, it should be noted that the sub-method 220 which is used for providing the reference data as described above, may in some embodiments not only be incorporated in method 200, but also may be implemented as a standalone method 220 for computing reference data for use in determining the DOA and SOP properties of an EM field sensed by a vector sensor. The reference data computed in this way may then be stored in the reference data repository, 120, and may be later on used/provided in the scope of operation method 200.

[0178] Reference is made to FIG. 4, which is a graphical illustration of the statistics (commutative distribution function) of matching scores obtained according to the technique of the present invention for the vector-sensor measurements <X> of waveforms of two cases: graph G1 shows the commutative distribution function of the maximal values of the matching scores calculated according to the technique of the present invention for the measured parametric EM fields <X> of respective waveforms in a first population of waveforms each associated with two or more wavefronts; and graph G2 shows the commutative distribution function of the maximal values of the matching scores calculated according to the technique of the present invention for the measured parametric EM fields <X> of respective waveforms in a second population of waveforms, each associated with a single wavefront. In other words, graph G1 shows the statistics of the matching scores obtained from a population of different signal measurements that are taken under obstructed conditions (e.g. multipath/interference conditions which may result from obstructions such as buildings and/or other reflecting/scattering objects), while graph G2 pertains to the statistics of the highest value matching scores obtained for a population of different signal measurements taken in unobstructed conditions (namely when no multipath/interference conditions exist).

[0179] Generally, the matching score reaches its maximum (e.g. unity) when the steering vector parameters of a certain reference data set is of the same direction and polarization i.e. {right arrow over (A)}={right arrow over (A)}(θ.sub.o, ϕ.sub.o, γ.sub.o, τ.sub.o) as the measured EM field, while the measured EM field is associated with a single wavefront. Indeed, in real life scenarios, the maximal matching score may degrade from unity. Nevertheless, actual values were found to be distributed in close vicinity to unity for a single wavefront propagating to the sensor, as shown in the dashed graph G2 in FIG. 4. However, as shown in graph G1, when observing the commutative distribution function, which resulted in obstructed vicinity (interference/multipath conditions), namely for sensed/measured waveforms having multiple wavefronts, the probability of low maximal values of the matching scores is significantly increased, and the probability to obtain high matching scores in this case becomes substantially low or negligible. To this end, the graphical illustration in FIG. 4 illustrates statistically that the technique of the present invention provides reliable discrimination between waveforms which are sensed/measured under obstructed/interference conditions and unobstructed/non-interference conditions. Or, in other words, it facilitates discrimination between the sensing of waveforms associated with multipath and non-multipath conditions. It should be noted that the threshold TH set/provided in operation 234 (e.g. by the threshold provider module 134) may be adjusted accordingly (e.g. based on such statistics or similar statistics), in order to obtain such discrimination between multipath conditions and non-multipath conditions, with desired reliability (namely to optimize the desired probability for false positive and false negative errors in such discrimination (also known as type I and type II statistical errors or alpha and beta errors). Considering the hypothesis that the sensed/measured waveform is subjected to multipath/interference (namely has multiple wavefronts), the probability of false positive error (type I) represents the probability that the highest matching score computed according to the technique for the present invention was below the threshold, thereby indicating that the sensed waveform has multiple wavefronts, while in practice it had only a single wavefront (no-multipath). In the same way, the probability of false negative error (type II) represents the probability that the highest matching score computed according to the technique of the present invention was above the threshold, thereby indicating that the sensed waveform has a single wavefront, while in practice it had multiple wavefronts (multipaths). Thus, the threshold conditions provided in 234, may be adjusted in accordance with the particular needs of a certain task so as to adjust/balance the probabilities for type I and type II errors in accordance with the statistics of the technique of the present invention (e.g. such as that exemplified in the non-limiting example of FIG. 4, or other statistics, which may be further adjusted/computed as per particular embodiments/implementations of the present invention).

[0180] For example, considering the particular example of FIG. 4, it may be realized that in this case assigning the threshold value to be 0.998 more than 80% of the obstructed/multipath population is filtered out (i.e. the type II error is 20%) while more than 50% of the unobstructed/single-path population (i.e. the type I error is 50%) is still valid. Accordingly, the threshold value TH can be tuned a priori or in real time in order for example to optimize the false alarm rate (type II error, obstructed population is considered to be unobstructed).

[0181] In this regard, it should be understood that further improvements in the above values of statistical errors are achieved according to the technique of the present invention by carrying out one or more of the following as described above: [0182] 1. Repeating method 200 several times (for several time frames/periods) to confirm that the measured signal indeed matches a certain particular data-set/steering vector parameters (DOA and SOP) with improved probability and reduced statistical errors; [0183] 2. Further filtration of allegeable matching candidate data sets, based on prior knowledge about the expected (e.g. roughly estimated) properties (e.g. DOA and/or SOP) of the waveform which is sensed/received (as for example described in relation to operation 241 above and performed by the SOP filter 141 described above). [0184] 3. A priori filtration of the data-sets, for which the matching score will be calculated in the first place based on prior knowledge about the expected (e.g. roughly estimated) properties (e.g. DOA and/or SOP) of the waveform which is sensed/received. Accordingly, the matching scores in this case will be only calculated for those data sets whose steering vectors' parameters match the expected properties of the sensed waveform, thereby reducing the probability of obtaining high matching scores for irrelevant data sets, and, in turn improving (reducing) the statistical errors indicated above.

SYSTEM AND METHOD FOR CHARACTERIZING PROPERTIES OF EM SIGNALS

Inventors

Cpc classification

Classification Explorer

G01S3/043

PHYSICS

Classification Explorer

G01S3/14

PHYSICS

Classification Explorer

G01R29/0892

PHYSICS

International classification

Classification Explorer

G01R29/08

PHYSICS

Classification Explorer

G01S3/04

PHYSICS

Abstract

Claims

Description