ESTIMATION OF STATISTICAL PARAMETERS OF A RICIAN CHANNEL IN THE PRESENCE OF MASKING BASED ON DEEP LEARNING

Description

technical field

[0001] The present invention relates to the field of wireless communications. More specifically, the invention relates to the estimation of statistical parameters of a propagation channel in order to perform a statistical estimation of a propagation channel. Previous technique

[0002] For some wireless communication networks, there is no fast feedback channel that allows for a reliable estimation of the instantaneous impulse response of a propagation channel at a transmitting node. In this context, optimal resource allocation adapted to propagation conditions is generally based on propagation channel statistics.

[0003] A frequently used statistical model for a propagation channel is the Rice channel. Conventional modeling of a Rice channel involves writing each coefficient of the propagation channel's impulse response as the sum of a deterministic term, the Line of Sight (hereafter referred to as LdV), and a random term following a centered normal complex distribution (Rayleigh component) corresponding to the non-line-of-sight component, known as non-LdV. Such a propagation channel is notably illustrated in the paper "A new simple model for land mobile satellite channels: first- and second-order statistics," by A. Abdi, W.C. Lau, M.-. Alouini, and M. Kaveh, published in IEEE Transactions on Wireless Communications, vol. 2, no. 3, pp. 519–528, May 2003. This statistical model accurately represents the distribution of a propagation channel amplitude when a direct Line of Sight exists between the transmitting node and a receiving node.In the model presented by Abdi et al., the LdV component is no longer deterministic but is itself subject to random fluctuations following a Nakagami-m distribution. An important indicator of the quality of the communication link is provided by a factor known as the Rice K factor. This K factor corresponds to a power ratio between the LdV component and the non-line-of-sight component between the transmitting node and the receiving node. This statistical propagation channel model encompasses the Rayleigh channel when K = 0 and the AWGN channel (for "Additive White Gaussian Noise") when K → +∞ as special cases. Thus, the higher the K factor, the better the link between the transmitting and receiving nodes. A solution for estimating the Rice K factor was proposed in C. Tepedelenlioglu et al.: "The Rice K factor: Estimation and performance analysis", IEEE Transactions on Wireless Communications, vol. 24, no. 5, May 1, 2003 (2003-05-01).

[0004] Knowledge of the K factor, as well as the attenuation of the received power relative to the transmitted power due to the distance between the transmitting and receiving nodes (called "pathloss"), allows for resource allocation adapted to propagation conditions. Estimating the K factor can become more complex in the presence of masking, which is caused by obstacles between the transmitting and receiving nodes. These obstacles generate random fluctuations in the LdV component value. This phenomenon is known as "shadowing." The paper by Abdi et al., for example, describes a Rice channel with Nakagami-m type shadowing.

[0005] There are many procedures in the literature for estimating the Rice K factor in the absence of “shadowing”. A relatively limited number of papers deal with the estimation of the K factor in the presence of “shadowing”. Among these papers, patent CN106850109B, the paper “Estimation of Rician K-Factor in the Presence of Nakagamim Shadowing for the LoS Component by G. Giunta, C. Hao and D. Orlando in IEEE Wireless Communications Letters, vol. 7, no. 4, pp. 550-553, Aug. 2018 » this document Giunta et al. and the document « Estimation of the Ricean $K$ Factor in the Presence of Shadowing, by X. Leturc, P. Ciblat and CJ Le Martret in IEEE Communications Letters, vol. 24, no. 1, pp. 108-112, Jan. 2020,” says the paper Leturc et al. These various papers seek to improve the estimation of the Rice K factor of the spreading channel model described in the paper Abdi et al.These documents thus describe conventional statistical methods for estimating Rice's K factor in the presence of "shadowing", such as the method of moments in patent CN106850109B. The variance of the estimation method is high, which is detrimental when using the factor estimation. K within the framework of an appropriate resource allocation.

[0006] The performance of the EM method is better than that of the method of moments, but this method is difficult to implement in an embedded system because it is iterative, with uncontrolled convergence that is potentially lengthy (several hundred iterations). Furthermore, in this EM method, each iteration involves special mathematical functions, such as a cylindrical parabolic function, for which embedded implementation is expensive. In addition, as shown in the paper by Leturc et al., the maximum likelihood estimator for the propagation channel model in the paper by Abdi et al. is not implementable with reasonable complexity due to the mathematical complexity of this channel model.

[0007] Patent EP3503649 describes a method and device for calculating statistical parameters of a propagation channel. This method comprises several steps executed by a processor within a communication link in the propagation channel. Thus, from a training sequence contained in the communication signal received by a receiving node, the communication link processor estimates the propagation channel and then its statistics. This information is transmitted to a resource allocator for resource allocation. In a final step, the resource manager allocates resources for the next transmission. These allocated resources include, for example, power, modulation scheme, and coding.Optimized resource allocation across the various nodes of the communication network maximizes network performance dynamically, allowing the network to continuously adapt to propagation conditions that change with node mobility. This optimized resource allocation is particularly useful for nodes in embedded systems, where propagation conditions are constantly evolving.

[0008] Patent EP3503649 describes the estimation of a factor as follows K Rice only in the absence of "shadowing". He also describes the "tracking" of the factor K based on tools derived from rupture detection.

[0009] The paper "Rician K-Factor Estimation Using Deep Learning" by M. Alymani, MH Alhazmi, A. Almarhabi, H. Alhazmi, A. Samarkandi, and Y.-D. Yao in the 2020 29th Wireless and Optical Communications Conference (WOCC), cited by Alymani et al., the paper "CNN Based Rician K factor estimation for non-stationary industrial fading channel" by G. Lu, Q. Zhang, X. Zhang, F. Shen, and F. Qin in the 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP), cited by Lu et al., and the paper "Rician K-factor Estimation based on Channel Quality Indicator in OFDM Systems using Neural Network" by Kun Wang, published in 2018 and available on arXiv, cited by Wang, all address the estimation of Rice's K factor using deep learning tools. However, these publications have limitations in their approach.

[0010] First, these documents are presented in a context free of shadowing. Second, all of these documents use classification tools based on Rice's K factor, so the estimation granularity is defined by a grid of allowed values ​​for classification. For example, in Wang's document, the classes considered range from 1 to 11 in increments of 1, but it can be useful to be able to estimate values ​​of the factor. K intermediate, indeed the performances between K = 1, K = 1.5 and K =Two are different. Furthermore, all three documents describe conventional neural network architectures from the literature without attempting to adapt them to the specifics of the estimation problem under consideration. In addition, these approaches all assume that the number of input samples for a neural network architecture is identical across scenarios, which is not necessarily the case in practice. Finally, Wang's document describes a "fully connected perceptron" architecture that is relatively simple, but this architecture can be considered inefficient in some cases because it does not account for permutation invariance of the input data.

[0011] Therefore, there is a need to propose a method for estimating the statistics of a propagation channel in a communication network in the presence of "shadowing", based on the estimation of the factor Kfrom the Rice channel and this from tools derived from "deep learning" whose complexity is controlled. Description of the invention

[0012] The present invention aims to address at least partially this need.

[0013] More specifically, the present invention aims to reduce the complexity of techniques based on conventional statistical methods.

[0014] This invention thus covers a method for estimating the statistics of a propagation channel, said propagation channel enabling the transmission of a communication signal between a transmitting node and a receiving node, said propagation channel being a Rice channel in the presence of masking, said communication signal comprising a learning sequence, said estimation method comprising: a step in determining a set of Nestimates of the propagation channel from said transmitted learning sequence N times by the emitting node; a pre-processing step of said set of estimates to obtain intermediate data; a processing step of the intermediate data to obtain processed intermediate data, said processing step being capable of taking into account a permutation invariance of said intermediate data, regardless of N ; a step of determining statistical parameters from the processed intermediate data; said intermediate data processing step and said statistical parameter determination step being carried out by an architecture of at least one neural network.

[0015] The invention thus provides a method for estimating the statistics of a propagation channel that is simple and practical compared to techniques based on conventional statistical methods. This estimation method is based on the estimation of statistical parameters, such as the factor K Rice's theory, particularly in the presence of shadowing. Furthermore, unlike the prior art, this invention proposes to perform regression rather than improved classification. Finally, the invention incorporates permutation invariance of the estimation problem considered for the design of the architecture of at least one neural network. This drastically accelerates convergence during the training of the neural network(s) of the architecture and improves the generalization capacity of this architecture.

[0016] In one particular embodiment, the intermediate data processing step includes: a processing step by a first neural network having an input and S outputs, said first neural network being capable of receiving as input intermediate data obtained from a real part and an imaginary part of each propagation channel estimate and of providing as output 2 N vectors of S inputs; a step of summing the 2 N vectors of S entries, said summons being commutative and applicable to all N.

[0017] In one particular embodiment, the pre-processing is a normalization of the set of N estimates.

[0018] In one particular embodiment, the estimation process includes a step of determining statistical information from the set of Nestimates, the said statistical information being intended to be transmitted to the architecture.

[0019] The invention thus allows the insertion of additional statistical information as input to the architecture.

[0020] In one particular embodiment, the step of determining statistical information includes an estimation of a factor K of the propagation channel.

[0021] In a particular embodiment, the estimation process includes a step of determining a variance of the estimation noise, said variance of the estimation noise being intended to be transmitted to the architecture, said variance of the noise being associated with the number N of estimates or to the square root of said number N.

[0022] In one particular embodiment, the architecture comprises four neural networks.

[0023] In a particular embodiment, the architecture is pre-trained during a training step, said training step comprising a first pre-conditioning step on synthesis data and a second adjustment step on data from at least one measurement campaign.

[0024] The first step aims to precondition the architecture, and more specifically the initial neural network, to begin training on real-world data with weights as close as possible to the optimal weights. This minimizes the penalty imposed by the small size of the databases from measurement campaigns. The second step refines these weights to ensure they are as close as possible to the optimal values. This two-stage training procedure optimizes the estimation of the propagation channel statistics.

[0025] In a particular embodiment, the propagation channel is a frequency-flat channel.

[0026] In one particular embodiment, the propagation channel is a frequency-selective channel.

[0027] Another object of the invention relates to a statistical estimation device in a communication network for the implementation of a statistical estimation method for a propagation channel according to an object of the invention, said propagation channel allowing the transmission of a communication signal between a transmitting node and a receiving node, said propagation channel being a Rice channel type channel in the presence of masking.

[0028] The proposed estimation device is well-suited to the problem at hand, unlike prior art approaches that use conventional architectures without adaptations. The device of the invention is also well-suited for estimating the statistical parameters of the propagation channel, with the possibility of including additional statistical information beyond just channel samples.

[0029] The present invention will be better understood upon reading the detailed description of embodiments taken by way of non-limiting examples and illustrated by the accompanying drawings, in which: [ Fig 1 ] there figure 1 illustrates a communication network between a sending node and a receiving node with a resource allocator; Fig 2 ] there figure 2 illustrates the different stages of a method for estimating the statistics of a propagation channel according to the invention; [ Fig 3 ] there figure 3 represents a device for estimating statistics for the implementation of the process of the figure 2 .

[0030] The invention is not limited to the embodiments and variants shown, and other embodiments and variants will be obvious to a person skilled in the art.

[0031] In the description that follows, we adopt the standard in which vectors are referenced in bold with respect to scalars.

[0032] There figure 1 illustrates a communication network 1 with a resource allocator.

[0033] This communication network 1 includes a transmitting node NE, a receiving node NR, and a terminal N GR having a resource management function.

[0034] The transmitting node NE and the receiving node NR communicate via a Rice channel propagation channel 10. The receiving node NR communicates with the terminal N GR via another propagation channel 20.

[0035] The NE transmitter node is suitable for transmitting N SA learning sequences at the receiver node NR. The receiver node NR is capable of estimating the propagation channel 10. These statistics are then transmitted to the terminal N GR so that it can determine a resource allocation. Resource allocation consists, for example, of determining a transmission power given the knowledge of the propagation channel statistics 10. In another embodiment, the receiver node NR and the terminal N GR are identical.

[0036] The method for estimating the statistics of propagation channel 10 is described in the figure 2 It includes: a step E1 of determining a set of N estimates ( , ..., , ... ) ; a pre-processing step E2 of all the N estimates ( , ..., , ... ); an E3 step for processing intermediate data; an E4 step for determining statistical information; an E5 step for determining the variance of the estimation noise; an E6 step for determining statistical parameters K, m ;

[0037] Step E1 is suitable for determining the set of N estimates , ..., , ... of the propagation channel from the learning sequences.

[0038] In the case of a frequency-flat channel, the impulse response of the propagation channel is constituted at each instant i of a single random complex coefficient h i which can be written according to this model: h i = ac i e jθ 0 + R i

[0039] Or ac i e iθ 0< corresponds to a part LdV and R i corresponds to a part nLdV. More precisely, a corresponds to the amplitude of the LdV component, c i is a random variable that follows a Nakagami-m distribution and represents the "shadowing" of the LdV component, θ 0 is the phase of the LdV component, and R i ∼ CN 0 , 2 σ h 2 Or CN 0 , 2 σ h 2 corresponds to the centered normal distribution and variance 2 σ h 2 . The probability density of c i can be written as follows: f c i x = 2 m m Γ m x 2 m − 1 e − mx 2 , ∀ x ≥ 0 .

[0040] We assume that the "shadowing" and the nLdV component vary independently from one measurement instant to another, that is, ∀ i ≠ j c i And c j are independent and identically distributed, just like R i And R j However, the approach proposed in the present invention can be extended to the case of block shadowing, that is, constant shadowing on T s consecutive samples.

[0041] It should be noted that the channel model considered includes the Rice channel without "shadowing" as a special case when m→ +∞.

[0042] It is assumed that the propagation channel is estimated, for example via the use of the training sequence, also called the pilot sequence. Thus, an estimated coefficient of the propagation channel can be written as follows: h ˜ i = h i + b i , Or b i ∼ CN 0 , 2 σ b 2 represents a complex, Gaussian white estimation noise, whose variance 2 σ b 2 may be known (for example via a radio equipment calibration process) or unknown.

[0043] It should be noted that the analytical expression of the likelihood of h̃ i is provided in the prior art and involves parabolic cylindrical functions, making the maximum likelihood estimator too complex to implement for use in embedded systems.

[0044] This model can be extended as follows to the case of a frequency-selective channel. The impulse response of a frequency-selective channel can be written h = [ h 0< ,..., h L- 1< ] , Or L represents the number of paths in the propagation channel. At each instant i, each input h i l of the vector h can be written h i l = a l c i l e jθ 0 l + R i l ,

[0045] where the LdV and nLdV components are independent between the different paths.

[0046] Also, the estimation of the channel's impulse response is noted as follows: h ˜ i = h ˜ i 0 , … , h ˜ i L − 1 .

[0047] In the context of the present invention, it is assumed that one has N Estimates of the propagation channel are used to estimate Rice's K factor. To adopt a common notation for both a flat channel and a frequency-selective channel, the set of available channel estimates will be denoted as := [ , ..., , ..., ], Or = h̃ i in the case of a frequency-flat channel, and = h̃ i in the case of a frequency-selective channel. The goal here is to estimate the different values ​​of K of each journey noted K̂ 0 , ..., K̂ L and potentially any other statistical channel parameter such as, for example, the different values ​​of the "shadowing" parameter m, denoted m̂ 0 , ..., m̂ L , from .

[0048] Note that the case of a frequency-selective channel can also be treated by considering each of the L paths as a frequency-flat channel.

[0049] In the following description, we focus on the case of a frequency-flat channel.

[0050] The estimation process also includes an E3 step for processing intermediate data. data1 obtained from the set of N estimates , ..., , ... Intermediate data data1 come from an E2 pre-processing of this set of N estimates , ..., , ... .

[0051] Step E3 of the intermediate data processing data1 is capable of taking into account a permutation invariance of the set of N estimates , ..., , ... .

[0052] Indeed, the estimation problem we are trying to solve is invariant under permutation, which means that the estimation of the values ​​of K the path of each journey is the same regardless of the permutation of the elements of Mathematically, this property can be written K̂ = f ( ) = f ( ϕ ( )) Or f represents the estimation procedure, and ϕ ( ) is a permutation of the elements of which can be written ϕ ( ) = [ , ..., ] Or ϕ i ≠ ϕ j ∀ i ≠ j, and ϕ i E [1 ,N ] ∀ i ∈ [1, N ] .

[0053] Note that the pre-processing in step E2 provides the intermediate data data1 in the example of the figure 2 is different from the treatment applied to take into account permutation invariance in step E3.

[0054] Preferably, this preprocessing step E2 is a normalization of the set of N estimates , ..., , ... .

[0055] Step E6 is suitable for determining statistical parameters K,m from the processed intermediate data data2. This intermediate data processed data2 are obtained here as output from step E3 of the intermediate data processing data1. Architecture A comprises several neural networks as illustrated in the figure 3 .

[0056] The data data3 are derived from step E4 of statistical information determination. This step E4 allows for a rough estimation of a factor K from the set of N estimates , ..., , ... This estimate of the factor K will help guide architecture A in its research.

[0057] The data data4 are a variance of the estimation noise determined during step E5. This step E5 corresponds to a calibration of the radio equipment performed beforehand. The data data4 may also contain the N of channel estimates. Alternatively, the data data4 contain the square root of the number N channel estimates. Transmitting such a number or the square root of this number improves the performance of the neural network architecture A.

[0058] The statistical estimation process of the figure 2 will be detailed in support of the description of a 30-point statistical estimation device illustrated in the figure 3 It is recalled that we are considering here a frequency-flat channel, that is to say Vi, = h̃ i

[0059] The statistical estimation device 30 includes: a module 301 for determining the set of N estimates , ..., , ... ; a pre-processing module 302 for said set of N estimates , ..., , ... ; a first neural network NN1; a 303 data storage module data'1 data from the first neural network NN1; a 304 data summation module data'1 derived from the first neural network NN1; a module 305 for determining statistical information data3 ; a second neural network NN2; a module 306 for determining the variance of estimation noise data4 ; a third neural network NN3; a fourth neural network NN4.

[0060] It should be noted from the outset that the neural networks NN1, NN2, NN3, NN4 are here deep neural networks.

[0061] Module 301 is suitable for providing the set of N estimates , ..., , ... . We note respectively And the real part and the imaginary part of each .

[0062] Two branches originate from this module 301: an upper branch and a central branch.

[0063] The upper branch corresponds to a calculation of statistical information. data3 by module 305 from the input data , ..., , ... . The calculation is as follows: the valuesµ 1, R And µ 1, I , which correspond respectively to the empirical averages of And an estimate K̂ MoM parameter K. This estimate can be obtained using the method of moments. These calculations correspond to step E4 of the determination of statistical information of the figure 2 .

[0064] We thus have three pieces of statistical information, which are provided as input to the second neural network NN2. This neural network NN2 includes S 2 outputs and allows data to be provided data5 to the fourth neural network NN4.

[0065] In the middle branch, each And is provided to the pre-processing module 302. This module 302 is adapted to implement the E2 pre-processing step of the figure 2 and thus provide the output data data1. This data data1 feed into the input of the first neural network NN1 to generate 2N output vectors of S 1 input forming the data data'1. In other words, we apply the first neural network NN1 in parallel to And of each This allows for a number of inputs that may vary from one scenario to another. The data data'1 This data is then stored in storage module 303. data'1 are then summed by the 304 summation module during a summation step of the 2N vectors of S 1 entries, said summation being commutative and applicable to all N.

[0066] In a lower branch, the variance of the data estimation noise data4 is determined by module 305 during step E5. This data data4 are given as inputs to the third neural network NN3. This data data4 also contain the N of channel estimates or the square root of that number N. The third neural network then includes two inputs, one for the variance of the estimation noise and one for the number N or its square root and S 3 outputs to generate data data6.

[0067] THE S 1 output from module 304, the S 2 outputs from the second neural network NN2 and the S The three outputs of the third neural network NN3 are concatenated, resulting in a vector of size S 1 + S 2 + S 3 . Finally, this vector was passed as input to the fourth neural network NN4 to provide the statistical parameters K,m.

[0068] Architecture A of neural networks NN1, NN2, NN3, NN4 is pre-trained during a training step which allows to better condition this architecture.

[0069] For the first neural network NN1 associated with the 304 summation module, the permutation invariance of the estimation problem is taken into account in the design of its components in order to drastically accelerate the training procedure and improve the algorithm's generalization capabilities. Indeed, considering permutation invariance eliminates the need to present the first neural network NN1 with every permutation of each element in the training set.

[0070] The number N of entries available to estimate the parameter KThe Rice channel can vary from one scenario to another (for example, depending on the amount of resources allocated to the sender, which will determine the number of channel estimates available), and therefore it is desirable to have an approach that allows for a single learning process that takes into account different values ​​of N. This avoids having to perform different training for each value of N possible in the first neural network, which would complicate the training procedure, and would require having to include as many first neural networks as there are values ​​of N considered.

[0071] The training stage of the first neural network NN1 includes a first pre-conditioning stage on synthetic data and a second fitting stage on data from at least one measurement campaign.

[0072] Learning on synthetic data has two advantages: firstly, it is possible to generate very large databases, and secondly, the data labels are perfectly known.

[0073] Learning from data from at least one measurement campaign has the advantage of limiting the risk of learning simulation artifacts that are not representative of real-world communication. Conventional estimators from the literature can be used to obtain the labels. Alternatively, real communications can be performed over the propagation channel, and their performance measured in terms of packet error rate or bit error rate to deduce the true value of the factor. K by performing a "mapping" between a theoretical error curve and a measured error curve.

[0074] The invention thus offers the following advantages: the use of "deep learning" for estimating the factor K of the Rice channel in the presence of "shadowing"; the selection of a neural network architecture well suited to the problem being addressed based on the identification of particular properties and constraints of this problem; an insertion of additional statistical information as input to the neural network architecture; a two-stage training procedure in a context of estimating the statistics of the propagation channel.

[0075] The solution proposed by the invention can be adapted to any other propagation channel model for which the use of conventional estimation tools is too complex and / or whose performance is insufficient.

[0076] The method of estimating propagation channel statistics can advantageously be implemented in a basic transmission station of a wireless communication network.

Claims

1. Estimation method for estimating statistics of a propagation channel (10), said propagation channel (10) allowing transmission of a communication signal between a transmitting node (NE) and a receiving node (NR), said propagation channel (10) being a Rician channel in the presence of masking, said communication signal comprising a learning sequence (LS), said estimation method comprising: - a step (E1) of determining a set of N estimations ( ... , , ... ) of the propagation channel (10) from said learning sequence (LS) transmitted N times by the transmitting node (NE); - a step of pre-processing (E2) said set of N estimations (, ..., , ... ) to obtain intermediate data (data1); - a step (E3) of processing the intermediate data (data1) to obtain processed intermediate data (data2), said processing step (E3) being able to take account of an invariance by permutation of said intermediate data (data1) whatever N; a step (E6) of determining the statistical parameters (K, m) from the processed intermediate data (data2); said step (E3) of processing intermediate data (data1) and said step (E6) of determining statistical parameters (K, m) being carried out by an architecture (A) of at least one neural network (NN1, NN2, NN3, NN4).

2. Estimation method according to claim 1, wherein the step (E3) of processing intermediate data (data1) comprises: - a step of processing by a first neutral network (NN1) having an input S1 outputs, said first neutral network (NN1) being able to receive, at the input, intermediate data obtained from a real part ( ) and an imaginary part () of each estimation () of the propagation channel (10) and to provide at the output 2N input vectors S1 (data'1); - a step of summing the 2N input vectors S1 (data'1), said summing being commutative and applicable for all N.

3. Estimation method according to any one of claims 1 or 2, wherein the pre-processing (E2) is a standardisation of the set of N estimations (, ... , , ... ).

4. Estimation method according to any one of claims 1 to 3, wherein said estimation method comprises a step (E4) of determining statistical information (data3) from the set of N estimations ( ... , , ... ), said statistical information (data3) being intended to be transmitted to the neural network architecture (A).

5. Estimation method according to claim 4, wherein the step (E4) of determining statistical information comprises an estimation of a factor K of the propagation channel (10).

6. Estimation method according to any one of claims 4 or 5, wherein said resource allocation method comprises a step (E5) of determining a variance of the estimation noise (data4), said variance of the estimation noise being intended to be transmitted to the architecture (A), said variance of the noise being associated with the number N of estimations or with the square root of said number N.

7. Estimation method according to any one of claims 1 to 6, wherein the architecture (A) comprises four neural networks (NN1, NN2, NN3, NN4).

8. Estimation method according to any one of claims 1 to 7, wherein the architecture (A) is trained beforehand during a training step, said training step comprising a first preconditioning step on synthesis data and a second adjustment step on data coming from at least one measuring campaign.

9. Estimation method according to any one of claims 1 to 8, wherein the propagation channel (10) is a flat frequency channel.

10. Estimation method according to any one of claims 1 to 8, wherein the propagation channel (10) is a selective frequency channel.

11. Estimation device for estimating statistics in a communication network (1) for implementing an estimation method for estimating statistics of a propagation channel (10) according to any one of claims 1 to 10, said propagation channel (10) allowing transmission of a communication signal between a transmitting node (NE) and a receiving node (NR), said propagation channel (10) being a Rician channel in the presence of masking.