Patent Grant
11463157
The present application is a National Phase of International Application Number PCT/EP2020/066059 filed Jun. 10, 2020, which designated the U.S. and claims priority benefits from French Application Number FR 1906354 filed Jun. 14, 2019, the entire contents of each of which are hereby incorporated by reference.
The present invention belongs to the field of air traffic surveillance by satellite. In particular, the invention relates to a method and a device for receiving from space ADS-B (acronym for “Automatic Dependent Surveillance-Broadcast”) messages emitted by aircrafts.
ADS-B is a surveillance system for air traffic control. Each aircraft equipped with ADS-B on-board equipment can periodically emit (for example every second) an ADS-B message which mainly contains an identifier, a position and a heading of the aircraft.
The ADS-B system was originally designed for communication between an aircraft and a ground station, or to surrounding aircrafts. The deployment of ground stations is impossible in oceanic areas and very delicate in certain land areas that are difficult to access or potentially having security risks. The need to ensure a homogeneous quality of surveillance throughout the flight path of the aircraft leads to considering the collection of ADS-B messages from space, which will also allow to open the new routes necessary to take account of the growth in air traffic.
The attenuation undergone by ADS-B messages received via a satellite is significant and causes a significant negative impact on the Signal-to-Noise Ratio (SNR). In the event of a collision of several ADS-B messages, that is to say when several ADS-B messages are emitted simultaneously by aircrafts present in the beam of a satellite antenna, conventional receiving devices are generally not capable of detecting and/or decoding said messages because the signal-to-noise ratio for each message is too low. This makes conventional devices unreliable and not applicable for surveillance means involving the safety of persons.
It is therefore necessary to set up specific means for detecting and collecting ADS-B messages. The use of a multi-beam antenna limits the effects of interference within each beam, because a reduction in the dimensions of a beam leads to a decrease in the number of collisions between messages emitted simultaneously by aircrafts present in the beam in question. However, with conventional receiving devices, the high number of beams to be used is unsatisfactory.
Patent applications WO 2008/134255 A2 and EP 2738972 A2 describe methods for receiving an ADS-B message. However, the performance of these methods is not sufficient to detect and synchronize an ADS-B message when the SNR level is particularly low.
The purpose of the present invention is to overcome all or part of the disadvantages of the prior art, in particular those set out above, by proposing a solution allowing to synchronize and decode an ADS-B message transported by a radio signal for which the signal-to-noise ratio is particularly low.
To this end, and according to a first aspect, provision is made by the present invention of a method for receiving an automatic dependent surveillance-broadcast (ADS-B) message. Such a message is carried by a radio signal and includes a preamble and a block of data encoded by symbols modulated by pulse position. The block of data includes a cyclic redundancy check (CRC) code. The reception method includes a synchronization phase for detecting the start of the message, and a decoding phase for recovering a bit stream corresponding to the block of data in the message.
The synchronization phase includes:
The decoding phase includes:
For low noise ratio levels (for example less than 13 dB), the synchronization phase of the reception method according to the invention has better detection performance of an ADS-B message than conventional methods based on a signal correlation.
The different steps of the decoding phase of the reception method according to the invention allow to optimize the parity matrix of the cyclic redundancy check code by reducing its density, so that it becomes possible to use a belief propagation algorithm to correct any errors during the decoding of the block of data in the message. A cyclic redundancy check code is conventionally not well adapted for the use of such an algorithm because the parity matrix associated therewith is too dense.
In particular embodiments, the invention may also include one or more of the following features, taken in isolation or in any technically possible combination.
In particular embodiments, during the synchronization phase, the selected candidate sequences correspond to the sequences including the largest absolute values of log-likelihood ratios among all the log-likelihood ratios of all the determined sequences.
In particular embodiments, during the synchronization phase, the candidate sequences selected correspond to the sequences for which the sums of the absolute values of the log-likelihood ratios are the greatest.
In particular embodiments, the starting sample is determined as the sample associated with the candidate sequence for which the calculated likelihood value is the greatest.
In particular embodiments, the synchronization phase of the reception method further includes a discrimination between different candidate sequences using parts of the preamble not including a pulse.
In particular embodiments, the discrimination between different candidate sequences includes a calculation of a sum, for each candidate sequence:
In particular embodiments, the decoding phase of the reception method further includes a discrimination between different possible bit streams obtained at the output of the belief propagation algorithm for the block of data in the message based on the Hamming distance between a possible bit stream and a bit stream corresponding to the sequence of log-likelihood ratios associated with the symbols of the block of data.
According to a second aspect, the present invention relates to a device for receiving an automatic dependent surveillance-broadcast (ADS-B) message. Such a message is carried by a radio signal and includes a preamble and a block of data encoded by symbols modulated by pulse position. The block of data includes a cyclic redundancy check (CRC) code. The receiving device includes a synchronization module configured to detect the start of the message, and a decoding module configured to recover a bit stream corresponding to the block of data in the message.
The synchronization module is configured for:
The decoding module is configured for:
The invention will be better understood upon reading the following description, given by way of non-limiting example, and made with reference to
In these figures, identical references from one figure to another designate identical or similar elements. For reasons of clarity, the elements shown are not necessarily on the same scale, unless otherwise indicated.
As indicated above, the present invention aims in particular at providing a solution allowing to synchronize and decode an ADS-B message transported by a radio signal for which the signal-to-noise ratio is particularly low.
The payload of the satellite 10 includes an antenna 11 configured to receive the radio signal carrying the ADS-B message. The antenna 11 may for example include, in a conventional manner, an array of antennas allowing to form several beams.
As explained previously, the distance separating the satellite 10 from the aircraft 12 and the potentially large number of collisions between the message emitted by the aircraft 12 and other messages emitted simultaneously by other aircrafts and picked up by the same beam 14 have the consequence that the signal-to-noise ratio (SNR) is generally much lower for a radio signal carrying an ADS-B message received by the satellite 10 than for the same radio signal received by a ground station 18 on a link 16 of aircraft-station communication, or for the same radio signal received by another aircraft 13 on an aircraft-aircraft communication link 17.
The payload of the satellite 10 includes a device for receiving an ADS-B message.
To this end, the synchronization module 21 and the decoding module 22 include for example a processing circuit (not shown in the figures), including for example one or more processors and storage means (magnetic hard disk, electronic memory, optical disc, etc.) wherein a computer program product is stored, in the form of a set of program code instructions to be executed to implement the steps of a method for receiving an ADS-B message. Alternatively or in addition, the processing circuit includes one or more programmable logic circuits (FPGA, PLD, etc.), and/or one or more specialized integrated circuits (ASIC), and/or a set of discrete electronic components, etc., adapted for implementing all or part of said steps.
The receiving device 20 is connected to the antenna 11 of the satellite and further includes a radio circuit (not shown in the figures) including equipment (amplifier, local oscillator, mixer, analog filter, analog/digital converter, etc.) known to the person skilled in the art, allowing said receiving device 20 to receive messages in the form of radio signals.
In other words, the radio-electric circuit, the synchronization module 21 and the decoding module 22 correspond to means of the receiving device 20 which are software (specific computer program product) and/or hardware (FPGA, PLD, ASIC, discrete electronic components, etc.) configured to implement the steps of a method for receiving an ADS-B message.
In the example considered, the synchronization module 21 and the module 22 for decoding a reception chain corresponding to an antenna beam are produced on programmable logic circuits of the FPGA type. The FPGA circuit can include several reception chains. In particular, in the example considered, the FPGA circuit includes nine reception chains corresponding respectively to nine different beams, and the payload of the satellite 10 includes four such FPGA circuits in order to support thirty-six different beams in all.
The synchronization phase 60 in particular includes the following steps:
The decoding phase 70 in particular includes the following steps:
As illustrated in
In the example considered, and as illustrated in
The radio signal carrying the ADS-B message is for example transmitted on a carrier frequency of 1090 MHz. A pulse then corresponds to a period of a duration of 0.5 μs during which the carrier frequency is emitted with a certain power. An absence of a pulse corresponds to a period during which the carrier frequency is not emitted. The radio signal received by the receiver device 20 is then brought back to baseband and sampled, in a conventional manner, by the synchronization module 21.
The radio signal sampling step 61 allows to obtain samples Ek as illustrated in
The objective of the synchronization phase 60 is to detect the sample, called the “starting sample” which corresponds to the first sample of the first symbol of the preamble of the ADS-B message. In the example shown in
It should be noted that a power detection step can precede the synchronization phase 60. A sliding time window of a predetermined duration is for example used. By
Fourier transform, it is then possible to determine, for said time window, a power spectral density in a frequency band of interest, for example in an 8 MHz wide band centered on the carrier frequency of 1090 MHz of the radio signal. If the power spectral density is greater than a predetermined threshold, then a synchronization phase 60 is triggered on the part of the signal comprised in this time window.
In step 62, for each sample of a predetermined number of successive samples, a sequence of log-likelihood ratios is determined. In
Each log-likelihood ratio corresponds to a ratio between the probability that a symbol of the signal received corresponds to the symbol of value 1 shown in
In the remainder of the description, the number of samples per symbol is denoted Ne (Ne is the oversampling factor, Ne=8 in the example considered). Thus, the element of index j in the sequence Si of index i corresponds to the log-likelihood ratio
LRRi,j, such that:
Pi,j1 is the probability that the symbol corresponding to the Ne samples Ei+(j−1)·Ne to Ei+j·Ne−1 correspond to a symbol of value 1.
Pi,j−1 is the probability that the symbol corresponding to the Ne samples Ei+(j−1)·Ne to Ei+j·Ne−1 correspond to a symbol of value −1.
where C is a normalization constant.
The positive or negative sign of the log-likelihood ratio LLRi,j indicates whether the observed symbol has a value of 1 or −1. The greater the absolute value of the log-likelihood ratio LLRi,j, the greater the confidence in the decision on the value of the symbol.
During this synchronization phase 60, the number of sequences Si considered corresponds to the number of successive samples considered. The number of log-likelihood ratios LLRi,j comprised in each sequence Si can for example correspond to the number Np of symbols of the preamble (Np=8). The number of sequences Si considered can for example be equal to Np×Ne, that is to say sixty-four in the example considered (Np×Ne=64).
In step 63, candidate sequences are selected from the various sequences Si thus determined. The selection is made based on the values of the log-likelihood ratios LLRi,j of said sequences. As indicated above, each candidate sequence Sj is associated with a candidate sample Ej which may correspond to the start of the message.
In particular embodiments, the selected candidate sequences correspond to the sequences including the largest absolute values of log-likelihood ratios among all the log-likelihood ratios of all the determined sequences.
In particular embodiments, the selected candidate sequences correspond to the sequences for which the sums of the absolute values of the log-likelihood ratios are the largest.
The number of candidate sequences selected during step 63 is not necessarily fixed. This number can be adapted according to the results obtained for the log-likelihood ratios of the different sequences. For example, a sequence Si is selected as a candidate sequence only if the maximum value among the absolute values of the log-likelihood ratios LLRi,j is greater than a predefined threshold, or if the sum of the absolute values of the log-likelihood ratios is greater than a predefined threshold. Alternatively, the number of selected candidate sequences can be predefined at a fixed value.
In step 64, for each candidate sequence thus selected, a likelihood value is calculated between said candidate sequence and a sequence of symbols expected for the preamble.
A log-likelihood ratio sequence Si can be written in the form:
Si=(LLRi,1, LLRi,2, LLRi,3, . . . , LLRi,Np)
The sequence Seq of symbols expected for the preamble can be written in the form:
Seq=(Seq1, Seq2, Seq3, . . . , SeqNp)=(1, 1, 0, −1, −1, 0, 0, 0)
The likelihood value LVi between a candidate sequence Si and the sequence Seq of symbols expected for the preamble can for example be calculated as follows:
In step 65, the starting sample corresponding to the start of the message is determined based on the likelihood values LVi thus calculated.
In particular embodiments, the starting sample is determined as being the sample associated with the candidate sequence Si for which the calculated likelihood value LVi is the greatest.
With such arrangements, the synchronization phase 60 allows to detect the start of an ADS-B message in 99% of cases when the signal-to-noise ratio is approximately 11 dB. Such a rate of detection of the start of an ADS-B message is only achieved by the conventional methods of the prior art for a signal-to-noise ratio greater than 13 dB. The method 50 according to the invention therefore allows a gain of 2 dB compared to the methods of the prior art for synchronizing an ADS-B message.
The parts of the preamble which do not include a pulse correspond on the one hand to the symbols #3, #6, #7 and #8 of the preamble as well as to the halves of the symbols #1, #2, #4 and #5 which do not include a pulse (see
This discrimination step 66 includes for example the calculation of a sum of the following elements for each candidate sequence:
The candidate sequences for which the sums thus calculated are the lowest are considered as the most likely. This additional discrimination step 63 allows to further improve the performance of the receiving device 20 during the synchronization phase 60.
Once the starting sample corresponding to the start of the ADS-B message is determined, the phase 70 of decoding the bit stream corresponding to the block of data in the message can start. The decoding phase 70 of the reception method 50 according to the invention includes in particular an error correction based on a belief propagation algorithm which uses the cyclic redundancy check (CRC) code.
The belief propagation algorithm, also known under the term sum-product message transmission (“Sum-Product algorithm”), is a known decoding algorithm allowing to recover the useful information of a received message. This is an iterative-type algorithm particularly well adapted to linear error correcting codes for which the parity matrix has a low density (that is to say a parity matrix whose elements are for the major part 0, and for the minor part 1, for example the proportion of 1 on all the elements of the matrix is of the order of 10−3). The belief propagation algorithm is particularly well adapted to LDPC (acronym for “Low-Density Parity Check code”) type codes.
A cyclic redundancy check (CRC) code is conventionally used to verify the integrity of transmitted data, and not to correct errors that have arisen in the transmission of data. A CRC code usually has a dense parity matrix, and such code is not well adapted for using a belief propagation algorithm. However, the decoding phase 70 of the reception method 50 according to the invention includes steps allowing to transform the parity matrix of the CRC code of an ADS-B message such that the use of the belief propagation algorithm becomes possible.
Conventionally, a CRC code is associated with a generator polynomial g(x) which is known to both the sender and the receiver of a message. For user data d(x), the parity bits p(x) of the CRC correspond to the remainder of the division of said user data by the generator polynomial: p(x)=d(x)/g(x). As a reminder, in the example considered, and as illustrated in
In the case of ADS-B, the CRC generator polynomial is:
g(x)=1+x3 +x10 +x12 +x13 +x14 +x15 +x16 +x17 +x18 +x19 +x20 +x21 +x22 +x23 +x24 [Math. 5]
It should be noted that we are working in the Galois field GF(2) including two elements, namely the elements 0 and 1. In GF(2), an addition corresponds to the operation “logical OR” (XOR) and a multiplication corresponds to the “logical AND” (AND) operation.
The number of bits in the block of data of an ADS-B message (Nb=112) is denoted Nb, and the number of user data bits included in said block of data (Nd=88) is denoted Nd.
The CRC generation matrix is the matrix G=[gi,j] such that for user data d=(d1, d2, . . . , dNd), the block of data b=(b1, b2, . . . , bNb) is such that b=d·G.
It is possible to determine the generation matrix G=[gi,j] of the CRC by solving the following system of linear equations by generating at least (Nb+1) different blocks of data:
The generation matrix G=[gi,j] includes Nd rows and Nb columns. G can be written as G=[I|P], where I is the Identity matrix, and P is a matrix including Nd rows and (Nb−Nd) columns.
The parity matrix H is the core of the generation matrix G, which means that G·HT=0 (HT is the transpose of the matrix H). In other words, for a block of data b=(b1, b2, . . . , bNb), we have b·HT=0 or H·bT=0. It is known that the parity matrix H associated with the CRC is the matrix H=[PT|I], where PT is the transpose of the matrix P. The parity matrix H of the CRC code includes (Nb−Nd) rows and Nb columns. The parity matrix H of the CRC code is shown in
The decoding phase 70 includes a step 71 of determining, from the starting sample, a sequence of log-likelihood ratios associated respectively with the symbols of the block of data in the message. If the starting sample is the sample EN0 of index N0, then the first sample of the first symbol of the block of data is the sample EN0+Np·Ne of index (N0+Np·Ne). The sequence includes one hundred and twelve (112) log-likelihood ratios associated respectively with the symbols of the block of data in the message. Each log-likelihood ratio is calculated in a manner similar to what was done previously for the symbols of the preamble during the synchronization phase 60.
The sequence of log-likelihood ratios associated respectively with the symbols of the block of data in the message is denoted SB. A log-likelihood ratio of index n of the sequence SB is denoted LLRn. Nb is the number of symbols in the block of data in the message (Nb=112). Then the sequence SB is written in the form SB=(LLR1, LLR2, LLR3, . . . , LLRNb) with:
The decoding phase 70 then includes a step 72 of ordering the sequence SB in ascending order in absolute value. In other words, the log-likelihood ratios LLRn of the sequence SB are ordered from the log-likelihood ratio having the lowest absolute value to the log-likelihood ratio having the greatest absolute value. In other words, the elements of the sequence SB are classified in ascending order of the confidence given to said element (as a reminder, the greater the absolute value of a log-likelihood ratio, the greater the confidence given to the decision on the value of the associated symbol, indicated by the sign of said ratio). The ordered sequence obtained is denoted SB′.
In step 73, the columns of the parity matrix H of the cyclic redundancy check code are permuted in a coherent manner with the permutations performed to order the sequence SB of log-likelihood ratios. This means that the columns of the parity matrix H are permuted in the same way that the elements of the sequence SB were permuted. In other words, if an element LLRk at position k in the sequence SB is at position k′ in the ordered sequence SB′ after the ordering step 72, then the column k of the matrix H becomes the column k′ of the permuted matrix after the permutation step 73. A mathematical operator corresponding to the permutation of the columns of the parity matrix H is denoted Π. The new position k′ of a column of the permuted matrix corresponding to the column initially at the position k in the parity matrix is such that k′=Π(k).
The decoding phase 70 then includes a step 74 of transforming the permuted matrix by reduction in scaled form of a sub-part of said permuted matrix, in order to obtain a matrix called “transformed matrix”. A matrix is said to be scaled if the number of zeros preceding the first non-zero value of a row increases row by row until there are possibly only zeros remaining.
For example, the square sub-matrix corresponding to the (Nb−Nd) rows and to the (Nb−Nd) first columns of the permuted matrix can be reduced to a scaled form by means of elementary row operations. Such a procedure is for example known as “Gaussian elimination”.
The decoding phase 70 then includes a step 75 of reverse permutation of the columns of the transformed matrix to return the columns to their initial order, in order to obtain a matrix called “optimized parity matrix”. In other words, the reverse operator Π−1 is applied, so that, after the reverse permutation step 75, the new position Π−1(k′) of a column of the optimized parity matrix corresponding to the column at position k′ in the transformed matrix is the initial position k of the column in the original parity matrix.
The matrix obtained is an optimized parity matrix insofar as its density has been reduced. The reduction in scaled form of part of the permuted parity matrix has indeed allowed to reduce the proportion of 1 among all the elements of the matrix.
The decoding phase 70 finally includes an application 76 of a belief propagation algorithm using the optimized parity matrix and the sequence SB of log-likelihood ratios to determine a bit stream corresponding to the block of data in the message. This algorithm is applied in a conventional manner.
Steps 72 to 75 of the decoding phase 70 are equivalent to cutting some connections between variable nodes (the log-likelihood ratios LLRn) and parity nodes (the parity equations) of a bipartite graph on which is based the belief propagation algorithm. Such a graph is illustrated in
With such arrangements, the decoding phase 70 allows to correctly decode the block of data of an ADS-B message in 99% of the cases when the signal-to-noise ratio is approximately 8.5 dB. Such a decoding rate of a block of data of an ADS-B message is only achieved by the conventional methods of the prior art for a signal-to-noise ratio of about 12 dB. The method 50 according to the invention therefore allows a gain of more than 3 dB compared to the methods of the prior art for decoding an ADS-B message.
Discrimination 77 is performed based on the Hamming distance between a candidate bit stream obtained at the output of the belief propagation algorithm, and a bit stream assumed corresponding to the sequence of log-likelihood ratios associated with the symbols of the block of data. Indeed, the belief propagation algorithm can result in several candidate bit streams satisfying the relationship H′·bT=0 (H′ being the optimized parity matrix and b being a candidate bit stream).
The Hamming distance between two bit streams of the same length is the number of bits for which the two bit streams are different.
The assumed bit stream corresponds to the sequence SB of log-likelihood ratios associated with the symbols of the block of data. It is obtained by considering that a log-likelihood ratio having a positive value corresponds to a bit of value 1 and that a log-likelihood ratio having a negative value corresponds to a bit of value 0.
For example, it is possible to eliminate a candidate bit stream (obtained at the output of the belief algorithm) which has a Hamming distance with the assumed bit stream (obtained from the sequence SB) greater than
(dmin being the minimum Hamming distance of the CRC code, and A
corresponding to the integer lower part of A).
According to another example, it is possible, for example, to choose as the result of the decoding phase 70 the candidate bit stream (obtained at the output of the belief algorithm) which has the smallest Hamming distance with the assumed bit stream (obtained from the sequence SB).
The above description clearly illustrates that, through its various features and their advantages, the present invention achieves the fixed objectives.
In particular, the synchronization phase 60 based on log-likelihood ratios for the symbols of the preamble allows to detect the start of an ADS-B message for a particularly low signal-to-noise ratio compared to what is possible to do with conventional methods for receiving an ADS-B message. Indeed, the conventional methods for synchronizing an ADS-B message are based on a conventional correlation between the received signal and a reference signal corresponding to the expected preamble. As the preamble is short and includes little power (the preamble includes only four pulses of 0.5 μs for a preamble duration of 8 μs), such a correlation can generate many false positives of detection when the signal-to-noise ratio is low (for example of the order of 8 dB).
In
On the other hand, the decoding phase 70 allows to take advantage of the CRC of the ADS-B message to correct any decoding errors of the block of data in the message. A CRC is conventionally not adapted for correcting decoding errors, but only for verifying the integrity of the decoded data. However, thanks to the various steps allowing to optimize the parity matrix of the CRC by reducing its density, the decoding phase 70 of the reception method 50 according to the invention allows to use a belief propagation algorithm based on said optimized parity matrix. Such arrangements allow to decode an ADS-B message even if it is received with a particularly low signal-to-noise ratio (for example of the order of 8 dB).
In
In general, it should be noted that the implementations and embodiments considered above have been described by way of non-limiting examples, and that other variants can therefore be considered.
In particular, different choices can be considered for the oversampling factor of the signal, for the number of successive samples considered in the synchronization phase 60, for the number of candidate sequences selected and the manner in which said candidate sequences are selected based on the values of the log-likelihood ratios they include, etc. Such choices correspond only to variants of the invention.
It should also be noted that in the receiving method 50 according to the invention, the synchronization phase 60 and the decoding phase 70 may be independent of each other. In other words, it is conceivable to implement a method for receiving a message which includes the synchronization phase 60 according to the invention, and which includes a decoding phase different from the decoding phase 70 according to the invention, for example a conventional decoding method which does not use a belief propagation algorithm. Similarly, it is conceivable to implement a method for receiving a message which includes the decoding phase 70 according to the invention, and which includes a synchronization phase different from the synchronization phase 60 according to the invention, for example a conventional synchronization method based on a signal correlation. The same applies for the synchronization module 21 and the detection module 22: these two modules are independent of each other. It is thus conceivable to produce a receiving device 20 which includes a synchronization module 21 according to the invention and a decoding module different from the decoding module 22 according to the invention. Similarly, it is conceivable to produce a receiving device 20 which includes a decoding module 22 according to the invention and a synchronization module different from the synchronization module 21 according to the invention.
It should also be noted that it can be considered to adapt the invention to other wireless communication protocols using pulse position modulation and a cyclic redundancy check code.
The invention was described by way of example for a device 20 for receiving an ADS-B message on board a payload of a satellite in LEO orbit. However, nothing excludes applying the invention to a satellite in MEO orbit, or even in GEO orbit, if the performances of the antennas of said satellites allow such a message to be received. Nothing excludes either applying the invention to other receivers such as for example a ground station or another aircraft.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1906354 | Jun 2019 | FR | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2020/066059 | 6/10/2020 | WO |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2020/249604 | 12/17/2020 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 9791562 | Stayton | Oct 2017 | B2 |
| 20180269955 | Dyson | Sep 2018 | A1 |
| Number | Date | Country |
|---|---|---|
| 2 365 355 | Sep 2011 | EP |
| 2 738 972 | Jun 2014 | EP |
| 2 738 972 | Jun 2017 | EP |
| 2008134255 | Nov 2008 | WO |
| 2008134255 | Nov 2008 | WO |
| WO-2012084956 | Jun 2012 | WO |
| Entry |
|---|
| English Translation of International Search Report for PCT/EP2020/066059, dated Sep. 14, 2020, 2 pages. |
| French International Search Report and Written Opinion of the ISA for PCT/EP2020/066059, dated Sep. 14, 2020, 12 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20220209849 A1 | Jun 2022 | US |
