WO2025193133A1

WO2025193133A1 - Communication device for providing computation method for radio channel estimation and method thereof

Info

Publication number: WO2025193133A1
Application number: PCT/SE2024/050234
Authority: WO
Inventors: Karl Werner; Rakesh Ranjan; Niklas JALDÉN
Original assignee: Telefonaktiebolaget LM Ericsson AB
Current assignee: Telefonaktiebolaget LM Ericsson AB
Priority date: 2024-03-14
Filing date: 2024-03-14
Publication date: 2025-09-18
Anticipated expiration: 2026-09-14

Abstract

34 ABSTRACT Embodiments of present disclosure provide UE (200) and method (300) for providing computation method for radio channel estimation. First computation method from the plurality of computation methods is selected for estimating the radio channel. The first computation method is selected according to method selection criterion. The first 5 computation method is applied over input signal obtained over wireless network (106) to compute the radio channel estimation for use in data reception through the communication device (1000). (FIG. 4) 10

Description

COMMUNICATION DEVICE FOR PROVIDING COMPUTATION METHOD FOR RADIO CHANNEL ESTIMATION AND METHOD THEREOF

TECHNICAL FIELD

The present disclosure relates generally to a field of a wireless network. More particularly, it relates to radio channel estimation in a communication device of the wireless network.

BACKGROUND

In order to demodulate an uplink transmission, it is necessary to estimate a radio channel for the uplink transmission. The estimated radio channel is used to correct phase and amplitude of the signals by coherently combining the signals from multiple antennas, in a frequency selective way for demodulating the uplink transmission. At present, in New Radio, NR, and Long-Term Evolution, LTE, the radio channel is estimated using Demodulation Reference Signals, DMRS. The DMRS are signals that are known to a receiver. The radio channel estimation is further used to determine radio channel quality, for the uplink transmission. For example, interleaving the DMRS in the physical resources refers to the practice of mixing different types of signals or data to allow efficient use of available resources in the wireless network for demodulating the uplink transmission. Further, the DMRS is used by the NR receiver to produce channel estimates for demodulation of the data received at a physical layer or decoding of a transport block/data packet. Additionally, terminals such as mobile phone have low output power as compared to base stations in the wireless network. Due to this low output power of the terminals an uplink coverage for the uplink transmission becomes challenging in absence of a channel estimate. The channel estimate is required for the uplink coverage. For example, the impact of the low output power of the terminals on the uplink coverage results in reduced network coverage, lower throughput etc. Therefore, the channel estimate for the uplink transmission is crucial for the demodulation of the uplink transmission.

Currently, in Orthogonal Frequency-Division Multiplexing, OFDM, demodulation is used as a simplified method for decoding a transport block in the uplink transmission for the NR and the LTE. In the OFDM demodulation, an input such as time domain signal is partitioned into OFDM symbols and transformed to a subcarrier representation using Discrete Fourier Transform, DFT. Further, extraction of the DM RS from a subset of the subcarrier is performed. Then, the radio channel estimation using the DMRS is performed which results in a channel estimate per subcarrier, transmit layer and receive rx-antenna/beam in the uplink transmission. Further, linear equalization of resource elements such as one subcarrier in one OFDM symbol carrying the data, is performed to obtain an output of equalized modulation symbols per resource element and the physical layer, for example, Multiple Input Multiple Output, MIMO layer. The linear equalization may be performed by setting equalizer weights to achieve minimum error in the output of equalized modulation symbols, wherein the linear equalization weights are selected according to the channel estimate. Soft bit computation and Low-Density Parity Check, LDPC decoding, or, in the case of LTE, turbo decoding may be used to compute cyclic redundancy check, CRC while decoding the transport block. The CRC allows the receiver to determine if the transport block was correctly decoded or not.

Use of Artificial Intelligence, Al, or Machine Learning, ML, based data-driven solutions is regarded as a key leverage to customize radio systems for diverse radio environments. This is especially important when complexity in Radio Access Networks, RANs, is growing for each new generation such as 5G and 6G. The learning capability of the Al provides better understanding of data by analysing an underlying relation between data components in current radio environments instead of human logics, symbolic modelling and analysis which may not be preferred when data is more complex.

The Al or ML enabled solutions essentially employ data-driven learning approaches where models learn data distribution and relationships between inputs and outputs without the need for understanding inherent complex processes. The Al and ML mainly rely on statistical techniques such as classification, clustering, Bayesian interference etc for employingthe data- driven learning approaches. On the contrary, existing/legacy solutions could be seen as model-driven where a solution is derived based on a simplified model of a problem. In the area of channel estimation, examples of this is the legacy algorithm based on Discrete Cosine Transform, DCT, or Akaike Information Criterion, AIC.

The data-driven learning approach forms a promising application of machine learning in the radio systems. The ML based solutions may be helpful in designing transceivers having better adaptability to radio environment based on prior performance prediction of the radio components. The ML based solutions may provide prediction of possible performance gains of radio components compared to general modelling or inaccurate modelling-based systems that otherwise lack analysis of complex data. Thus, the ML based solutions may contribute in reducing product design cycles with more general modelling.

SUMMARY

The ML based solution for channel estimation potentially has better performance than legacy or existing methods. However, the ML models come with a risk of non-generalizability of data and thus may not give desired channel estimation when considered as a standalone solution in a new site or a cell where there is less data, and the ML model may not be well trained. For example, if a ML model is trained in a given environment it may not perform well if the environment is changed. With a traditional/existing, model-based approach, more robustness may be expected. The ML models characteristically also bring higher complexities leading to costlier hardware and higher energy consumption.

Therefore, there is a need for an improved method for radio channel estimation.

It is therefore an object of the present disclosure to provide a communication device and a method for radio channel estimation adapted to mitigate, alleviate, or eliminate all or at least some of the above-discussed drawbacks of presently known solutions.

This and other objects are achieved by means of the communication device and a method defined in the appended claims. The term exemplary is in the present context to be understood as serving as an instance, example or illustration.

According to a first aspect of the present disclosure, a computer implemented method implemented in a communication device of a wireless network for radio channel estimation is disclosed. The method comprises obtaining at least one input signal over the wireless network. The method comprises selecting a first computation method from a plurality of computation methods usable in the radio channel estimation. The first computation method is selected according to a method selection criterion. The method comprises applying the selected first computation method over said obtained at least one input signal to compute the radio channel estimation for use in data reception through the communication device.

Optionally, the at least one input signal comprises at least one of a reference signal and data symbols. Optionally, the method selection criterion comprises one of: past results of the radio channel estimation, made using at least one computation method, optionally including the first computational method, from the plurality of computation methods, past CRC results, a computational load associated with the at least one computation method. Link performance for the radio channel, characteristics of the at least one input signal, said characteristics comprising at least one of a Signal-to-noise ratio, SNR, a Block Error rate, and a spatial distribution of the at least one input signal. Latency in communication.

Optionally, the method comprises decoding a data packet received through the communication device to obtain a decoded data packet, by usingthe radio channel estimation computed by using the first computation method. The method comprises identifying whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, identifying whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet comprises performing an error check for the decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied overthe decoded data packetto obtain a CRC result forthe decoded data packet. Further, identifying that the decoded data packet is a successfully decoded data packet when the CRC result indicates a no CRC error; or that the decoded data packet is a not successfully decoded data packet when the CRC result indicates a CRC error .

Optionally, the method comprises updating the method selection criterion to obtain an updated method selection criterion according to the identification of whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, the method comprises selecting a second computation method from said plurality of computation methods when the decoded data packet has been identified as the not successfully decoded data packet. Further, the method comprises computing the radio channel estimation based on an application of the second computation method over the at least one input signal. The method comprises decoding the data packet received through the communication device to obtain a decoded data packet using the radio channel estimation computed through the second computation method. Further, identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet. Updating the method selection criterion to obtain an updated method selection criterion, according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, the method comprises selecting a second computation method from said plurality of computation methods when the decoded data packet decoded using the radio channel estimation computed through the first computation method is identified as a successfully decoded data packet. The method comprises computing the radio channel estimation based on an application of the second computation method over the at least one input signal. Further, decoding the data packet received through the communication device to obtain the decoded data packet using the radio channel estimation computed through the second computation method. Identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet. Updating the method selection criterion to obtain the updated method selection criterion according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, the identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet comprises performing an error check for the decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied over the decoded data packet to obtain a CRC result of the decoded data packet. Further, identifying the decoded data packet as a successfully decoded data packet when the CRC result indicates a no CRC error and as a not successfully decoded data packet when the CRC result comprises a CRC error . Further, updating the method selection criterion to obtain the updated method selection criterion, according to the identification of whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet using the radio channel estimation computed through the second computation method.

Optionally, the method comprises training the second computation method according to the updated method selection criterion.

Optionally, the method comprises comparing the CRC result obtained for the first computation method with the CRC result obtained for the second computation method. Updating the method selection criterion to obtain the updated method selection criterion based on the comparison.

Optionally, the plurality of computation methods comprise a model-driven method comprising an Akaike information criterion, AIC, method.

Optionally, the plurality of computation methods comprise a data-driven method comprising a machine learning, ML, method.

According to a second aspect of the present disclosure, a communication device for radio channel estimation in a wireless network is disclosed. The communication device comprises a controlling circuitry is arranged for obtaining at least one input signal over the wireless network. The controlling circuitry is arranged for selecting a first computation method from a plurality of computation methods usable in the radio channel estimation. The first computation method is selected according to a method selection criterion. Further, the controlling circuitry is arranged for applying the selected first computation method over said obtained at least one input signal to compute the radio channel estimation of use in data reception through the communication device.

According to a third aspect of the present disclosure, there is provided a computer program product comprising a non-transitory computer readable medium, having thereon a computer program comprising program instructions. The computer program is loadable into a data processing unit and configured to cause execution of the method according to the first, second and third aspects when the computer program is run by the data processing unit.

Some embodiments disclosed herein have one or more of the following advantages: - The proposed communication device and method provides a selection of an appropriate computation method to be used often for correctly estimating the radio channel by using a CRC check on decoded data. In particular, the data is decoded using the estimated radio channel.

- If the communication device uses the first computation method, such as a model- driven method, and if the model-driven method results in successful decoding of the data in 90% of the cases, then the second computation method, such as a complex ML model, may be needed only 10% of the time on an average for decoding of the data. The second computation method is characteristically the more complex computation method only for more than 10 percent of error rate in the data reception through the communication device, while the first will be used on all data receptions when the error rate is less or more than 10 %.

- Further, another computation method, i.e., a ML model learns from criterion selection data and may predict better radio components through the radio channel estimation. Such improved radio channel estimation may result in adaptability to radio channel conditions on a specific site for the downlink transmission.

- The present communication device and method increases a probability of decoding of the transport block by using the ML method. The ML based method provides the radio channel estimation with higher accuracy and less errors as the ML based method provides improved analysis of the input signal. Such improved radio channel estimation is beneficial for the performance of the communication device, as it allows more efficient use of radio resources through the communication device.

- The present communication device and method provides the selection of the computation method to provide a balance in computational complexity of the computational method and performance of the computation method in terms of throughput, robustness, reliability etc. Since the computation method is selected to compute the radio channel estimation and then the selection criteria are updated based on results of the decoding through the radio channel estimation, such update of the selection criterion for further selection of the computation method provides an optimization in computational load for channel estimation and link performance using machine learning. Other advantages may be readily apparent to one having skill in the art. Certain embodiments may have none, some, or all of the recited advantages.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing will be apparent from the following more particular description of the example embodiments, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the example embodiments.

FIG. 1 discloses a wireless communication system according to some examples;

FIG. 2 is a schematic block diagram illustrating an example communication device for radio channel estimation, according to some embodiments;

FIGs. 3A-3G are flowcharts illustrating example steps for a method for radio channel estimation, according to some embodiments;

FIG. 4 provides an example flow of selection of computation method for the radio channel estimation, according to some embodiments;

FIG. 5 provides an example of spatial beams used in the selection for selecting the computation method, according to some embodiments;

FIG. 6 provides an example implementation of the method for radio channel estimation, according to some embodiments; and

FIG. 7 discloses an example computing environment according to some embodiments.

DETAILED DESCRIPTION

Aspects of the present disclosure will be described more fully hereinafter with reference to the accompanying drawings. The apparatus and methods disclosed herein can, however, be realized in many different forms and should not be construed as being limited to the aspects set forth herein. Like numbers in the drawings refer to like elements throughout.

The terminology used herein is for the purpose of describing particular aspects of the disclosure only and is not intended to limit the invention. It should be emphasized that the term "comprises/comprising" when used in this specification is taken to specify the presence of stated features, integers, steps, or components, but does not preclude the presence or addition of one or more other features, integers, steps, components, or groups thereof. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

Embodiments of the present disclosure will be described and exemplified more fully hereinafter with reference to the accompanying drawings. The solutions disclosed herein can, however, be realized in many different forms and should not be construed as being limited to the embodiments set forth herein.

It will be appreciated that when the present disclosure is described in terms of a platform and a method, it may also be embodied in one or more processors and one or more memories coupled to the one or more processors, wherein the one or more memories store one or more programs that perform the steps, services and functions disclosed herein when executed by the one or more processors.

FIG. 1 discloses an example wireless communication system 100. Although the subject matter described herein may be implemented in any appropriate type of system using any suitable components, the embodiments disclosed herein are described in relation to a wireless communication system/wireless network, such as the example wireless communication system 100 described in FIG. 1.

The wireless communication system 100 may comprise and/or interface with any type of communication, telecommunication, data, cellular, and/or radio network or other similar type of system. In some embodiments, the wireless communication system 100 may be configured to operate according to specific standards or other types of predefined rules of procedures. Thus, particular embodiments of the wireless communication system 100 may implement communication standards, such as, but not limited to, global system for mobile communications, GSM, universal mobile telecommunications system, UMTS, long term evolution, LTE, and/or other suitable 2G, 3G, 4G, or 5G standards, wireless local area network, WLAN, standards such as, IEEE 802.11 standards, and/or any other appropriate wireless communication standards, such as, worldwide interoperability for microwave access, WiMax, Bluetooth, Z-Wave and/or ZigBee standards.

For simplicity, as depicted in FIG. 1, the wireless communication system 100 comprises a communication device 1000 for example, a User Equipment, UE 200, or a network node 800 arranged in a wireless communication network/network 106. The UE 200 and the network node 800 operate together in order to provide wireless connections in the wireless communication system 100. The network 106 may comprise one or more backhaul networks, core networks, IP networks, public switched telephone networks, PSTNs, packet data networks, optical networks, wide-area networks, WANs, local area networks, LANs, wireless local area networks, WLANs, wired networks, wireless networks, metropolitan area networks, and other networks to enable communication between devices (for example, wireless devices and network node).

The network node 800 may refer to equipment capable, configured, arranged, and/or operable to communicate directly or indirectly with the UE 200 and/or with other network nodes or equipment in the wireless communication system 100 to enable and/or provide wireless access to the UE 200 and/or to perform other functions (for example, administration) in the wireless communication system 100. Examples of the network node 800 may include, but not limited to, access points, APs (for example, radio access points), base stations, BSs (for example, radio base stations, nodeBs, evolved NodeBs, eNBs, new radio, NR, nodes (gNBs), or the like). The BSs may be categorized based on an amount of coverage the BSs provide (or, stated different, their transmit power level) and may then also be referred to as femto BSs, pico BSs, micro BSs, macro BSs. The BS may be a relay node or a relay donor node controlling a relay.

The UE 200 may refer to a device capable, configured, arranged and/or operable to communicate wirelessly with the network node 800 and/or other wireless devices.

In some examples, the UE 200 may include one or more of: computing devices, wireless devices, ultra-low power wireless devices, Internet of Things, loT, devices, and so on.

Examples of the computing devices may include, but not limited to, a smart phone, a mobile phone, a cell phone, a voice over Internet Protocol, IP, VoIP, phone, a wireless local loop phone, a desktop computer, a personal digital assistant, PDA, a wireless camera, a gaming console or device, a wearable terminal device, a wireless endpoint, a mobile station, a tablet, a laptop, a laptop-embedded equipment, LEE, a laptop-mounted equipment, LME, a smart device, a wireless customer-premise equipment, CPE, a vehicle- mounted wireless terminal device, and so on. It should be understood that the UE 200 may not be limited to the above-described wireless devices. The UE 200 may be extended to other wireless devices of different classes or categories providing different services while supporting, for example, Enhanced Mobile Broadband, eMBB, massive Machine-Type Communication, MTC, Ultra-Reliable Low Latency Communication, URLLC, Time Sensitive Networking, TSN, or the like.

In the wireless communication system 100, the network node 800 and the UE 200 are connected through for example, 3GPP 5G core network where specific network services and operations are provided through software components called network functions, NFs. The wireless communication system 100 hosts large scale applications.

In many scenarios, the Machine Learning, ML, based solutions provide statistical analysis of data and identifies the data distribution in radio systems and thus the ML based methods may be preferred for channel estimation over legacy or existing methods which relies on conventional approach of problem modelling without considering complexity of data. Though ML models are preferred for the above reasons, however, the ML models come with a risk of non-generalizability of data and thus may not give desired channel estimation when considered as a standalone solution in a new site or a cell where there is less data, and the ML model may not be well trained. For example, if a ML model is trained in a given environment it may not perform well if the environment is changed. With a traditional/existing, model-based approach, more robustness may be expected. The ML models also bring higher complexities leading to costlier hardware and higher energy consumption.

Thus, the present disclosure enables the wireless communication network/network 106, the network node 800 and the UE 200, examples of the communication device 1000 to enable one of the UE 200 or the node 800 for radio channel estimation to reduce costlier hardware and higher energy consumption during the radio channel estimation.

The present communication device and method provides the selection of the computation method to provide a balance in computational complexity of the computational method and performance of the computation method in terms of throughput, robustness, reliability etc. Since the computation method is selected to compute the radio channel estimation and then update the selection criteria based on results of the decoding through the radio channel estimation, such update in the selection criterion for further selection of the computation method provides an optimization in computational load for channel estimation and link performance using machine learning.

FIG. 2 is a schematic block diagram illustrating an example communication device 1000 for radio channel estimation for receiving or transmitting data through the communication device 1000.

The communication device 1000 comprises a processor 1004 arranged to execute the instructions stored in a non-transitory computer readable medium i.e., memory 1002. The communication device 1000 comprises a controlling circuitry 1006 communicatively coupled to the processor 1004 for selecting a first computation method from a plurality of computation methods, usable in the radio channel estimation for data communication through the communication device 1000.

The communication device 1000 further comprises a transceiver 1008 communicatively coupled to the controlling circuitry 1006 for selecting the first computation method from the plurality of computation methods, usable in the radio channel estimation for receiving or transmitting the data through the communication device 1000, for example through the UE 200 or the node 800.

The controlling circuitry 1006 is arranged for obtaining at least one input signal over the wireless network 106. Further, the controlling circuitry 1006 is arranged for selecting the first computation method from the plurality of computation methods to be used in the radio channel estimation. The first computation method is selected according to the method selection criterion. Thereafter, the controlling circuitry 1006 is arranged for applying the selected first computation method over said obtained at least one input signal to compute the radio channel estimation for use in data reception through the communication device 1000.

Optionally, the at least one input signal comprises at least one of a reference signal and data symbols. Optionally, the method selection criterion comprises one of: past results of the radio channel estimation, made using at least one computation method, optionally including the first computational method, from the plurality of computation methods. A computational load associated with the at least one computation method. Link performance for the radio channel, characteristics of the at least one input signal, said characteristics comprising at least one of a Signal-to-noise ratio, SNR, a Block Error rate, and a spatial distribution of the at least one input signal and latency in communication.

Optionally, the controlling circuitry 1006 is arranged for decoding a data packet received through the communication device to obtain a decoded data packet, by using the radio channel estimation computed by using the first computation method. Further, the controlling circuitry 1006 is arranged for identifying whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, identifying whetherthe decoded data packet is a successfully decoded data packet or a not successfully decoded data packet comprises performing an error check for the decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied overthe decoded data packetto obtain a CRC result forthe decoded data packet. Further, identifying that the decoded data packet is a successfully decoded data packet when the CRC result indicates a no CRC error; or that the decoded data packet is a not successfully decoded data packet when the CRC result indicates a CRC error.

Optionally, the controlling circuitry 1006 is arranged for updating the method selection criterion to obtain an updated method selection criterion according to the identification of whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, the controlling circuitry 1006 is arranged for selecting a second computation method from said plurality of computation methods when the decoded data packet has been identified as the not successfully decoded data packet. Further, the controlling circuitry 1006 is arranged for computing the radio channel estimation based on an application of the second computation method over the at least one input signal. The controlling circuitry 1006 is arranged for decoding the data packet received through the communication device to obtain the decoded data packet using the radio channel estimation computed through the second computation method. Further, identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet. The controlling circuitry 1006 is arranged for updating the method selection criterion to obtain an updated method selection criterion, according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, the controlling circuitry 1006 is arranged for selecting a second computation method from said plurality of computation methods when the decoded data packet decoded using the radio channel estimation computed through the first computation method is identified as the successfully decoded data packet. Further, the controlling circuitry 1006 is arranged for computing the radio channel estimation based on an application of the second computation method over the at least one input signal. Further, decoding the data packet received through the communication device to obtain the decoded data packet using the radio channel estimation computed through the second computation method. The controlling circuitry 1006 is arranged for identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet. The controlling circuitry 1006 is arranged for updating the method selection criterion to obtain the updated method selection criterion according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

Optionally, the identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet comprises performing an error check for the decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied over the decoded data packet to obtain a CRC result of the decoded data packet. Further, the controlling circuitry 1006 is arranged for identifying the decoded data packet as a successfully decoded data packet when the CRC result indicates a no CRC error and as a not successfully decoded data packet when the CRC result indicates a CRC error. Further, the controlling circuitry 1006 is arranged for updating the method selection criterion to obtain the updated method selection criterion, according to the identification of whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet using the radio channel estimation computed through the second computation method.

Optionally, the controlling circuitry 1006 is arranged for training the second computation method according to the updated method selection criterion.

Optionally, the controlling circuitry 1006 is arranged for comparing the CRC result obtained for the first computation method with the CRC result obtained for the second computation method. The controlling circuitry 1006 is arranged for updating the method selection criterion to obtain the updated method selection criterion based on the comparison.

Figs 3A-3G shows schematic block diagram of a method 300, also referred as a computer implemented method 300 implemented in the communication device 1000 of the wireless network 106 for the radio channel estimation to be used in data reception through the communication device 1000.

At step 302, the method 300 provides obtaining at least one input signal over the wireless network 106.

At step 304, the method 300 provides selecting the first computation method from the plurality of computation methods usable in the radio channel estimation. The first computation method is selected according to the method selection criterion. At step 306, the method 300 provides applying the selected first computation method over the said obtained at least one input signal to compute the radio channel estimation for use in data reception through the communication device 1000.

In an example, FIG. 3B shows additional steps of the method 300 implemented in the communication device 1000 of the wireless network 106 for the radio channel estimation to be used in data reception through the communication device 1000.

At step 308, the method 300 provides decoding the data packet received through the communication device 1000 to obtain the decoded data packet, by using the radio channel estimation computed by using the first computation method.

At step 310, the method 300 provides identifying whether the decoded data packet is the successfully decoded data packet or the not successfully decoded data packet.

In an example, FIG. 3C shows steps of the method 300 implemented in the communication device 1000 of the wireless network 106 for the radio channel estimation to be used in data reception through the communication device 1000.

At step 312, the method 300 provides performing the error check for the decoded data packet through the at least one error check method applied over the decoded data packet. The at least one error check method comprises the CRC applied over the decoded data packet to obtain the CRC result for the decoded data packet.

At step 314, the method 300 provides identifying that the decoded data packet is a successfully decoded data packet when the CRC result indicates a no CRC error or that the decoded data packet is not a successfully decoded data packet when the CRC result indicates a CRC error.

For example, all commonly used CRCs employ two elements to indicate the CRC result. The two elements are usually called 0 and 1. In an example, the decoded data packet is a successfully decoded data packet when the CRC result is 1 or "no CRC error". In line with this example, the decoded data packet is a not successfully decoded data packet when the CRC result is 0 or "CRC error".

At step 316, the method 300 provides updating the method selection criterion to obtain the updated method selection criterion according to the identification of whether the decoded data packet is the successfully decoded data packet or the not successfully decoded data packet.

In an example, FIG. 3D shows steps of the method 300 implemented in the communication device 1000 of the wireless network 106 for the radio channel estimation to be used in data reception through the communication device 1000.

At step 318, the method 300 provides selecting the second computation method from the plurality of computation methods when the decoded data packet has been identified as the not successfully decoded data packet.

At step 320, the method 300 provides computing the radio channel estimation based on the application of the second computation method over the at least one input signal.

At step 322, the method 300 provides decoding the data packet received through the communication device 1000 to obtain the decoded data packet using the radio channel estimation computed through the second computation method.

At step 324, the method 300 provides identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is the successfully decoded data packet or the not successfully decoded data packet.

At step 326, the method 300 provides updating the method selection criterion to obtain the updated method selection criterion, according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is the successfully decoded data packet or the not successfully decoded data packet.

At step 344, the method 300 provides training the second computation method according to the updated method selection criterion.

In an example, FIG. 3E shows steps of the method 300 implemented in the communication device 1000 of the wireless network 106 for the radio channel estimation to be used in data reception through the communication device 1000.

At step 328, the method 300 provides selecting the second computation method from the said plurality of computation methods when the decoded data packet decoded using the radio channel estimation computed through the first computation method is identified as the successfully decoded data packet.

At step 330, the method 300 provides computing the radio channel estimation based on the application of the second computation method over the at least one input signal.

At step 332, the method 300 provides decoding the data packet received through the communication device 1000 to obtain the decoded data packet using the radio channel estimation computed through the second computation method.

At step 334, the method 300 provides identifying whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is the successfully decoded data packet or the not successfully decoded data packet.

At step 336, the method 300 provides updating the method selection criterion to obtain the updated method selection criterion according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is the successfully decoded data packet or the not successfully decoded data packet.

In an example, FIG. 3F shows steps of the method 300 implemented in the communication device 1000 of the wireless network 106 for the radio channel estimation to be used in data reception through the communication device 1000.

At step 338, the method 300 provides performing the errorcheckforthe decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises the CRC applied over the decoded data packet to obtain the CRC result of the decoded data packet.

At step 340, the method 300 provides identifying the decoded data packet as the successfully decoded data packet when the CRC result indicates a no CRC error and as the not successfully decoded data packet when the CRC result indicates a CRC error. At step 342, the method 300 provides updating the method selection criterion to obtain the updated method selection criterion, according to the identification of whether the decoded data packet is the successfully decoded data packet or the not successfully decoded data packet using the radio channel estimation computed through the second computation method.

In an example, FIG. 3G shows steps of the method 300 implemented in the communication device 1000 of the wireless network 106 for the radio channel estimation to be used in data reception through the communication device 1000.

At step 346, the method 300 provides comparing the CRC result obtained for the first computation method with the CRC result obtained for the second computation method.

At step 348, the method 300 provides updating the method selection criterion to obtain the updated method selection criterion based on the comparison.

In an example, the proposed communication device 1000 and the proposed method 300 provides selection between two channel estimation methods, i.e., the first and the second computation method, for the radio channel estimation. In another example, the proposed communication device 1000 and the proposed method 300 allow optimization of computational load for the radio channel estimation and link performance using the second computation method/ML method.

In an example, the selection of either the first or the second computation method for the radio channel estimation at each given time or link, is based according to the method selection criterion, alternatively referred as a criterion function. This criterion function may be updated based on historical data related to the computation methods. In an example, the proposed method 300 implemented in the computing device 1000 may either select one or both the computation methods to decode the single transport block. In case both the computation method, i.e., the first and the second computation method are used, they may operate on the same data but may produce two resulting decoded transport blocks, respectively. Further, based on the CRC performed on the decoded transport blocks, the proposed method 300 may provide action related to updating the criterion of the radio channel estimation methods, i.e., the first and the second computation method.

In an example, upon applying the selected computation method, if the error is found in one computation method, for example the first computation method, then another channel estimation method/second computation method may be applied on the same data. Upon applying the second computation method the CRC may be performed by the proposed method 300 again and if the CRC passes on the second computation method, the transport block is sent to higher layers (not discussed in detail here), and the selection criterion function is updated. In an example, acknowledgement, ACK, negative acknowledgment, NACK or retransmission request may be used on a first decoded transport block even when a second decoding is performed to provide a second decoded transport block. This is performed to save latency. In another example, the ACK, NACK, or retransmission request may be sent for the second or final decoding attempt.

In another example, the criterion function is optimized in such a manner that for all scenarios where the probability of obtaining correct decoded transport block is above a certain threshold value, the first computation method is applied on the at least one input signal. While the second computation method is applied only when the first computation method provides error or failed decoding/no successful decoding.

In another example, the second computation method is used/applied on the at least one input signal to compute the radio channel estimation even if the first computation method provides a correct CRC. This is performed in order to train the second computation method, when the second computation method is ML model observed deviating from the dataset, i.e., the at least one input signal, that is used to train the ML method/model.

In an example, the proposed communication device 1000 and the proposed method 300 provides the combination of the first computation method and the second computation method for the radio channel estimation. This is to be understood that the proposed communication device 1000 and the proposed method 300 may provide a combination of more number of computation methods for the radio channel estimation further used for decoding the data packet or a transport block. The proposed method 300 implemented through the communication device 1000 is described below using a selection of two independent computation methods, i.e., the first and the second computation method, from the plurality of computation methods. However, it should be understood for a person skilled in the art that it may be extended to three or more parallel methods. If not expressed otherwise, the first computation method or model is the model-driven method, for example, the classical AIC method/model or Discrete Cosine Transform, DCT, whereas the second computation method or model is the data-driven method, for example, the ML trained method/model. In an example, the ML method/model may comprise a supervised ML model or an unsupervised ML model.

A few examples of the implementation of the proposed method 300 are described by simplified examples in detail below. The proposed method 300 may pertain to decoding of Physical Uplink Shared Channel, PUSCH, i.e., NR or LTE, at the node 800, i.e., gNB side. However, in an example, the proposed method 300 be equally well applied to Physical Downlink Shared Channel, PDSCH decoding on the UE 200 side. Further, in another example, the proposed method 300 may be applied to New Radio, NR, Physical Uplink Control Channel, PUCCH, or Long-term Evolution, LTE PUCCH and 6G corresponding channels.

In an example a FIG. 4 shows a flowchart providing additional details of the proposed method 300 on how processing chain may be configured to allow two radio channel estimation computation methodology, i.e., the first computation method and the second computation method, to operate on same data, i.e., the at least one input signals obtained over the wireless network 106. In an example, the at least one input signal is transmitted through one or more receiving ports (not shown in FIG. 4) arranged in the wireless network 106.

As shown in FIG. 4, the proposed method 300 may provide at step 402, an "initial processing" common to both estimators used in the first computation method and the second computation method and need to be performed once for the transport block. The initial processing may comprise, example Orthogonal frequency-division multiplexing, OFDM, demodulation, the Demodulation Reference Signal, DMRS, extraction. After the step 404 of the "criterion" which refers to the method selection function, at step 406, the first computation method, method 1 is applied and data packet is decoded and at step 410, the error check is performed for the decoded data packet using CRC to identify whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet by using the first computation method. In an example, when the CRC error is found and the CRC result of the CRC error is 0, at step 408, the second computation method, method 2 is selected and applied over the at least one input signal and data packet is decoded. At step 412, the CRC error check is performed for the decoded data packet using the CRC to identify whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet by using the channel estimation of the second computation method.

In an example, each of the first and the second computation methods may also compute equalization, softbit generation, Low density parity check, LDPC decoding, along with the CRC computation. Note that all these steps from 402 to 412, i.e., initial processing, criterion, method 1 and 2, and check, depends on the radio channel estimate and thus needs to be performed on an output on each channel estimator which may result in two different error values if both the first and the second models were used. Further, the criterion block i.e., the block of selection function is where the first computation method/model or the second computation method/model or both the first and the second computation method/model is used on the at least one input signal. In some examples the criterion may depend on past outcomes of the CRC error checks for each of the first computation method and the second computation method.

In another example, the model selection criterion of selection function is addressed. The model selection function may be related to signal-to-noise ratio, SNR values. Further, consider that the AIC method/orthe first computation method has lower computation complexity than that of the ML method or the second computation method. In such a scenario, the proposed method 300 may provide an acceptable balance between computation load and throughput that may be achieved, for example at Y dB. In this example, the selection function is a simple SNR threshold where SNR> YdB results in selection of the first computation method for the radio channel estimation and for SNR< YdB the radio channel estimation method 2, i.e., the second computation method/ML method is used.

In an alternative example, the selection function may be more complex, constituting of more than one parameter being considered. On such example is spatial distribution of power. FIG. 5 illustrates a direction 500 of 32 spatial beams, and colored beams illustrating the spatial power distribution of the dataset used for training the estimator, or ML based radio channel estimator that applies the ML model for the radio channel estimation. Suppose that the SNR is Z dB which would in previous example result in method 2 being used. However, the beam with the strongest received power is beam: 8, which is not well represented in the training data set for method 2, hence method 1 is selected as the primary estimator.

In another example FIG. 6 shows a high-level block diagram 600 of the implementation of the proposed method 300 forthe monitoring of application of the first orthe second computation method. After the step 602 of initial processing and criterion selection at step 604, both the first and the second computation method may be applied at steps 606 and 608 for the radio channel estimation and then at steps 610 and 614, the outputs of the first computation method and the computation second method are sent for further processing for the CRC error check i.e., cross checking at step 612. At step 616, the radio channel estimation is transmitted to higher layers for further data communication if the CRC error is within allowable limits based on the comparison between the CRC result of the first computation method and the CRC result of the second computation method.

In an example, in selected situations where the first computation method/model is used and it results in CRC decoded correctly, then the second computation method/model is used as well to update/train the ML radio channel estimator. Herein the received data may or may not be artificially perturbed with for example noise for training purposes. Cross checking of the radio channel estimation results between both the first and the second computation method may be applied, as discussed above in step 616.

In an example, the cross checking at step 616 may be applied through decoding results, i.e., CRCs. For example, if CRC checks for the second computation method/model but not for the first computation method/model, actions such as updating the criterion, i.e., updating SNR thresholds, are carried out. In an alternative, if CRC checks for both the first and the second computation methods, then a soft bit distance metric between decoded and actual, i.e., reencoded, bits may also be used to determine which of the first or the second computation method/model resulted in the appropriate radio channel estimation. Further, the radio channel estimation results may be used to re-train or reset the second computation method/model. The proposed method 300 provides switching to the second computation method/model in the meantime when the error is not within permissible limits, raising an alarm regarding the error of the first computation method and increasing a counter. In another example, computation complexity of the ML model is also addressed through the proposed method 300. Initially, the first computation method/model is used always. Further, the second computation method/model is used if the CRC fails for the first computation method, or according to another criterion, the criterion threshold is adaptively updated by increasing the SNR threshold. By performing this, the ML model/the second computation method only consumes resources when it is needed.

Process of selection of the computation method may be time consuming which may be a drawback. In particular, the process of decoding the same data again suing the second computation method. Such latency drawback may be mitigated if retransmission requests, or NACKs are sent already before decoding of the transport block by using the second computation method/model. Alternatively, this updating of training is only performed for traffic that is known not to be latency sensitive. In one example, the second model is used more often for decoding retransmissions compared to new transmissions.

FIG. 7 illustrates an example-computing environment 700 implementing the communication device 1000 and the methods 300-A to 300-E as shown in FIGs. 3A-3G for providing selection of the computation method for radio channel estimation. As depicted in FIG. 7, the computing environment 700 comprises at least one data processing module 706 that is equipped with a control module 702 and an Arithmetic Logic Unit (ALU) 704, a plurality of networking devices 708 and a plurality Input output, I/O devices 710, a memory 712, a storage 714. The data processing module 706 may be responsible for implementing the platform and method described in FIGs 3A-3G, 4 and 6 respectively. For example, the data processing module 706 in some embodiments is equivalent to the controlling circuitry 1006 of the platform described above in conjunction with FIGs 3A-3G, 4 and 6. The data processing module 706 is capable of executing software instructions stored in memory 712. The data processing module 706 receives commands from the control module 702 in order to perform its processing. Further, any logical and arithmetic operations involved in the execution of the instructions are computed with the help of the ALU 704.

The computer program is loadable into the data processing module 706, which may, for example, be comprised in an electronic apparatus (such as the platform). When loaded into the data processing module 706, the computer program may be stored in the memory 712 associated with or comprised in the data processing module 706. According to some embodiments, the computer program may, when loaded into and run by the data processing module 706, cause execution of method steps according to, for example, any of the methods illustrated in FIGs 3A-3G, 4 and 6, or otherwise described herein.

The overall computing environment 700 may be composed of multiple homogeneous and/or heterogeneous cores, multiple CPUs of different kinds, special media and other accelerators.

Further, the plurality of data processing modules 706 may be located on a single chip or over multiple chips.

The algorithm comprising of instructions and codes required for the implementation are stored in either the memory 712 or the storage 714 or both. At the time of execution, the instructions may be fetched from the corresponding memory 712 and/or storage 714 and executed by the data processing module 706.

In case of any hardware implementations various networking devices 708 or external I/O devices 710 may be connected to the computing environment to support the implementation through the networking devices 708 and the I/O devices 710. The embodiments disclosed herein can be implemented through at least one software program running on at least one hardware device and performing network management functions to control the elements. The elements shown in FIG 7 include blocks which can be at least one of a hardware device, or a combination of hardware device and software module.

Claims

1. A computer implemented method (300) for radio channel estimation, said computer- implemented method (300) being implemented in a communication device (1000) of a wireless network (106), the method (300) comprising: obtaining (302) at least one input signal over the wireless network (106); selecting (304) a first computation method from a plurality of computation methods usable in the radio channel estimation, wherein the first computation method is selected according to a method selection criterion; and applying (306) the selected first computation method over said obtained at least one input signal to compute the radio channel estimation for use in data reception through the communication device (1000).

2. The method (300) according to claim 1, wherein the at least one input signal comprises at least one of a reference signal and data symbols.

3. The method (300) according to any of the preceding claims, wherein the method selection criterion comprises at least one of: past results of the radio channel estimation, made using at least one computation method, optionally the first computational method, from the plurality of computation methods; past CRC results, a computational load associated with the at least one computation method; link performance for the radio channel, characteristics of the at least one input signal, said characteristics comprising at least one of a Signal-to-noise ratio, SNR, a Block Error rate, and a spatial distribution of the at least one input signal; and latency in communication.

4. The method (300) according to any of the preceding claims, comprising: decoding (308) a data packet received through the communication device (1000) to obtain a decoded data packet, by using the radio channel estimation computed by using the first computation method; and identifying (310) whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

5. The method (300) according to claim 4, wherein the identifying (310) whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet comprises: performing (312) an error check for the decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied over the decoded data packet to obtain a CRC result for the decoded data packet; and identifying (314) that the decoded data packet is a successfully decoded data packet when the CRC result indicates a no CRC error; or that the decoded data packet is a not successfully decoded data packet when the CRC result indicates a CRC error .

6. The method (300) according to any of the claims 3 to 5, comprising: updating (316) the method selection criterion to obtain an updated method selection criterion according to the identification of whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

7. The method (300) according to any of the claims 3 to 6, comprising: selecting (318) a second computation method from said plurality of computation methods when the decoded data packet has been identified as not successfully decoded data packet; computing (320) the radio channel estimation based on an application of the second computation method over the at least one input signal; decoding (322) the data packet received through the communication device (1000) to obtain a decoded data packet using the radio channel estimation computed through the second computation method; identifying (324) whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet; and updating (326) the method selection criterion to obtain an updated method selection criterion, according to the identification of whetherthe decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

8. The method (300) according to any of the claim 3 to 6, comprising: selecting (328) a second computation method from said plurality of computation methods when the decoded data packet decoded using the radio channel estimation computed through the first computation method is identified as the successfully decoded data packet; computing (330) the radio channel estimation based on an application of the second computation method over the at least one input signal; decoding (332) the data packet received through the communication device (1000) to obtain the decoded data packet using the radio channel estimation computed through the second computation method; identifying (334) whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet; and updating (336) the method selection criterion to obtain the updated method selection criterion according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

9. The method (300) according to any of the claims 7 or 8, wherein the identifying (324, 334) whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet comprises: performing (338) an error check for the decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied over the decoded data packet to obtain a CRC result of the decoded data packet; identifying (340) the decoded data packet as a successfully decoded data packet when the CRC result indicates a no CRCerror and as a not successfully decoded data packet when the CRC result indicates a CRC error; and updating (342) the method selection criterion to obtain the updated method selection criterion, according to the identification of whetherthe decoded data packet is a successfully decoded data packet or a not successfully decoded data packet using the radio channel estimation computed through the second computation method.

10. The method (300) according to any of the claims 7 to 9, comprising: training (344) the second computation method according to the updated method selection criterion.

11. The method (300) according to any of the claim 3 to 10, comprising: comparing (346) the CRC result obtained for the first computation method with the CRC result obtained for the second computation method; and updating (348) the method selection criterion to obtain the updated method selection criterion based on the comparison.

12. The method (300) according to any of the preceding claims, wherein the plurality of computation methods comprise a model-driven method comprising an Akaike information criterion, AIC, method.

13. The method (300) according to any of the preceding claims, wherein the plurality of computation methods comprise a data-driven method comprising a machine learning, ML, method.

14. A communication device (1000) for radio channel estimation in a wireless network, the communication device (1000) comprising: a controlling circuitry (1106) arranged for: obtaining (302) at least one input signal over the wireless network (106); selecting (304) a first computation method from a plurality of computation methods usable in the radio channel estimation, wherein the first computation method is selected accordingto a method selection criterion; and applying (306) the selected first computation method over said obtained at least one input signal to compute the radio channel estimation for use in data reception through the communication device (1000).

15. The communication device (1000) according to claim 14, wherein the at least one input signal comprises at least one of a reference signal and data symbols.

16. The communication device (1000) according to claim 14, wherein the method selection criterion comprises: past results of the radio channel estimation, made using at least one computation method, optionally including the first computational method, from the plurality of computation methods; a computational load associated with the at least one computation method; link performance for the radio channel, characteristics of the at least one input signal, said characteristics comprising at least one of a Signal-to-noise ratio, SNR, a Block Error rate, and a spatial distribution of the at least one input signal; and latency in communication.

17. The communication device (1000) according to any of the claims 14-16, wherein the controlling circuitry (1106) is arranged for: decoding (308) a data packet received through the communication device (1000) to obtain a decoded data packet, by using the radio channel estimation computed using the first computation method; and identifying (310) whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

18. The communication device (1000) according to any of the claims 14-17, wherein the controlling circuitry (1106) is arranged for: performing (312) an error check for the decoded data packet through at least one error check method applied over the decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied over the decoded data packet to obtain a CRC result for the decoded data packet; and identifying (314) that the decoded data packet is a successfully decoded data packet when the CRC result indicates a no CRC error ; or that the decoded data packet is a not successfully decoded data packet when the CRC result indicates a CRC error.

19. The communication device (1000) according to any of the claims 14-18, wherein the controlling circuitry (1106) is arranged for: updating (316) the method selection criterion to obtain an updated method selection criterion according to the identification of whether the decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

20. The communication device (1000) according to any of the claims 14-19, wherein the controlling circuitry (1106) is arranged for: selecting (318) a second computation method from said plurality of computation methods when the decoded data packet has been identified as the not successfully decoded data packet; computing (320) the radio channel estimation based on an application of the second computation method over the at least one input signal; decoding (322) the data packet received through the communication device (1000) to obtain the decoded data packet using the radio channel estimation computed through the second computation method; identifying (324) whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet; and updating (326) the method selection criterion to obtain an updated method selection criterion, according to the identification of whetherthe decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

21. The communication device (1000) according to any of the claims 14-20, wherein the controlling circuitry (1106) is arranged for: selecting (328) a second computation method from said plurality of computation methods when the decoded data packet decoded using the radio channel estimation computed through the first computation method is identified as the successfully decoded data packet; computing (330) the radio channel estimation based on an application of the second computation method over the at least one input signal; decoding (332) the data packet received through the communication device (1000) to obtain the decoded data packet using the radio channel estimation computed through the second computation method; identifying (334) whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet; and updating (336) the method selection criterion to obtain the updated method selection criterion according to the identification of whether the decoded data packet decoded using the radio channel estimation computed through the second computation method is a successfully decoded data packet or a not successfully decoded data packet.

22. The communication device (1000) according to any of the claims 14-21, wherein the controlling circuitry (1106) is arranged for: performing (338) an error check for the decoded data packet decoded through the second computation method, through at least one error check method applied overthe decoded data packet, wherein the at least one error check method comprises a Cyclic Redundancy Check, CRC, applied over the decoded data packet to obtain a CRC result of the decoded data packet; identifying (340) the decoded data packet as a successfully decoded data packet when the CRC result indicates a no CRC error and as a not successfully decoded data packet when the CRC result indicates a CRC error ; and updating (342) the method selection criterion to obtain the updated method selection criterion, according to the identification of whetherthe decoded data packet is a successfully decoded data packet or a not successfully decoded data packet.

23. The communication device (1000) according to any of the claims 14-22, wherein the controlling circuitry (1106) is arranged for: training (344) the second computation method according to the updated method selection criterion.

24. The communication device (1000) according to any of the claims 14-23, wherein the controlling circuitry (1106) is arranged for: comparing (346) the CRC result obtained for the first computation method with the CRC result obtained for the second computation method; and updating (348) the method selection criterion to obtain the updated method selection criterion based on the comparison.

25. The communication device (1000) according to any of the claims 14-24, wherein the plurality of computation methods comprise a model-driven method comprising an Akaike information criterion, AIC, method.

26. The communication device (1000) according to any of the claims 14-25, wherein the plurality of computation methods comprise a data-driven method comprising a machine learning, ML, method.

27. A computer program product comprising a non-transitory computer readable medium, having thereon a computer program comprising program instructions, the computer program is loadable into a data processing unit and configured to cause execution of the method according to any of claims 1 through 13 when the computer program is run by the data processing unit.