Method and device for transmitting and receiving uplink signal

Abstract

The invention relates to a method and a device for transmitting and receiving uplink signals. In one aspect, there is provided a method performed by a user equipment (UE) in a communication system, comprising: determining a first processing method of a plurality of methods of obtaining complex valued information associated with channel state information (CSI), the plurality of methods comprising: a first method of obtaining complex valued information based on a bit sequence, the bit sequence is obtained by performing bit sequence generation and channel coding on the CSI, and a second method for obtaining complex value information based on the CSI, and the CSI does not perform bit sequence generation and channel coding; processing the CSI based on the first processing method to obtain complex value information associated with the CSI; and sending an uplink channel or signal based on the complex value information.

Description

Technical Field

[0001] The present application relates to the field of wireless communication technology, and more specifically, to a method and apparatus for sending and receiving uplink signals in a wireless communication network. Background Art

[0002] Considering the development of wireless communication from generation to generation, these technologies have been developed mainly for services targeting people, such as voice calls, multimedia services and data services. With the commercialization of the fifth generation (5G) communication system, the number of connected devices is expected to grow exponentially. These will be increasingly connected to the communication network. Examples of the Internet of Things can include vehicles, robots, drones, home appliances, displays, smart sensors connected to various infrastructures, construction machinery and factory equipment. Mobile devices are expected to develop in various forms (such as augmented reality glasses, virtual reality headsets and holographic devices). Efforts have been made to develop improved 6G communication systems so that various services can be provided by connecting hundreds of billions of devices and things in the sixth generation (6G) era. For these reasons, 6G communication systems are called super 5G systems.

[0003] The 6G communication system, which is expected to be commercialized around 2030, will have a peak data rate of tera (1,000 gigabits) per second and a radio latency of less than 100 μsec, thus being 50 times the data rate of the 5G communication system and having 1 / 10 of its radio latency.

[0004] In order to achieve such high data rates and ultra-low latency, the implementation of 6G communication systems in the terahertz band (e.g., the 95 GHz to 3 THz band) has been considered. It is expected that since the path loss and atmospheric absorption in the terahertz band are more serious than those in the millimeter wave (mmWave) band introduced in 5G, technologies that can ensure the signal transmission distance (i.e., coverage) will become more critical. As the main technology to ensure coverage, it is necessary to develop radio frequency (RF) elements, antennas, and new waveforms with better coverage than orthogonal frequency division multiplexing (OFDM), beamforming and massive multiple input multiple output (MIMO), full-dimensional MIMO (FD-MIMO), array antennas, and multi-antenna transmission technologies such as massive antennas. In addition, new technologies for improving signal coverage in the terahertz band, such as metamaterial-based lenses and antennas, orbital angular momentum (OAM), and reconfigurable smart surfaces (RIS), have been under discussion.

[0005] Furthermore, to improve spectrum efficiency and overall network performance, the following technologies have been developed for 6G communication systems: full-duplex technology for enabling uplink and downlink transmissions to simultaneously use the same frequency resources; network technologies that utilize satellites, high-altitude platform stations (HAPS), and other systems in an integrated manner; improved network structures to support mobile base stations and enable network operation optimization and automation; dynamic spectrum sharing technology with conflict avoidance based on spectrum usage prediction; the use of artificial intelligence (AI) in wireless communications to improve overall network operations by leveraging AI from the design stage of 6G development and internalizing end-to-end AI support functions; and next-generation distributed computing technologies that overcome the computing power limitations of user equipment (UE) by leveraging ultra-high-performance communication and computing resources available on the network, such as mobile edge computing (MEC) and the cloud. Furthermore, efforts are continuing to strengthen connectivity between devices, optimize networks, promote the softwareization of network entities, and increase the openness of wireless communications by designing new protocols to be used in 6G communication systems, developing mechanisms to implement hardware-based security environments and secure data usage, and developing technologies to maintain privacy.

[0006] Research and development of 6G communication systems, including hyperconnectivity for both human-to-machine (P2M) and machine-to-machine (M2M), are expected to bring about the next hyperconnected experience. Specifically, services such as truly immersive extended reality (XR), high-fidelity mobile holograms, and digital replicas are expected to be provided through 6G communication systems. Furthermore, services such as remote surgery for enhanced safety and reliability, industrial automation, and emergency response will be provided through 6G communication systems, enabling the technology to be applied in a variety of fields such as industry, healthcare, automobiles, and home appliances. Summary of the Invention

[0007] At least one embodiment of the present disclosure provides a method performed by a user equipment (UE) in a communication system, including:

[0008] Determining a first processing method among a plurality of methods for obtaining complex-valued information associated with channel state information (CSI), wherein the plurality of methods include:

[0009] A first method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on CSI, and

[0010] A second method for obtaining complex-valued information based on CSI, wherein the CSI is not subjected to bit sequence generation and channel coding;

[0011] Processing the CSI based on the first processing method to obtain complex-valued information associated with the CSI;

[0012] An uplink channel or signal is sent based on the complex-valued information.

[0013] In one implementation, sending an uplink channel or signal based on the complex-valued information includes:

[0014] The complex-valued information is mapped onto an uplink channel or signal and sent.

[0015] In one implementation, the method further includes: receiving high-layer signaling, where the high-layer signaling includes first information, where the first information includes at least one of the following:

[0016] Information about the first treatment method;

[0017] information of at least one parameter set associated with the second method;

[0018] The method further includes: receiving physical layer signaling, where the physical layer signaling includes second information of a parameter set in the at least one parameter set.

[0019] In one implementation, the first information includes at least one of an index, a name, and a table of at least one parameter set associated with the second method.

[0020] In one implementation, the second information includes information associated with at least one of an index, name, modulation order, signal-to-noise ratio (SNR), signal-to-interference-plus-noise ratio (SINR), SINR interval, SINR upper bound, SINR lower bound, output type, output length, and output format corresponding to the parameter set.

[0021] In one implementation, the second method includes a method based on a neural network model, which further includes: determining a parameter set of the neural network based on the second information, wherein the parameter set includes at least one of the following: the type of model, the structure of the model, the number of layers of the model, the number of neurons of the model, the activation function of the model, an input length equal to the number of elements in the channel state information, the output format of the model, and the weight of the model.

[0022] In one implementation, the output format of the model further includes at least one of the following: output data format, output data structure, number of output elements, priority of output elements, and index of output elements.

[0023] In one implementation, the method further includes: before the UE maps the complex-valued information associated with the CSI onto an uplink channel or signal, the UE multiplexing the uplink channel or signal to carry the complex-valued information associated with the second uplink control information;

[0024] The multiplexing method includes combining the complex value information associated with the CSI and the complex value information associated with the second uplink control information and mapping them onto an uplink channel or signal;

[0025] The complex value information associated with the second uplink control information may be obtained by one of the following methods:

[0026] A third method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a process of bit sequence generation and channel coding on second uplink control information; and a fourth method for obtaining complex-valued information based on the second uplink control information, wherein bit sequence generation and channel coding are not performed on the second uplink control information.

[0027] In one implementation, the method further includes: before the UE maps the complex-valued information associated with the CSI to an uplink channel or signal, the UE obtains matched complex-valued information based on the complex-valued information and the number of time-frequency resource units associated with the reported CSI, wherein the number of complex-valued symbols included in the matched complex-valued information is the same as the number of time-frequency resource units associated with the reported CSI.

[0028] In one implementation, the obtained matched complex-valued information may also include: the UE matches the output elements of the neural network model with the time-frequency resource units associated with the reported CSI in descending order of priority according to the priority of the output elements, to obtain a matched complex-valued symbol sequence with priority.

[0029] In one implementation, the obtained matched complex value information may also include: the UE matches the output elements with the time-frequency resource units associated with the reported CSI in sequence according to the index of the output element of the neural network model in the order of the index, and obtains the matched complex value symbol sequence related to the index.

[0030] In one implementation, the method further includes: determining, based on the content and / or format of the uplink control information, a first processing method among multiple methods for obtaining complex-valued information associated with the CSI.

[0031] In one implementation, the method further includes the UE reporting a capability of the UE to the base station, where the capability of the UE includes at least one of the following:

[0032] The first capability indicates that the UE has the capability of the first method;

[0033] The second capability indicates that the UE has the capability of the second method;

[0034] The third capability indicates that the UE has the capabilities of the first method and the second method.

[0035] At least one embodiment of the present disclosure provides a method performed by a base station in a communication system, including:

[0036] Sending high-layer signaling to the UE, where the high-layer signaling includes first information, where the first information includes at least one of the following: relevant information of the first processing method and information of at least one parameter set associated with the second method;

[0037] sending physical layer signaling to the UE, where the physical layer signaling includes second information of one parameter set among the at least one parameter set,

[0038] The first processing method is one of multiple methods for obtaining complex-valued information associated with CSI, wherein the multiple methods include:

[0039] A first method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on CSI, and

[0040] A second method for obtaining complex-valued information based on CSI, wherein the CSI does not undergo bit sequence generation and channel coding.

[0041] In one implementation, the second information includes information associated with at least one of an index, name, modulation order, SNR, SINR, SINR interval, SINR upper bound, SINR lower bound, output type, output length, and output format corresponding to the parameter set.

[0042] In one implementation, the second method includes a method based on a neural network model, which further includes: determining a parameter set of the neural network based on the second information, wherein the parameter set includes at least one of the following: the type of model, the structure of the model, the number of layers of the model, the number of neurons of the model, the activation function of the model, an input length equal to the number of elements in the channel state information, the output format of the model, and the weight of the model.

[0043] In one implementation, the output format of the model further includes at least one of the following: output data format, output data structure, number of output elements, priority of output elements, and index of output elements.

[0044] In one implementation, the method further includes:

[0045] Receive UE capabilities reported by the UE, where the UE capabilities include at least one of the following:

[0046] The first capability indicates that the UE has the capability of the first method;

[0047] The second capability indicates that the UE has the capability of the second method;

[0048] The third capability indicates that the UE has the capabilities of the first method and the second method.

[0049] At least one embodiment of the present disclosure provides a user equipment (UE) in a communication system, including:

[0050] a transceiver configured to transmit and / or receive signals;

[0051] A controller is configured to control the UE to execute the method according to at least one embodiment of the present disclosure.

[0052] At least one embodiment of the present disclosure provides a base station in a communication system, including:

[0053] a transceiver configured to transmit and / or receive signals;

[0054] A controller is configured to control the base station to execute the method according to at least one embodiment of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 An example wireless network according to an embodiment of the present disclosure is shown;

[0056] Figure 2 An example base station according to an embodiment of the present disclosure is shown;

[0057] Figure 3 An example user device according to an embodiment of the present disclosure is shown;

[0058] Figure 4 A schematic diagram illustrating a method for a UE in a communication system to report uplink control information according to at least one embodiment of the present disclosure is shown;

[0059] Figure 5 A schematic diagram showing a method and solution for determining and obtaining complex-valued information representing CSI by a UE in a communication system according to the present disclosure using first and second configuration information;

[0060] Figure 6 A schematic diagram showing a UE in a communication system according to the present disclosure obtaining complex-valued information based on a CSI bit sequence and CSI reconstruction at a base station;

[0061] Figure 7 A schematic diagram showing a neural network model used by a first training method in a communication system according to the present disclosure is shown;

[0062] Figure 8A schematic diagram showing a neural network structure for obtaining a bit sequence carrying a CSI feature vector at a UE and reconstructing CSI at a base station in a communication system according to the present disclosure is shown;

[0063] Figure 9 A schematic diagram showing a method for a UE in a communication system according to the present disclosure to obtain complex-valued information using a neural network is shown;

[0064] Figure 10 A schematic diagram of an autoencoder including a neural network model for obtaining CSI complex-valued information in a communication system according to the present disclosure is shown;

[0065] Figure 11 A schematic diagram illustrating acquisition of CSI complex value information at a UE and CSI reconstruction at a base station in a communication system according to the present disclosure is shown;

[0066] Figure 12 A schematic diagram showing a neural network structure for obtaining CSI complex value information at the UE side and reconstructing CSI at the base station side in a communication system according to the present disclosure is shown;

[0067] Figure 13 shows a schematic diagram of training an autoencoder according to the present disclosure;

[0068] Figure 14 A schematic diagram showing complex information of a UE multiplexing the uplink channel or signal to carry second uplink control information in a communication system according to the present disclosure is shown;

[0069] Figure 15 A schematic diagram of a circular buffer for fast matching in a communication system according to the present disclosure is shown;

[0070] Figure 16 A schematic diagram showing mapping of complex symbols in complex information to time-frequency resource units indicated by a configuration by a user in a communication system according to the present disclosure is shown;

[0071] Figure 17 A schematic diagram showing mapping of pilot-related complex symbols to time-frequency resource units by a user in a communication system according to the present disclosure is shown;

[0072] Figure 18 A schematic diagram showing a user in a communication system according to the present disclosure obtaining complex value information including pilot signals associated with time-frequency resources according to a time division multiplexing method;

[0073] Figure 19 A schematic diagram showing a user in a communication system according to the present disclosure obtaining complex value information including pilot signals associated with time-frequency resources according to a frequency division multiplexing method;

[0074] Figure 20A schematic diagram showing an example of a method for a user to report CSI in a communication system according to the present disclosure;

[0075] Figure 21 A schematic diagram showing a method for a UE in a communication system according to the present disclosure to obtain complex-valued information related to uplink control information is shown;

[0076] Figure 22 A schematic diagram showing an identification table including N neural network models in a communication system according to the present disclosure is shown;

[0077] Figure 23 A schematic diagram showing an information table in which a UE in a communication system according to the present disclosure determines two complex value acquisition methods according to first configuration information;

[0078] Figure 24 A schematic diagram showing a method in which a UE in a communication system according to the present disclosure determines an information table including two methods according to an instruction of first configuration information;

[0079] Figure 25 The PUCCH process according to at least one embodiment of the present disclosure is shown;

[0080] Figure 26 A method for implementing a JSCM model through training according to at least one embodiment of the present disclosure is shown.

[0081] Figure 27 A schematic structural diagram of a user equipment according to at least one embodiment of the present disclosure is shown;

[0082] Figure 28 A schematic structural diagram of a base station according to at least one embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0083] Before proceeding to the following specific embodiments, it may be advantageous to set forth the definitions of certain words and phrases used throughout the patent document. The term "connect" and its derivatives refer to any direct or indirect communication between two or more elements, regardless of whether those elements are in physical contact with each other. The terms "send," "receive," and "communicate," and their derivatives encompass both direct and indirect communication. The terms "include," "comprise," and "includes," and their derivatives, mean to include without limitation. The term "or" is inclusive, meaning and / or. The phrase "associated with," and its derivatives, means to include, be included within, be interconnected with, include, be included within, be connected to or connected with, be coupled to or coupled with, communicate with, collaborate with, be interwoven, juxtaposed, be close to, be bound to or bound with, have, have an attribute of, have a relationship to ... or have a relationship to ... etc. The term "controller" means any device, system, or part thereof that controls at least one operation. Such a controller can be implemented in hardware or in a combination of hardware and software and / or firmware. The functionality associated with any particular controller, whether local or remote, can be centralized or distributed. The phrase "at least one of" when used with a list of items means that different combinations of one or more of the listed items can be used, and only one item in the list may be required. For example, "at least one of A, B, and C" includes any of the following combinations: A, B, C, A and B, A and C, B and C, and only A, only B, and only C. Similarly, the term "set" means one or more. Thus, a set of items can be a single item or a set of two or more items.

[0084] Moreover, various functions as described below can be implemented or supported by one or more computer programs, each of which is formed by computer-readable program code and embodied in a computer-readable medium. The terms "application" and "program" refer to one or more computer programs, software components, instruction sets, processes, functions, objects, categories, instances, related data, or a part thereof that are suitable for being implemented in a suitable computer-readable program code. The phrase "computer-readable program code" includes any type of computer code, including source code, object code, and executable code. The phrase "computer-readable medium" includes any type of medium that can be accessed by a computer, such as a read-only memory (ROM), random access memory (RAM), hard drive, compact disc (CD), digital video disc (DVD), or any other type of memory. "Non-transient" computer-readable medium excludes wired, wireless, optical, or other communication links that transmit instantaneous electrical signals or other signals. Non-transient computer-readable medium includes a medium in which data can be permanently stored and a medium in which data can be stored and rewritten later, such as a rewritable optical disc or erasable memory device.

[0085] Definitions for certain other words and phrases are provided throughout this patent document. Those skilled in the art should understand that in many, if not most instances, such definitions apply to prior, as well as future uses of such defined words and phrases.

[0086] The figures and various embodiments used to describe the principles of the present disclosure are included herein for illustration only and should not be construed in any way as limiting the scope of the present disclosure. In addition, those skilled in the art will appreciate that the principles of the present disclosure can be implemented in any appropriately arranged wireless communication system.

[0087] The following Figures 1 to 28 Various embodiments of the present disclosure are described as being implemented in a wireless communication system. Figures 1 to 28 The description is not meant to imply physical or architectural limitations to the manner in which different embodiments may be implemented. Different embodiments of the present disclosure may be implemented in any suitably arranged communications system.

[0088] Figure 1 An example wireless network according to an embodiment of the present disclosure is shown. Figure 1 The embodiment of the wireless network shown in FIGURE 1 is for illustration only. Other embodiments of the wireless network 100 may be used without departing from the scope of this disclosure.

[0089] like Figure 1 As shown, the wireless network includes base station (gNB or gNodeB) 101, gNB 102, and gNB 103. gNB 101 communicates with gNB 102 and gNB 103. gNB 101 also communicates with at least one network 130, such as the Internet, a proprietary Internet Protocol (IP) network, or other data network.

[0090] gNB 102 provides wireless broadband access to network 130 for a plurality of first user equipment (UEs) within coverage area 120 of gNB 102. The plurality of first UEs include UE 111, which may be located at a small business (SB); UE 112, which may be located at an enterprise (E); UE 113, which may be located at a WiFi hotspot (HS); UE 114, which may be located at a first residence (R1); UE 115, which may be located at a second residence (R2); and UE 116, which may be a mobile device (M) such as a cellular phone, a wireless laptop, a wireless personal digital assistant (PDA), etc. gNB 103 provides wireless broadband access to network 130 for a plurality of second UEs within coverage area 125 of gNB 103. The plurality of second UEs include UE 115 and UE 116, as well as subscriber stations (SS, e.g., UEs) 117, 118, and 119. In some embodiments, one or more of gNBs 101-103 may communicate with each other and UEs 111-116 using existing wireless communication technologies, and one or more of UEs 111-119 may communicate directly with each other (e.g., UEs 117-119) using other existing or proposed wireless communication technologies.

[0091] Depending on the network type, the term "base station" or "BS" can refer to any component (or collection of components) configured to provide wireless access to a network, such as a transmission point (TP), a transceiver point (TRP), an enhanced (or "evolved") base station (eNodeB or eNB), a 5G base station (gNB), a macro cell, a femto cell, a Wireless Fidelity (WiFi) access point (AP), or other wireless-capable device. A base station can provide wireless access according to one or more wireless communication protocols, such as 3GPP 5G New Radio (NR), Long Term Evolution (LTE), LTE-Advanced (LTE-A), High Speed Packet Access (HSPA), Wi-Fi 802.11a / b / g / n / ac, etc. For convenience, various names of base station-type devices and functions are used interchangeably in this patent document to refer to network infrastructure components that provide wireless access to remote terminals. Moreover, depending on the network type, the term "user equipment" (UE) can refer to any component such as a mobile station (MS), a subscriber station (SS), a remote terminal, a wireless terminal, a receiving point, or a user device. For convenience, various names of user equipment type devices and functions are used interchangeably in this patent document to refer to a remote wireless device that wirelessly accesses a BS, regardless of whether the UE is a mobile device (such as a mobile phone or smartphone) or a device generally considered to be a fixed device (such as a desktop computer or vending machine).

[0092] Dashed lines illustrate the approximate extents of coverage areas 120 and 125, which are shown as generally circular for purposes of illustration and explanation only. It should be clearly understood that coverage areas associated with a gNB, such as coverage areas 120 and 125, may have other shapes, including irregular shapes, depending on the configuration of the gNB and variations in the wireless environment associated with natural and man-made obstacles.

[0093] As described in more detail below, one or more of UEs 111-119 include circuitry, programming, or a combination thereof. In certain embodiments, one or more of gNBs 101-103 include circuitry, programming, or a combination thereof.

[0094] although Figure 1 An example of a wireless network is shown, but Figure 1 Various changes may be made. For example, wireless network 100 may include any number of gNBs and any number of UEs in any suitable arrangement. Moreover, gNB 101 may communicate directly with any number of UEs and provide those UEs with wireless broadband access to network 130. Similarly, each gNB 102-103 may communicate directly with network 130 and provide the UEs with direct wireless broadband access to network 130. In addition, gNBs 101, 102, and / or 103 may provide access to other or additional external networks, such as an external telephone network or other types of data networks.

[0095] Figure 2 An example base station according to an embodiment of the present disclosure is shown. Figure 2 The embodiment of the gNB 102 shown in FIGURE 1 is for illustration only, and Figure 1 gNBs 101 and 103 may have the same or similar configurations. However, gNBs appear in a variety of configurations, and Figure 2 The scope of this disclosure is not limited to any particular implementation of a gNB.

[0096] like Figure 2 As shown in FIG, gNB 102 includes multiple antennas 200a-200n, multiple radio frequency (RF) transceivers 201a-201n, transmit (TX) processing circuitry 203, and receive (RX) processing circuitry 204. gNB 102 also includes a controller / processor 205, memory 206, and a backhaul or network interface 207.

[0097] RF transceivers 201a-201n receive incoming RF signals from antennas 200a-200n, such as signals transmitted by UEs in network 100. RF transceivers 201a-201n downconvert the incoming RF signals to generate intermediate frequency (IF) or baseband signals. The IF or baseband signals are sent to RX processing circuitry 204, which generates processed baseband signals by filtering, decoding, and / or digitizing the baseband or IF signals. RX processing circuitry 204 sends the processed baseband signals to controller / processor 205 for further processing.

[0098] The TX processing circuitry 203 receives analog or digital data (such as voice data, web data, email, or interactive video game data) from the controller / processor 205. The TX processing circuitry 203 encodes, multiplexes, and / or digitizes the outgoing baseband data to generate processed baseband or IF signals. The RF transceivers 201a-201n receive the outgoing processed baseband or IF signals from the TX processing circuitry 203 and up-convert the baseband or IF signals into RF signals that are transmitted via the antennas 201a-201n.

[0099] The controller / processor 205 may include one or more processors or other processing devices that control the overall operation of the gNB 102. For example, the controller / processor 205 may control the reception of forward channel signals and the transmission of reverse channel signals by the RF transceivers 201a-201n, the RX processing circuitry 204, and the TX processing circuitry 203 in accordance with well-known principles. The controller / processor 205 may also support additional functionality, such as more advanced wireless communication functionality.

[0100] For example, the controller / processor 205 may support beamforming or directional routing operations, in which outgoing signals from the multiple antennas 200a-200n are weighted differently to effectively steer the outgoing signals in a desired direction. Any of a variety of other functions may be supported in the gNB 102 by the controller / processor 205.

[0101] Controller / processor 205 is also capable of executing programs and other processes, such as an operating system (OS), located in memory 206. Controller / processor 205 may move data into or out of memory 206 as needed to execute processes.

[0102] The controller / processor 205 is also connected to a backhaul or network interface 207. The backhaul or network interface 207 allows the gNB 102 to communicate with other devices or systems over a backhaul connection or over a network. The interface 207 can support communication over any suitable wired or wireless connection(s). For example, when the gNB 102 is implemented as part of a cellular communication system (such as one that supports 5G, LTE, or LTE-A), the interface 207 can allow the gNB 102 to communicate with other gNBs over a wired or wireless backhaul connection. When the gNB 102 is implemented as an access point, the interface 207 can allow the gNB 102 to communicate over a wired or wireless local area network or over a wired or wireless connection to a larger network (such as the Internet). The interface 207 includes any suitable structure that supports communication over a wired or wireless connection, such as Ethernet or an RF transceiver.

[0103] Memory 206 is connected to controller / processor 205. A portion of memory 206 may include random access memory (RAM), and another portion of memory 206 may include flash memory or other read-only memory (ROM).

[0104] although Figure 2 An example of gNB 102 is shown, but the Figure 2 For example, gNB 102 may include any number of Figure 2 As a specific example, the access point may include multiple interfaces 207, and the controller / processor 205 may support routing functionality to route data between different network addresses. As another specific example, although shown as including a single instance of the TX processing circuitry 203 and a single instance of the RX processing circuitry 204, the gNB 102 may include multiple instances of each (such as one for each RF transceiver). For example, Figure 2 The various components in may be combined, further subdivided, or omitted, and additional components may be added according to specific needs.

[0105] Figure 3 An example user device according to an embodiment of the present disclosure is shown. Figure 3 The embodiment of UE 116 shown in FIGURE 1 is for illustration only, and Figure 1 UEs 111-115 and 117-119 may have the same or similar configurations. However, UEs may appear in a variety of configurations, and Figure 3 The scope of this disclosure is not limited to any particular implementation of the UE.

[0106] like Figure 3As shown in FIG, UE 116 includes an antenna 301, a radio frequency (RF) transceiver 302, a TX processing circuit 303, a microphone 304, and a receive (RX) processing circuit 305. UE 116 also includes a speaker 306, a controller or processor 307, an input / output (I / O) interface (IF) 308, a touch screen display 310, and a memory 311. The memory 311 includes an OS 312 and one or more applications 313.

[0107] RF transceiver 302 receives incoming RF signals from antenna 301, transmitted by a gNB of network 100. RF transceiver 302 downconverts the incoming RF signals to generate an IF or baseband signal. The IF or baseband signal is sent to RX processing circuitry 305, which generates a processed baseband signal by filtering, decoding, and / or digitizing the baseband or IF signal. RX processing circuitry 305 sends the processed baseband signal to speaker 306 (such as for voice data) or processor 307 for further processing (such as for web browsing data).

[0108] The TX processing circuit 303 receives analog or digital voice data from the microphone 304 or other outgoing baseband data (such as web data, email, or interactive video game data) from the processor 307. The TX processing circuit 303 encodes, multiplexes, and / or digitizes the outgoing baseband data to generate a processed baseband or IF signal. The RF transceiver 302 receives the outgoing processed baseband or IF signal from the TX processing circuit 303 and up-converts the baseband or IF signal into an RF signal that is transmitted via the antenna 301.

[0109] The processor 307 may include one or more processors or other processing devices and executes the OS 312 stored in the memory 311 to control the overall operation of the UE 116. For example, the processor 307 may control the reception of forward channel signals and the transmission of reverse channel signals by the RF transceiver 302, the RX processing circuit 305, and the TX processing circuit 303 according to well-known principles. In some embodiments, the processor 307 includes at least one microprocessor or microcontroller.

[0110] Processor 307 is also capable of executing other processes and programs located in memory 311, such as processes for CSI reporting on uplink channels. Processor 307 can move data into or out of memory 311 as needed for the executed processes. In some embodiments, processor 307 is configured to execute application 313 based on OS 312 or in response to signals received from the gNB or operator. Processor 307 is also coupled to I / O interface 309, which provides UE 116 with the ability to connect to other devices such as laptops and portable computers. I / O interface 309 is the communication path between these accessories and processor 307.

[0111] Processor 307 is also connected to touch screen display 310. A user of UE 116 may use touch screen display 310 to enter data into UE 116. Touch screen display 310 may be a liquid crystal display, a light emitting diode display, or other display capable of rendering text and / or at least limited graphics, such as from a website.

[0112] The memory 311 is connected to the processor 307. A portion of the memory 311 may include RAM, and another portion of the memory 311 may include flash memory or other ROM.

[0113] although Figure 3 An example of a UE 116 is shown, but the Figure 3 Make various changes. For example, Figure 3 The various components in the can be combined, further subdivided, or omitted, and additional components can be added according to specific needs. As a specific example, the processor 307 can be divided into multiple processors, such as one or more central processing units (CPUs) and one or more graphics processing units (GPUs). Moreover, although Figure 3 The UE 116 is shown configured as a mobile phone or smartphone, but the UE may be configured to operate as other types of mobile or stationary devices.

[0114] Modern wireless communication systems use three steps to enable user equipment (UE) to report uplink control information (including channel state-related information, scheduling requests, acknowledgments, etc.), including: 1) First, use a bit generation method (also known as a source coding method, a quantization method, etc.) to convert the uplink control information into bit information, so as to obtain bit information and reduce the inherent redundancy in the uplink control information, thereby reducing the amount of data that needs to be transmitted; 2) Then, channel coding (also known as error correction coding) is performed on the obtained bit information, so as to implement error correction control, so that when errors occur in the information during channel transmission, errors can be corrected and data can be recovered; 3) Finally, the bit information after channel coding is modulated to obtain complex-valued information representing the uplink control information (or called complex information, complex-valued modulation symbols, complex-valued modulated symbols, complex-valued symbols, symbols, etc., these descriptions are used interchangeably in this disclosure), so as to realize the transmission of information payload on the carrier signal.

[0115] For any uplink control information, the existing system usually needs to execute the above three steps separately to realize reporting. However, for some uplink control information (such as channel state related information, etc.), it is not the optimal solution for the wireless communication system to execute the above three steps to realize reporting. According to some embodiments of the present disclosure, through some technical means, the system can use only one or two steps of the above steps to achieve the same function as executing the above three steps, and achieve higher reporting performance. On the other hand, the inventors realized that the existing system does not support method switching for the content of uplink control information, but can only use the above three steps related processing for all types of control information for reporting, so the performance is not high. According to some embodiments of the present disclosure, the corresponding processing method can be selected according to the type of control information to be reported, so as to achieve a function with higher reporting performance.

[0116] It should be understood that the technical problems that can be solved by the present disclosure are not limited to the problems mentioned in the above description, but can include any technical problems that can be solved based on the essence and principles of the technical content described throughout this document, all of which are within the scope of the present disclosure. In addition, the present disclosure does not necessarily solve all of the technical problems explicitly listed.

[0117] Beneficial effects

[0118] The present disclosure provides a method for a UE in a communication system to report uplink control information (e.g., channel state information (CSI)). According to the method of at least one embodiment of the present disclosure, the efficiency and reliability of communication can be improved. The improvement in communication efficiency can be manifested in that, under the same signal to interference plus noise ratio (SINR) conditions, the method according to at least one embodiment of the present disclosure can utilize fewer communication resources compared to traditional methods to achieve information reporting with the same accuracy, such as reporting of channel state information. The improvement in communication reliability can be manifested in that, under the same SINR conditions, the method according to at least one embodiment of the present disclosure can achieve information reporting with higher accuracy, such as reporting of channel state information, by utilizing the same communication resources compared to traditional methods. In addition, in some aspects, according to the method of at least one embodiment of the present disclosure, a corresponding processing method can be selected according to the type of control information to be reported to obtain copy information representing the control information for transmission, thereby further improving performance.

[0119] The embodiments of the present disclosure will be described in more detail below with reference to examples.

[0120] Example 1

[0121] Figure 4 A schematic diagram illustrates a method 400 for a UE in a communication system to report uplink control information according to at least one embodiment of the present disclosure. For ease of description, the following description primarily uses channel state information (CSI) as an example of uplink control information to be reported by the UE. It should be understood that the description using CSI as an example is merely exemplary and does not limit the type of uplink control information applicable to the method to CSI alone. The method can also cover situations where various other types of uplink control information are reported.

[0122] like Figure 4 As shown, method 400 includes steps 401 to 403:

[0123] Step 401: The UE determines a first processing method among multiple methods for obtaining complex value information associated with channel state information (CSI), wherein the multiple methods include:

[0124] A first method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on CSI, and

[0125] A second method for obtaining complex-valued information based on CSI, wherein the CSI is not subjected to bit sequence generation and channel coding;

[0126] Step 402: The UE processes the CSI based on the first processing method to obtain complex value information associated with the CSI;

[0127] Step 403: The UE sends an uplink channel or signal based on the complex value information.

[0128] According to the method of the embodiment of the present disclosure, it can be determined to use different methods to process CSI to obtain complex value information for transmission.

[0129] Various aspects of the present disclosure are described in more detail below.

[0130] In some examples, the "channel state information (CSI)" may include one of a vector, a matrix, or a tensor including an estimated channel impulse response (CIR) in at least one dimension. For example, the at least one dimension may include at least one dimension in the time domain, the frequency domain, and the spatial domain. The granularity of the channel impulse response per unit in the time domain dimension may include symbol, slot, subframe, frame, etc.; the granularity of the channel impulse response per unit in the frequency domain dimension may include subcarrier, resource block, multiple resource blocks of fixed length, subband, wideband, etc.; the granularity of the channel impulse response per unit in the spatial domain dimension may include antenna, antenna port, antenna panel, beam, etc. The channel impulse response in the spatial domain dimension may include a channel impulse response associated with a UE or a channel impulse response associated with a base station. The CSI may include channel impulse responses in both the UE and base station spatial dimensions. Taking CSI as an example, a tensor of channel impulse response including four dimensions: time domain, frequency domain, UE spatial domain, and base station spatial domain, the granularity units of CSI in the time domain, frequency domain, UE spatial domain, and base station spatial domain are symbol, subcarrier, UE antenna port, and base station antenna port, respectively. The number of samples of channel impulse response of CSI in the time domain, frequency domain, UE spatial domain, and base station spatial domain are T, F, RX, and TX, respectively. Then, CSI is a tensor of dimension [T, F, RX, TX]. The dimensions of the CSI may also include [T, F, TX, RX], [T, TX, F, RX], [TX, T, F, RX], [T, RX, F, TX], [RX, T, F, TX], [F, T, RX, TX], etc. Taking the CSI as an example, which is 14 time domain symbols, 936 frequency domain subcarriers, 4 receive antenna ports, and 64 transmit antenna ports, the CSI is a tensor with a dimension of [T, F, RX, TX] = [14, 936, 4, 64].

[0131] For example, the estimated channel impulse response may include a channel impulse response obtained based on a pilot signal. The pilot signal may include a demodulation reference signal (DMRS), a sounding reference signal (SRS), a channel state information reference signal (CSI-RS), etc. The pilot signal used to obtain the estimated channel impulse response may be configured by a base station. The method for obtaining the channel impulse response based on the pilot signal may include a channel estimation (CE) method, including least squares channel estimation (LS CE), minimum mean square error channel estimation (MMSE CE), maximum likelihood channel estimation (ML CE), Kalman filter channel estimation (KF CE), etc. The channel state information may also include an estimated channel impulse response vector with only one dimension, or an estimated channel impulse response tensor with no less than three dimensions. The channel state information may also be referred to as a raw channel matrix or an estimated raw channel matrix.

[0132] In some examples, the "channel state information (CSI)" may also include an autocorrelation matrix of an estimated channel impulse response matrix. The autocorrelation matrix may be obtained by multiplying the estimated channel impulse response matrix by a conjugate matrix of the estimated channel impulse response matrix.

[0133] In some examples, the "channel state information (CSI)" may also be channel state information after a signal processing-related transformation. For example, the channel state information after a signal processing-related transformation is obtained by performing a signal processing-related transformation on an estimated channel impulse response matrix, where the signal processing-related transformation may include one of the following methods: Fourier transform, Laplace transform, wavelet transform, Hilbert-Huang transform, bispectral transform, etc. For example, the channel state information after a signal processing-related transformation may include delay domain or angle domain information of the channel state, where the information is obtained by performing a Fourier transform on the estimated channel impulse response matrix.

[0134] In some examples, the "channel state information (CSI)" may also include a precoding matrix (Precoding matrix) including precoding weights in at least one dimension in the time domain, frequency domain, and spatial domain. For example, the precoding matrix includes no less than one eigenvector. The eigenvector may include a representation in the spatial domain or a representation in the angle-delay domain. For example, the precoding matrix may be obtained based on an estimated channel impulse response matrix by a precoding method, wherein the precoding method may include zero-forcing precoding (ZF precoding), block diagonal precoding (BD precoding), minimum mean square error precoding (MMSE precoding), TH precoding (Tomlinson-Harashima precoding), codebook-based precoding, etc. The codebook in the codebook-based precoding may be obtained by a discrete Fourier transform (DFT) matrix. For example, the channel state information may also include a precoding vector with only one dimension, or a precoding tensor with no less than three dimensions.

[0135] In some examples, the "channel state information (CSI)" may also include one of the following: precoding matrix indication (PMI), channel quality indication (CQI), CSI-RS resource indication (CRI), SS / PBCH resource block indication (SSBRI), layer indication (LI), rank indication (RI), reference signal received power (RSRP), etc.

[0136] In some examples, the "complex-valued information associated with CSI" may include information that characterizes or carries CSI, including one of a sequence, vector, matrix, or tensor of at least one complex-valued modulation symbol. The complex-valued modulation symbol is represented by a complex value within a fixed range. The complex-valued information may be discrete or continuous. Specifically, the complex-valued modulation symbol in the discrete complex-valued information is represented by quantizing the range to a finite fixed value, and the complex-valued modulation symbol in the continuous complex-valued information is represented by any value within the range. For example, the complex-valued modulation symbol may be represented by The two numerical values are represented, and the method is called BPSK modulation. The complex value information formed by the method is discrete. The complex value modulation symbol can be in the range Any numerical value in indicates that the complex-valued information constructed by the method is continuous.

[0137] In one embodiment, the method for UE to report CSI is to report a precoding matrix indication. For example, the UE and the base station have the same precoding codebook, which includes no less than one eigenvector carrying a mark, and the base station obtains the precoding matrix by mapping the obtained precoding matrix indication to the corresponding eigenvector. Compared with the above method, the method proposed in the present invention can directly generate complex-valued information from the original CSI through a method and report it to the base station. Compared with the above method, the reporting method proposed in the present invention can have a finer reporting granularity, which can make the reported information more accurate, thereby improving the performance of the system. The method of directly reporting the original CSI proposed in the present invention can give the base station greater flexibility to select the precoding method, rather than being limited to the method provided in the codebook, so that there can be a higher performance ceiling.

[0138] In some examples, the "determining a first processing method among multiple methods for obtaining complex-valued information associated with channel state information (CSI)" includes: the UE determines, based on an indication in configuration information, a first processing method among multiple methods for obtaining complex-valued information representing CSI. Specifically, the UE includes multiple methods and information related to each method, such as an identifier. In this disclosure, for convenience of description, "identifier" is used as an example of information related to a method, scheme, or table. It is understood that this is merely exemplary, and the "identifier" of a method, scheme, or table can also be replaced with other information or expressions that can be used to represent, indicate, or determine the method, scheme, or table, such as "name," "type," "indication," "identifier," "index," etc. The indication is an identifier of a method, and the UE determines the selected method by matching the indication with the identifiers of multiple methods. The identifier can include an index, name, etc. of the method. The configuration information can include one of higher-layer signaling, MAC layer signaling, and physical layer signaling, such as radio resource control (RRC) and downlink control information (DCI).

[0139] In some examples, the configuration information may include an uplink shared channel configuration (also referred to as PUSCH-Config) or an uplink control channel configuration (also referred to as PUCCH-Config) in RRC. The configuration information may also include downlink control signaling (DCI).

[0140] In some examples, each method in the "multiple methods" can be implemented using no less than one scheme. The scheme represents an implementation method based on a parameter set, wherein the parameter set needs to be determined or configured by the UE. For example, when the method is the first method, the scheme may include a modulation and coding scheme (MCS), wherein the required parameter set includes a modulation order (Modulation Order) and a target coding rate (Targetcode Rate). For example, when the method is the second method, the parameter set of the scheme may include at least one of the following parameters: model type, number of model layers, number of model neurons, model activation function, model weight, input format, output format, etc.

[0141] In some examples, determining the first processing method further includes: the UE determining, based on the indication of the configuration information, one of the multiple methods for obtaining complex-valued information representing the CSI, and a scheme for executing the method. For example, the UE includes information about multiple methods, and includes multiple schemes and identifiers of each scheme in each method of the UE; the indication includes a method identifier and a scheme identifier. After the UE determines the method based on the method identifier in the indication, the UE determines the selected scheme by matching the scheme identifier in the indication with the identifiers of the multiple schemes included in the UE. The scheme identifier may include the index, name, characteristics, etc. of the scheme.

[0142] Figure 5 A schematic diagram shows a method and scheme for a UE in a communication system according to the present disclosure to determine and obtain complex value information representing CSI using first and second configuration information.

[0143] In some examples, determining the first processing method further includes: the UE determining, based on an indication of the first configuration information, one of multiple methods for obtaining complex-valued information representing the CSI; and the UE determining, based on an indication of the second configuration information, a scheme for executing the method. For example, the UE determines, based on a method identifier indicated by the radio resource control information, that the method for obtaining the complex-valued information representing the CSI is the second method; thereafter, the UE determines, based on a scheme identifier indicated by the downlink control information, a scheme for executing the method, wherein the scheme is a neural network model and includes the following parameters: model type, number of model layers, number of model neurons, model activation function, model weights, input format, and output format.

[0144] In some examples, the first method may include: the UE generates an uplink control information bit sequence to obtain a bit sequence related to the uplink control information; the UE segments the bit sequence and adds parity check to each bit sequence block to obtain a bit sequence with an attached CRC (bits after CRC attachment); the UE performs channel coding on the bit sequence with the attached CRC to obtain a channel-coded bit sequence; the UE modulates the channel-coded bit sequence according to a configured modulation order to obtain complex-valued information including no less than one complex-valued modulation symbol.

[0145] In some examples, the first method may further include: the UE generates an uplink control information bit sequence to obtain a bit sequence related to the uplink control information; the UE cuts the bit sequence and adds a parity check at the end of each bit sequence block to obtain a bit sequence with a CRC attached; the UE performs bit interleaving on the bit sequence with the CRC attached to obtain an interleaved bit sequence; the UE performs channel coding on the interleaved bit sequence according to a polarization sequence to obtain a channel-coded bit sequence; the UE modulates the channel-coded bit sequence according to a configured modulation order to obtain complex-valued information including at least one complex-valued modulation symbol.

[0146] In some examples, the first method may include: the UE generates an uplink control information bit sequence to obtain a bit sequence related to the uplink control information; the UE adds a parity check at the end of the bit sequence to obtain a CRC-attached bit sequence; the UE segments the CRC-attached bit sequence and optionally adds an additional CRC sequence at the end of each bit sequence block; the UE performs channel coding on the segmented bit sequence blocks according to a low-density parity check pattern to obtain a channel-coded bit sequence; the UE modulates the channel-coded bit sequence according to a configured modulation order to obtain complex-valued information including at least one complex-valued modulation symbol. In some examples, the first method may also include: the UE obtains the channel-coded bit sequence through a neural network model; the UE modulates the channel-coded bit sequence according to the configured modulation order to obtain complex-valued information including at least one complex-valued modulation symbol. The method for the UE to obtain the channel-coded bit sequence through a neural network model may include: the UE inputs the CSI into a neural network model to obtain an output; the UE inputs the output of the neural network model into a quantizer to obtain a channel-coded bit sequence. The neural network model of the UE may include an encoder portion of an autoencoder.

[0147] Figure 6 A schematic diagram is shown of obtaining complex-valued information based on a CSI bit sequence at a UE end and CSI reconstruction at a base station end in a communication system according to the present disclosure.

[0148] In some examples, the UE is configured with an encoder portion and a quantizer (Quantizer) in an autoencoder, and the base station is configured with a decoder portion in the same autoencoder and a dequantizer (De-Quantizer) paired with the quantizer in the UE. The method for obtaining the encoder and decoder is to train the autoencoder they constitute. The quantizer is used to obtain a bit sequence through a floating-point vector, and the dequantizer is used to obtain a floating-point vector (floating-point vector) through a bit sequence. The quantizer and the dequantizer are paired with each other to indicate that the bit sequence generated by the quantizer can be restored to a floating-point vector by the dequantizer, and the restored floating-point vector is the same as the floating-point vector. The same is the same in the case of loss of quantization accuracy.

[0149] A method for a UE to obtain complex information carrying CSI may include: the UE obtaining a floating-point number vector carrying CSI based on a configured encoder; the UE obtaining a bit sequence carrying CSI based on a quantizer and the floating-point number vector; and the UE obtaining the complex information carrying CSI by modulating the bit sequence carrying CSI. A method for a base station to obtain CSI based on the complex information carrying CSI may include: the UE demodulating the complex information carrying CSI to obtain a bit sequence carrying CSI; the UE obtaining a floating-point number vector carrying CSI based on a dequantizer and the bit sequence carrying CSI; and obtaining CSI based on the floating-point number vector carrying CSI and a configured decoder.

[0150] Figure 7 A schematic diagram of a neural network model used by a first training method in a communication system according to the present disclosure is shown.

[0151] In some examples, the UE is configured with an encoder portion and quantizer from an autoencoder, and the base station is configured with a decoder portion and dequantizer from the same or corresponding autoencoder. The encoder and decoder are obtained by training an autoencoder consisting of an encoder, a channel network, and a decoder, for example, jointly or separately. When training the autoencoder, the method for inputting CSI into the autoencoder for forward propagation includes: inputting the CSI into the encoder to obtain a floating-point vector representing the CSI; inputting the floating-point vector into the quantizer to obtain a bit sequence representing the CSI; inputting the bit sequence representing the CSI into the channel network to obtain a bit sequence that has passed through the channel; inputting the bit sequence that has passed through the channel into the dequantizer to obtain a floating-point vector that has passed through the channel; and inputting the floating-point vector that has passed through the channel into the decoder to obtain reconstructed CSI. The channel network is a neural network that distorts input information by simulating a real channel. The input of the channel network is bit information, and the output is bit information that has passed through the channel. For example, the channel network may include a neural network that simulates a binary symmetric channel or a binary erasure channel, which functions to flip each bit in the bit information according to a probability corresponding to the SINR of the channel. When training the autoencoder, updating the autoencoder weights by gradient descent further includes: locking the weights of the quantizer, the channel network, and the dequantizer. Locking means that the weights will not be updated.

[0152] In some examples, the method for training the paired autoencoder includes: inputting CSI into the autoencoder for forward propagation to obtain reconstructed CSI output by the decoder; calculating a loss function based on the CSI input to the autoencoder and the reconstructed CSI output by the decoder; and updating the weights of the autoencoder using the loss function by gradient descent according to a backpropagation method. The loss function may include mean square error (MSE), normalized mean square error (NMSE), cosine similarity (GCS), etc.

[0153] Figure 8 A schematic diagram of a neural network structure for obtaining a bit sequence carrying a CSI feature vector at a UE and reconstructing CSI at a base station in a communication system according to the present disclosure is shown.

[0154] In some examples, the UE is configured with an encoder portion and quantizer in an autoencoder, and the base station is configured with a decoder portion and dequantizer in the same or corresponding autoencoder. The structure of the autoencoder may include an encoder and a decoder each including multiple transformer layers. For example, the encoder input may include CSI feature vectors for N subbands. The CSI feature vector is input into the encoder, and after processing through a linear layer, a multi-layer transformer, and a linear layer, a floating-point vector carrying the CSI feature vector is obtained. The floating-point vector is then input into the quantizer to obtain a bit sequence carrying the CSI feature vector. Before inputting the CSI feature vector into the encoder, the UE may obtain the CSI feature vector by performing SVD decomposition based on the estimated original channel matrix. The decoder has the same structure as the encoder, and the bit sequence carrying the CSI feature vector is input into the dequantizer to obtain a floating-point vector carrying the CSI feature vector. The floating-point vector is then input into the decoder, and after processing through a linear layer, a multi-layer transformer, and a linear layer, a reconstructed CSI feature vector is obtained.

[0155] Figure 9 A schematic diagram shows a method for a UE in a communication system according to the present disclosure to obtain complex-valued information by using a neural network.

[0156] In some examples, the second method may include obtaining complex-valued information based on the CSI and a neural network model. For example, the method of obtaining the complex-valued information using the neural network model may include converting the data format and structure of the CSI to obtain a CSI input consistent with the data format and structure of a neural network input layer; inputting the converted CSI into the neural network to obtain a neural network output related to the complex-valued information; and obtaining the complex-valued information based on the neural network output.

[0157] Among them, the structure of the "neural network model" includes but is not limited to auto-encoder, denoising autoencoder, variational autoencoder, adversarial neural network (GAN), diffusion model, multi-layer perceptron (MLP), convolutional neural network (CNN), deep neural network (DNN), recurrent neural network (RNN), restricted Boltzmann machine (RBM), graph neural network (GNN), deep belief network (DBN), bidirectional recursive deep neural network (BRDNN), neural network with self-attention mechanism (such as transformer), etc.

[0158] In some examples, the complex-valued information obtained from the complex-valued information obtained based on the CSI without performing bit sequence generation and channel coding may not carry verification information.

[0159] In some examples, the complex-valued information obtained in the method of obtaining complex-valued information based on CSI without bit sequence generation and channel coding may carry check information. The check information may include a cyclic redundancy check (CRC). The method of adding check information to the complex-valued information may include at least one of the following: adding a CRC before inputting the CSI into the neural network; and adding a CRC to the complex-valued information obtained based on the output of the neural network.

[0160] In some examples, the data format of CSI in the "obtaining CSI input consistent with the data format and structure of the neural network input layer by transforming the data format and structure of CSI" may include complex numbers, and the data format of the neural network model may include floating-point numbers. The method of transforming the data format of CSI may include separating the real part and the imaginary part of the complex number in CSI to obtain two floating-point numbers. The data structure of CSI may include one of a vector, a matrix, and a tensor, and the structure of the neural network input may also include one of a vector, a matrix, and a tensor. The method of transforming the data structure of CSI is to perform dimensionality conversion on CSI according to the data structure of the neural network input layer. The dimensional conversion may also be referred to as permute, permutation, transpose, rearrange, reararment, alteration, etc. Taking CSI as an example, which is a channel impulse response including four dimensions: time domain, frequency domain, UE spatial domain, and base station spatial domain, and the structure of the neural network input is a vector, the number of samples of the channel impulse response of CSI in the time domain, frequency domain, UE spatial domain, and base station spatial domain are T, F, RX, and TX, respectively, forming a tensor with a dimension of [T, F, RX, TX]. The method for obtaining CSI input consistent with the data format and structure of the neural network input layer includes: first, converting the data format of the CSI from a complex number to a floating point number, forming a tensor with a dimension of [T, F, RX, TX, RI], where RI = 2 carries a floating point number representing the real part and imaginary part of the complex-valued symbol; then, converting the dimension of the tensor with the dimension of [T, F, RX, TX, RI] and reducing it to a vector including T*F*RX*TX*RI elements.

[0161] In some examples, the data format of the neural network output may also include complex numbers, the data structure of the neural network output may include one of a sequence, a vector, a matrix, and a tensor, and the structure of the complex-valued information may include one of a sequence, a vector, a matrix, and a tensor. The method of obtaining complex-valued information based on the output of the neural network may include obtaining complex-valued information from the output of the neural network using a data format and structure transformation method. The method of transforming the data format of the neural network output may include combining floating-point numbers into complex numbers. The method of structurally transforming the neural network output is to perform dimensional conversion on the neural network output according to the specified complex-valued information data format. Taking the example that the neural network output is a vector and the complex-valued information includes two dimensions, time domain and frequency domain, the length of the neural network output vector is T out *F out *RI, where RI=2, the method for obtaining complex value information includes: first, converting the dimension of the neural network output vector into a vector with dimension [T out ,F out ,RI]; then, the two floating point numbers in each group of data of the RI dimension in the above tensor are used as the real part and the imaginary part to form a complex number, and a tensor of dimension [T out ,F out ], the tensor is a tensor containing T out *F out The complex-valued information of the complex-valued modulation symbols.

[0162] Figure 10 A schematic diagram of an autoencoder including a neural network model for acquiring CSI complex-valued information in a communication system according to the present disclosure is shown.

[0163] In some examples, the neural network model used by the UE to obtain CSI complex-valued information may include an encoder part in an autoencoder. The autoencoder includes three parts: an encoder, a channel network, and a decoder, and each part of the autoencoder may include a neural network with at least one layer. The input of the encoder is CSI, and the output is complex-valued information. The input of the decoder is complex-valued information passing through the channel, and the output is CSI. The channel network is a neural network that distorts input information by imitating a real channel, and the input of the channel network is complex-valued information, and the output is complex-valued information passing through the channel. For example, the channel network may include a neural network that simulates a Gaussian channel, which acts to add Gaussian noise to the complex-valued information.

[0164] In some examples, a method for a base station to obtain CSI may include: the base station receives a radio frequency signal and extracts complex-valued information carrying CSI from the radio frequency signal; the base station inputs the copied information into a decoder used to reconstruct CSI, and obtains CSI from the decoder output. The radio frequency signal received by the base station is a radio frequency signal transmitted by a UE and carrying complex-valued CSI information, and the decoder used by the base station to reconstruct CSI is a decoder paired with an encoder used by the UE to obtain the complex-valued information. The pairing means that the decoder used to reconstruct CSI by the base station and the encoder used by the UE to obtain the complex-valued CSI information are trained and acquired through pairing.

[0165] Figure 11 A schematic diagram of obtaining CSI complex value information at a UE and reconstructing CSI at a base station in a communication system according to the present disclosure is shown.

[0166] In some examples, the UE is configured with an encoder portion of an autoencoder, and the base station is configured with a decoder portion of the same or corresponding autoencoder. The method for acquiring the encoder and decoder is to train the autoencoder consisting of the encoder, decoder, and channel network, for example, jointly or separately. The method for the UE to acquire complex information carrying CSI may include: the UE acquiring the complex information carrying CSI based on the configured encoder, and the base station acquiring CSI based on the received complex information based on the configured decoder. The UE acquires the complex information carrying CSI from the encoder output without a quantizer, and the base station acquires CSI from the decoder output without a dequantizer. The advantage of not using a quantizer and dequantizer is that quantization error is eliminated. The UE acquires the complex information carrying CSI from the encoder output without additional modulation, and the base station acquires CSI from the decoder output without additional demodulation. No additional modulation and demodulation are required because the encoder and decoder have modulation and demodulation capabilities. The encoder output is a modulated complex-valued modulation symbol, and the decoder output is a demodulated floating-point vector carrying CSI.

[0167] Figure 12 A schematic diagram of a neural network structure for acquiring CSI complex value information at the UE side and reconstructing CSI at the base station side in a communication system according to the present disclosure is shown.

[0168] In some examples, the UE is configured with an encoder part in an autoencoder, and the base station is configured with a decoder part in the same or corresponding autoencoder. The structure of the autoencoder may include an encoder and a decoder each including a plurality of transformer layers. For example, the input of the encoder may include CSI feature vectors on N subbands. By inputting the CSI feature vector into the encoder, complex-valued information carrying the CSI feature vector is obtained after processing through a linear layer, a multi-layer transformer, and a linear layer. Before inputting the CSI feature vector into the encoder, the method for the UE to obtain the CSI feature vector may include performing SVD decomposition based on the estimated original channel matrix. The decoder has the same structure as the encoder, and by inputting the complex-valued information carrying the CSI feature vector into the decoder, a reconstructed CSI feature vector is obtained after processing through a linear layer, a multi-layer transformer, and a linear layer.

[0169] In some examples, the method for training the paired autoencoder includes: inputting CSI into the autoencoder for forward propagation to obtain reconstructed CSI output by the decoder; calculating a loss function based on the CSI input to the autoencoder and the reconstructed CSI output by the decoder; and updating the weights of the autoencoder using the loss function by gradient descent according to a backpropagation method. The loss function may include mean square error (MSE), normalized mean square error (NMSE), cosine similarity (GCS), etc.

[0170] Figure 13 A schematic diagram of training an autoencoder according to the present disclosure is shown.

[0171] In some examples, when training the autoencoder in pairs, the method of inputting CSI into the autoencoder for forward propagation includes: inputting the CSI into the encoder to obtain complex-valued information representing the CSI; inputting the complex-valued information into the channel network to obtain complex-valued information that has passed through the channel; and inputting the complex-valued information that has passed through the channel into the decoder to obtain reconstructed CSI. The channel network may function to add a random number representing the Gaussian channel to each complex-valued symbol in the input complex-valued information and output the result. When training the autoencoder in pairs, updating the autoencoder's weights by gradient descent also includes: locking the weights of the channel network. Locking means that the weights of the channel network will not be updated.

[0172] In some examples, the auto-encoder may be trained at the UE. For example, the UE trains the auto-encoder based on the acquired CSI and reports the trained decoder to the base station.

[0173] In some examples, the autoencoder can also be trained at the base station. For example, the UE reports acquired CSI to the base station, which then trains the autoencoder based on the acquired CSI and sends the trained autoencoder to the UE. The process of the UE reporting CSI to the base station may include: the UE acquiring CSI and storing it in a buffer; the UE collecting a certain amount of CSI to form a CSI dataset; and the UE reporting the CSI dataset.

[0174] In some examples, the autoencoder may also be trained offline. For example, an autoencoder is obtained by an offline training method, and the encoder and decoder are deployed on the UE and the base station, respectively. In some examples, the "uplink channel or signal" may include a physical channel for carrying uplink control information or a physical channel for carrying uplink shared information. The physical channel for carrying uplink control information may include a physical uplink control channel (also known as a Physical Uplink Control Channel or PUCCH). The physical channel for carrying uplink shared information may include a physical uplink shared channel (also known as a Physical Uplink Shared Channel or PUSCH).

[0175] Figure 14 A schematic diagram showing complex information of a UE multiplexing the uplink channel or signal to carry second uplink control information in a communication system according to the present disclosure is shown.

[0176] In some examples, before the UE maps the complex-valued information to an uplink channel or signal and sends it, it also includes: the UE multiplexes the complex-valued information of the uplink channel or signal to carry the complex-valued information of the second uplink control information. The multiplexing method may include, first, based on the complex-valued information related to the CSI and the complex-valued information related to the second uplink control information, obtaining the multiplexed complex-valued information, and then mapping the multiplexed complex-valued information to the uplink channel or signal. The complex-valued information of the second uplink control information may include one of the following methods, including: a method of obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on the second uplink control information, or a method of obtaining complex-valued information based on the second uplink control information, wherein the second uplink control information is not subjected to bit sequence generation and channel coding. The CSI may also be referred to as the first uplink control information.

[0177] In some examples, the "second uplink control information" may include at least one of the following information: channel state information (CSI), a scheduling request (SR), and an acknowledgement (ACK). The second uplink control information may be the same as the first uplink control information. For example, the first uplink control information is CSI, and the second uplink control information is also CSI. The second uplink control information may be different from the first uplink control information. For example, the first uplink control information is CSI, and the second uplink control information is a scheduling request or an acknowledgement.

[0178] In some examples, before obtaining the complex-valued information of the second uplink control information, the method further includes: the UE determines, according to the indication of the configuration information, one of multiple methods for obtaining the complex-valued information representing the second uplink control information. In some examples, each method in the "multiple methods for representing the complex-valued information of the second uplink control information" can be implemented using at least one scheme. For example, when the method is "a method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing bit sequence generation and channel coding on the second uplink control information", the scheme may include a modulation and coding scheme (MCS), wherein the required parameters include a modulation order and a target code rate. For example, when the method is "a method for obtaining complex-valued information based on the second uplink control information, wherein the second uplink control information is not subjected to bit sequence generation and channel coding", the parameters that may be included in the scheme include at least one of the following: model type, number of model layers, number of model neurons, model activation function, model weight, input format, output format, etc.

[0179] In some examples, the method for "obtaining the complex-valued information of the second uplink control information" may be different from the method for obtaining the complex-valued information of the first uplink control information. For example, the second uplink control information is obtained by performing channel coding and modulation to obtain the complex-valued information, while the first uplink control information is obtained by not performing at least one of the channel coding and modulation steps.

[0180] In some examples, the number of complex-valued symbols included in the "multiplexed complex-valued information" is the same as the sum of the number of complex symbols in the complex-valued information related to the first uplink control information and the complex-valued information related to the second uplink control information. The multiplexed complex-valued information includes a first part and a second part of complex symbols, the first part of complex symbols being the complex symbols in the first or second uplink control information, and the second part of complex symbols being the complex symbols in the first or second uplink control information other than the first part of complex symbols. For example, the complex symbols included in the first uplink control information are The complex symbols included in the second uplink control information are The first part of the multiplexed complex value information is the complex symbol included in the first uplink control information, and the multiplexed complex value information is g0, g1, g2, ..., g G-1 ,in, When i=0,1,…,G (1) -1, When i=G (1) ,G (1) +1,…,G (1) +G (2) -1.

[0181] In some examples, the "multiplexed complex-valued information" further includes obtaining the multiplexed complex-valued information according to different priorities. The first uplink control information and the second uplink control information used to obtain the multiplexed complex-valued information are configured to have different priorities, and the first portion of complex symbols included in the multiplexed complex-valued information are the complex symbols in the uplink control information with a higher priority, and the second portion of complex symbols are the complex symbols in the uplink control information with a lower priority. Taking the example of the first uplink control information having a higher priority than the second uplink control information, the first portion of complex symbols included in the multiplexed complex-valued information are the complex symbols in the first uplink control information, and the second portion of complex symbols are the complex symbols in the second uplink control information.

[0182] In some examples, before the UE maps the complex-valued information to an uplink channel or signal and sends it, it also includes: the UE multiplexes the uplink channel or signal to carry complex-valued information of other information. The other information may include data information. The multiplexing method may include, first, obtaining the multiplexed complex-valued information based on the complex-valued information related to the uplink control information and the complex-valued information related to the other information, and then mapping the multiplexed complex-valued information to the uplink channel or signal. Obtaining the complex-valued information of the other information may include one of the following methods, including: a method of obtaining the complex-valued information by performing channel coding and modulation based on the other information, or a method of obtaining the complex-valued information without performing at least one of the channel coding and modulation steps based on the other information.

[0183] In some examples, the "uplink channel or signal" may include a physical uplink control channel (PUCCH) or a physical uplink shared channel (PUSCH).

[0184] In some examples, before the UE maps the complex-valued information to an uplink channel or signal, the method further includes: the UE obtains matched complex-valued information based on the complex-valued information and the number of time-frequency resource units used to report uplink control information, wherein the number of complex-valued symbols contained in the matched complex-valued information is the same as the number of time-frequency resource units used to report uplink control information. The matching method may also be referred to as rate matching, symbol selection, symbol-level rate matching, etc. The time-frequency resource unit may include a resource element (RE).

[0185] For the sake of convenience, the following content will refer to "time-frequency resource unit used to report uplink control information" as "resource unit", and the following content will refer to "the number of time-frequency resource units used to report uplink control information" as "the number of resource units".

[0186] In some examples, the method for the UE to obtain "the number of time-frequency resource units used to report uplink control information" may include: the UE obtains it according to the configured time-frequency resource information and pilot information. Among them, the pilot may include a demodulation reference signal (DM-RS). Specifically, the UE obtains the number of symbols related to the uplink control information, the number of physical resource blocks (PRBs), the spreading parameter (spreading factor) and the position information of the pilot according to the configuration. The UE obtains the number of time-frequency resource units that can be used for uplink control information in each symbol and each PRB based on the position information of the pilot, and obtains the number of resource units associated with reporting the uplink control information based on the above information. The number of symbols related to the uplink control information is N, the number of physical resource blocks (PRBs) is N, and the spreading parameter (spreading factor) is N respectively. symb,UCI 、N PRB 、N SF For example, the number of time-frequency resource units that can be used for uplink control information in each symbol and each PRB is 8 or 12, and the number of resource units associated with reporting uplink control information is 8*N symb,UCI *N PRB / N SF or 16*N symb,UCI *N PRB / N SF .

[0187] In some examples, the method for the UE to obtain the matched complex-valued information based on the complex-valued information and the number of time-frequency resource units used to report uplink control information may include: the UE writes the complex-valued information into a circular buffer, wherein the length of the circular buffer is equal to the number of complex symbols in the complex-valued information; the UE reads the complex symbols from the circular buffer according to the configured method to obtain the matched complex-valued information. The configured method may include: the UE extracts the complex symbols from the circular buffer according to the configured starting position and / or reading order to obtain the matched complex-valued information. The number of complex-valued symbols contained in the matched complex-valued information is the same as the number of resource units. The starting position, reading order, and number of resource units may be obtained based on the configuration information.

[0188] In some examples, the "circular buffer" is characterized in that the first complex symbol of the complex value information stored in the circular buffer is adjacent to the last complex symbol. The complex value information includes E complex symbols e0, e1, e2, ..., e E-1 For example, the length of the circular buffer is E, and in the circular buffer e E-1 The next symbol is e0. Taking the example of extracting E+3 symbols from the e0 symbol in the above circular buffer, the obtained complex value information after matching is e0, e1, e2, ..., e E-1 ,e0,e1,e2.

[0189] Figure 15 A schematic diagram of a circular buffer for fast matching in a communication system according to the present disclosure is shown.

[0190] In some examples, the UE reads complex symbols from the circular buffer according to a configured method, and the method for obtaining the matched complex value information may include, when the length of the complex information is greater than the number of resource units, calculating the difference between the length of the complex information and the number of resource units, and using the difference as the starting position of the circular buffer reading, extracting complex symbols in sequence until complex symbols equal to the number of resource units are obtained. This method may be called puncturing. The advantage of using the puncturing method to achieve rate matching is that when the priority of the complex symbols in the circular buffer increases with position, the priority of the complex symbols retained by the UE through puncturing is higher than the priority of the lost complex symbols. The UE improves the effectiveness of the signal recovered by the receiving end by sending complex symbols with higher priority. With complex information y0, y1, y2,…, y N-1 , which includes N complex-valued modulation symbols and the number of resource units is E, where N>E. For example, the UE first converts the complex information y0, y1, y2, ..., y N-1Store it in the circular buffer, and then read it from the NEth complex-valued modulation symbol in the circular buffer to obtain the matched complex-valued information e0, e1, e2, ..., e E-1 For example, the method is to use k=0 to E-1, and e k =y k+N-E .

[0191] In some examples, the UE reads complex symbols from the circular buffer according to a configured method, and the method for obtaining the matched complex value information may include, when the length of the complex information is greater than the number of resource units, starting from the first complex symbol of the circular buffer, extracting the complex symbols in sequence until the number of complex symbols equal to the number of resource units is obtained. This method may be called shortening. The advantage of using the shortening method to achieve rate matching is that when the priority of the complex symbols in the circular buffer decreases with position, the priority of the complex symbols retained by the UE through puncturing is higher than the priority of the lost complex symbols. The UE improves the effectiveness of the signal recovered by the receiving end by sending complex symbols with higher priority. With complex information y0, y1, y2,…, y N-1 , which includes N complex-valued modulation symbols and the number of resource units is E, where N>E. For example, the UE first converts the complex information y0, y1, y2, ..., y N-1 Store in the circular buffer, and then read from the first complex-valued modulation symbol in the circular buffer to obtain the matched complex-valued information e0, e1, e2, ..., e E-1 For example, the method is to use k=0 to E-1, and e k =y k .

[0192] In some examples, the UE reads complex symbols from the circular buffer according to a configured method, and the method for obtaining the matched complex value information may include, when the length of the complex information is less than the number of resource units, starting from the first complex symbol of the circular buffer, extracting the complex symbols in a forward or reverse order until the number of complex symbols equal to the number of resource units is obtained. This method may be called repetition. The reason why this method is called repetition is that when the complex symbols are extracted sequentially and the entire circular buffer is traversed for the first time, the number of extracted complex symbols does not reach the required rate matching output sequence length. At this time, it still starts from the first complex symbol of the circular buffer and extracts complex symbols in sequence until the number of complex symbols equal to the number of resource units is obtained. The advantage of using the repetition method to achieve rate matching is that when the priority of the complex symbols in the circular buffer increases or decreases with position, the UE achieves diversity gain of the high-priority complex symbols by repeatedly sending complex symbols with high priority, thereby improving the accuracy of the signal recovered by the receiving end. With complex information y0, y1, y2,…, y N-1, which includes N complex-valued modulation symbols and the number of resource units is E. Taking N < E as an example, the UE first stores the complex information y0, y1, y2, …, y N-1 in a cyclic buffer, and then starts reading from the first complex-valued modulation symbol in the cyclic buffer to obtain the matched complex information e0, e1, e2, …, e E-1 . For example, the method is to set k from 0 to E - 1, and set e k = y mod(k,N) .

[0193] In some examples, before the UE maps the complex information to an uplink channel or signal, it further includes: the UE multiplexes the uplink channel or signal to carry second uplink control information to obtain multiplexed complex information; the UE obtains the matched complex information according to the multiplexed complex information and the number of time-frequency resource units used to report uplink control information, where the number of complex symbols included in the matched complex information is the same as the number of time-frequency resource units used to report uplink control information.

[0194] In some examples, before the UE maps the complex information to an uplink channel or signal, it further includes: the UE performs interleaving processing on the complex symbols in the complex information according to the configuration information to obtain interleaved complex information. For example, the method of performing interleaving processing on the complex symbols in the complex information includes: the UE determines the interleaving order according to the uplink signal related configuration information; the UE determines the interleaving mapping pattern according to the interleaving order and performs interleaving mapping on the second sequence.

[0195] For example, the second sequence of the UE includes a total of E complex symbols e0, e1, e2, …, e E-1 , and these E complex symbols are interleaved into a complex symbol sequence f0, f1, f2, …, f E-1 . The UE determines the interleaving order as Q through the uplink signal related configuration information m , then the method for the UE to obtain the complex symbol sequence f0, f1, f2, …, f E-1 through interleaving is as follows in pseudocode:

[0196]

[0197] In some examples, the method for the UE to map the complex information to an uplink channel or signal and transmit it includes: the UE maps the complex symbols in the complex information to the time-frequency resource units indicated by the configuration to obtain complex information associated with the time-frequency resource; the UE obtains a radio frequency signal and transmits it according to the complex information associated with the time-frequency resource.

[0198] Figure 16A schematic diagram is shown of mapping complex symbols in complex information to time-frequency resource units indicated by a configuration by a user in a communication system according to the present disclosure.

[0199] In some examples, the UE maps the complex-valued symbols in the complex-valued information to the time-frequency resource units indicated by the configuration, and the method for obtaining the complex-valued information associated with the time-frequency resources includes: the UE multiplies each complex symbol in the complex-valued information by an amplitude scaling factor, and maps the complex symbols in sequence to all qualified (unreserved) resource units. The purpose of multiplying each complex symbol by the amplitude scaling factor is to make each symbol meet the specified transmit power. The qualified resource units meet the following requirements: the resource units are located on the resource blocks allocated to the UE for transmission; the resource units are not allocated to be associated with the demodulation reference signal (DM-RS). The order of mapping the complex symbols to the resource units is to map them to qualified frequency domain resource units in sequence on a time domain symbol in the order of the serial numbers, and when all the frequency domain resources on a time domain symbol are filled, the frequency domain resource units on the next time domain symbol are mapped. For example, the time-frequency resource unit is represented by (k, l), where k represents the subcarrier number and l represents the time domain symbol number. The UE first maps the complex symbols to the unallocated resource units in ascending order of sequence number k, and then, when all the frequency domain resources on a time domain symbol are filled, maps the resources on the next time domain symbol in ascending order of sequence number l.

[0200] Figure 17 A schematic diagram is shown of mapping pilot-related complex symbols to time-frequency resource units by a user in a communication system according to the present disclosure.

[0201] In some examples, before the UE obtains and transmits the radio frequency signal based on the complex value information associated with the time-frequency resources, the process also includes: the UE generates a signal sequence for the pilot based on the configuration information; the UE maps the signal sequence for the pilot to the time-frequency resource unit related to the pilot indicated in the time-frequency resource information. The pilot may include a reference signal (DM-RS) for demodulation. For example, the UE first generates a signal sequence for the pilot based on the configuration information, multiplies each complex symbol of the signal sequence for the pilot by an amplitude scaling factor, and maps them sequentially to the resource units reserved for the pilot. Multiplying each complex symbol by the amplitude scaling factor is to ensure that each symbol meets the specified transmit power. Taking the time-frequency resource unit represented by (k, l), where k represents the subcarrier number and l represents the time domain symbol number, as an example, when the resources reserved for the pilot are distributed in the time domain, the UE generates a signal sequence z(m) for the pilot, and maps the complex symbols in the signal sequence z(m) for the pilot starting from z(0) to the resource unit reserved for the pilot in ascending order of sequence number l=0, 2, 4,...

[0202] Figure 18 The present invention shows a schematic diagram of a user in a communication system according to the present disclosure obtaining complex value information including pilot signals associated with time-frequency resources according to a time division multiplexing method.

[0203] In some examples, before the UE obtains and transmits the radio frequency signal based on the complex-valued information associated with the time-frequency resources, the method further includes: the UE maps the complex-valued information and the pilot signal sequence to the time-frequency resource unit indicated by the configuration according to the time division multiplexing method to obtain the complex-valued information including the pilot associated with the time-frequency resources. The complex-valued information can be obtained by a method that does not perform the channel coding and modulation steps based on the uplink control information. The method that does not perform the channel coding and modulation steps may include a method for obtaining the complex-valued information based on a neural network model. The time division multiplexing method may include a method for mapping the complex-valued information and the pilot signal sequence on different time domain symbols. For example, through the time division multiplexing method, the complex-valued information and the signal sequence are mapped to time-frequency resource units with the same subcarrier sequence number and different time domain symbol sequence numbers. Taking the time-frequency resource unit represented by (k, l), where k represents the subcarrier number and l represents the time domain symbol number, the UE maps the complex symbols in the signal sequence z(m) used for the pilot to the time domain symbol number l∈L in the order of frequency domain first and then time domain. z On the resource unit, the UE maps the complex symbols in the complex information to the time domain symbols in the order of frequency domain first and then time domain. f On the resource unit. Among them, L z represents the set of time domain symbol numbers used for pilot, Lf represents a set of time domain symbol numbers for complex information, and satisfies L z ∪L f Includes all time domain symbol numbers assigned to uplink control information related to complex information. For example, L z =[0,2],L f =[1,3], the UE maps the pilot signal sequence z(m) on the time-frequency resource unit (k,l) with the time domain symbol number l=0,2, and the UE maps the complex information with the time domain symbol number l=1,3.

[0204] Figure 19 The present invention shows a schematic diagram of a user in a communication system according to the present disclosure obtaining complex value information including pilot signals associated with time-frequency resources according to a frequency division multiplexing method.

[0205] In some examples, before the UE obtains and transmits the radio frequency signal based on the complex-valued information associated with the time-frequency resources, the process also includes: the UE maps the complex-valued information and the pilot signal sequence to the time-frequency resource unit indicated by the configuration according to the frequency division multiplexing method to obtain the complex-valued information including the pilot associated with the time-frequency resources. The complex-valued information can be obtained by a method that does not perform the channel coding and modulation steps based on the uplink control information. The method that does not perform the channel coding and modulation steps may include a method for obtaining the complex-valued information based on a neural network model. The frequency division multiplexing method may include a method for mapping the complex-valued information and the pilot signal sequence onto different frequency domain resources. For example, through the frequency division multiplexing method, the complex-valued information and the signal sequence are mapped to time-frequency resource units with different subcarrier numbers and the same time domain symbol number. Taking the time-frequency resource unit represented by (k, l), where k represents the subcarrier number and l represents the time domain symbol number, the UE maps the complex symbols in the signal sequence z(m) used for the pilot to the frequency domain subcarrier number k∈k in the order of frequency domain first and then time domain. z On the resource unit, the UE maps the complex symbols in the complex information to the time domain symbols in the order of frequency domain first and then time domain. f On the resource unit. Among them, K z represents the set of subcarrier numbers used for pilot, L f represents a set of subcarrier numbers used for complex information, and satisfies K z ∪K f Includes all subcarrier numbers allocated to uplink control information related to complex information.

[0206] In some examples, the method in which the UE maps the complex information and the pilot signal sequence to the time-frequency resource unit indicated by the configuration according to the frequency division multiplexing method also includes that the UE maps the pilot signal sequence to the time-frequency resource unit in a comb-like form. The mapping in the comb-like form indicates that there are more than one time-frequency resource units of the same number associated with the complex information between the time-frequency resource units mapped by any two consecutive pilot symbols in the frequency domain. Taking the example that there are 3 time-frequency resource units associated with the complex information between the time-frequency resource units mapped by any two consecutive pilot symbols in the frequency domain, the UE generates a pilot sequence z for each time domain symbol l l (m), and sequentially replace the signal sequence z used for pilot in the order of sequence number k=3m+1 l The complex symbol in (m) starts from z l (0) Start mapping to the resource unit reserved for the pilot.

[0207] In some examples, the method for the UE to obtain and transmit a radio frequency signal based on the complex value information associated with the time-frequency resources includes: the UE first obtains a time-domain continuous orthogonal frequency division multiplexing (OFDM) baseband signal based on the complex value information associated with the time-frequency resources; then, the UE obtains a radio frequency signal by modulating and upconverting the time-domain baseband signal to a carrier frequency; finally, the UE transmits the radio frequency signal on the corresponding antenna port.

[0208] In some examples, the method for the UE to obtain and transmit a radio frequency signal based on the complex value information associated with the time-frequency resources also includes: the UE maps the complex value information and the pilot signal sequence to the time-frequency resource unit indicated by the configuration according to a time division multiplexing method to obtain the complex value information including the pilot associated with the time-frequency resources; the UE obtains the time domain continuous orthogonal frequency division multiplexing baseband signals corresponding to the complex value information and the pilot information according to the complex value information associated with the time-frequency resource unit and the pilot information associated with the time-frequency resource unit, and merges the baseband signals in the time domain; thereafter, the UE obtains the radio frequency signal by modulating and up-converting the time domain baseband signal to the carrier frequency; finally, the UE transmits the radio frequency signal on the corresponding antenna port.

[0209] Figure 20 A schematic diagram showing an example of a method for users to report CSI in a communication system according to the present disclosure is shown.

[0210] In some examples, the CSI reported by the UE may include one of a vector, a matrix, or a tensor including an estimated channel impulse response in at least one dimension. For example, the CSI is a tensor including channel impulse responses in four dimensions: time domain, frequency domain, UE's spatial domain, and base station's spatial domain. The granularity units of the CSI in the time domain, frequency domain, UE's spatial domain, and base station's spatial domain are symbol, subcarrier, UE's antenna port, and base station's antenna port, respectively. The number of samples of the channel impulse response in the time domain, frequency domain, UE's spatial domain, and base station's spatial domain are T, F, RX, and TX, respectively. Then, the CSI is a tensor with a dimension of [T, F, RX, TX].

[0211] The UE determines, based on the indication of the first configuration information, one of multiple methods for obtaining complex-valued information representing the CSI, and determines, based on the indication of the second configuration information, a scheme for executing the method. For example, the UE determines, based on the method identifier in the first configuration information, a method for obtaining complex-valued information based on the CSI, wherein the CSI does not undergo bit sequence generation and channel coding, and the UE determines, based on the scheme identifier in the second configuration information, a neural network model for obtaining the complex information.

[0212] The UE obtains complex-valued information based on the CSI and the determined neural network model. For example, the method of obtaining complex-valued information using a neural network model may include: obtaining a CSI input that is consistent with the data format and structure of the neural network input layer by transforming the data format and structure of the CSI; inputting the transformed CSI into the neural network to obtain a neural network output related to the complex-valued information; and obtaining complex-valued information based on the output of the neural network. For example, the CSI is a tensor of dimension [T, F, RX, TX]. First, the data format of the CSI is converted from a complex number to a floating point number to form a tensor of dimension [T, F, RX, TX, RI], where RI=2 carries a floating point number representing the real and imaginary parts of the complex-valued symbol; then, the tensor of dimension [T, F, RX, TX, RI] is dimensionally transformed to reduce the dimension to a vector including T*F*RX*TX*RI elements; then, the vector is input into the neural network to obtain a vector of length T. out *F out *RI neural network output vector, where RI = 2; the neural network output vector is dimensionally transformed and upgraded to a vector with a dimension of [T out ,F out ,RI]; Finally, the two floating-point numbers in each group of data of the RI dimension in the above tensor are used as the real part and the imaginary part to form a complex number, and a tensor of dimension [T out ,F out ], the tensor is a tensor containing T out *F outThe complex-valued information of the complex-valued modulation symbols.

[0213] The UE multiplexes the uplink channel or signal to carry complex information of second uplink control information, wherein the second uplink control information may include at least one of the following information: channel state information (CSI), scheduling request (SR), and acknowledgement (ACK). For example, the second information is a scheduling request.

[0214] The UE determines, based on the indication of the first configuration information, one of multiple methods for obtaining complex-valued information representing the second uplink control information, and determines, based on the indication of the second configuration information, a scheme for executing the method. For example, based on the method identifier in the first configuration information, the UE determines a method for obtaining the complex-valued information based on a bit sequence, where the bit sequence is obtained by performing bit sequence generation and channel coding on the second uplink control information. The UE determines, based on the scheme identifier in the second configuration information, parameters required for executing the method, including a modulation order and a target coding rate.

[0215] The UE obtains complex information carrying the second uplink control information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on the second uplink control information. The UE generates the uplink control information bit sequence. For example, the UE obtains a bit sequence carrying the second uplink control information; the UE cuts the bit sequence and adds a parity check at the end of each bit sequence block to obtain a bit sequence with a CRC attached; the UE performs bit interleaving on the bit sequence with the CRC attached to obtain an interleaved bit sequence; the UE performs channel coding on the interleaved bit sequence according to a polarization sequence to obtain a channel-coded bit sequence; the UE modulates the channel-coded bit sequence according to the configured modulation order to obtain second complex information including at least one complex-valued modulation symbol.

[0216] The UE multiplexes the uplink channel or signal to carry complex information of the second uplink control information. For example, the UE concatenates the last complex symbol of the second complex information with the first complex symbol of the first complex information to obtain multiplexed complex information of length N.

[0217] The UE obtains the matched complex value information based on the multiplexed complex value information and the number of time-frequency resource units used to report the uplink control information, wherein the number of complex symbols contained in the matched complex value information is the same as the number of time-frequency resource units used to report the uplink control information. For example, the UE obtains the number of resource units E based on the configured time-frequency resource information and DM-RS information, and the multiplexed complex information is y0, y1, y2, ..., y N-1 , which includes N complex-valued modulation symbols, where N>E, the UE determines to perform rate matching by puncturing. The UE first converts the complex information y0,y1,y2,…,yN-1 Store it in the circular buffer, and then read it starting from the NEth complex-valued modulation symbol in the circular buffer to obtain the matched complex-valued information e0, e1, e2, ..., e E-1 For example, the method is to use k=0 to E-1, and e k =y k+N-E .

[0218] The UE interleaves the complex symbols in the matched complex value information according to the configuration information to obtain the interleaved complex value information. For example, the matched complex value information of the UE includes a total of E complex value modulation symbols e0, e1, e2, ..., e E-1 , UE determines the interleaving order as Q through configuration information m , UE according to the interleaving order Q m Obtain the interleaved complex information, which includes complex-valued modulation symbols f0, f1, f2, ..., m E-1 .

[0219] The UE maps the interleaved complex information and the pilot signal sequence to the time-frequency resource unit indicated by the configuration according to the frequency division multiplexing method, and obtains the complex information including the pilot associated with the time-frequency resource. For example, the time-frequency resource unit is represented by (k, l), where k represents the subcarrier number and l represents the time domain symbol number. The UE maps the complex symbols in the signal sequence z(m) used for the pilot to the frequency domain subcarrier number k∈K in the order of frequency domain first and then time domain. z On the resource unit, the UE will interleave the complex information f0,f2,f2,…,f E-1 The complex symbols in the frequency domain are mapped to the time domain symbols in the order of first frequency domain and then time domain, and the serial number is k∈K f On the resource unit. Among them, K z represents the set of subcarrier numbers used for pilot, L f represents a set of subcarrier numbers used for complex information, and satisfies K z ∪K f Includes all subcarrier numbers allocated to uplink control information related to complex information.

[0220] The UE obtains a time-domain continuous orthogonal frequency division multiplexing baseband signal according to the complex value information associated with the time-frequency resource.

[0221] The UE obtains a radio frequency signal by modulating the time domain baseband signal and up-converting the signal to a carrier frequency.

[0222] The UE transmits the radio frequency signal on the corresponding antenna port.

[0223] Example 2

[0224] In some examples, a UE determines, based on an indication in configuration information, one of multiple methods for obtaining complex-valued information representing CSI and a scheme for executing the method. For example, the UE has information about multiple methods, and each method of the UE includes multiple schemes, and each method of the UE includes an identification table of the multiple schemes. The identification table includes at least one identifier for each scheme. Based on the identification table, the UE determines the selected scheme by matching the scheme identifier in the indication with the corresponding identifier in the UE identification table. It should be understood that, for ease of description, the term "identification table" is used with respect to the schemes included in the method, and is intended to refer to a table or parameter set used to represent at least one scheme corresponding to the method. Through this "identification table," each scheme or at least one parameter set corresponding to the method can be identified or determined. The term "identification table" is not intended to be limiting, and other terms with the same or similar meanings may be used, such as "index table," "parameter set," "scheme list," and so on. Similarly, it is understood that "identification" may be replaced by other terms, such as "index," "name," "number," "related information," and so on.

[0225] In some examples, the "UE determines one of the multiple methods for obtaining complex value information representing CSI and a scheme for executing the method according to the indication of the configuration information" may also include: the UE determines one of the multiple methods for obtaining complex value information representing CSI and an identification table of multiple schemes according to the indication of the first configuration information; the UE determines the scheme for executing the method according to the identification table of multiple schemes according to the indication of the second configuration information.

[0226] Figure 21 A schematic diagram shows a method for a UE in a communication system according to the present disclosure to obtain complex-valued information related to uplink control information.

[0227] In some examples, the "UE determines one of the multiple methods for obtaining complex-valued information representing CSI according to the indication of the configuration information, and a scheme for executing the method" may also include: the UE determines an identification table including multiple schemes according to the indication of the first configuration information, and implicitly determines one of the multiple methods for obtaining complex-valued information representing CSI according to the table; the UE determines the scheme for executing the method according to the identification table of multiple schemes according to the indication of the second configuration information.

[0228] In some examples, the identifiers of the multiple schemes contained in the identification table may include: index, name, number, modulation order, target code rate, spectrum efficiency, signal-to-noise ratio (SNR), signal-to-interference-plus-noise ratio (SINR), SINR interval, SINR upper bound, SINR lower bound, output type, output format, output length, etc. The SINR interval includes an SINR upper bound and a SINR lower bound, which respectively represent the maximum and minimum SINRs in which the corresponding scheme can work. The output length represents the number of complex-valued symbols in the complex-valued information related to the uplink control information obtained by the corresponding method. The output type represents the data type of each element in the information related to the uplink control information obtained by the corresponding scheme, and the data type may include one of a complex number, a floating point number, and an integer.

[0229] Taking the example of a UE obtaining complex-valued information based on CSI without bit sequence generation and channel coding through a neural network model, first, the UE determines an identification table including multiple schemes based on the first configuration information, and the table includes the identifications of N neural network models, and the UE also includes the structure and weights of the N neural network models corresponding to the table. The identification in the table includes the model number, SINR lower bound, and output length. Secondly, the UE determines the neural network model based on the identification in the second configuration information, and the identification may include at least one of the model number, SINR lower bound, and output length. Finally, the UE obtains complex-valued information based on the CSI without bit sequence generation and channel coding according to the neural network model determined by the identification.

[0230] Figure 22 A schematic diagram of an identification table including N neural network models in a communication system according to the present disclosure is shown.

[0231] In some examples, the "UE determines one of the multiple methods for obtaining complex value information representing CSI according to the indication of the configuration information, and the scheme for executing the method" may also include: the UE determines at least two identification tables including multiple schemes according to the indication of the first configuration information; the UE determines a method and a corresponding identification table according to the indication of the second configuration information, and then determines the scheme for executing the corresponding method according to the table. For example, the UE determines two information tables of complex value acquisition methods according to the first configuration information, and the first identification table implicitly determines the method by which the UE obtains the complex value information based on the CSI without bit sequence generation and channel coding through a neural network model, and the second identification table implicitly determines the method by which the UE obtains the complex value information based on the bit sequence obtained in the process of bit sequence generation and channel coding of the CSI. The identifiers contained in the first identification table include the model number, SINR lower bound, and output length, and the identifiers in the second identification table include the scheme number, modulation order, target coding rate, and spectrum efficiency.

[0232] Figure 23 A schematic diagram shows an information table in which a UE in a communication system according to the present disclosure determines two complex value acquisition methods according to first configuration information.

[0233] In some examples, the "UE determines, based on the indication of the configuration information, one of multiple methods for obtaining complex-valued information representing CSI, and a scheme for executing the method" may also include: the UE determines, based on the indication of the first configuration information, an information table including no less than two methods, and including multiple schemes for each method; and the UE determines, based on the indication of the second configuration information, a method and a scheme for executing the corresponding method. The identification table may include the number of each scheme. For example, the identification table includes two methods, the first method includes N1 schemes, and the second method includes N2 schemes. The table includes a total of N1+N2 schemes and the scheme number corresponding to each scheme.

[0234] For example, the UE determines an information table including two methods based on the indication of the first configuration information. The first method may include a method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on the CSI; the second method may include a method for obtaining complex-valued information based on the CSI, wherein the CSI does not perform bit sequence generation and channel coding. The identifiers in the identification table include a scheme number, SINR lower bound, output length, modulation order, target coding rate, and spectrum efficiency. The UE determines a method and a scheme for executing the corresponding method based on the scheme number indicated by the second configuration information.

[0235] Figure 24A schematic diagram is shown in which a UE in a communication system according to the present disclosure determines an information table including two methods according to an instruction of first configuration information.

[0236] In some examples, the "UE determines one of the multiple methods for obtaining complex-value information representing CSI according to the indication of the configuration information, and a scheme for executing the method" may also include: the UE determines an identification table including multiple schemes according to the table name indicated by the RRC, and implicitly determines one of the multiple methods for obtaining complex-value information representing the CSI according to the table; the UE determines the scheme for executing the method according to the identification table according to the scheme index indicated by the DCI.

[0237] In some examples, the UE may also determine, based on the indication of the configuration information, one of multiple methods for obtaining complex-valued information representing the first uplink control information and a scheme for executing the method, and simultaneously determine one of multiple methods for obtaining complex-valued information representing the second uplink control information and a scheme for executing the method. The first and second uplink control information may include CSI, a scheduling request, an acknowledgement, and the like. For example, based on the indication of the first configuration information, the UE determines a first identification table including multiple schemes for obtaining complex-valued information representing the CSI and a second identification table including multiple schemes for obtaining complex-valued information representing the second uplink control information; and based on the indication of the second configuration information, the UE determines, based on the identification tables of the multiple schemes, a scheme for executing the method. The UE determining, based on the indication of the second configuration information, the scheme for executing the method based on the identification tables of the multiple schemes further includes: the UE first determines, based on the indication of the second configuration information, one of the first and second identification tables; and then, based on the indication of the second configuration information and the determined identification table, the UE determines, based on the indication of the second configuration information, a scheme for executing the method. Among them, the UE determines the scheme for executing the method according to the indication of the second configuration information and the identification table of multiple schemes, which can also include: the UE first determines one of the first and second identification tables according to the content of the uplink control information; the UE then determines the scheme for executing the method according to the indication of the second configuration information and the determined identification table.

[0238] For example, the UE determines, based on the first configuration information, a first identification table for obtaining complex-valued information of CSI (first uplink control information). The method corresponding to the first identification table is a method based on a neural network model without performing bit sequence generation and channel coding. The UE also determines a second identification table for obtaining complex-valued information of a scheduling request (second uplink control information). The method corresponding to the second identification table is a method for obtaining complex-valued information based on a bit sequence obtained by performing bit sequence generation and channel coding on the CSI. The UE selects the first or second identification table based on the content of the obtained uplink control information, and then determines a scheme for executing the method based on the indication of the second configuration information and the determined identification table.

[0239] In some examples, the UE may also determine, based on the indication of the configuration information, one of multiple methods for obtaining complex-valued information and a scheme for executing the method. The method determined by the UE and the scheme for executing the method may be used to obtain complex-valued information of the first uplink control information, and may also be used to obtain complex-valued information representing the second uplink control information. The uplink control information may include one of CSI, a scheduling request, and a confirmer. For example, based on the indication of the configuration information, the UE may determine an information table including at least one method and including multiple schemes for each method; and based on the indication of the second configuration information, the UE may determine a method and a scheme for executing the corresponding method.

[0240] For example, the UE determines an information table including two methods according to the indication of the configuration information. The first method may include a method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a process of bit sequence generation and channel coding on the CSI; the second method may include a method for obtaining complex-valued information based on the CSI, wherein the CSI does not perform bit sequence generation and channel coding. The identifiers in the identification table include a scheme number, SINR lower bound, output length, modulation order, target coding rate, and spectrum efficiency. The UE determines a method and a scheme for executing the corresponding method according to the scheme number indicated by the second configuration information. When the uplink control information is CSI, the UE determines a scheme in the first method according to the second configuration information; when the uplink control information is a scheduling request or confirmer, the UE determines a scheme in the second method according to the second configuration information.

[0241] In some examples, before the UE determines one of the multiple methods for obtaining complex-valued information representing channel state information (CSI), the method further includes: the UE reports capability information, wherein the capability information includes the UE's ability to obtain complex-valued information based on CSI without bit sequence generation and channel coding. The base station can determine configuration information based on the UE capability information, wherein the configuration information includes a complex-valued information method for instructing the UE to represent channel state information (CSI). The benefit of the UE reporting capability information is that the base station can maximize the UE's capabilities based on the capability information by appropriately configuring the complex-valued acquisition method, thereby increasing the efficiency and stability of the UE reporting uplink control information.

[0242] In some examples, the capability information reported by the UE to the base station includes an identifier corresponding to the capability. The capability includes at least one of the following:

[0243] A first capability indicates that the UE has the capability of a first method for obtaining complex-valued information based on a bit sequence, where the bit sequence is obtained by performing bit sequence generation and channel coding on the CSI;

[0244] The second capability indicates that the UE has the capability of obtaining complex-valued information based on the CSI using the second method, wherein the CSI does not undergo bit sequence generation and channel coding.

[0245] The third capability indicates that the UE has both the ability to use the first method to obtain complex-valued information based on a bit sequence, where the bit sequence is obtained by performing bit sequence generation and channel coding on the CSI, and the ability to use the second method to obtain complex-valued information based on the CSI, where the CSI does not perform bit sequence generation and channel coding.

[0246] The identifier corresponding to the capability may include an index and a name of the capability.

[0247] Example 3

[0248] In some examples, a method for a UE to obtain complex-valued information based on CSI without bit sequence generation and channel coding may include: obtaining the complex-valued information based on the CSI and a neural network model. For example, the method for obtaining the complex-valued information using the neural network model may include: transforming the data format and structure of the CSI to obtain a CSI input consistent with the data format and structure of a neural network input layer; inputting the transformed CSI into the neural network to obtain a neural network output related to the complex-valued information; and obtaining the complex-valued information based on the output of the neural network.

[0249] In some examples, the CSI data format in the "transforming the CSI data format and structure to obtain a CSI input consistent with the data format and structure of the neural network input layer" may include one of a complex number, a floating point number, and an integer. The data format of the neural network model may also include one of a complex number, a floating point number, and an integer. The method for transforming the CSI data format may include at least one of the following: combining floating point numbers into complex numbers; separating the real and imaginary parts of the complex number to obtain two floating point numbers; quantizing the floating point number into an integer; and dequantizing the integer into a floating point number.

[0250] For example, when the data format of the elements in the CSI is consistent with the data format of the neural network model, the neural network model includes an input length equal to the number of elements in the CSI. When the data format of the elements in the CSI is complex and the data format output by the neural network model is not complex, the neural network model includes an input length twice the number of elements in the CSI. When the data format of the elements in the CSI is not complex and the data format output by the neural network model is complex, the neural network model includes an input length half the number of elements in the CSI.

[0251] In some examples, the data format of the neural network output may also include one of a complex number, a floating point number, and an integer. The method of obtaining complex-valued information based on the output of the neural network may include one of the following: when the output of the neural network is a complex number, the method of obtaining the complex-valued information may include performing a structural conversion on the output of the neural network according to the structure of the complex-valued information; when the output of the neural network is a floating point number, the method of obtaining the complex-valued information may include first forming the floating point number into a complex number, and then performing a structural conversion on the obtained floating point number according to the structure of the complex-valued information; when the output of the neural network is an integer number, the method of obtaining the complex-valued information may include first mapping the obtained integer number into a floating point number according to a mapping relationship between integers and floating point numbers, then forming the obtained floating point number into a complex number, and finally performing a structural conversion on the obtained floating point number according to the structure of the complex-valued information.

[0252] In some examples, before inputting the CSI into the neural network, the UE may also pre-process the CSI to obtain the pre-processed CSI. The method for the UE to pre-process the CSI may include performing a signal processing-related transform on the CSI to obtain the CSI after the signal processing-related transform. For example, performing a signal processing-related transform on the CSI may include one of the following methods: Fourier transform, Laplace transform, wavelet transform, Hilbert-Huang transform, bispectral transform, etc. The advantage of performing a signal processing-related transform on the CSI is that the CSI can be transformed into a form that is convenient for subsequent processing or processing by a neural network model. The method for the UE to pre-process the CSI may also include performing redundant correlation processing on the CSI to obtain the CSI after the redundant correlation processing.

[0253] For example, a method for performing redundant correlation processing on CSI may include setting a redundant threshold and setting the elements in the CSI that are smaller than the redundant threshold to zero. The benefit of performing redundant correlation processing on CSI is to reduce the redundant information in the CSI, thereby reducing the total amount of information that needs to be sent. Among them, the process of pre-processing the CSI by the UE may also include normalizing or standardizing the CSI to obtain normalized or standardized uplink control information. For example, normalizing or standardizing the uplink control information may include one of the following methods: maximum and minimum normalization, z-score normalization, log function normalization, inverse tangent function normalization, L2 norm normalization, etc. The benefit of normalizing or standardizing the CSI is to limit the data to a fixed range, thereby eliminating the adverse effects of singular samples on the overall data, thereby reducing the loss of neural network model performance. Taking the original channel matrix estimated by CSI as an example: the UE first performs a Fourier transform on the channel state information to obtain the delay domain channel state information; then, the UE sets a redundancy threshold and sets the symbols in the delay domain channel state information below the redundancy threshold to zero; finally, the UE performs maximum and minimum normalization on the redundantly processed channel state information. In this example, through signal processing and redundancy processing, the UE compresses the redundant information in the time domain of the source sequence, thereby reducing the total amount of information that needs to be sent. In this example, through maximum and minimum normalization processing, the UE limits the compressed data to a fixed range, thereby eliminating the adverse effects of singular samples on the overall data, and thus eliminating the loss of neural network model performance.

[0254] In some examples, the method of "inputting the transformed CSI into a neural network to obtain a neural network output related to complex-valued information" by the UE may also include: combining the configuration information related to the uplink control information with the CSI after structural conversion, and inputting the information into the neural network to obtain a neural network output related to the complex-valued information. The configuration information related to the uplink control information may include at least one of the following: number, SINR, SINR interval, SINR upper bound, SINR lower bound, number of time-frequency resources, and position of time-frequency resources. Combining the configuration information related to the uplink control information with the CSI may include one of the following methods: directly splicing the configuration information related to the uplink control information with the CSI, and using the configuration information related to the uplink control information as the position encoding of the CSI. The advantage of combining the configuration information related to the uplink control information with the CSI is that the complex information output by the neural network model is associated with the uplink control information configuration.

[0255] Example 4

[0256] In some examples, the method for a UE to obtain matched complex-valued information may further include: the UE matching the outputs with resource units in order of priority based on the priorities of the neural network model outputs to obtain prioritized matched complex-valued symbol information. The priorities of the neural network model outputs include the priority weight of each output neuron in the neural network. The method for matching the outputs with the number of resource units in order of priority may include: the UE writing the complex-valued information into a circular buffer based on the priority weights of the output neurons associated with the complex symbols in the complex-valued information, in descending order or in ascending order, wherein the length of the circular buffer is equal to the number of complex symbols in the complex-valued information; and the UE reading the complex symbols from the circular buffer according to a configured method to obtain the matched complex-valued information. The configured method may include: the UE extracting the complex symbols from the circular buffer according to a configured starting position and / or reading order to obtain the matched complex-valued information. The number of complex-valued symbols contained in the matched complex-valued information is the same as the number of time-frequency resource units associated with the reported uplink control information. The starting position, reading order, and number of resource units may be obtained based on configuration information.

[0257] In some examples, when the UE extracts complex symbols from a circular buffer according to a configured starting position and / or reading order, the method for determining the starting position and reading order may include the UE determining the starting position and reading order of the complex symbols extracted from the circular buffer by comparing the length of the complex-valued information with the number of resource units and based on the priority of the complex symbols in the complex-valued information. For example, if the length of the complex-valued information is less than the number of resource units and the priority of the complex symbols in the circular buffer decreases with the position, the UE extracts the complex symbols by extracting the complex symbols from the circular buffer in forward order by a repetitive method; if the length of the complex-valued information is less than the number of resource units and the priority of the complex symbols in the circular buffer increases with the position, the UE extracts the complex symbols by extracting the complex symbols from the circular buffer in reverse order by a repetitive method; if the length of the complex-valued information is greater than the number of resource units and the priority of the complex symbols in the circular buffer decreases with the position, the UE extracts the complex symbols by extracting the complex symbols from the circular buffer by a shortening method; if the length of the complex-valued information is greater than the length of the number of resource units and the priority of the complex symbols in the circular buffer increases with the position, the UE extracts the complex symbols from the circular buffer by a puncturing method.

[0258] For example, the UE determines the neural network model z based on the relevant configuration information, and obtains complex-valued information of length N using the neural network model z and the uplink control information. The complex symbols in the obtained complex-valued information are prioritized from largest to smallest. Simultaneously, the UE determines the number of resource units E based on the relevant configuration information, with N > E. Based on this information, the UE determines to implement rate matching by extracting complex symbols from the circular buffer using a shortened method.

[0259] In some examples, the method for the UE to obtain the matched complex-valued information may further include: the UE matches the output of the time-frequency resource units associated with the reported uplink control information in order of the index output by the neural network model, thereby obtaining the matched complex-valued information associated with the index. The method for sequentially matching the output with the number of resource units in order of the index may include: the UE writes the complex-valued information into a circular buffer according to the index of the output neuron associated with the complex symbols in the complex-valued information, in the order of the index from large to small or from small to large, wherein the length of the circular buffer is equal to the number of complex symbols in the complex-valued information; and the UE reads the complex symbols from the circular buffer according to the configured method to obtain the matched complex-valued information.

[0260] In some examples, before the UE maps the complex-valued information to an uplink channel or signal, it also includes: the UE multiplexing the uplink channel or signal to carry second uplink control information with different priorities to obtain the multiplexed complex-valued information; the UE determines the starting position and reading order of extracting the complex symbols from the circular buffer by comparing the length of the complex-valued information with the number of resource units and according to the priority of the complex symbols in the complex-valued information, and obtains the multiplexed and matched complex-valued information through the circular buffer. Wherein, the multiplexing method may include obtaining the multiplexed complex-valued information based on the complex-valued information related to the first uplink control information and the complex-valued information related to the second uplink control information. Wherein, the method for obtaining the multiplexed complex-valued information based on the first complex-valued information and the second complex-valued information may include: the UE determines the priority relationship between the first and second uplink control information, and merges the first complex-valued information with the second complex-valued information according to the priority relationship. For example, if the priority of the complex symbols of the first complex-valued information decreases with position, the second complex-valued information sequence with a higher priority is concatenated with the starting complex symbol of the first complex-valued information, and the second complex-valued information sequence with a lower priority is concatenated with the last complex symbol of the first complex-valued information. If the priority of the complex symbols in the first complex-valued information increases with position, the second complex-valued information sequence with a higher priority is concatenated with the last complex symbol of the first complex-valued information, and the second complex-valued information sequence with a lower priority is concatenated with the starting complex symbol of the first complex-valued information. The advantage of performing the merging of the first complex-valued information and the second complex-valued information in this way is that it avoids the loss of the high-priority information sequence during rate matching, thereby ensuring the validity of the information obtained by the receiving end.

[0261] Taking the example of a first complex-valued information including CSI and a second complex-valued information including a scheduling request, the UE obtains the first complex-valued information based on the CSI without bit sequence generation and channel coding, and obtains the second complex-valued information based on the bit sequence of the second uplink control information obtained through the bit sequence generation and channel coding process. The UE then determines that the priority of the complex symbols in the first complex-valued information decreases with position, and determines that the priority of the second complex-valued information is higher than that of the first complex-valued information. Therefore, the UE concatenates the second complex-valued information with the starting complex symbol of the first complex-valued information to obtain the multiplexed complex-valued information. The length of the multiplexed complex-valued information is N, and the UE determines through configuration that the number of resource units is E, where N>E. The priorities of the complex symbols in the multiplexed complex-valued information are arranged from largest to smallest, and the UE determines to perform rate matching through a shortening method. For example, the UE stores the complex symbols in the multiplexed complex-valued information sequentially in a circular buffer, and extracts complex symbols sequentially starting from the first complex symbol in the circular buffer until a number of complex symbols equal to the number of resource units is obtained.

[0262] In some examples, the neural network model with priority output used by the UE to obtain CSI complex-valued information may include an encoder part in an autoencoder. The autoencoder includes three parts: an encoder, a channel network, and a decoder, and each part of the autoencoder may include a neural network with no less than one layer. The input of the encoder is CSI, and the output is complex-valued information with priority. The method for obtaining the encoder is to modify the output of the encoder part according to the priority during the training process of the autoencoder to obtain complex-valued information with some complex-valued symbols lost. The complex-valued information with some symbols lost may include setting the symbol at the lost position to 0 or replacing it with a random number.

[0263] Among them, the method for determining the position and number of the lost partial complex-valued symbols may include matching a loss probability for each output neuron of the encoder part, and determining whether the complex-valued symbol corresponding to the output is lost for each output according to the matched loss probability. For example, the method for determining whether there is a loss is to generate a random number for each output neuron, and determine whether there is a loss by comparing the random number and the loss probability. For example, the loss probability of each output neuron is arranged from large to small according to the priority of the output neuron. During the training process, the method for implementing the above-mentioned determination of the position and number of the lost partial complex-valued symbols and executing it may include adding a drop-out function after each output neuron of the encoder part, and setting the weight of the drop-out function according to the loss probability.

[0264] Among them, the method for determining the position and number of lost complex-valued symbols may also include arranging the complex-valued symbols according to the priority of the output neurons, generating a random number of complex-valued symbol losses, and discarding the arranged complex-valued symbols in order from small to large priority to generate the said number.

[0265] Figure 25 A schematic diagram illustrating a PUCCH procedure (which may be referred to as an EJ-PUCCH procedure) and a first PUCCH procedure according to at least one embodiment of the present disclosure is shown.

[0266] The PUCCH procedure in some examples of the present disclosure may also be referred to as EJ-PUCCH. Figure 25As shown, the EJ-PUCCH process does not include independent CSI compression, channel coding and modulation modules. The JSCM module is included and is responsible for generating complex-valued modulation symbols. In the EJ-PUCCH process, symbol-level multiplexing, rate matching and interleaving are included after the JSCM module. A direct way to enable EJ-PUCCH is to retain the mechanisms of these three modules in the traditional PUCCH and simply adjust the target from bits to symbols. However, reusing existing principles may not fully realize the potential of JSCM, and more customized EJ-PUCCH designs are expected to optimize JSCM. In addition, a reverse procedure is required on the receiver side, and corresponding UE capability reporting, model scheduling, model monitoring and a strong backoff mechanism can be included to ensure system reliability.

[0267] Figure 26 A method for implementing JSCM model training according to at least one embodiment of the present disclosure is shown.

[0268] In some examples, the JSCM model can be used to directly generate complex-valued modulation symbols, thus omitting the quantizer and dequantizer. The CSI is fed forward to the UE's encoder, which generates a set of floating-point numbers. These floating-point numbers are then combined to form complex-valued modulation symbols. After passing through the channel, the base station receives the complex-valued modulation symbols and decomposes them back into floating-point numbers, which are then passed to the decoder for CSI reconstruction. The JSCM model is implemented using a Transformer network architecture.

[0269] In some examples, three additional functions are added between the encoder and decoder during the training phase: a defect simulation module, a noisy channel simulation module, and a prioritized noise simulation module. The defect simulation module multiplexes noise into the encoder output, enabling the AI model to learn to handle channel estimation errors and RF imperfections. This noise is derived from a Richan distribution with a mean of 1 and a variance that is inversely proportional to the SNR. This design aims to simulate channel estimation errors that are negatively correlated with the SNR.

[0270] In some examples, a noisy channel simulation module can be added after the defect simulation module to simulate the uplink channel experienced by the PUCCH. This allows the model to learn how to handle channel distortion caused by additive white Gaussian noise (AWGN). AWGN is considered in this location because fading is eliminated by equalization, and the remaining error is handled by the defect simulation module.

[0271] In some examples, a prioritized noise simulation module can be added after the noisy channel simulation module. This module teaches the model to prioritize outputs by adding customized Gaussian noise to each output neuron of the encoder, weighted according to their likelihood of being punctured or repeated during rate matching. Symbols with a higher probability of being punctured receive a larger weight. During this module's pass, each floating-point number in the collection is supplemented with Gaussian noise that is repeated with its priority. This results in a JSCM model whose output neurons have customized priorities, ensuring that complex-valued modulation symbols that are likely to be punctured carry less important information. Thus, the downlink CSI is fed forward sequentially through the encoder, the imperfect simulation module, the noisy channel simulation module, the prioritized noise simulation module, and the decoder, where the weights are updated in a direction that minimizes the NMSE and GCS losses.

[0272] In some examples, the "defect simulation module" can also be implemented by simulating residual fading and RF defects from channel estimation errors. The residual fading can be obtained by the following formula:

[0273]

[0274] Among them, s PUCCH,RS is the signal before it goes through the defect simulation module, It can be a vector obtained by a sine function, and the step size of the sine function can be related to the number of paths in the channel. CE It can be a constant representing the residual channel fading degree obtained based on the channel estimation capability, n RS It can be obtained from a Gaussian distribution, which represents the noise intensity on the reference signal. The RF defect can be simulated by an RF power amplifier model. For example, the RF power amplifier model can be the Rapp RF amplifier model, which is expressed as:

[0275]

[0276] Among them, G, V SAT , p is the power amplifier parameter, which can be a fixed value or obtained from a random distribution. For example, the parameter can be obtained from a Gaussian distribution, G, V SAT , the means of the Gaussian distributions corresponding to p are 31.6228, 79.339, and 3, and the variances are 1, 2, and 0.2, respectively. Γ FFT () and Γ IFFT () represent fast Fourier transform and inverse fast Fourier transform respectively. It can also be through a signal that has undergone residual fading To obtain, the specific formula is:

[0277]

[0278] In some examples, the noise channel simulation module and the priority noise simulation module can be jointly implemented as follows:

[0279]

[0280] Where α is a configurable constant related to the degree of rate matching, β p Is a PUCCH,RS A vector of multiple probabilities of the same dimension, used to represent s PUCCH,RS The priority weight of the element in s, n is a PUCCH,RS A vector containing noise of the same dimension, where each noise value can be obtained through a Gaussian distribution.

[0281] In some examples, the noise channel simulation module and the priority noise simulation module may also be implemented based on the signal passing through the defect simulation module, and the specific implementation method is as follows:

[0282]

[0283] It is understandable that the above method described in the embodiments of the present disclosure is also applicable to other signal sources, such as image signals, voice signals, video signals, sampled analog signals, image, voice or video signals after source encoding, etc. as signal sources.

[0284] Figure 27 FIG2 shows a schematic diagram of the structure of a user equipment 2500 according to at least one embodiment of the present disclosure. Figure 27 The user equipment 2500 includes a transceiver 2501 and a controller 2502. The transceiver 2501 is configured to transmit data or signals and receive data or signals. The controller 2502 is coupled to the transceiver 2501 and is configured to perform control so that the user equipment 2500 performs the method according to the embodiment of the present disclosure. In one implementation, the user equipment 2500 may further include a memory (not shown) having computer-executable instructions stored therein. When the instructions are executed by the controller 2502, the user equipment 2500 may perform at least one method corresponding to the above-mentioned embodiments of the present disclosure.

[0285] Figure 28 FIG2 shows a schematic structural diagram of a base station 2600 according to at least one embodiment of the present disclosure. Figure 28The base station 2600 includes a transceiver 2601 and a controller 2602. The transceiver 2601 is configured to transmit data or signals and receive data or signals. The controller 2602 is coupled to the transceiver 2601 and is configured to perform control so that the base station 2600 performs the method according to the embodiment of the present disclosure. In one implementation, the base station 2600 may further include a memory (not shown) having computer-executable instructions stored therein. When the instructions are executed by the controller 2602, the base station 2600 may perform at least one method corresponding to the above-mentioned embodiments of the present disclosure.

[0286] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0287] Those skilled in the art will appreciate that the present invention includes devices for performing one or more of the operations described herein. These devices may be specially designed and manufactured for the desired purpose, or they may include known devices found in general-purpose computers. These devices have computer programs stored therein, which are selectively activated or reconfigured. Such computer programs may be stored on a device (e.g., a computer) readable medium or on any type of medium suitable for storing electronic instructions and coupled to a bus, including but not limited to any type of disk (including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks), ROM (Read-Only Memory), RAM (Random Access Memory), EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory, magnetic cards, or optical cards. In other words, a readable medium includes any medium that can be used by a device (e.g., a computer) to store or transmit information in a form that can be read.

[0288] Those skilled in the art will appreciate that each block in these structural diagrams and / or block diagrams and / or flow charts, as well as combinations of blocks in these structural diagrams and / or block diagrams and / or flow charts, can be implemented using computer program instructions. Those skilled in the art will appreciate that these computer program instructions can be provided to a general-purpose computer, a specialized computer, or a processor of other programmable data processing methods for implementation, thereby executing the schemes specified in the blocks or multiple blocks of the structural diagrams and / or block diagrams and / or flow charts disclosed in the present invention through the processor of the computer or other programmable data processing method.

[0289] Those skilled in the art will appreciate that the steps, measures, and schemes in the various operations, methods, and processes discussed in the present invention may be interchanged, modified, combined, or deleted. Furthermore, other steps, measures, and schemes in the various operations, methods, and processes discussed in the present invention may also be interchanged, modified, rearranged, decomposed, combined, or deleted. Furthermore, steps, measures, and schemes in the prior art that are similar to those disclosed in the present invention may also be interchanged, modified, rearranged, decomposed, combined, or deleted.

[0290] The above descriptions are only partial embodiments of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and modifications without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method performed by a user equipment (UE) in a communication system, comprising: Determining a first processing method among a plurality of methods for obtaining complex-valued information associated with channel state information (CSI), wherein the plurality of methods include: A first method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on CSI, and A second method for obtaining complex-valued information based on CSI, wherein the CSI is not subjected to bit sequence generation and channel coding; Processing the CSI based on the first processing method to obtain complex-valued information associated with the CSI; An uplink channel or signal is sent based on the complex-valued information.

2. The method according to claim 1, wherein The sending of an uplink channel or signal based on the complex-valued information includes: The complex-valued information is mapped onto an uplink channel or signal and sent.

3. The method according to claim 1, further comprising: Receive high-layer signaling, where the high-layer signaling includes first information, where the first information includes at least one of the following: Information about the first treatment method; information of at least one parameter set associated with the second method; The method further includes: receiving physical layer signaling, where the physical layer signaling includes second information of a parameter set in the at least one parameter set.

4. The method according to claim 3, wherein: The first information includes at least one of an index, a name, and a table of at least one parameter set associated with the second method.

5. The method according to claim 3, wherein: The second information includes information associated with at least one of an index, name, modulation order, signal-to-noise ratio (SNR), signal-to-interference-plus-noise ratio (SINR), SINR interval, SINR upper bound, SINR lower bound, output type, output length, and output format corresponding to the parameter set.

6. The method according to claim 3, wherein: The second method includes a method based on a neural network model, and the method further includes: determining a parameter set of the neural network based on the second information, wherein the parameter set includes at least one of the following: the type of model, the structure of the model, the number of layers of the model, the number of neurons of the model, the activation function of the model, an input length equal to the number of elements in the channel state information, the output format of the model, and the weight of the model.

7. The method according to claim 6, wherein: The output format of the model also includes at least one of the following: output data format, output data structure, number of output elements, priority of output elements, and index of output elements.

8. The method according to claim 1, wherein The method further includes: before the UE maps the complex-valued information associated with the CSI onto an uplink channel or signal, the UE multiplexing the uplink channel or signal to carry the complex-valued information associated with the second uplink control information; The multiplexing method includes combining the complex value information associated with the CSI and the complex value information associated with the second uplink control information and mapping them onto an uplink channel or signal; The complex value information associated with the second uplink control information is obtained by one of the following methods: A third method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a process of bit sequence generation and channel coding on second uplink control information; and a fourth method for obtaining complex-valued information based on the second uplink control information, wherein bit sequence generation and channel coding are not performed on the second uplink control information.

9. The method according to claim 7, further comprising: Before the UE maps the complex-valued information associated with the CSI to an uplink channel or signal, the UE obtains matched complex-valued information based on the complex-valued information and the number of time-frequency resource units associated with the reported CSI, wherein the number of complex-valued symbols included in the matched complex-valued information is the same as the number of time-frequency resource units associated with the reported CSI.

10. The method according to claim 9, wherein obtaining the matched complex value information comprises: The UE matches the output elements with the time-frequency resource units associated with the reported CSI in order from large to small according to the priorities of the output elements of the neural network model, and obtains a matched complex-valued symbol sequence with priority.

11. The method according to claim 9, wherein obtaining the matched complex value information comprises: The UE matches the output elements of the neural network model with the time-frequency resource units associated with the reported CSI in sequence according to the indexes of the output elements of the neural network model, and obtains a matched complex-valued symbol sequence related to the indexes.

12. The method according to claim 1, further comprising: Based on the content and / or format of the uplink control information, a first processing method among multiple methods for obtaining complex-valued information associated with the CSI is determined.

13. A method performed by a base station in a communication system, comprising: Sending high-layer signaling to the UE, where the high-layer signaling includes first information, where the first information includes at least one of the following: relevant information of the first processing method and information of at least one parameter set associated with the second method; sending physical layer signaling to the UE, where the physical layer signaling includes second information of one parameter set among the at least one parameter set, The first processing method is one of multiple methods for obtaining complex-valued information associated with CSI, wherein the multiple methods include: A first method for obtaining complex-valued information based on a bit sequence, wherein the bit sequence is obtained by performing a bit sequence generation and channel coding on CSI, and A second method for obtaining complex-valued information based on CSI, wherein the CSI does not undergo bit sequence generation and channel coding.

14. A user equipment (UE) in a communication system, comprising: a transceiver configured to transmit and / or receive signals; A controller is configured to control the UE to execute the method according to any one of claims 1-12.

15. A base station in a communication system, comprising: a transceiver configured to transmit and / or receive signals; A controller is configured to control the base station to perform the method according to claim 13.

Images