EP4168937A1 - Procédé et système d'adaptation d'un réseau de neurones utilisé dans un réseau de télécommunications - Google Patents
Procédé et système d'adaptation d'un réseau de neurones utilisé dans un réseau de télécommunicationsInfo
- Publication number
- EP4168937A1 EP4168937A1 EP21739158.0A EP21739158A EP4168937A1 EP 4168937 A1 EP4168937 A1 EP 4168937A1 EP 21739158 A EP21739158 A EP 21739158A EP 4168937 A1 EP4168937 A1 EP 4168937A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- network
- parameters
- neural network
- cnr
- snr
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0499—Feedforward networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/09—Supervised learning
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/391—Modelling the propagation channel
- H04B17/3913—Predictive models, e.g. based on neural network models
Definitions
- the invention relates to the general field of telecommunications. More specifically, the invention relates to the field of signal processing using so-called artificial neural networks in telecommunications networks.
- FIG. 1 represents a communication network of the state of the art, for example a cellular communication network, in which an NR neural network can be implemented.
- the network comprises at least one mobile terminal device UE, connected via a radio communication channel CN to a device BS of the base station type.
- the terminal UE transmits a radio signal x (t) to the base station BS
- the base station BS will receive a radio signal y (t) different from the transmitted signal x (t).
- the transmitted signal x (t) undergoes alterations due to its propagation on the radio channel CN.
- the base station implements a network function in the form of an NR neural network to estimate, from the received signal y (t), the signal x (t) issued by the UE terminal.
- the CN radio channel is modeled by choosing the functions of the different neurons, and by training the NR neural network so that it determines the parameters (P weight and bias) of each neuron during a learning phase. of the NR network.
- the neural network NR receives a plurality of signals y '(t) corresponding respectively to transmitted signals x' (t) belonging to a set of known sequences.
- the neural network is able to estimate a transmitted signal x (t) for a new received signal y (t).
- the CN radio channel evolves, it is necessary to adapt (relearn) the NR neural network to take into account the evolution of the CN channel and improve the estimation of the signal x (t).
- the adaptation of the neural network requires time and resources in terms of memory and computational capacity. The adaptation is all the longer and more expensive as the neural network model is complex.
- the network function which performs the adaptation of a neural network transmits the parameters of this neural network to other network functions which use the neural network.
- the station BS following the adaptation of the neural network NR implemented by the base station BS, the station BS sends the new parameters (functions, weight and bias) of its NR network to d other network equipment, for example at the UE terminal.
- the UE terminal uses the new parameters to initialize its own neural network.
- One solution can consist in using a neural network of less complex architecture, so that its adaptation is faster, easier and less expensive, but such a network has poorer expressiveness.
- a complexity of a neural network is for example defined in terms of the number of parameters, of the number of neurons, of the number of layers.
- the expressiveness of the neural network NR represents its capacity to approximate the signal processing function implemented, for example the equalization function. The latter corrects the received signal to facilitate demodulation. This modification is made according to the CN channel.
- a poorer expressivity therefore affects the reliability of the estimate of the transmitted signal x (t).
- Another solution may consist in taking into account, during the adaptation, only certain parameters, for example the weights of a limited number of neurons, by freezing the weights associated with other neurons. This solution is not satisfactory because it requires knowing the weights to be frozen. The equalization applied to the received signal to estimate the transmitted signal x (t) is thus less reliable.
- the invention relates to a method of adapting the parameters of a first neural network used in a communication network to implement a signal processing function by a device, to process an input signal in order to get an output signal.
- the process comprises steps of:
- a third neural network configured to determine a transfer function of the parameters of a second neural network to the parameters of the first neural network, the second neural network being less complex than the first neural network and being also used to implement the signal processing function, the learning of the first and second neural networks having been carried out by the same input and output signals, the transfer function making it possible to deduce parameters of the first network from parameters of the second neural network;
- the invention relates to a system for adapting the parameters of a first neural network used in a communication network to implement a signal processing function by a device for processing an input signal in order to '' obtain an output signal, the system comprising:
- a device called a “third device”, configured to perform a learning process of a third neural network configured to determine a transfer function of the parameters of a second neural network to the parameters of the first neural network, the second network being less complex than the first network and being also used to implement the signal processing function, the learning of the first and second neural networks having been carried out by the same input and output signals, the transfer function making it possible to deriving parameters of the first network from parameters of the second neural network;
- a device called a "second device”, configured to adapt, following a detection of an evolution of said processing function, the parameters of the second neural network by means of input signals associated with a learning sequence;
- first device configured to adapt parameters of the first neural network using the adapted parameters of the second network and the transfer function.
- the adaptation system according to the invention implements the adaptation method according to the invention.
- the architecture of the first neural network is more complex than that of the second neural network.
- the time required for learning the first network is greater than that required for learning the second network.
- the expressiveness of the first network is better than that of the second network.
- the first and second neural networks are configured to perform the same signal processing function.
- this signal processing function corresponds to the estimation of the signal transmitted by a transmitting device on a communication channel to a receiving device.
- the input signal corresponds to the signal received by the receiving equipment item and the output signal then corresponds to the estimate of the transmitted signal.
- the first and second neural networks model the equalization function of the communication channel, the modeling by the first neural network being more reliable.
- the second neural network can be considered as an approximation of the first neural network.
- the first neural network (called complex) is configured with more parameters than the second neural network (called simple) thus allowing it to perform more complex signal processing operations to implement the processing function of signal.
- this involves in particular compensating for the effects of complex non-linearity of the power amplifiers or the effects linked to its propagation on the channel, such as a power loss, a rotation of phase, masking and a frequency or time shift.
- the two neural networks are trained with the same input signals (for example the signals received by the receiving equipment) and the same signals targets (for example the signals emitted by the sending equipment).
- the transfer function according to the invention makes it possible to determine the parameters of the complex network from the parameters of the simple network.
- the adaptation of the complex network is based on this transfer function to deduce, from the adapted parameters of the simple network, the new parameters of the complex network.
- the transfer function according to the invention is an application of a set of real numbers of dimension n, Rn, to a set of real numbers of dimension m, Rm, where n and m are strictly positive integers, and m is greater than n.
- the transfer function is a signal processing function, for example an arithmetic function.
- the third neural network for implementing the transfer function is trained with parameters of the second neural network as inputs.
- the target of the third neural network is to determine the set of parameters of the first neural network as if the first had been trained directly from signals transmitted over the communication channel.
- the third neural network learns how to get the complex network from the simple network.
- the proposed technique makes it possible to reduce the time, memory and computational capacity required for the adaptation of the complex neural network. Indeed, the adaptation of the parameters of the complex network only uses the parameters of the simple neural network and the transfer function performed by the third neural network as defined previously. The adaptation of the complex neural network is thus faster.
- the proposed technique also makes it possible to reduce the consumption of resources necessary for the device implementing the first neural network by making it possible to deport the adaptation of this first network to another device, without requiring transmission of the data necessary for this. adaptation.
- the first device which performs the adaptation of the parameters of the complex neural network has more resources in terms of memory and computational capacity than the second device which performs the adaptation of the parameters of the network. of simple neurons.
- the first (or respectively the second) device also performs the initial learning of the first (or respectively of the second) neural network.
- the adaptation of the parameters of the first neural network, the adaptation of the parameters of the second neural network, and / or the determination of the transfer function can be performed by the same device.
- the proposed method further comprises, after the adaptation of the parameters of the second neural network as a function of the input signals associated with the learning sequence and the adaptation of the parameters of the first network of neurons using the adapted parameters of the second network and the transfer function, following the detection of the evolution of the processing function, a step of complementary adaptation of the adapted parameters of the first network as a function of the associated input signals to this learning sequence.
- the execution of a said neural network can be carried out by a device other than that which carries out the learning and / or the adaptation of the parameters of this neural network.
- the device that performs the learning or adaptation sends the parameters of the neural network to the device that performs it.
- the invention has an advantageous application within the framework of the standardization of the nature of the exchanges between the equipments of a communication network for the implementation of neural networks.
- the second device is a base station, for example of the eNodeB, advanced eNodeB or gNodeB type and the first device is a server at the heart of the communication network, for example a server of the "datacenter" type.
- the third device is a server of the communication network, which may in particular be the same as the first device.
- the invention also relates to a computer program on a recording medium, this program being capable of being implemented in a computer or one of the devices of the proposed adaptation system.
- This program comprises instructions adapted to the implementation of an adaptation method as described above, when the program is executed by a computer.
- the program can use any programming language, and be in the form of source code, object code, or intermediate code between source code and code.
- object such as in a partially compiled form, or in any other desirable form.
- the invention also relates to an information medium or a recording medium readable by a computer, and comprising instructions of the computer program as mentioned above.
- the information or recording medium can be any entity or device capable of storing the program.
- the medium may include a storage means, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or else a magnetic recording means, for example a floppy disk or a disk. hard, or flash memory.
- the information or recording medium can be a transmissible medium such as an electrical or optical signal, which can be conveyed via an electrical or optical cable, by radio link, by optical link without wire or by other means.
- the program according to the invention can in particular be downloaded from an Internet type network.
- the information or recording medium can be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the method according to the invention.
- FIG. 1 already described, illustrates an architecture of a communication network in which a neural network is used according to a method of the state of the art
- FIG. 2 is an architecture of a NET communication network in which a method for adapting the parameters of a neural network is implemented according to a particular embodiment
- FIG. 3 is a flowchart representing steps of an adaptation method implemented according to a particular embodiment
- FIG. 4 represents a functional architecture, according to a particular embodiment, of a system for adapting the parameters of a neural network
- FIG. 5 shows a hardware architecture of a device of the adaptation system according to a particular embodiment.
- FIG. 2 is an architecture of a NET communication network in which a method for adapting the parameters of a neural network is implemented according to a particular embodiment.
- the NET network is a cellular communication network, for example of the 3G, 4G or 5G type.
- the proposed method can be implemented in communication networks based on other technologies.
- the NET communication network is an optical network.
- the communication network NET comprises at least one terminal UE of a user such as a mobile telephone and at least one base station BS of the eNodeB or gNodeB type.
- a radio communication channel CN connects the terminal UE to the base station BS.
- the NET network also includes two DC and DT servers of the datacenter type.
- the base station BS and the servers DC and DT form an adaptation system SYS in accordance with the invention.
- a radio signal x (t) transmitted by the terminal UE to the base station BS undergoes alterations in the CN channel, for example complex non-linearity effects of the power amplifiers or effects linked to its propagation on the channel, such as a weakening of its amplitude (its power), masking, a phase rotation of the symbols included in the signal, a frequency shift, sampling desynchronization, interference with other transmitted signals on neighboring channels, etc.
- the base station receives a signal y (t) different from the transmitted signal x (t).
- the BS base station has the architecture of a computer. It is configured to implement an SNR neural network.
- the base station BS is configured to perform the learning phase of the SNR network, to adapt the parameters P2 (t) of this SNR network and also to execute it (inference phase) once the learning phase ( or adaptation) completed.
- This SNR neural network makes it possible to execute a signal processing function, such as an equalization function to compensate for the effects of the CN channel.
- the base station BS is also configured to run a CNR neural network of higher complexity than that of the SNR network.
- the DC server is configured to train this complex CNR neural network. This CNR neural network makes it possible to perform a signal processing function, such as an equalization function to compensate for the effects of the CN channel.
- the SNR and CNR networks are both intended to estimate the signal x (t) transmitted by the terminal UE from the signal y (t) received by the base station BS (also called the input signal).
- the two networks SNR and CNR perform signal processing operations to implement the equalization function.
- the expressiveness of the CNR complex network is better than that of the simple SNR network.
- the two networks SNR and CNR are driven by the same set of input signals y (t) and output signals x (t). In a particular embodiment, the two networks use an identical error function.
- the DT server is configured to perform the training of a third neural network T which is configured to determine a transfer function of the parameters P2 (t) of the simple network SNR to the parameters Pl (t) of the complex network CNR.
- the transfer function of the network T makes it possible to deduce the parameters Pl (t) of the complex network CNR from the parameters P2 (t) of the simple network SNR.
- the DT server sends the T network implementing the transfer function to the DC server.
- the base station BS performs an adaptation of the simple SNR network and sends the new parameters P2 (t) of the SNR network to the DC server.
- the server DC uses the network T and the adapted parameters P2 (t) to determine the new parameters Pl (t) of the complex network CNR and adapt them.
- BS base stations, DT and DC servers are described as devices. Their network function can also be implemented by virtual functions (VNF for “Virtual Network Function”) running on equipment.
- VNF Virtual Network Function
- FIG. 3 is a flowchart representing the steps of an adaptation method, implemented according to a particular embodiment, by the SYS system described with reference to FIG. 2.
- the terminal UE transmits a training sequence seql comprising symbols allowing the base station BS, as receiver, to estimate the channel of CN communication.
- This seql learning sequence can be fully or partially known by the receiving device, as can its statistical properties.
- This transmitted sequence seql comprises all of the target signals x (t) for the receiver device BS.
- a sequence of deterministic symbols is of the Zadoff-Chu type.
- An example of such a sequence is defined in specification 3GPP TS 38.211 “NR; Physical channels and modulation (Release 15) »vl5.8.0.
- the base station BS learns the parameters P2 (t) of the simple network SNR.
- the terminal UE sends target signals x (t), corresponding to the training sequence seql, and the base station BS receives signals y (t), called input signals, which correspond to the signals x (t) following their alteration by the CN channel.
- this learning E020 iteratively updates the parameters of the simple SNR network by backpropagation of the gradient by minimizing a cost function (also called an error function) based on the quality of the reconstruction by the model of the known sequence seql.
- the base station BS sends the parameters P2 (t) to the server DT.
- the base station BS sends the training sequence seql and the signals y (t) associated with this training sequence that it has received to the server DC.
- the base station BS sends only the signals y (t) associated with this training sequence that it has received to the server DC.
- the server DC has information relating to the training sequence, for example a training sequence number, or else stores the training sequence.
- the server DC learns the parameters Pl (t) of the complex network CNR. For example, this learning E050 iteratively updates the parameters of the complex network by back-propagation of the gradient while minimizing the cost function based on the quality of the reconstruction by the model of the known sequence seql.
- the server DC sends the parameters Pl (t) to the server DT.
- the server DC sends during a step E062 to the base station BS the parameters Pl (t) of the learned complex network CNR.
- the BS base station is capable of executing the complex CNR network, but not of carrying out its training.
- the base station BS is able to execute the simple SNR network.
- the base station BS estimates a signal x (t) having been sent by the terminal UE from an input signal, the received signal y (t) coming from the terminal .
- the base station uses the CNR complex network for estimation E064 because its expressiveness is greater. When it does not have the necessary resources to run the complex CNR network, it uses the simple SNR network.
- the estimated signal is noted x1 (t) or x2 (t) depending on the neural network used for the estimation, CNR or SNR respectively.
- step E070 which can be implemented in parallel, before or after steps E062 and E064, the server DT executes a phase of learning the parameters of the neural network T, which implements the transfer function.
- This transfer function makes it possible to determine the parameters Pl (t) from the parameters P2 (t).
- the DT server received the parameters Pl (t) and P2 (t) during steps E060 and E030 respectively.
- the server DT sends the server DC the parameters of the trained network T.
- the terminal UE transmits during a step E100 target signals x (t), corresponding to another known learning sequence seq2 at the base station BS, for example a Zadoff-Chu type sequence.
- the base station BS receives signals y (t) which correspond to the signals x (t) of this other training sequence seq2 following their alteration by the CN channel.
- the base station BS detects an evolution of the CN channel, for example, by a drop in performance observed after an inference by the simple network SNR.
- the base station BS adapts the previous parameters P2 (t) of the simple network SNR from the signals y (t) associated with this other training sequence seq2 that it has received. It is emphasized here that the learning sequences seq1 and seq2 can be identical.
- the base station BS sends the adapted parameters P2 (t) of the simple network SNR to the server DC.
- the server DC adapts the complex network CNR by using the transfer function of the network T and the adapted parameters P2 (t) of the network simple SNR.
- the server DC determines the adapted parameters Pl (t) of the complex network CNR.
- the server DC sends the adapted parameters Pl (t) of the complex network CNR to the base station BS.
- the base station BS preferably executes the complex network CNR to implement the equalization function of the radio channel CN and to estimate the signals x (t) which have been transmitted by the EU terminal. If the resources of the base station BS do not allow it to execute the complex network CNR (limited memory or calculation capacity), it uses, during step E200, the simple network SNR to implement the function of equalization.
- the neural network CNR or SNR implement an equalization function of the received signal y (t) by multiplying its power by two to estimate the signal x (t).
- the CNR neural network can compensate for other more complex effects and implement the equalization function with better reliability.
- the parameters P2 (t) of the simple SNR network are adapted (E120) following each detection (E100) of an evolution of the equalization function.
- the parameters Pl (t) of the complex network CNR can be adapted (E160) less often, for example every five or ten adaptations of the parameters P2 (t) of the simple network SNR.
- the parameters Pl (t) of the complex network CNR are adapted (E160) if it is determined that the difference between the results of the two networks SNR and CNR exceeds a certain threshold.
- the difference can be observed by the base station BS, which sends, if the difference exceeds the threshold, a request to the server DC to adapt the parameters Pl (t) of the complex network CNR.
- the base station BS is able to adapt the complex network CNR.
- the server DT sends the network T to the base station BS.
- the base station can locally adapt the complex network CNR using the network T and the new parameters P2 (t) of the simple network SNR.
- the base station BS implements the adaptation step E160 instead of the server DC, which avoids the exchanges E140 and E180 with the server DC.
- the base station BS is configured to run complex and simple neural networks but is not configured to train or train them. to adapt them.
- the proposed SYS system then comprises another device configured to train and adapt the simple SNR network and transmit its parameters to the base station so that the latter can execute it.
- the DC server uses at during an optional step E170 (shown in dotted lines in FIG. 3) the input signals y (t) associated with the learning sequence seq2 for a complementary adaptation of the current parameters Pl (t).
- the base station BS transmits the training sequence seq2 and the signals y (t) associated with the sequence seq2 that it received and which were used to adapt (E120) the neural network SNR. This embodiment makes it possible to optimize the adapted parameters Pl (t) of the complex network CNR.
- the base station BS sends only the signals y (t) associated with this training sequence that it has received to the server DC.
- the server DC has information relating to the training sequence, for example a training sequence number, or else stores the training sequence.
- the SNR and CNR neural networks are configured to implement an equalization function.
- the SNR and CNR neural networks are configured to implement a signal processing function to process an input signal (y (t)) in order to obtain an output signal (x (t)).
- the signal processing function can include time and / or frequency synchronization between the transmitter and the receiver. To maintain this synchronization, it is necessary to perform a certain number of time / frequency drift measurements.
- FIG. 4 represents a functional architecture, according to an embodiment of the invention, of the proposed adaptation system SYS, described with reference to FIGS. 2 and 3.
- the SYS system consists of the base station BS, the DC server and the DT server. These devices respectively comprise SNR_m, CNR_m and T_m modules which are configured to learn (E020, E050, E070) the SNR, CNR and T neural networks, and adapt their parameters (E120, E160, E170).
- the BS base station has an exec module configured to run (E064, E220) the SNR simple network and / or the CNR complex network.
- Each of the devices of the SYS system includes a COM communication module configured to exchange (E030, E060, E080, E090, E140 and E180) neural networks as described previously with reference to figure 3.
- the base station BS further comprises a CNR_m module configured to adapt the complex network CNR.
- the base station BS and the server DT form a single device.
- the DT and DC servers form a single device.
- each device BS, DC and DT of the adaptation system SYS has the hardware architecture of a computer, as illustrated in FIG. 5.
- each of the BS, DC and DT devices comprises in particular a processor 7, a random access memory 8, a read only memory 9, a non-volatile flash memory 10 in a particular embodiment, as well as communication means. 11. Such means are known per se and are not described in more detail here.
- the read only memory 9 of the device BS, DC, DT constitutes a recording medium according to the invention, readable by the processor 7 and on which is recorded here a computer program Prog according to the invention.
- the memory 10 of the device BS, DC, DT makes it possible to record variables used for the execution of the steps of the method of adapting a neural network as described above, such as CNR neural networks, SNR and T, and the parameters Pl (t) and P2 (t).
- the computer program Prog defines functional and software modules here, configured to adapt the parameters of a neural network from a transfer function and parameters of another network of less complexity. These functional modules are based on and / or control the hardware elements 7-11 of the device BS, DC, DT DG mentioned above.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| FR2006325A FR3111454A1 (fr) | 2020-06-17 | 2020-06-17 | Procédé et système d'adaptation d'un réseau de neurones utilisé dans un réseau de télécommunications |
| PCT/FR2021/051004 WO2021255362A1 (fr) | 2020-06-17 | 2021-06-03 | Procédé et système d'adaptation d'un réseau de neurones utilisé dans un réseau de télécommunications |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| EP4168937A1 true EP4168937A1 (fr) | 2023-04-26 |
Family
ID=72709511
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| EP21739158.0A Pending EP4168937A1 (fr) | 2020-06-17 | 2021-06-03 | Procédé et système d'adaptation d'un réseau de neurones utilisé dans un réseau de télécommunications |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US20230351156A1 (fr) |
| EP (1) | EP4168937A1 (fr) |
| FR (1) | FR3111454A1 (fr) |
| WO (1) | WO2021255362A1 (fr) |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11551083B2 (en) * | 2019-12-17 | 2023-01-10 | Soundhound, Inc. | Neural network training from private data |
-
2020
- 2020-06-17 FR FR2006325A patent/FR3111454A1/fr not_active Withdrawn
-
2021
- 2021-06-03 WO PCT/FR2021/051004 patent/WO2021255362A1/fr not_active Ceased
- 2021-06-03 EP EP21739158.0A patent/EP4168937A1/fr active Pending
- 2021-06-03 US US18/002,094 patent/US20230351156A1/en active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| FR3111454A1 (fr) | 2021-12-17 |
| US20230351156A1 (en) | 2023-11-02 |
| WO2021255362A1 (fr) | 2021-12-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP4272125B1 (fr) | Procédés et dispositifs pour geler ou adapter des paramètres d'un réseau de neurones utilisé dans un réseau de télécommunications | |
| EP0734133A1 (fr) | Système dégalisation pour récepteur multiporteur | |
| EP3825919A1 (fr) | Méthode d'association d'équipements d'utilisateurs dans un réseau cellulaire au moyen d'un apprentissage par renforcement multi-agent | |
| WO2009156705A2 (fr) | Procède d'estimation de la qualité de focalisation d'un signal pre-egalise par retournement temporel | |
| EP4117178A1 (fr) | Dispositif de linéarisation, par prédistorsion numérique, d'un amplificateur de puissance d'un système de communication | |
| EP3979717A1 (fr) | Méthode d'association d'équipements d'utilisateurs dans un réseau cellulaire selon une politique d'association transférable | |
| EP2427991B1 (fr) | Procede de pre-egalisation par retournement temporel en fdd | |
| EP1330091B1 (fr) | Procédé de sélection d'une position d'une fenêtre FFT dans un récepteur COFDM | |
| EP1878185B1 (fr) | Procédé de codage d'un signal multiporteuse de type OFDM/OQAM utilisant des symboles à valeurs complexes, signal, dispositifs et programmes d'ordinateur correspondants | |
| Li et al. | Model-free end-to-end deep learning of joint geometric and probabilistic shaping for optical fiber communication in IM/DD system | |
| EP4168937A1 (fr) | Procédé et système d'adaptation d'un réseau de neurones utilisé dans un réseau de télécommunications | |
| EP0080544B1 (fr) | Procédé de réception d'un signal de données en modulation à double bande latérale-porteuses en quadrature | |
| FR3094164A1 (fr) | Procédé d’obtention d’un segment de données par un dispositif client apte à communiquer avec une pluralité de réseaux de diffusion de contenu | |
| EP2641369B1 (fr) | Procede de communication dans un reseau cooperatif | |
| EP3539259B1 (fr) | Procédé et dispositif d'actualisation d'un modèle prédictif d'une variable relative à un terminal mobile | |
| CA2374095A1 (fr) | Procede et dispositif d'estimation d'un canal de propagation a partir de ses statistiques | |
| FR2985397A1 (fr) | Procede de pre-codage en tdd | |
| EP4391464A2 (fr) | Procédé d'amélioration de la détermination de l'estimée d'un canal de télécommunication, dispositif récepteur associé | |
| WO2025073504A1 (fr) | Procédé, dispositif et système d'adaptation d'une transmission de données à la suite d'un changement d'attachement | |
| EP1206045A1 (fr) | Procédé de correction de l'erreur de frequence | |
| WO2025073505A1 (fr) | Procédé, dispositif et système de transmission de données d'un terminal lors d'un changement d'attachement | |
| FR3067541A1 (fr) | Emission et reception d'un flux de donnees | |
| FR2809575A1 (fr) | Recepteur, procede de reception, et milieu d'enregistrement qui enregistre un programme pour recevoir des signaux de donnees | |
| WO2006117268A1 (fr) | Procédé de décodage itératif d'un signal ofdm/oqam utilisant des symboles à valeurs complexes, dispositif et programme d'ordinateur correspondants | |
| EP4158811A1 (fr) | Procede et dispositif de quantification de donnees representatives d'un signal radio reçu par une antenne radio d'un reseau mobile |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
| PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
| 17P | Request for examination filed |
Effective date: 20230102 |
|
| AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
| DAV | Request for validation of the european patent (deleted) | ||
| DAX | Request for extension of the european patent (deleted) | ||
| RAP3 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: ORANGE |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
| 17Q | First examination report despatched |
Effective date: 20250205 |