US-12627536-B2 - Radio receiver

US12627536B2US 12627536 B2US12627536 B2US 12627536B2US-12627536-B2

Abstract

According to an example embodiment, a radio receiver is configured to: obtain a data array including a plurality of elements, wherein each element in the plurality of elements in the data array corresponds to a sub-carrier in a plurality of subcarriers, to a timeslot in a time interval, and to an antenna stream; implement a machine learning model including at least one neural network and a transformation, wherein the transformation includes at least one multiplicative layer or equalisation; and input data into the machine learning model, wherein the data includes at least the data array; wherein the machine learning model is configured to, based on the data, output an output array representing values of the plurality of elements in the data array, wherein the values include bits or symbols. A radio receiver, a method and a computer program product are disclosed.

Inventors

Mikko Johannes Honkala
Dani Johannes Korpi
Janne Matti Juhani Huttunen
Vesa STARCK

Assignees

NOKIA SOLUTIONS AND NETWORKS OY

Dates

Publication Date: 20260512
Application Date: 20210921
Priority Date: 20201008

Claims (15)

1 . A radio receiver, comprising: at least one processor; and at least one memory storing instructions that, when executed with the at least one processor, cause the radio receiver to: obtain a data array comprising a plurality of elements, wherein an element in the plurality of elements in the data array corresponds to a subcarrier in a plurality of subcarriers, to a timeslot in a time interval, and to an antenna stream; implement a machine learning model comprising at least one neural network and a transformation, wherein the transformation comprises at least one multiplicative layer to multiply together elements in the data array, wherein the multiplicative layer is trained in conjunction with the at least one neural network of the machine learning model, and wherein output of the multiplicative layer functions as input for the at least one neural network; and input data into the machine learning model, wherein the data comprises at least the data array; wherein the machine learning model is configured to, based on the input data, output an output array representing values of the plurality of elements in the data array, wherein the values comprise bits or symbols.
2 . The radio receiver according to claim 1 , wherein the machine learning model comprises a neural network component before the transformation and a neural network component after the transformation.
3 . The radio receiver according to claim 2 , wherein the instructions, when executed with the at least one processor, cause the radio receiver to perform processing between elements in the data array corresponding to at least one of different subcarriers or to different timeslots.
4 . The radio receiver according to claim 1 , wherein the machine learning model comprises equalisation that comprises maximal-ratio combining.
5 . The radio receiver according to claim 1 , wherein the instructions, when executed with the at least one processor, cause the multiplicative layer to complex conjugate at least one of the elements to be multiplied before the multiplication.
6 . The radio receiver according to claim 1 , wherein the instructions, when executed with the at least one processor, cause the multiplicative layer to scale at least one of an imaginary part or a real part separately from another of at least one element in the data array.
7 . The radio receiver according to claim 1 , wherein the instructions, when executed with the at least one processor, cause the multiplicative layer to perform a sparse expansion on the data array.
8 . The radio receiver according to claim 1 , wherein the at least one neural network comprises a convolutional neural network.
9 . The radio receiver according to claim 1 , wherein the instructions, when executed with the at least one processor, cause the radio receiver to: obtain a reference signal array representing a reference signal configuration applied during the time interval, and wherein the data further comprises the reference signal array; or obtain a reference signal array representing a reference signal configuration applied during the time interval and compute a channel estimate based on the data array and the reference signal array, and wherein the data further comprises the channel estimate.
10 . The radio receiver according to claim 9 , wherein the reference signal array comprises a plurality of channels, wherein a channel corresponds to a layer of a multiple-input and multiple-output transmission.
11 . The radio receiver according to claim 1 , wherein the instructions, when executed with the at least one processor, cause the radio receiver to: receive raw data using a plurality of samples during the time interval; and perform a Fourier transform on the received raw data, producing the data array.
12 . A client device comprising the radio receiver according to claim 1 .
13 . A network node device comprising the radio receiver according to claim 1 .
14 . A method, comprising: obtaining a data array comprising a plurality of elements, wherein an element in the plurality of elements in the data array corresponds to a subcarrier in a plurality of subcarriers, to a timeslot in a time interval, and to an antenna stream; implementing a machine learning model comprising at least one neural network and a transformation, wherein the transformation comprises at least one multiplicative layer to multiply together elements in the data array, wherein the multiplicative layer is trained in conjunction with the at least one neural network of the machine learning model, and wherein output of the multiplicative layer functions as input for the at least one neural network; and inputting data into the machine learning model, wherein the data comprises at least the data array; wherein the machine learning model is configured to, based on the input data, output an output array representing values of the plurality of elements in the data array, wherein the values comprise bits or symbols.
15 . A non-transitory program storage device readable with an apparatus, tangibly embodying a program of instructions executable with the apparatus for performing: obtaining a data array comprising a plurality of elements, wherein an element in the plurality of elements in the data array corresponds to a subcarrier in a plurality of subcarriers, to a timeslot in a time interval, and to an antenna stream; implementing a machine learning model comprising at least one neural network and a transformation, wherein the transformation comprises at least one multiplicative layer to multiply together elements in the data array, wherein the multiplicative layer is trained in conjunction with the at least one neural network of the machine learning model, and wherein output of the multiplicative layer functions as input for the at least one neural network; and inputting data into the machine learning model, wherein the data comprises at least the data array; wherein the machine learning model is configured to, based on the input data, output an output array representing values of the plurality of elements in the data array, wherein the values comprise bits or symbols.

Description

TECHNICAL FIELD The present application generally relates to the field of wireless communications. In particular, the present application relates to a radio receiver device for wireless communication, and related methods and computer programs. BACKGROUND Radio receiver algorithms may comprise blocks that are based on mathematical and statistical algorithms. These algorithms may be developed and programmed manually, which may be labour intensive. For example, it requires large amounts of manual labour to implement different reference signal configurations. Receiver algorithms designed this way may perform adequately for most channel conditions but may not give the best possible performance for any of the channels. It may also be difficult to estimate how algorithms developed based on theoretical channel conditions align with actual physical channel conditions. Multiple-input and multiple-output (MIMO) radio receivers can introduce additional challenges, since the nature of MIMO detection can require separating multiple overlapping spatial streams. SUMMARY The scope of protection sought for various example embodiments of the invention is set out by the independent claims. The example embodiments and features, if any, described in this specification that do not fall under the scope of the independent claims are to be interpreted as examples useful for understanding various example embodiments of the invention. An example embodiment of a radio receiver comprises at least one processor and at least one memory comprising computer program code. The at least one memory and the computer program code are configured to, with the at least one processor, cause the radio receiver to: obtain a data array comprising a plurality of elements, wherein each element in the plurality of elements in the data array corresponds to a subcarrier in a plurality of subcarriers, to a timeslot in a time interval, and to an antenna stream; implement a machine learning model comprising at least one neural network and a transformation, wherein the transformation comprises at least one multiplicative layer or equalisation; and input data into the machine learning model, wherein the data comprises at least the data array; wherein the machine learning model is configured to, based on the data, output an output array representing values of the plurality of elements in the data array, wherein the values comprise bits or symbols. The radio receiver may be able to, for example, efficiently receiver the transmission. The machine learning model can be trained to receive the transmission with a high degree of flexibility. An example embodiment of a radio receiver comprises means for performing: obtain a data array comprising a plurality of elements, wherein each element in the plurality of elements in the data array corresponds to a subcarrier in a plurality of subcarriers, to a timeslot in a time interval, and to an antenna stream; implement a machine learning model comprising at least one neural network and a transformation, wherein the transformation comprises at least one multiplicative layer or equalisation; and input data into the machine learning model, wherein the data comprises at least the data array; wherein the machine learning model is configured to, based on the data, output an output array representing values of the plurality of elements in the data array, wherein the values comprise bits or symbols. In an example embodiment, alternatively or in addition to the above-described example embodiments, the machine learning model comprises a neural network component before the transformation and a neural network component after the transformation. The neural network component before the transformation may, for example, process the data to be more easily processable for the transformation and the neural network component after the transformation may, for example, perform bit/symbol detection. In an example embodiment, alternatively or in addition to the above-described example embodiments, the neural network component before the transformation is configured to perform processing between elements in the data array corresponding to different subcarriers and/or to different timeslots. Thus, the neural network component before the transformation may mix different resource elements. This may improve the performance of the transformation and/or of the radio receiver. In an example embodiment, alternatively or in addition to the above-described example embodiments, the equalisation comprises maximal-ratio combining. The radio receiver may be able to, for example, efficiently perform the equalisation. In an example embodiment, alternatively or in addition to the above-described example embodiments, the multiplicative layer is configured to multiply together elements in the data array layers. Thus, the multiplicative layer may be able to perform multiplication between elements, which may be difficult to implement using neural networks. In an example e