WO2009156501A1

WO2009156501A1 - Method of sending by a sender of a data set destined for a receiver

Info

Publication number: WO2009156501A1
Application number: PCT/EP2009/058027
Authority: WO
Inventors: Ghaïs EL ZEIN; Ali Khaleghi
Original assignee: Institut National Des Sciences Appliquees; Centre National De La Recherche Scientifique (Cnrs)
Priority date: 2008-06-27
Filing date: 2009-06-26
Publication date: 2009-12-30
Also published as: FR2933251A1; FR2933251B1

Abstract

The invention relates to a sender (EM), of a data set destined for a receiver. The sender comprises: - means for determining, on the basis of information representative of an existing channel between the sender and the receiver (RE), a first ordered group of coefficients of a first time reversal filter (M0), - means for determining at least one second ordered group of coefficients of at least one second time reversal filter (M1), the values of coefficients of each second group being the first group’s coefficient values offset in a direction, - means (SP) for dividing the data set into at least two data streams, - means for filtering the signals representative of each data stream by a determined filter, each filter being different from each other filter, - means (Ante) for sending the sum of the filtered streams.

Description

A method of transmitting by a transmitter a set of data to a receiver.

The present invention relates to a method of transmission by a transmitter of a set of data to a receiver.

In telecommunication systems based on the transfer of electromagnetic waves by a transmitter, the electromagnetic waves received by the receiver are not identical to those transferred by the transmitter.

Because of the multiple paths associated with the different reflections of an electromagnetic wave on obstacles, when a transmitter sends an electromagnetic wave, the receiver receives a plurality of electromagnetic waves shifted in time. New telecommunication methods take advantage of this plurality of received electromagnetic waves. Among these methods, that of time reversal, or Time Reversal in English terminology, transmitted electromagnetic waves is particularly promising. Classically, in a learning phase of the channel, the transmitter emits an electromagnetic wave, for example in the form of a pulse, the receiver records the electromagnetic pulses received, transfers information representative of the pulses received to the transmitter that calculates a time reversal filter so as to temporally return the transmitted signal.

The representative signals of data to be transmitted are filtered by the time reversal filter and then transmitted to the receiver. The transmitter thus transfers a plurality of electromagnetic pulses so that the receiver receives an electromagnetic pulse. In time-reversing telecommunication systems using Bipolar Pulse Amplitude Modulation (BPAM), the data bits are transmitted according to the following formula: x (t) = U _n C (J) where C (t) represents the time reversal filter, d _n = \ if the transmitted bit has the value 1 and d _n = -1 if the transmitted bit has the value 0.

The time reversal filter C (t) is defined by C {i) = Ah {-i) where h {-i) is the response of the truncated channel, A is the amplification factor to obtain a constant power at the output of the filter. The response of the channel is truncated to reduce the transmission time T _s of a symbol while ensuring that the majority of the signal energy is received. The signal x (t) propagates in the channel according to the same paths as used during the learning of the channel. The signal received by the receiver is represented by the following formula: y (t) = d Ah (-t) * h (t) + n (t) _Λ ~ _,. , ,, ,,. ,,,

¹ or R _hfι (t) represents the autocorrelation of the response

= d _J AR _hh (t) + n (t) of the channel and the truncated channel response and n (t) represents the noise. The rate of time-reversing telecommunication systems using bipolar pulse amplitude modulation is low.

The object of the invention is to solve the drawbacks of the prior art by proposing a method and a device for transmitting electromagnetic signals to a receiver that offers a transmission rate greater than those proposed by the state of the art. the technique.

To this end, according to a first aspect, the invention proposes a method of sending, by a transmitter, a set of data to a receiver, characterized in that the method comprises the steps of: determination, based on information representative of an existing channel between the transmitter and the receiver, of ordered coefficients of a first time reversal filter,

determining, for at least one second time reversal filter, ordered coefficients of the second time reversal filter, the ordered coefficient values of the second time reversal filter being the values of the coefficients of the first time reversal filter shifted in one direction,

- division of the dataset into at least two data streams,

filtering the signals representative of each data stream by a determined time reversal filter, each time reversal filter filtering the signals representative of a stream being different from each other time reversal filter filtering the signals representative of another stream,

- issuing the sum of the filtered flows.

Correlatively, the present invention relates to a transmitter of a set of data to a receiver, characterized in that the transmitter comprises:

means for determining, based on information representative of a channel existing between the transmitter and the receiver, a first ordered group of coefficients of a first time reversal filter,

means for determining at least a second ordered group of coefficients of at least one second time reversal filter, the coefficient values of each second group being the values of the coefficients of the first group shifted in one direction,

means for dividing the data set into at least two data streams; means for filtering the signals representative of each data stream by a determined time reversal filter, each time reversal filter filtering the signals representative of the data stream; a flow being different from each other time reversal filter filtering the signals representative of another flow,

means for transmitting the sum of the filtered flows. Thus, the transmission rate is greater than those proposed by the state of the art.

According to a particular embodiment of the invention, the data set is divided into an even number M of data streams. According to a particular embodiment of the invention, the data set is divided into an odd number M of data streams and a data stream is filtered by the first time reversal filter and each other data stream is filtered by a second time reversal filter. The inventors have noticed that by dividing the data set into an odd number of data streams, the system performance is improved.

According to one particular embodiment of the invention, (MI) / 2 second time reversal filters have coefficients whose values are shifted in a first direction and (M1) / 2 time reversal filters have coefficients whose values are shifted according to a second direction different from the first direction.

Thus, the performances are improved.

According to one particular embodiment of the invention, at least one time reversal filter filtering a data stream has coefficients whose values are the values of the coefficients of another time reversal filter filtering data shifted in one direction and the minus a time reversal filter aa coefficients whose value is zero more than the other time reversal filter.

Thus, by setting coefficients to zero, performance is improved. According to one particular embodiment of the invention, the coefficients whose value is zero more than the other time reversal filter are at the position of the last non-zero coefficients of the coefficients of the other time reversal filter ordered according to the direction. offset.

According to a particular embodiment of the invention, each data stream is multiplied by a scale factor determined from the coefficients of the time reversal filter filtering the data stream.

According to a particular embodiment of the invention, the value of a is determined by selecting a value of a for which a signal-to-average interference ratio is greater than or equal to a given value and for which a variance of the signal-to-interference ratio is less than a maximum variance value of the signal-to-interference ratio.

Thus, the determined offset taking into account the interference, the amount of data transmitted according to the present invention is increased significantly without the signal to interference ratio being degraded too significantly. The characteristics of the invention mentioned above, as well as others, will appear more clearly on reading the following description of an exemplary embodiment, said description being made in relation to the attached drawings, among which: FIG. . 1 represents the architecture of the telecommunication system in which the present invention is implemented; Figs. 2 show examples of values of the coefficients taken by the coefficients of the time reversal filters according to the present invention; FIG. 3 represents an algorithm for determining an offset of coefficient values according to the present invention.

Fig. 1 represents the architecture of the telecommunication system in which the present invention is implemented.

In the telecommunication system according to the present invention, an EM transmitter comprising at least one antenna Ante transfers data to a receiver RE comprising at least one antenna Antr.

The emitter EM emits a test electromagnetic wave such as a pulse through its antenna Ante. The receiver RE measures the response of the channel h (t). The response of the channel h (t) is representative of the multiple paths, propagation delays and attenuations of the transmitted electromagnetic signal. The receiver RE transfers information representative of the response of the channel h (t) to the transmitter EM.

The data set, to be transmitted by the transmitter EM is divided into a plurality M of data flow _LJ5 d _2j to dM by a parallel series converter SP with A /> 1.

Fig. 1 represents an example of an emitter in which M = 3. Each data stream d_i _j, C, r, respectively is filtered by a filter C_ i (t), Co (t), C ₁ (X) of time reversal to obtain signals X _-1 (X), xo ( t), X ₁ (t) of different shapes that depend on the values of the filter coefficients C _-1 (X), C ₀ (t) and C ₁ (O.

At each filter, a normalization of the power A _m is carried out so that the power of the signals at the output of the filters is constant.

A = - ^ _{= ^ =} where 11.11 represents the Frobenius norm operation.

The filtering C _-1 (X) and the normalization of the power Ai are executed by the module MI, the filtering C ₀ (t) and the normalization of the power AQ are executed by the module MO and the filtering Ci (t) and the normalization of the power Ai are executed by the module M1.

The M filtered and normalized flows are summed by an adder add to form a signal X (t). From the summation, the power of the resulting signal X (t) is modified. So as to keep a transmission power P ₀ the same as that of a conventional transmitter in which no flow division is performed, the signal X (t) can be normalized by the β Norm module according to a factor such as n _

If we consider a time reversal filter C (t), it has ordered coefficients and is represented by the following formula:

L-I

C (t) = ^' Ya _ι δ (t-lτ _o ) where L is the length of the filter expressed in number of

coefficients, ai is the amplitude of the coefficient at position I QH T ₀ is the delay associated with the coefficient of rank /. τ ₀ depends on the maximum sampling frequency or filter bandwidth. At least part of the filters C _-1 (t), Co (t) ... Ci (t) is obtained from a first time reversal filter consisting of a first set of coefficients determined from information representative of the existing channel between the transmitter EM and the receiver RE.

Each of the M filters C _-1 (t), Co (t) ... C ₁ (X) has coefficients whose values are the values of the coefficients of the first time reversal filter shifted in one direction, where each of the MI filters C _-1 (t), Co (t) ... Ci (t) a coefficients whose values are the values of the coefficients of the first time reversal filter shifted in at least one direction, the Mth filter being the first filter of time reversal. It should be noted here that part of the MI or M filters has coefficients whose values are the values of the coefficients of the first time reversal filter shifted in one direction, the other part of MI or M filters have coefficients whose values are the values of the coefficients of the first time reversal filter shifted in another direction. A reversal filter whose values are the values of the coefficients of the first time reversal filter shifted in one direction can be expressed as follows: C _m (t) = F {C {t), p) where F represents the operation of shifting the values of the coefficients by a number of positions / ?. p is defined by p = m -a where a is an optimized value as described later with reference to FIG. 3.

In other words, a reversal filter whose values are the values of coefficients of another second time reversal filter offset in one direction can be expressed as follows: C _m (t) = F (C ( t), a) where F represents the operation of shifting the values of the coefficients by a number of positions a.

The determination of a is performed by the shift deciding module Mdec.

The offset function can be expressed as follows: for a shift to the right

for a shift to the left

The offset of the coefficient values of the filter C _n (t) = F (C (t), p) is executed by a shift module Decm. m can be selected from one of these integers:

Alternatively, if M is even, m can be chosen from one of these integers:

. \ (M) (M) _Λ (M) _Λ (M

Negative values correspond to a shift to the left, positive values to a shift to the right.

For example, siM = 5, the offsets are /? = {-2α, -α, 0, α, 2α}. For example, if M = 4, the offsets are /? = {-2α, -α, α, 2α}, or /? = {-2α, -α, 0, α}.

The receiver RE receives, according to the present invention, M pulses Im-I, ImO and ImI in the transmission time T _s which is in the state of the art the duration of transmission of a symbol. Each pulse Im-I, ImO and ImI is respectively representative of a symbol.

The amplitude of the pulses Im-I, ImO and ImI is less than the case in which a single pulse is received. The amplitude of the pulses is a function of the offset value / ?. The amplitude of the pulse Imm is given by: LIp

-J Λ _T = ^d ^A l Σ ^a ² I ι ^ -P ⁶⁾ - if right shift

[M

Peak signal (m, p) = -

1 [ ^{L ~ ι} I

-7 = <J 4UΣ ^α / ² \ ^δ ( ^{t +} P)> if shift to the left

where i / Vf is representative of the normalization by the factor β at the emitter EM. The amplitude of the pulse Imm depends on the amplitudes of the (L- /?) Coefficients.

Figs. 2 show examples of values of the coefficients taken by the coefficients of the time reversal filters according to the present invention.

Fig. 2a represents the amplitudes of the ordered coefficients of a time reversal filter of length four. The first coefficient has an amplitude a _0, the second coefficient has a magnitude _ls the third coefficient has a magnitude ₂ and the fourth coefficient has a magnitude _3. The coefficients of FIG. 2a are the coefficients of a first time reversal filter obtained from information representative of the channel existing between the transmitter EM and the receiver RE.

Fig. 2b represents the amplitudes of the ordered coefficients of a time reversal filter of length four. The first coefficient has an amplitude between two predetermined values or preferably equal to a zero amplitude, the second coefficient has an amplitude ao, the third coefficient has an amplitude ai and the fourth coefficient has an amplitude a ₂ . The coefficients of FIG. 2b are the coefficients of a second time reversal filter whose values are the values of the coefficients of the first time reversal filter shifted to the right and the last coefficient starting from the right has a zero value.

Fig. 2c represents the amplitudes of the ordered coefficients of a time reversal filter of length four. The first and second coefficients have an amplitude between two predetermined values or preferably equal to a zero amplitude, the third coefficient has an amplitude ao and the fourth coefficient has an amplitude ai. The coefficients of FIG. 2c are the coefficients of a second time reversal filter whose values are the values of the coefficients of the first time reversal filter shifted to the right by two positions or are the coefficients of a second time reversal filter whose values are the values of the coefficients of the second time reversal filter of FIG. 2b shifted to the right of a position. The last coefficient from the straight line whose value is non-zero in FIG. 2b takes a zero value in FIG. 2c. Fig. 2d represents the amplitudes of the ordered coefficients of a time reversal filter of length four. The first coefficient has a magnitude _ls has the second coefficient has a magnitude _2, the third coefficient has an amplitude a ₃ and the fourth coefficient has an amplitude between two predetermined values or preferably equal to zero amplitude. The coefficients of FIG. 2d are the coefficients of a second time reversal filter whose values are the values of the coefficients of the first time reversal filter shifted to the left and the last coefficient starting from the left has a zero value.

Fig. 2e represents the amplitudes of the ordered coefficients of a time reversal filter of length four. The first coefficient has a magnitude _2, the second coefficient has a magnitude _3, the third and fourth coefficients have an amplitude between two predetermined values or preferably equal to zero amplitude.

The coefficients of FIG. 2e are the coefficients of a second time reversal filter whose values are the values of the coefficients of the first time reversal filter shifted to the left by two positions or are the coefficients of a second time reversal filter whose values are the values of the coefficients of the second time reversal filter of FIG. 2d shifted to the left of a position. The last coefficient from the left whose value is non-zero in FIG. 2d takes a zero value in FIG. 2nd.

Fig. 3 represents an algorithm for determining an offset of coefficient values according to the present invention.

In step E300, the module Mdec determines, from information representative of the channel existing between the transmitter EM and the receiver RE, ordered coefficients of a first time reversal filter.

In the next step E301, the module Mdec determines the length L of the first time reversal filter as a function of the channel.

In the next step E302, the module Mdec initializes α and n to zero.

In the next step E303, the module Mdec selects the minimum average value of the signal-to-interference ratio (SIR ₀ ) and the value of the maximum variance (L2) of the signal-to-interference ratio.

In the next step E304, the module Mdec selects M.

In the following step E305, the module Mdec calculates the value of a according to the following formula: a = Int {(a + ή) L} where Int represents the integer part. In the following step E306, the module Mdec calculates for each value taken by m, the offsets p = m -a.

In the next step E307, the module Mdec calculates the average interference signal ratio SIR resulting for the signal X (t) normalized. In the next step E308, the module Mdec checks SiSIR ₀ <SIR. If so, the module Mdec goes to step E310. If not, the module Mdec goes to step E309.

In step E309, the module Mdec increments n according to the following formula: n = ^{n +} ] 4 _{nr \} and returns to step E305. In step E310, the module Mdec checks whether the variance of the signal-to-interference ratio SIR is less than L2. If so, the module Mdec goes to step E311. If not, the module Mdec goes to step E309. In step E311, the module Mdec stores the value of a. Of course, the present invention is not limited to the embodiments described herein, but encompasses, on the contrary, any variant within the scope of those skilled in the art and particularly the combination of different embodiments of the present invention.

Claims

1) A method of sending, by an emitter, a set of data to a receiver, characterized in that the method comprises the steps of:

determination, based on information representative of an existing channel between the transmitter and the receiver, of ordered coefficients of a first time reversal filter,

determining, for at least one second time reversal filter, ordered coefficients of the second time reversal filter, the ordered coefficient values of the second time reversal filter being the values of the coefficients of the first time reversal filter shifted in one direction, - division of the dataset into at least two data streams,

filtering the signals representative of each data stream by a determined time reversal filter, each time reversal filter filtering the signals representative of a stream being different from each other time reversal filter filtering the signals representative of another stream, - issuing the sum of the filtered flows.

2) Method according to claim 1, characterized in that the data set is divided into an even number M of data streams.

3) Method according to claim 1, characterized in that the data set is divided into an odd number M of data streams and in that a data stream is filtered by the first time reversal filter and each other stream of data is filtered by a second time reversal filter.

4) Method according to claim 3, characterized in that [MX] II second time reversal filters have coefficients whose values are shifted in a first direction and [MY] II time reversal filters have coefficients whose values are shifted in a second direction different from the first direction.

5) Method according to any one of claims 1 to 4, characterized in that at least one time reversal filter filtering a data stream has coefficients whose values are the values of the coefficients of another filter time reversal filtering data shifted in one direction and the at least one time reversal filter aa coefficients whose value is zero more than the other time reversal filter. 6) Process according to claim 5, characterized in that the coefficients whose value is zero more than the other time reversal filter are at the position of the last non-zero coefficients of the coefficients of the other ordered time reversal filter. according to the shift direction. 7) Method according to any one of claims 1 to 6, characterized in that each data stream is multiplied by a scale factor determined from the coefficients of the time reversal filter filtering the data stream.

8) Method according to any one of claims 5 to 7, characterized in that the value of a is determined by selecting a value of a for which a signal ratio over average interference is greater than or equal to a given value and for which a The signal to interference ratio variance is less than a maximum variance value of the signal-to-interference ratio.

9) Transmitter, a set of data to a receiver, characterized in that the transmitter comprises: - determination means, from information representative of a channel existing between the transmitter and the receiver a first ordered group of coefficients of a first time reversal filter,

means for dividing the data set into at least two data streams,

means for filtering the signals representative of each data stream by a determined time reversal filter, each time reversal filter filtering the signals representative of a flow being different from each other time reversal filter filtering the signals representative of a other stream,

means for transmitting the sum of the filtered flows.