Method for predicting the channel between a transmitter/receiver and a connected vehicle

Abstract

Predicting a channel between a transceiver and a connected vehicle having a “main antenna”, dedicated to exchanges of payload data with the transceiver, and a “predictor antenna”, placed in front of the main antenna to predict the radio channel dealt with by the main antenna when reaching the current position of the predictor antenna. The method includes: selecting, using an estimate of the vehicle's speed and acceleration, for a multiplet of channel samples measured at the main antenna, a multiplet of channel samples measured at the predictor antenna, each sample of the predictor antenna being selected to correspond to a sample of the main antenna subsequently measured at the same position; calculating a criterion associating multiplets of samples measured at the main and predictor antennas; and selecting samples of the predictor antenna using a speed/acceleration pair optimizing the criterion, to predict the channel between the transceiver and main antenna.

Claims

1. A channel prediction method performed by a network entity, the method comprising: predicting a radio channel between a transceiver and a connected vehicle, said vehicle comprising at least one antenna, known as “main antenna”, dedicated to exchanges of payload data with said transceiver, and at least one other antenna, known as “predictor antenna”, placed in front of the main antenna, an estimate of the radio channel at a current position of the predictor antenna making it possible to predict the radio channel that the main antenna will be dealing with when the main antenna reaches the current position of the predictor antenna, the predicting comprising: selecting, on the basis of an initial estimate of a speed and of an acceleration of the vehicle, for a given multiplet of channel samples measured at the main antenna, a multiplet of channel samples measured at the predictor antenna, each sample of the predictor antenna being selected such that the sample corresponds to a given sample of the main antenna subsequently measured at the same position, assuming that the vehicle has moved at said estimated speed and said estimated acceleration, calculating a criterion associating said multiplet of samples measured at the main antenna with said multiplet of samples measured at the predictor antenna, determining a speed/acceleration pair that optimizes said criterion, and, in order to predict the channel between the transceiver and the main antenna, selecting samples of the predictor antenna by way of said determined speed/acceleration pair.

2. The channel prediction method as claimed in claim 1, wherein the method comprises the network entity, for at least one pair of multiplets comprising a multiplet
h.sub.m=[h.sub.m(0), . . . ,h.sub.m(N−1)] of N samples h.sub.m(k) measured at the main antenna, and a multiplet
h.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1)] of N samples h.sub.p(k) measured at the predictor antenna, the following steps: choosing an integer X>1, and subdividing each multiplet h.sub.m and h.sub.p into X sub-multiplets each comprising N.sub.x=N/X samples, so as to obtain the pairs of sub-multiplets
h.sub.m,i=[h.sub.m((i−1)N.sub.x), . . . ,h.sub.m(iN.sub.x−1)]
and
h.sub.p,i=[h.sub.p((i−1)N.sub.x), . . . ,h.sub.p(iN.sub.x−1)], where i={1, . . . , X}, subtracting k.sub.0 first samples from each sub-multiplet h.sub.m,i and k.sub.0 last samples from each sub-multiplet h.sub.p,i, where k.sub.0 is the number of samples measured during the time it takes for the main antenna to reach the position of the predictor antenna, on the basis of an initial conjecture of the speed and of the acceleration of the vehicle, estimating an average speed for each pair of sub-multiplets by maximizing a correlation between the sub-multiplets of each pair, obtaining the multiplet of speeds v={v.sub.i=1, . . . , v.sub.i=X} representing the estimated speeds for the samples numbered
k={N.sub.x/2,3N.sub.x/2, . . . ,(2X−1)N.sub.x/2}, and obtaining, from said multiplet of speeds v, an estimated acceleration a.sub.c,est and an estimated speed v.sub.0,est of the vehicle, where v.sub.0,est is equal to the estimated speed of the vehicle at the instant of the measurement of the sample k=0.

3. The channel prediction method as claimed in claim 1, wherein the method comprises, for at least one pair of multiplets comprising a multiplet
h.sub.m=[h.sub.m(0), . . . ,h.sub.m(N−1)] of N samples h.sub.m(k) measured at the main antenna, and a multiplet
h.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1)] of N samples h.sub.p(k) measured at the predictor antenna, the network entity performing the following steps: subtracting k.sub.0 first samples measured at the main antenna and k.sub.0 last samples measured at the predictor antenna, where k.sub.0 is the number of samples measured during the time it takes for the main antenna to reach the position of the predictor antenna, on the basis of an initial conjecture of the speed and of the acceleration of the vehicle, so as to obtain the multiplets of samples
h′.sub.m=[h.sub.m(k.sub.0), . . . ,h.sub.m(N−1)]
and
h′.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1−k.sub.0)], choosing an integer X>1, and subdividing each multiplet h′.sub.m and h′.sub.p into X sub-multiplets of N.sub.x=(N−k.sub.0)/X samples, so as to obtain the pairs of sub-multiplets
h′.sub.m,i=[h.sub.m(k.sub.0+(i−1)N.sub.x), . . . ,h.sub.m(k.sub.0+iN.sub.x−1)]
and
h′.sub.p,i=[h.sub.p((i−1)N.sub.x), . . . ,h.sub.p(iN.sub.x−1)], where i={1, . . . , X}, estimating, for each such pair of sub-multiplets, an average speed by maximizing a correlation between the sub-multiplets of each pair, so as to obtain a series of estimated speeds v={v.sub.i=1, . . . , v.sub.i=x} representing the estimated speeds for samples numbered
k={(k.sub.0+N.sub.x)/2,(k.sub.0+3N.sub.x)/2, . . . ,(k.sub.0+(2X−1)N.sub.x)/2}, obtaining, from these speed estimates v, an estimated acceleration a.sub.c=a.sub.c,est and an estimated speed v.sub.0,est of the vehicle, and deducing therefrom the estimated speed
v.sub.0=v.sub.avg,init+v.sub.0,est of the vehicle at the instant of the measurement of the sample k=0, where v.sub.avg,init is the average speed conjectured over the time interval during which the N samples are measured.

4. The channel prediction method as claimed in claim 1, wherein the method comprises the network entity performing the following steps: calculating a time interval Δ.sup.+(t) representing the time it will take for the main antenna to reach the location where the predictor antenna is located at the instant t, calculating a time interval Δ.sup.−(t) representing the time elapsed since the predictor antenna was located at the location where the main antenna is located at the instant t, deducing therefrom shifts Δ.sub.k.sup.+(k) and Δ.sub.k.sup.−(k) in terms of numbers of samples k corresponding respectively to Δ.sup.+(t) and Δ.sup.−(t), removing, in a multiplet of N samples measured at the main antenna, samples with a number less than k.sub.0=Δ.sub.k.sup.+(0), for various conjectured pairs (v.sub.0, a.sub.c), determining, by way of the values of Δ.sup.+(t), Δ.sup.−(t), Δ.sub.k.sup.+(k) and Δ.sub.k.sup.−(k), which samples measured at the main antenna should be associated with which samples measured at the predictor antenna, and calculating an average correlation between a multiplet of samples measured at the main antenna and the associated multiplet of samples measured at the predictor antenna, and determining which pair (v.sub.0, a.sub.c) maximizes said average correlation.

5. The channel prediction method as claimed in claim 1, further comprising: the network entity receiving the channel samples measured at the main antenna; and the network entity receiving the channel samples measured at the predictor antenna.

6. The channel prediction method as claimed in claim 5, further comprising: the connected vehicle measuring the channel samples at the main antenna; and the connected vehicle measuring the channel samples at the predictor antenna.

7. The channel prediction method as claimed in claim 5, further comprising: receiving an antenna signal from the main antenna; and pre-equalizing the antenna signal using the prediction of the radio channel.

8. A network entity used in prediction of a radio channel between a transceiver and a connected vehicle, said vehicle comprising at least one antenna, known as “main antenna”, dedicated to exchanges of payload data with said transceiver, and at least one other antenna, known as “predictor antenna”, placed in front of the main antenna, an estimate of the radio channel at a current position of the predictor antenna making it possible to predict the radio channel that the main antenna will be dealing with when the main antenna reaches the current position of the predictor antenna, wherein the network entity comprises: a processor; and a non-transitory computer-readable medium comprising instructions stored thereon or incorporated therein, which when executed by the processor configure the network entity to: select, on the basis of an initial estimate of a speed and of an acceleration of the vehicle, for a given multiplet of channel samples measured at the main antenna, a multiplet of channel samples measured at the predictor antenna, each sample of the predictor antenna being selected such that the sample corresponds to a given sample of the main antenna subsequently measured at the same position, assuming that the vehicle has moved at said estimated speed and said estimated acceleration, calculate a criterion associating said multiplet of samples measured at the main antenna with said multiplet of samples measured at the predictor antenna, determine a speed/acceleration pair that optimizes said criterion, and, in order to predict the radio channel between the transceiver and the main antenna, select samples of the predictor antenna by way of said determined speed/acceleration pair.

9. The network entity as claimed in claim 8, wherein the instructions further configure the network entity to, for at least one pair of multiplets comprising a multiplet
h.sub.m=[h.sub.m(0), . . . ,h.sub.m(N−1)] of N samples h.sub.m(k) measured at the main antenna, and a multiplet
h.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1)] of N samples h.sub.p(k) measured at the predictor antenna: choose an integer X>1, and subdividing each multiplet h.sub.m and h.sub.p into X sub-multiplets each comprising N.sub.x=N/X samples, so as to obtain the pairs of sub-multiplets
h.sub.m,i=[h.sub.m((i−1)N.sub.x), . . . ,h.sub.m(iN.sub.x−1)]
and
h.sub.p,i=[h.sub.p((i−1)N.sub.x), . . . ,h.sub.p(iN.sub.x−1)], where i={1, . . . , X}, subtract k.sub.0 first samples from each sub-multiplet h.sub.m,i, and k.sub.0 last samples from each sub-multiplet h.sub.p,i, where k.sub.0 is the number of samples measured during the time it takes for the main antenna to reach the position of the predictor antenna, on the basis of an initial conjecture of the speed and of the acceleration of the vehicle, estimate an average speed for each pair of sub-multiplets by maximizing a correlation between the sub-multiplets of each pair, obtain the multiplet of speeds v={v.sub.i=1, . . . , v.sub.i=x} representing the estimated speeds for the samples numbered
k={N.sub.x/2,3N.sub.x/2, . . . ,(2X−1)N.sub.x/2}, and obtain, from said multiplet of speeds v, an estimated acceleration a.sub.c,est and an estimated speed v.sub.0,est of the vehicle, where v.sub.0,est is equal to the estimated speed of the vehicle at the instant of the measurement of the sample k=0.

10. The network entity as claimed in claim 8, the instructions further configure the network entity to, for at least one pair of multiplets comprising a multiplet
h.sub.m=[h.sub.m(0), . . . ,h.sub.m(N−1)] of N samples h.sub.m(k) measured at the main antenna, and a multiplet
h.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1)] of N samples h.sub.p (k) measured at the predictor antenna: subtract k.sub.0 first samples measured at the main antenna and k.sub.0 last samples measured at the predictor antenna, where k.sub.0 is the number of samples measured during the time it takes for the main antenna to reach the position of the predictor antenna, on the basis of an initial conjecture of the speed and of the acceleration of the vehicle, so as to obtain the multiplets of samples
h′.sub.m=[h.sub.m(k.sub.0), . . . ,h.sub.m(N−1)]
and
h′.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1−k.sub.0)], choose an integer X>1, and subdividing each multiplet h′.sub.m and h′.sub.p into X sub-multiplets of N.sub.x=(N−k.sub.0)/X samples, so as to obtain the pairs of sub-multiplets
h′.sub.m,i=[h.sub.m(k.sub.0+(i−1)N.sub.x), . . . ,h.sub.m(k.sub.0+iN.sub.x−1)]
and
h′.sub.p,i=[h.sub.p((i−1)N.sub.x), . . . ,h.sub.p(iN.sub.x−1)], where i={1, . . . , X}, estimate, for each such pair of sub-multiplets, an average speed by maximizing a correlation between the sub-multiplets of each pair, so as to obtain a series of estimated speeds v={v.sub.i=1, . . . , v.sub.i=x} representing the estimated speeds for samples numbered
k={(k.sub.0+N.sub.x)/2,(k.sub.0+3N.sub.x)/2, . . . ,(k.sub.0+(2X−1)N.sub.x)/2}, obtain, from these speed estimates v, an estimated acceleration a.sub.c=a.sub.c,est and an estimated speed v.sub.0,est of the vehicle, and deduce therefrom the estimated speed
v.sub.0=v.sub.avg,init+v.sub.0,est of the vehicle at the instant of the measurement of the sample k=0, where v.sub.avg,init is the average speed conjectured over the time interval during which the N samples are measured.

11. The network entity as claimed in claim 8, wherein the instructions further configure the network entity to: calculate a time interval Δ.sup.+(t) representing the time it will take for the main antenna to reach the location where the predictor antenna is located at the instant t, calculate a time interval Δ.sup.−(t) representing the time elapsed since the predictor antenna was located at the location where the main antenna is located at the instant t, deduce therefrom shifts Δ.sub.k.sup.+(k) and Δ.sub.k.sup.−(k) in terms of numbers of samples k corresponding respectively to Δ.sup.+(t) and Δ.sup.−(t), remove, in a multiplet of N samples measured at the main antenna, samples with a number less than k.sub.0=Δ.sub.k.sup.+(0), for various conjectured pairs (v.sub.0, a.sub.c), determine, by way of the values of Δ.sup.+(t), Δ.sup.−(t), Δ.sub.k.sup.+(k) and Δ.sub.k.sup.−(k), which samples measured at the main antenna should be associated with which samples measured at the predictor antenna, and calculate an average correlation between a multiplet of samples measured at the main antenna and the associated multiplet of samples measured at the predictor antenna, and determine which pair (v.sub.0, a.sub.c) maximizes said average correlation.

12. The network entity as claimed in claim 8, wherein the network entity is accommodated in a base station or in an MSC (Mobile Switching Center) of a cellular communication network.

13. A communication network comprising: at least one network entity as claimed in claim 8, and at least one vehicle equipped with at least one main antenna and with at least one predictor antenna.

14. A non-transitory computer-readable medium comprising instructions stored thereon, which when executed by a processor of a network entity configure the network entity to: predict a radio channel between a transceiver and a connected vehicle, said vehicle comprising at least one antenna, known as “main antenna”, dedicated to exchanges of payload data with said transceiver, and at least one other antenna, known as “predictor antenna”, placed in front of the main antenna, an estimate of the radio channel at a current position of the predictor antenna making it possible to predict the radio channel that the main antenna will be dealing with when the main antenna reaches the current position of the predictor antenna, the predicting comprising: selecting, on the basis of an initial estimate of a speed and of an acceleration of the vehicle, for a given multiplet of channel samples measured at the main antenna, a multiplet of channel samples measured at the predictor antenna, each sample of the predictor antenna being selected such that the sample corresponds to a given sample of the main antenna subsequently measured at the same position, assuming that the vehicle has moved at said estimated speed and said estimated acceleration, calculating a criterion associating said multiplet of samples measured at the main antenna with said multiplet of samples measured at the predictor antenna, determining a speed/acceleration pair that optimizes said criterion, and, in order to predict the channel between the transceiver and the main antenna, selecting samples of the predictor antenna by way of said determined speed/acceleration pair.

Description

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

(1) Other aspects and advantages of the invention will become apparent on reading the detailed description below of particular embodiments, which are given by way of nonlimiting example.

(2) These embodiments implement, in the context of a predictor antenna method, the following main steps.

(3) a) By using a time window of channel estimates made at the predictor antenna and the main antenna, estimating the effective speed and the effective acceleration of the vehicle.

(4) b) Once the speed of the vehicle is known as a function of time, it is possible to determine, taking into account the distance d between the two antennas, the preceding instant at which the predictor antenna was located at the current position of the main antenna; it is then possible to select, from the set of channel measurements made in the past at the predictor antenna, the one that was performed at this previous instant.

(5) When a base station then exchanges payload data with the main antenna, the signal transporting these payload data will advantageously be pre-equalized by way of the channel measurement that was performed by the predictor antenna at the position where the main antenna is currently located.

(6) If the acceleration a.sub.c is non-zero, then the time interval Δt involved in equation (1) above is variable over time. To simplify the disclosure, it will be assumed below that the acceleration a.sub.c of the vehicle is constant during the series of measurements under consideration.

(7) A first embodiment will now be described.

(8) The steps of this embodiment are implemented for at least one pair of multiplets comprising a multiplet of N samples
h.sub.m=[h.sub.m(0), . . . ,h.sub.m(N−1)]
measured at the main antenna, and a multiplet of N samples
h.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1)]
measured at the predictor antenna.

(9) In a first step, an integer X>1 is chosen freely, and each multiplet h.sub.m and h.sub.p is subdivided into X sub-multiplets each comprising
N.sub.x=N/X
samples. The pairs of sub-multiplets
h.sub.m,i=[h.sub.m((i−1)N.sub.x), . . . ,h.sub.m(iN.sub.x−1)]
and
h.sub.p,i=[h.sub.p((i−1)N.sub.x), . . . ,h.sub.p(iN.sub.x−1)]
are thus obtained, where i={1, . . . , X}.

(10) In a second step, the k.sub.0 first samples are subtracted from each sub-multiplet h.sub.m,i and the k.sub.0 last samples are subtracted from each sub-multiplet h.sub.p,i (and therefore Xk.sub.0 samples from each multiplet h.sub.m and from each multiplet h.sub.p), where k.sub.0 is the number of samples measured during the time it takes for the main antenna to reach the position of the predictor antenna, on the basis of an initial conjecture of the speed and of the acceleration of the vehicle.

(11) In a third step, the average speed is estimated for each pair of sub-multiplets by maximizing the correlation between the sub-multiplets of this pair.

(12) Finally, in a fourth step, an estimated acceleration a.sub.c,est and an estimated speed v.sub.0,est of the vehicle are obtained, for example by way of a linear regression, from the multiplet of speeds
v={v.sub.i=1, . . . ,v.sub.i=X}
representing the estimated speeds for the samples numbered
k={N.sub.x/2,3N.sub.x/2, . . . ,(2X−1)N.sub.x/2}.

(13) It should be noted that the speed v.sub.0 of the vehicle at the instant of the measurement of the sample k=0 is simply equal to v.sub.0,est.

(14) A second embodiment will now be described.

(15) Just as for the first embodiment, the steps of this embodiment are implemented for at least one pair of multiplets comprising a multiplet of N samples
h.sub.m=[h.sub.m(0), . . . ,h.sub.m(N−1)]
measured at the main antenna, and a multiplet of N samples
h.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1)]
measured at the predictor antenna.

(16) The samples numbered
k={0, . . . ,N−1}
are collected at time intervals T.sub.s (sampling time interval). In a first step, the k.sub.0 first samples measured at the main antenna and the k.sub.0 last samples measured at the predictor antenna, where k.sub.0 is the number of samples measured during the time it takes for the main antenna to reach the position of the predictor antenna, are subtracted on the basis of an initial conjecture of the initial speed v.sub.0,init and of the acceleration a.sub.c,init. k.sub.0 is thus the integer closest to the ratio

(17) $\frac{d}{T_s.Math.v_{\mathrm{avg},\mathrm{init}}},$
where d is the distance between the main antenna and the predictor antenna, and
v.sub.avg,init=v.sub.0,init+a.sub.c,initNT.sub.s/2  (2)
is the average speed conjectured over the time interval during which the N samples are measured.

(18) The values chosen for v.sub.0,init and a.sub.c,init could for example be zero, or be the estimates resulting from the analysis of the previous dataset. The samples at the front main antenna k.sub.0 are not associated with any sample at the predictor antenna at the same position, and therefore cannot be used. The multiplets of samples
h′.sub.m=[h.sub.m(k.sub.0), . . . ,h.sub.m(N−1)]
and
h′.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1−k.sub.0)] are thus obtained,
each comprising (N−k.sub.0) samples.

(19) In a second step, an integer X>1 is chosen freely, and each multiplet h′.sub.m and h′.sub.p is subdivided into X sub-multiplets of
N.sub.x=(N−k.sub.0)/X
samples. The pairs of sub-multiplets
h′.sub.m,i=[h.sub.m(k.sub.0+(i−1)N.sub.x), . . . ,h.sub.m(k.sub.0+iN.sub.x−1)]
and
h′.sub.p,i=[h.sub.p((i−1)N.sub.x), . . . ,h.sub.p(iN.sub.x−1)] are thus obtained,
where i={1, . . . , X}.

(20) In a third step, the average speed is estimated for each such pair of sub-multiplets by maximizing the correlation between the sub-multiplets of this pair. This then gives a multiplet
v={v.sub.i=1, . . . ,v.sub.i=X}
representing the estimated speeds for the samples numbered
k={(k.sub.0+N.sub.x)/2,(k.sub.0+3N.sub.x)/2, . . . ,(k.sub.0+(2X−1)N.sub.x/2}
which correspond to the middle of each sub-multiplet.

(21) Finally, in a fourth step, an estimated acceleration a.sub.c=a.sub.c,est and an estimated speed v.sub.0,est of the vehicle are obtained, for example by way of a linear regression, from these speed estimates v.

(22) The estimated speed
v.sub.0=v.sub.avg,init+v.sub.0,est  (3)
of the vehicle at the instant of the measurement of the sample k=0 is deduced therefrom.

(23) A third embodiment will now be described.

(24) To this end, for each channel sample measured at the main antenna, a channel sample measured at the predictor antenna is first of all chosen on the basis of an initial estimate of the speed and of the acceleration. The sample of the predictor antenna is chosen so as to match a sample measured at the same position as the sample of the main antenna if the vehicle has moved at the estimated speed and the estimated acceleration. A mechanical model translates between the time domain and the spatial domain. The average correlation between the samples of the main antenna and the samples of the predictor antenna is then calculated, and then the speed/acceleration pair that optimizes a correlation coefficient between samples is determined.

(25) The optimum speed/acceleration pair may be determined using a matrix search method by testing all realistic pairs, or using a gradient algorithm that estimates the gradient on the matrix and iteratively finds the value on the matrix that optimizes said given criterion.

(26) In the context of this third embodiment, it is possible more specifically to proceed as follows:

(27) The initial speed v.sub.0 at the start of a series of measurements, and the speed v(t) at an instant t, satisfy
v(t)=v.sub.0+a.sub.ct.  (4)

(28) Since the acceleration is assumed to be non-zero, it is necessary to define two time intervals. The first time interval, denoted Δ.sup.+(t), is defined as the time it will take for the main antenna to reach the location where the predictor antenna is located at the instant t (movement d of the main antenna starting at the instant t). The second time interval, denoted Δ.sup.−(t), is defined as the time elapsed since the predictor antenna was located at the location where the main antenna is located at the instant t (movement d of the predictor antenna finishing at the instant t).

(29) These two time intervals are deduced from quadratic equations derived from equation (4), thereby giving (taking v(t)>0)

(30) $\begin{array}{cc}{\Delta }^{+}\left(t\right)=\frac{\sqrt{{\left(v_0+a_{c}t\right)}^2+2a_{c}d}-v_0-a_{c}t}{a_c}& \left(5\right)\\ \mathrm{and}& \\ {\Delta }^{-}\left(t\right)=\frac{v_0-a_{c}t-\sqrt{{\left(v_0+a_{c}t\right)}^2+2a_{c}d}}{a_c},& \left(6\right)\end{array}$
where v.sub.0 is the speed at the instant t=0 corresponding to the first sample of each series of measurements.

(31) The shifts Δ.sub.k.sup.+(k) and Δ.sub.k.sup.−(k) in terms of numbers of samples k corresponding respectively to Δ.sup.+(t) and Δ.sup.−(t) are deduced therefrom: Δ.sub.k.sup.+(k) is the integer closest to Δ.sup.+(kT.sub.s)/T.sub.s, and Δ.sub.k.sup.−(k) is the integer closest to Δ.sup.− (kT.sub.s)/T.sub.s, where T.sub.s is the sampling interval.

(32) It will be noted that, in a multiplet
h.sub.m=[h.sub.m(0), . . . ,h.sub.m(N−1)]
of N samples h.sub.m(k) measured at the main antenna, samples with a number less than k.sub.0=Δ.sub.k.sup.+(0) are not associated with any sample at the predictor antenna at the current position of the main antenna, and therefore cannot be used, such that the number of usable samples is equal to
N.sub.k=N−k.sub.0.  (7)
Using equations (5), (6) and the above values of the shifts Δ.sub.k.sup.+(k) and Δ.sub.k.sup.−(k), it is determined which samples should be associated for various conjectured pairs (v.sub.0, a.sub.c).

(33) The average correlation ĉ(v.sub.0, a.sub.c) between a multiplet of samples h.sub.m measured at the main antenna and the associated multiplet
h.sub.p=[h.sub.p(0), . . . ,h.sub.p(N−1)]
of N samples h.sub.p(k) measured at the predictor antenna, for time intervals Δ.sup.+(t) and Δ.sup.−(t) that may vary over time, is given by

(34) $\begin{array}{cc}\hat{c}\left(v_0,a_c\right)=\frac{1}{N_k}{.Math.}_{k=k_0}^{N-1}{\hat{c}}_k\left(v_0,a_c\right),& \left(8\right)\\ \mathrm{where}& \\ {\hat{c}}_k\left(v_0,a_c\right)=h_m\left(k{T}_s\right)h_p^{*}\left(\left(k-{\Delta }_k^{-}\left(k\right)\right)T_s\right).& \left(9\right)\end{array}$

(35) The effective value of the pair (v.sub.0, a.sub.c) is finally obtained by determining which pair maximizes this average correlation. Specifically, for any estimate of the speed v.sub.0 and of the acceleration a.sub.c that would not correspond to the effective values, the correlation would be calculated based on measurements taken at different positions; the correlation calculated with erroneous values is less than the correlation calculated on the basis of the effective value of the pair (v.sub.0, a.sub.c).

(36) It will finally be noted that, in the case of modulation of the radio signal using frequency multiplexing, the speed and the acceleration should be the same for all of the sub-carriers; the optimization algorithm should therefore preferably, for better efficiency, include a weighted sum of the correlation over all of the sub-carriers.

(37) The invention may be implemented within an entity, for example a base station or an MSC, of a cellular communication network, by way of software components and/or hardware components.

(38) The software components may be integrated into a conventional computer program for network node management. It is for this reason, as indicated above, that the present invention also relates to a computing system. This computing system includes, as is conventional, a central processing unit that uses signals to control a memory, and also an input unit and an output unit. Moreover, this computing system may be used to execute a computer program that includes instructions for implementing any one of the channel prediction methods according to the invention.

(39) Specifically, another subject of the invention is a computer program able to be downloaded from a communication network and comprising instructions for executing the steps of a channel prediction method according to the invention when it is executed on a computer. This computer program may be stored on a computer-readable medium and may be able to be executed by a microprocessor.

(40) This program may use any programming language, and may be in the form of source code, object code, or intermediate code between source code and object code, such as in a partially compiled form, or in any other desirable form.

(41) Another subject of the invention is an irremovable, or partially or fully removable computer-readable information medium that includes instructions of a computer program as mentioned above.

(42) The information medium may be any entity or device capable of storing the program. For example, the medium may comprise a storage means, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or a magnetic recording means, such as a hard disk, or else a USB stick (“USB flash drive”).

(43) Moreover, the information medium may be a transmissible medium such as an electrical or optical signal, which may be routed via an electrical or optical cable, by radio or by other means. The computer program according to the invention may in particular be downloaded from an Internet network.

(44) As a variant, the information medium may be an integrated circuit in which the program is incorporated, the circuit being designed to execute or to be used in the execution of any one of the channel prediction methods according to the invention.

(45) Although the present disclosure has been described with reference to one or more examples, workers skilled in the art will recognize that changes may be made in form and detail without departing from the scope of the disclosure and/or the appended claims.