Method and device for semantic communication in wireless communication system

Abstract

A method performed by a first wireless device according to an embodiment of the present specification comprises the steps of: segmenting each of semantic features related to source data into one or more attributes on the basis of a shared knowledge base; generating a plurality of semantic representation maps on the basis of the attributes of the semantic features and a relation weight vector; transmitting information on the relation weight vector to a second wireless device; and transmitting a plurality of semantic streams based on the plurality of semantic representation maps to the second wireless device. Each semantic representation map comprises i) the semantic features and ii) one or more attributes mapped to each of the semantic features on the basis of the relation weight vector. A semantic feature to which each of attributes related to error occurring among the attributes is mapped is different for each semantic representation map.

Description

Method and device for semantic communication in a wireless communication system

[0001] The present invention relates to a method and device for semantic communication in a wireless communication system.

[0002] Mobile communication systems were developed to provide voice services while ensuring user activity. However, they have expanded beyond voice to include data services. Currently, explosive growth in traffic is leading to resource shortages and users are demanding faster services, necessitating a more advanced mobile communication system.

[0003] Next-generation mobile communication systems must support explosive data traffic growth, dramatically increasing data rates per user, a vastly increased number of connected devices, ultra-low end-to-end latency, and high energy efficiency. To achieve these goals, various technologies are being studied, including dual connectivity, massive multiple input multiple output (MIMO), in-band full duplex, non-orthogonal multiple access (NOMA), super wideband support, and device networking.

[0004] Meanwhile, in semantic communication, a source transmits a semantic stream based on semantic feature(s) associated with source data to a destination. At this time, the destination infers semantic features from the semantic stream based on a shared knowledge base.

[0005] The existing semantic communication method described above is prone to errors, which can lead to the following problems: If an error occurs in the semantic stream received by the destination, completely different semantic feature(s) may be inferred from the stream, rather than the semantic feature(s) related to the source data.

[0006] The purpose of this specification is to propose a method to address the aforementioned issues. Specifically, the purpose of this specification is to propose a method for a destination to successfully infer semantic features from a semantic stram in which an error has occurred.

[0007] A method performed by a first wireless device in a wireless communication system according to one embodiment of the present disclosure comprises the steps of: dividing each of semantic features associated with source data into one or more attributes based on a shared knowledge base; generating a plurality of semantic representation maps based on attributes of the semantic features and a relation weight vector; transmitting information about the relation weight vector to a second wireless device; and transmitting a plurality of semantic streams based on the plurality of semantic representation maps to the second wireless device.

[0008] Each semantic representation map comprises i) the semantic features and ii) one or more attributes mapped to each of the semantic features based on the relationship weight vector.

[0009] The semantic features to which each attribute related to error occurring among the above attributes is mapped are different for each semantic representation map.

[0010] Based on the shared knowledge base, attributes based on the plurality of semantic streams can be decoded by the second wireless device. Semantic features can be determined through semantic feature reasoning based on the decoded attributes.

[0011] Among the semantic features to which the decoded attributes are mapped, the semantic features may be determined based on at least one of the number of times each attribute is decoded and / or the relationship weight associated with the shared knowledge base.

[0012] Based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes are mapped: the determined semantic features may include a semantic feature to which an attribute having a high decoding count among the different semantic features is mapped.

[0013] Based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes having the same decoding count are mapped: the determined semantic features may include semantic features determined based on relationship weights related to the shared knowledge base among the different semantic features.

[0014] Each semantic representation map can be a knowledge graph represented based on nodes and edges. The nodes can be based on a single semantic feature or attribute. The edges can represent the relationship between a single semantic feature and an attribute.

[0015] The above relationship weight vector may include multiple weight values. Each weight may be a value representing a relationship between i) one semantic feature and ii) one or more attributes mapped to the one semantic feature.

[0016] A first wireless device operating in a wireless communication system according to another embodiment of the present disclosure includes one or more transceivers, one or more processors, and one or more memories connected to the one or more processors and storing instructions.

[0017] The instructions are characterized in that, based on being executed by the one or more processors, the one or more processors are set to perform all steps of any one of the methods performed by the first wireless device.

[0018] A device according to another embodiment of the present disclosure comprises one or more memories and one or more processors functionally connected to the one or more memories.

[0019] Said one or more memories are characterized in that they store instructions that, based on being executed by said one or more processors, cause said one or more processors to perform all steps of any one of the methods performed by said first wireless device.

[0020] In another embodiment of the present disclosure, one or more non-transitory computer-readable media store instructions.

[0021] The instructions executable by the one or more processors are characterized by configuring the one or more processors to perform all steps of any one of the methods performed by the first wireless device.

[0022] In another embodiment of the present disclosure, a method performed by a second wireless device in a wireless communication system includes the steps of receiving information about a relation weight vector associated with generation of a plurality of semantic representation maps from a first wireless device, receiving a plurality of semantic streams based on the plurality of semantic representation maps from the first wireless device, decoding attributes based on the plurality of semantic streams based on a shared knowledge base, and determining semantic features through semantic feature reasoning based on the decoded attributes.

[0023] It is characterized in that, among the semantic features to which the decoded attributes are mapped, the semantic features are determined based on at least one of the number of times each attribute is decoded and / or the relationship weight related to the shared knowledge base.

[0024] Based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes are mapped: the determined semantic features may include a semantic feature to which an attribute having a high decoding count among the different semantic features is mapped.

[0025] Based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes having the same decoding count are mapped: the determined semantic features may include semantic features determined based on relationship weights related to the shared knowledge base among the different semantic features.

[0026] Based on the shared knowledge base, each of the semantic features associated with the source data of the first wireless device can be divided into one or more attributes. Based on the attributes of the semantic features and the relationship weight vector, the plurality of semantic representation maps can be generated.

[0027] Each semantic representation map may include i) semantic features associated with the source data, and ii) attributes mapped to each of the semantic features associated with the source data based on the relationship weight vector. The semantic feature to which each of the attributes associated with an error occurring among the attributes is mapped may be different for each semantic representation map.

[0028] A second wireless device operating in a wireless communication system according to another embodiment of the present disclosure includes one or more transceivers, one or more processors, and one or more memories coupled to the one or more processors and storing instructions.

[0029] The instructions are characterized in that, based on being executed by the one or more processors, the one or more processors are set to perform all steps of the method performed by the second wireless device.

[0030] According to an embodiment of the present specification, a plurality of semantic streams are transmitted based on a plurality of semantic representation maps generated by varying the positions of features related to error occurrence among features mapped to semantic features related to source data.

[0031] Accordingly, a diversity effect can be achieved based on the above-described plurality of semantic streams. That is, the probability of successful reception of the semantic stream(s) related to the source data can be improved.

[0032] Furthermore, even if errors occur in the semantic stream, they can be corrected based on decoded attributes at the destination. In other words, semantic features related to the source data can be normally inferred from the error-prone semantic stream, supporting error-resilient semantic communication.

[0033] The effects that can be obtained from this specification are not limited to the effects mentioned above, and other effects that are not mentioned can be clearly understood by a person having ordinary skill in the technical field to which this specification belongs from the description below.

[0034] The accompanying drawings are intended to aid understanding of the present specification and may provide embodiments of the present specification along with detailed descriptions. However, the technical features of the present specification are not limited to specific drawings, and the features disclosed in each drawing may be combined with each other to form new embodiments. Reference numerals in each drawing may indicate structural elements.

[0035] Figure 1 is a drawing showing an example of a communication system applicable to this specification.

[0036] Figure 2 is a drawing showing an example of a wireless device applicable to this specification.

[0037] FIG. 3 is a diagram illustrating a method for processing a transmission signal applicable to the present specification.

[0038] FIG. 4 is a drawing showing another example of a wireless device applicable to this specification.

[0039] FIG. 5 is a drawing showing an example of a mobile device applicable to this specification.

[0040] Figure 6 is a diagram showing physical channels applicable to this specification and a signal transmission method using them.

[0041] Figure 7 is a diagram showing an example of a perceptron structure.

[0042] Figure 8 is a diagram showing an example of a multilayer perceptron structure.

[0043] Figure 9 is a diagram showing an example of a deep neural network.

[0044] Figure 10 is a diagram showing an example of a convolutional neural network.

[0045] Figure 11 is a diagram showing an example of a filter operation in a convolutional neural network.

[0046] Figure 12 shows an example of a neural network structure in which a recurrent loop exists.

[0047] Figure 13 shows an example of the operating structure of a recurrent neural network.

[0048] Figure 14 is a diagram illustrating a level-based communication model to which the embodiment proposed in this specification can be applied.

[0049] FIG. 15 is a conceptual diagram illustrating semantic communication according to one embodiment of the present specification.

[0050] FIG. 16 illustrates a transmission structure for semantic communication to which a method according to one embodiment of the present specification can be applied.

[0051] FIG. 17 illustrates a transmission technique of semantic features to which a method according to one embodiment of the present specification can be applied.

[0052] FIG. 18 illustrates a semantic error correction technique to which a method according to one embodiment of the present specification may be applied.

[0053] FIG. 19 illustrates image samples used for learning to perform fine segmentation according to one embodiment of the present disclosure.

[0054] Figure 20 illustrates a semantic stream generation process according to one embodiment of the present specification.

[0055] FIG. 21 illustrates decoding, semantic feature reasoning operations at a destination according to one embodiment of the present specification, and cases in which errors occur in the operations.

[0056] FIG. 22 illustrates a semantic stream transmission process based on multiple semantic representation maps according to one embodiment of the present specification.

[0057] Figure 23 illustrates a process for generating multiple semantic representation maps according to one embodiment of the present specification.

[0058] FIG. 24 illustrates a feature decoding and semantic feature inference process for a semantic stream performed based on multiple semantic representation maps according to one embodiment of the present specification.

[0059] FIG. 25 illustrates a semantic feature transmission and reception procedure based on multiple semantic representation maps according to one embodiment of the present specification.

[0060] FIG. 26 is a flowchart illustrating a method performed by a first wireless device in a wireless communication system according to one embodiment of the present specification.

[0061] FIG. 27 is a flowchart illustrating a method performed by a second wireless device in a wireless communication system according to another embodiment of the present specification.

[0062] The following embodiments combine the components and features of this specification in a predetermined form. Each component or feature may be considered optional unless explicitly stated otherwise. Each component or feature may be implemented without being combined with other components or features. Furthermore, some components and / or features may be combined to form embodiments of this specification. The order of operations described in the embodiments of this specification may be changed. Some components or features of one embodiment may be included in another embodiment or may be replaced with corresponding components or features of another embodiment.

[0063] In the description of the drawings, procedures or steps that may obscure the gist of the present specification are not described, and procedures or steps that can be understood by a person skilled in the art are also not described.

[0064] Throughout the specification, when a part is said to "comprising" (or including) a certain component, this does not mean that other components are excluded, but rather that other components can be included, unless specifically stated otherwise. In addition, terms such as "...part," "...unit," and "module" described in the specification mean a unit that processes at least one function or operation, which may be implemented by hardware, software, or a combination of hardware and software. In addition, the words "a" or "an," "one," "the," and similar related words may be used in the context of describing this specification (especially in the context of the claims below) to include both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context.

[0065] The embodiments of this specification have been described with a focus on the data transmission and reception relationship between a base station and a mobile station. Here, the base station is understood as a terminal node of a network that directly communicates with the mobile station. Certain operations described herein as being performed by the base station may, in some cases, be performed by an upper node of the base station.

[0066] That is, in a network consisting of multiple network nodes including a base station, various operations performed for communication with a mobile station may be performed by the base station or other network nodes other than the base station. In this case, the term 'base station' may be replaced by terms such as fixed station, Node B, eNB (eNode B), gNB (gNode B), ng-eNB, advanced base station (ABS), or access point.

[0067] Additionally, in the embodiments of the present specification, the term terminal may be replaced with terms such as user equipment (UE), mobile station (MS), subscriber station (SS), mobile subscriber station (MSS), mobile terminal, or advanced mobile station (AMS).

[0068] Additionally, a transmitter refers to a fixed and / or mobile node that provides data or voice services, and a receiver refers to a fixed and / or mobile node that receives data or voice services. Therefore, for uplink, a mobile station can be the transmitter, and a base station can be the receiver. Similarly, for downlink, a mobile station can be the receiver, and a base station can be the transmitter.

[0069] Embodiments of the present specification may be supported by standard documents disclosed in at least one of wireless access systems, such as IEEE 802.xx system, 3rd Generation Partnership Project (3GPP) system, 3GPP Long Term Evolution (LTE) system, 3GPP 5G (5th generation) NR (New Radio) system and 3GPP2 system, and in particular, embodiments of the present specification may be supported by 3GPP TS (technical specification) 38.211, 3GPP TS 38.212, 3GPP TS 38.213, 3GPP TS 38.321 and 3GPP TS 38.331 documents.

[0070] Furthermore, the embodiments of this specification may be applied to other wireless access systems and are not limited to the aforementioned systems. For example, they may also be applicable to systems implemented after the 3GPP 5G NR system, and are not limited to a specific system.

[0071] That is, obvious steps or parts not described in the embodiments of this specification may be explained by reference to the above documents. In addition, all terms disclosed in this specification may be explained by the above standard documents.

[0072] Hereinafter, preferred embodiments according to the present specification will be described in detail with reference to the accompanying drawings. The detailed description set forth below, together with the accompanying drawings, is intended to illustrate exemplary embodiments of the present specification and is not intended to represent the only embodiments in which the technical components of the present specification may be implemented.

[0073] Additionally, specific terms used in the embodiments of this specification are provided to aid in understanding of this specification, and the use of these specific terms may be changed to other forms without departing from the technical spirit of this specification.

[0074] The following technology can be applied to various wireless access systems such as CDMA (code division multiple access), FDMA (frequency division multiple access), TDMA (time division multiple access), OFDMA (orthogonal frequency division multiple access), and SC-FDMA (single carrier frequency division multiple access).

[0075] In order to make the following description clear, the following description is based on a 3GPP communication system (e.g., LTE, NR, etc.), but the technical idea of ​​the present invention is not limited thereto. LTE may refer to technology after 3GPP TS 36.xxx Release 8. Specifically, LTE technology after 3GPP TS 36.xxx Release 10 may be referred to as LTE-A, and LTE technology after 3GPP TS 36.xxx Release 13 may be referred to as LTE-A pro. 3GPP NR may refer to technology after TS 38.xxx Release 15. 3GPP 6G may refer to technology after TS Release 17 and / or Release 18. "xxx" refers to a standard document detail number. LTE / NR / 6G may be collectively referred to as a 3GPP system.

[0076] For background information, terms, abbreviations, etc. used in this specification, reference may be made to standard documents published prior to the invention of the present invention. For example, reference may be made to the 36.xxx and 38.xxx standard documents.

[0077] Communication systems applicable to this specification

[0078] Although not limited thereto, the various descriptions, functions, procedures, proposals, methods and / or operational flowcharts disclosed herein may be applied to various fields requiring wireless communication / connectivity (e.g., 5G) between devices.

[0079] Hereinafter, more specific examples will be provided with reference to the drawings. In the drawings / descriptions below, the same drawing reference numerals may represent identical or corresponding hardware blocks, software blocks, or functional blocks, unless otherwise described.

[0080] FIG. 1 is a diagram illustrating an example of a communication system applicable to the present specification. Referring to FIG. 1, a communication system (100) applicable to the present specification includes a wireless device, a base station, and a network. Here, a wireless device refers to a device that performs communication using a wireless access technology (e.g., 5G NR, LTE) and may be referred to as a communication / wireless / 5G device. Although not limited thereto, the wireless device may include a robot (100a), a vehicle (100b-1, 100b-2), an XR (extended reality) device (100c), a hand-held device (100d), a home appliance (100e), an IoT (Internet of Things) device (100f), and an AI (artificial intelligence) device / server (100g). For example, the vehicle may include a vehicle equipped with a wireless communication function, an autonomous vehicle, a vehicle capable of performing vehicle-to-vehicle communication, etc. Here, the vehicles (100b-1, 100b-2) may include unmanned aerial vehicles (UAVs) (e.g., drones). The XR devices (100c) include augmented reality (AR) / virtual reality (VR) / mixed reality (MR) devices, and may be implemented in the form of head-mounted devices (HMDs), head-up displays (HUDs) installed in vehicles, televisions, smartphones, computers, wearable devices, home appliances, digital signage, vehicles, robots, etc. The portable devices (100d) may include smartphones, smart pads, wearable devices (e.g., smartwatches, smart glasses), computers (e.g., laptops, etc.), etc. The home appliances (100e) may include TVs, refrigerators, washing machines, etc. The IoT devices (100f) may include sensors, smart meters, etc.For example, the base station (120) and the network (130) may also be implemented as wireless devices, and a specific wireless device (120a) may act as a base station / network node to other wireless devices.

[0081] Wireless devices (100a to 100f) can be connected to a network (130) via a base station (120). AI technology can be applied to the wireless devices (100a to 100f), and the wireless devices (100a to 100f) can be connected to an AI server (100g) via a network (130). The network (130) can be configured using a 3G network, a 4G (e.g., LTE) network, a 5G (e.g., NR) network, etc. The wireless devices (100a to 100f) can communicate with each other via the base station (120) / network (130), but can also communicate directly (e.g., sidelink communication) without going through the base station (120) / network (130). For example, vehicles (100b-1, 100b-2) can communicate directly (e.g., V2V (vehicle to vehicle) / V2X (vehicle to everything) communication). In addition, IoT devices (100f) (e.g., sensors) can communicate directly with other IoT devices (e.g., sensors) or other wireless devices (100a to 100f).

[0082] Wireless communication / connection (150a, 150b, 150c) can be established between wireless devices (100a to 100f) / base stations (120), and base stations (120) / base stations (120). Here, the wireless communication / connection can be established through various wireless access technologies (e.g., 5G NR) such as uplink / downlink communication (150a), sidelink communication (150b) (or D2D communication), and base station-to-base station communication (150c) (e.g., relay, IAB (integrated access backhaul)). Through the wireless communication / connection (150a, 150b, 150c), the wireless device and base station / wireless device, and base stations and base stations can transmit / receive wireless signals to / from each other. For example, the wireless communication / connection (150a, 150b, 150c) can transmit / receive signals through various physical channels. To this end, at least some of the various configuration information setting processes for transmitting / receiving wireless signals, various signal processing processes (e.g., channel encoding / decoding, modulation / demodulation, resource mapping / demapping, etc.), and resource allocation processes may be performed based on various proposals of this specification.

[0083] Communication systems applicable to this specification

[0084] FIG. 2 is a diagram illustrating an example of a wireless device applicable to this specification.

[0085] Referring to FIG. 2, the first wireless device (200a) and the second wireless device (200b) can transmit and receive wireless signals via various wireless access technologies (e.g., LTE, NR). Here, {the first wireless device (200a), the second wireless device (200b)} can correspond to {the wireless device (100x), the base station (120)} and / or {the wireless device (100x), the wireless device (100x)} of FIG. 1.

[0086] A first wireless device (200a) includes one or more processors (202a) and one or more memories (204a), and may further include one or more transceivers (206a) and / or one or more antennas (208a). The processor (202a) controls the memories (204a) and / or the transceivers (206a), and may be configured to implement the descriptions, functions, procedures, proposals, methods, and / or operational flowcharts disclosed herein. For example, the processor (202a) may process information in the memory (204a) to generate first information / signals, and then transmit a wireless signal including the first information / signals via the transceivers (206a). In addition, the processor (202a) may receive a wireless signal including second information / signals via the transceivers (206a), and then store information obtained from signal processing of the second information / signals in the memory (204a). The memory (204a) may be connected to the processor (202a) and may store various information related to the operation of the processor (202a). For example, the memory (204a) may perform some or all of the processes controlled by the processor (202a), or may store software code including instructions for performing the descriptions, functions, procedures, proposals, methods, and / or operational flowcharts disclosed herein. Here, the processor (202a) and the memory (204a) may be part of a communication modem / circuit / chip designed to implement wireless communication technology (e.g., LTE, NR). The transceiver (206a) may be connected to the processor (202a) and may transmit and / or receive wireless signals via one or more antennas (208a). The transceiver (206a) may include a transmitter and / or a receiver. The transceiver (206a) may be used interchangeably with an RF (radio frequency) unit. In this specification, wireless device may also mean a communication modem / circuit / chip.

[0087] The second wireless device (200b) includes one or more processors (202b), one or more memories (204b), and may further include one or more transceivers (206b) and / or one or more antennas (208b). The processor (202b) controls the memories (204b) and / or the transceivers (206b), and may be configured to implement the descriptions, functions, procedures, proposals, methods, and / or operational flowcharts disclosed herein. For example, the processor (202b) may process information in the memory (204b) to generate third information / signals, and then transmit a wireless signal including the third information / signals via the transceivers (206b). In addition, the processor (202b) may receive a wireless signal including fourth information / signals via the transceivers (206b), and then store information obtained from signal processing of the fourth information / signals in the memory (204b). The memory (204b) may be connected to the processor (202b) and may store various information related to the operation of the processor (202b). For example, the memory (204b) may perform some or all of the processes controlled by the processor (202b), or may store software code including instructions for performing the descriptions, functions, procedures, proposals, methods, and / or operational flowcharts disclosed herein. Here, the processor (202b) and the memory (204b) may be part of a communication modem / circuit / chip designed to implement wireless communication technology (e.g., LTE, NR). The transceiver (206b) may be connected to the processor (202b) and may transmit and / or receive wireless signals via one or more antennas (208b). The transceiver (206b) may include a transmitter and / or a receiver. The transceiver (206b) may be used interchangeably with an RF unit. In this specification, wireless device may also mean a communication modem / circuit / chip.

[0088] Hereinafter, hardware elements of the wireless device (200a, 200b) will be described in more detail. Although not limited thereto, one or more protocol layers may be implemented by one or more processors (202a, 202b). For example, one or more processors (202a, 202b) may implement one or more layers (e.g., functional layers such as physical (PHY), media access control (MAC), radio link control (RLC), packet data convergence protocol (PDCP), radio resource control (RRC), and service data adaptation protocol (SDAP)). One or more processors (202a, 202b) may generate one or more Protocol Data Units (PDUs) and / or one or more Service Data Units (SDUs) according to the descriptions, functions, procedures, proposals, methods, and / or operational flowcharts disclosed herein. One or more processors (202a, 202b) may generate messages, control information, data or information according to the descriptions, functions, procedures, proposals, methods and / or operational flowcharts disclosed herein. One or more processors (202a, 202b) may generate signals (e.g., baseband signals) including PDUs, SDUs, messages, control information, data or information according to the functions, procedures, proposals and / or methods disclosed herein and provide the signals to one or more transceivers (206a, 206b). One or more processors (202a, 202b) may receive signals (e.g., baseband signals) from one or more transceivers (206a, 206b) and obtain PDUs, SDUs, messages, control information, data or information according to the descriptions, functions, procedures, proposals, methods and / or operational flowcharts disclosed herein.

[0089] One or more processors (202a, 202b) may be referred to as a controller, a microcontroller, a microprocessor, or a microcomputer. One or more processors (202a, 202b) may be implemented by hardware, firmware, software, or a combination thereof. For example, one or more application specific integrated circuits (ASICs), one or more digital signal processors (DSPs), one or more digital signal processing devices (DSPDs), one or more programmable logic devices (PLDs), or one or more field programmable gate arrays (FPGAs) may be included in one or more processors (202a, 202b). The descriptions, functions, procedures, proposals, methods, and / or operational flowcharts disclosed herein may be implemented using firmware or software, and the firmware or software may be implemented to include modules, procedures, functions, etc. The descriptions, functions, procedures, suggestions, methods and / or operation flowcharts disclosed in this specification may be implemented using firmware or software configured to perform one or more processors (202a, 202b) or stored in one or more memories (204a, 204b) and executed by one or more processors (202a, 202b). The descriptions, functions, procedures, suggestions, methods and / or operation flowcharts disclosed in this specification may be implemented using firmware or software in the form of codes, instructions and / or sets of instructions.

[0090] One or more memories (204a, 204b) may be coupled to one or more processors (202a, 202b) and may store various forms of data, signals, messages, information, programs, codes, instructions, and / or commands. The one or more memories (204a, 204b) may be configured as read only memory (ROM), random access memory (RAM), erasable programmable read only memory (EPROM), flash memory, hard drives, registers, cache memory, computer readable storage media, and / or combinations thereof. The one or more memories (204a, 204b) may be located internally and / or externally to the one or more processors (202a, 202b). Additionally, the one or more memories (204a, 204b) may be coupled to the one or more processors (202a, 202b) via various technologies, such as wired or wireless connections.

[0091] One or more transceivers (206a, 206b) can transmit user data, control information, wireless signals / channels, etc., as mentioned in the methods and / or flowcharts of this specification, to one or more other devices. One or more transceivers (206a, 206b) can receive user data, control information, wireless signals / channels, etc., as mentioned in the descriptions, functions, procedures, proposals, methods and / or flowcharts of this specification, from one or more other devices. For example, one or more transceivers (206a, 206b) can be coupled to one or more processors (202a, 202b) and can transmit and receive wireless signals. For example, one or more processors (202a, 202b) can control one or more transceivers (206a, 206b) to transmit user data, control information, or wireless signals to one or more other devices. Additionally, one or more processors (202a, 202b) may control one or more transceivers (206a, 206b) to receive user data, control information, or wireless signals from one or more other devices. Additionally, one or more transceivers (206a, 206b) may be coupled to one or more antennas (208a, 208b), and one or more transceivers (206a, 206b) may be configured to transmit and receive user data, control information, wireless signals / channels, or the like, as referred to in the descriptions, functions, procedures, proposals, methods, and / or operational flowcharts disclosed herein, via one or more antennas (208a, 208b). In the present specification, one or more antennas may be multiple physical antennas or multiple logical antennas (e.g., antenna ports). One or more transceivers (206a, 206b) can convert received user data, control information, wireless signals / channels, etc. from RF band signals to baseband signals in order to process the received user data, control information, wireless signals / channels, etc. using one or more processors (202a, 202b).One or more transceivers (206a, 206b) may convert user data, control information, wireless signals / channels, etc. processed by one or more processors (202a, 202b) from baseband signals to RF band signals. For this purpose, one or more transceivers (206a, 206b) may include an (analog) oscillator and / or filter.

[0092] FIG. 3 is a diagram illustrating a method for processing a transmission signal applied to the present specification. For example, the transmission signal may be processed by a signal processing circuit. At this time, the signal processing circuit (300) may include a scrambler (310), a modulator (320), a layer mapper (330), a precoder (340), a resource mapper (350), and a signal generator (360). At this time, as an example, the operations / functions of FIG. 3 may be performed in the processors (202a, 202b) and / or the transceivers (206a, 206b) of FIG. 2. Furthermore, as an example, the hardware elements of FIG. 3 may be implemented in the processors (202a, 202b) and / or the transceivers (206a, 206b) of FIG. 2. For example, blocks 310 to 350 may be implemented in the processor (202a, 202b) of FIG. 2, and block 360 may be implemented in the transceiver (206a, 206b) of FIG. 2, and are not limited to the above-described embodiments.

[0093] The codeword can be converted into a wireless signal through the signal processing circuit (300) of FIG. 3. Here, the codeword is an encoded bit sequence of an information block. The information block may include a transport block (e.g., a UL-SCH transport block, a DL-SCH transport block). The wireless signal may be transmitted through various physical channels (e.g., a PUSCH, a PDSCH) of FIG. 6. Specifically, the codeword can be converted into a bit sequence scrambled by a scrambler (310). The scramble sequence used for scrambling is generated based on an initialization value, and the initialization value may include ID information of the wireless device, etc. The scrambled bit sequence can be modulated into a modulation symbol sequence by a modulator (320). The modulation scheme may include pi / 2-binary phase shift keying (pi / 2-BPSK), m-phase shift keying (m-PSK), m-quadrature amplitude modulation (m-QAM), etc.

[0094] A complex modulation symbol sequence can be mapped to one or more transmission layers by a layer mapper (330). The modulation symbols of each transmission layer can be mapped to the corresponding antenna port(s) by a precoder (340) (precoding). The output z of the precoder (340) can be obtained by multiplying the output y of the layer mapper (330) by an N*M precoding matrix W. Here, N is the number of antenna ports, and M is the number of transmission layers. Here, the precoder (340) can perform precoding after performing transform precoding (e.g., discrete Fourier transform (DFT) transform) on the complex modulation symbols. In addition, the precoder (340) can perform precoding without performing transform precoding.

[0095] The resource mapper (350) can map modulation symbols of each antenna port to time-frequency resources. The time-frequency resources can include multiple symbols (e.g., CP-OFDMA symbols, DFT-s-OFDMA symbols) in the time domain and multiple subcarriers in the frequency domain. The signal generator (360) generates a wireless signal from the mapped modulation symbols, and the generated wireless signal can be transmitted to another device through each antenna. To this end, the signal generator (360) can include an inverse fast Fourier transform (IFFT) module, a cyclic prefix (CP) inserter, a digital-to-analog converter (DAC), a frequency uplink converter, and the like.

[0096] The signal processing process for receiving signals in a wireless device can be configured in reverse order of the signal processing process (310-360) of FIG. 3. For example, a wireless device (e.g., 200a, 200b of FIG. 2) can receive wireless signals from the outside through an antenna port / transceiver. The received wireless signals can be converted into baseband signals through a signal restorer. For this purpose, the signal restorer can include a frequency downlink converter, an analog-to-digital converter (ADC), a CP remover, and a fast Fourier transform (FFT) module. Thereafter, the baseband signal can be restored to a codeword through a resource demapper process, a postcoding process, a demodulation process, and a descrambling process. The codewords can be restored to the original information blocks through decoding. Accordingly, a signal processing circuit (not shown) for a received signal may include a signal restorer, a resource de-mapper, a postcoder, a demodulator, a de-scrambler, and a decoder.

[0097] Wireless device structure applicable to this specification

[0098] FIG. 4 is a diagram illustrating another example of a wireless device to which the present specification applies.

[0099] Referring to FIG. 4, the wireless device (400) corresponds to the wireless devices (200a, 200b) of FIG. 2 and may be composed of various elements, components, units, and / or modules. For example, the wireless device (400) may include a communication unit (410), a control unit (420), a memory unit (430), and additional elements (440). The communication unit may include a communication circuit (412) and a transceiver(s) (414). For example, the communication circuit (412) may include one or more processors (202a, 202b) and / or one or more memories (204a, 204b) of FIG. 2. For example, the transceiver(s) (414) may include one or more transceivers (206a, 206b) and / or one or more antennas (208a, 208b) of FIG. 2. The control unit (420) is electrically connected to the communication unit (410), the memory unit (430), and the additional elements (440) and controls the overall operation of the wireless device. For example, the control unit (420) may control the electrical / mechanical operation of the wireless device based on the program / code / command / information stored in the memory unit (430). In addition, the control unit (420) may transmit information stored in the memory unit (430) to an external device (e.g., another communication device) via a wireless / wired interface through the communication unit (410), or store information received from an external device (e.g., another communication device) via a wireless / wired interface in the memory unit (430).

[0100] The additional element (440) may be configured in various ways depending on the type of the wireless device. For example, the additional element (440) may include at least one of a power unit / battery, an input / output unit, a driving unit, and a computing unit. Although not limited thereto, the wireless device (400) may be implemented in the form of a robot (Fig. 1, 100a), a vehicle (Fig. 1, 100b-1, 100b-2), an XR device (Fig. 1, 100c), a portable device (Fig. 1, 100d), a home appliance (Fig. 1, 100e), an IoT device (Fig. 1, 100f), a digital broadcasting terminal, a hologram device, a public safety device, an MTC device, a medical device, a fintech device (or a financial device), a security device, a climate / environmental device, an AI server / device (Fig. 1, 140), a base station (Fig. 1, 120), a network node, etc. Wireless devices may be mobile or stationary depending on the use / service.

[0101] In FIG. 4, various elements, components, units / parts, and / or modules within the wireless device (400) may be entirely interconnected via a wired interface, or at least some may be wirelessly connected via a communication unit (410). For example, within the wireless device (400), the control unit (420) and the communication unit (410) may be wired, and the control unit (420) and a first unit (e.g., 430, 440) may be wirelessly connected via the communication unit (410). In addition, each element, component, unit / part, and / or module within the wireless device (400) may further include one or more elements. For example, the control unit (420) may be composed of a set of one or more processors. For example, the control unit (420) may be composed of a set of a communication control processor, an application processor, an electronic control unit (ECU), a graphics processing processor, a memory control processor, etc. As another example, the memory unit (430) may be composed of RAM, DRAM (dynamic RAM), ROM, flash memory, volatile memory, non-volatile memory, and / or a combination thereof.

[0102] Mobile devices to which this specification applies

[0103] FIG. 5 is a drawing illustrating an example of a mobile device to which the present specification applies.

[0104] Figure 5 illustrates an example of a mobile device applicable to the present specification. The mobile device may include a smartphone, a smart pad, a wearable device (e.g., a smart watch, smart glasses), or a portable computer (e.g., a laptop, etc.). The mobile device may be referred to as a mobile station (MS), a user terminal (UT), a mobile subscriber station (MSS), a subscriber station (SS), an advanced mobile station (AMS), or a wireless terminal (WT).

[0105] Referring to FIG. 5, the portable device (500) may include an antenna unit (508), a communication unit (510), a control unit (520), a memory unit (530), a power supply unit (540a), an interface unit (540b), and an input / output unit (540c). The antenna unit (508) may be configured as a part of the communication unit (510). Blocks 510 to 530 / 540a to 540c correspond to blocks 410 to 430 / 440 of FIG. 4, respectively.

[0106] The communication unit (510) can transmit and receive signals (e.g., data, control signals, etc.) with other wireless devices and base stations. The control unit (520) can control components of the portable device (500) to perform various operations. The control unit (520) can include an AP (application processor). The memory unit (530) can store data / parameters / programs / codes / commands required for operating the portable device (500). In addition, the memory unit (530) can store input / output data / information, etc. The power supply unit (540a) supplies power to the portable device (500) and can include a wired / wireless charging circuit, a battery, etc. The interface unit (540b) can support connection between the portable device (500) and other external devices. The interface unit (540b) can include various ports (e.g., audio input / output ports, video input / output ports) for connection with external devices. The input / output unit (540c) can input or output video information / signals, audio information / signals, data, and / or information input from a user. The input / output unit (540c) may include a camera, a microphone, a user input unit, a display unit (540d), a speaker, and / or a haptic module.

[0107] For example, in the case of data communication, the input / output unit (540c) obtains information / signals (e.g., touch, text, voice, image, video) input by the user, and the obtained information / signals can be stored in the memory unit (530). The communication unit (510) can convert the information / signals stored in the memory into wireless signals, and transmit the converted wireless signals directly to other wireless devices or to a base station. In addition, the communication unit (510) can receive wireless signals from other wireless devices or base stations, and then restore the received wireless signals to the original information / signals. The restored information / signals can be stored in the memory unit (530) and then output in various forms (e.g., text, voice, image, video, haptic) through the input / output unit (540c).

[0108] Physical channels and general signal transmission

[0109] FIG. 6 is a diagram illustrating physical channels applicable to this specification and a signal transmission method using them.

[0110] When a terminal is powered on again from a powered-off state or newly enters a cell, it performs an initial cell search operation, such as synchronizing with the base station, in step S611. To this end, the terminal receives a primary synchronization channel (P-SCH) and a secondary synchronization channel (S-SCH) from the base station to synchronize with the base station and obtain information such as the cell ID.

[0111] After that, the terminal can obtain broadcast information within the cell by receiving a physical broadcast channel (PBCH) signal from the base station. Meanwhile, the terminal can check the downlink channel status by receiving a downlink reference signal (DL RS) in the initial cell search phase. After completing the initial cell search, the terminal can obtain more specific system information by receiving a physical downlink control channel (PDCCH) and a physical downlink shared channel (PDSCH) based on the physical downlink control channel information in step S612.

[0112] Thereafter, the terminal may perform a random access procedure such as steps S613 to S616 to complete connection to the base station. To this end, the terminal may transmit a preamble through a physical random access channel (PRACH) (S613) and receive a random access response (RAR) for the preamble through a physical downlink control channel and a physical downlink shared channel corresponding thereto (S614). The terminal may transmit a physical uplink shared channel (PUSCH) using scheduling information in the RAR (S615) and perform a contention resolution procedure such as receiving a physical downlink control channel signal and a physical downlink shared channel signal corresponding thereto (S616).

[0113] A terminal that has performed the procedure described above can then perform reception of a physical downlink control channel signal and / or a physical downlink shared channel signal (S617) and transmission of a physical uplink shared channel (PUSCH) signal and / or a physical uplink control channel (PUCCH) signal (S618) as a general uplink / downlink signal transmission procedure.

[0114] Control information transmitted from a terminal to a base station is collectively referred to as uplink control information (UCI). UCI includes hybrid automatic repeat and request acknowledgment / negative ACK (HARQ-ACK / NACK), scheduling request (SR), channel quality indication (CQI), precoding matrix indication (PMI), rank indication (RI), and beam indication (BI) information. UCI is generally transmitted periodically through PUCCH, but depending on the embodiment (e.g., when control information and traffic data must be transmitted simultaneously), it may be transmitted through PUSCH. In addition, the terminal may transmit UCI aperiodically through PUSCH upon request / instruction from the network.

[0115] 6G communication system

[0116] The 6G (wireless communication) system aims to achieve (i) very high data rates per device, (ii) a very large number of connected devices, (iii) global connectivity, (iv) very low latency, (v) low energy consumption for battery-free IoT devices, (vi) ultra-reliable connectivity, and (vii) connected intelligence with machine learning capabilities. The vision of the 6G system can be divided into four aspects: "intelligent connectivity," "deep connectivity," "holographic connectivity," and "ubiquitous connectivity," and the 6G system can satisfy the requirements as shown in Table 1 below. In other words, Table 1 is a table showing the requirements of the 6G system.

[0117]

[0118] At this time, the 6G system may have key factors such as enhanced mobile broadband (eMBB), ultra-reliable low latency communications (URLLC), massive machine type communications (mMTC), AI integrated communication, tactile internet, high throughput, high network capacity, high energy efficiency, low backhaul and access network congestion, and enhanced data security.

[0119] Core implementation technology of 6G systems

[0120] Artificial Intelligence

[0121] The most crucial and newly introduced technology for 6G systems is AI. 4G systems did not involve AI. 5G systems will support partial or very limited AI. However, 6G systems will fully support AI for automation. Advances in machine learning will create more intelligent networks for real-time communications in 6G. Incorporating AI into communications can streamline and improve real-time data transmission. AI can use numerous analyses to determine how complex target tasks should be performed. In other words, AI can increase efficiency and reduce processing delays.

[0122] Time-consuming tasks such as handover, network selection, and resource scheduling can be performed instantly using AI. AI can also play a crucial role in machine-to-machine (M2M), machine-to-human, and human-to-machine communications. Furthermore, AI can facilitate rapid communication in brain-computer interfaces (BCIs). AI-based communication systems can be supported by metamaterials, intelligent structures, intelligent networks, intelligent devices, intelligent cognitive radios, self-sustaining wireless networks, and machine learning.

[0123] Recent attempts to integrate AI into wireless communication systems have focused on the application layer, network layer, and especially deep learning in wireless resource management and allocation. However, this research is increasingly evolving to the MAC layer and physical layer, with attempts to combine deep learning with wireless transmission, particularly at the physical layer. AI-based physical layer transmission refers to the application of AI-based signal processing and communication mechanisms, rather than traditional communication frameworks, in the fundamental signal processing and communication mechanisms. For example, this may include deep learning-based channel coding and decoding, deep learning-based signal estimation and detection, deep learning-based MIMO mechanisms, and AI-based resource scheduling and allocation.

[0124] Machine learning can be used for channel estimation and channel tracking, as well as for power allocation and interference cancellation in the physical layer of the downlink (DL). Furthermore, machine learning can be used for antenna selection, power control, and symbol detection in MIMO systems.

[0125] Below, we will look at machine learning in more detail.

[0126] Machine learning refers to a series of operations that train machines to perform tasks that humans can or cannot perform. Machine learning requires data and a learning model. Data learning methods in machine learning can be broadly categorized into three types: supervised learning, unsupervised learning, and reinforcement learning.

[0127] Neural network training aims to minimize output errors. It involves repeatedly inputting training data into a neural network, calculating the neural network output and target error for the training data, and backpropagating the neural network error from the output layer to the input layer to update the weights of each node in the neural network to reduce the error.

[0128] Supervised learning uses labeled training data, while unsupervised learning may not have labeled training data. For example, in the case of supervised learning for data classification, the training data may be data in which each training data category is labeled. Labeled training data is input to a neural network, and the error can be calculated by comparing the output (categories) of the neural network with the training data labels. The calculated error is backpropagated through the neural network in the backward direction (i.e., from the output layer to the input layer), and the connection weights of each node in each layer of the neural network can be updated through backpropagation. The amount of change in the connection weights of each updated node can be determined by the learning rate. The neural network's calculation of the input data and the backpropagation of the error can constitute a learning cycle (epoch). The learning rate can be applied differently depending on the number of iterations of the neural network's learning cycle. For example, in the early stages of training a neural network, a high learning rate can be used to quickly allow the network to achieve a certain level of performance, thereby increasing efficiency. In the later stages of training, a low learning rate can be used to increase accuracy.

[0129] Learning methods may vary depending on the characteristics of the data. For example, if the goal is to accurately predict data transmitted by a transmitter in a communication system, supervised learning is preferable to unsupervised learning or reinforcement learning.

[0130] The learning model corresponds to the human brain, and the most basic linear model can be thought of, but the machine learning paradigm that uses highly complex neural network structures, such as artificial neural networks, as learning models is called deep learning.

[0131] The neural network cores used in learning methods are mainly divided into deep neural networks (DNN), convolutional deep neural networks (CNN), and recurrent boltzmann machines (RNN).

[0132] An artificial neural network is an example of a network of multiple perceptrons.

[0133] Referring to Fig. 7, when an input vector x=(x1,x2,...,xd) is input, the entire process of multiplying each component by a weight (W1,W2,...,Wd), adding up all the results, and then applying the activation function σ() is called a perceptron. A large-scale artificial neural network structure can extend the simplified perceptron structure illustrated in Fig. 7 to apply the input vector to perceptrons of different dimensions. For convenience of explanation, input values ​​or output values ​​are called nodes.

[0134] Meanwhile, the perceptron structure illustrated in Fig. 7 can be explained as consisting of a total of three layers based on input and output values. An artificial neural network in which there are H perceptrons of dimension (d+1) between the 1st layer and the 2nd layer, and K perceptrons of dimension (H+1) between the 2nd layer and the 3rd layer can be expressed as in Fig. 8.

[0135] The layer where the input vector is located is called the input layer, the layer where the final output value is located is called the output layer, and all layers located between the input layer and the output layer are called hidden layers. The example in Fig. 8 shows three layers, but when counting the number of layers in an actual artificial neural network, the input layer is excluded, so it can be viewed as a total of two layers. An artificial neural network is composed of perceptrons, which are basic blocks, connected in two dimensions.

[0136] The aforementioned input, hidden, and output layers can be applied jointly not only to multilayer perceptrons but also to various artificial neural network structures, such as CNNs and RNNs, which will be described later. The greater the number of hidden layers, the deeper the artificial neural network. The machine learning paradigm that uses sufficiently deep artificial neural networks as learning models is called deep learning. Furthermore, the artificial neural network used for deep learning is called a deep neural network (DNN).

[0137] The deep neural network illustrated in Figure 9 is a multilayer perceptron consisting of eight hidden layers and eight output layers. The multilayer perceptron structure is referred to as a fully-connected neural network. In a fully-connected neural network, there is no connection between nodes located in the same layer, and there is a connection only between nodes located in adjacent layers. DNN has a fully-connected neural network structure and is composed of a combination of multiple hidden layers and activation functions, and can be usefully applied to identify correlation characteristics between inputs and outputs. Here, the correlation characteristic can mean the joint probability of inputs and outputs. Figure 9 is a diagram illustrating an example of a deep neural network.

[0138] Meanwhile, depending on how multiple perceptrons are connected to each other, various artificial neural network structures different from the aforementioned DNN can be formed.

[0139] In DNN, nodes within a single layer are arranged vertically in a one-dimensional manner. However, Fig. 10 can assume a case where nodes are arranged two-dimensionally, with w nodes in width and h nodes in height (the convolutional neural network structure of Fig. 10). In this case, since a weight is added to each connection in the connection process from one input node to the hidden layer, a total of h×w weights must be considered. Since there are h×w nodes in the input layer, a total of h2w2 weights are required between two adjacent layers.

[0140] Figure 10 is a diagram showing an example of a convolutional neural network.

[0141] The convolutional neural network of Fig. 10 has a problem in that the number of weights increases exponentially according to the number of connections. Therefore, instead of considering the connections of all modes between adjacent layers, it assumes that there are small filters, and performs weighted sum and activation function operations on the overlapping portions of the filters, as in Fig. 10.

[0142] Each filter has a weight corresponding to its size, and weight learning can be performed to extract and output a specific feature on the image as a factor. In Fig. 10, a 3×3 filter is applied to the upper left 3×3 region of the input layer, and the output value resulting from performing weighted sum and activation function operations on the corresponding node is stored in z22.

[0143] The above filter performs weighted sum and activation function operations while moving horizontally and vertically at a certain interval while scanning the input layer, and places the output value at the current filter location. This operation method is similar to the convolution operation for images in the field of computer vision, so a deep neural network with this structure is called a convolutional neural network (CNN), and the hidden layer generated as a result of the convolution operation is called a convolutional layer. In addition, a neural network with multiple convolutional layers is called a deep convolutional neural network (DCNN).

[0144] Figure 11 is a diagram showing an example of a filter operation in a convolutional neural network.

[0145] In the convolutional layer, the number of weights can be reduced by calculating a weighted sum that includes only the nodes located in the area covered by the filter, starting from the node where the current filter is located. This allows a single filter to focus on features within a local area. Accordingly, CNNs can be effectively applied to image data processing where physical distance in a two-dimensional area is an important criterion for judgment. Meanwhile, CNNs can apply multiple filters immediately before the convolutional layer, and can generate multiple output results through the convolution operation of each filter.

[0146] Meanwhile, depending on the data properties, there may be data for which sequence characteristics are important. Considering the length variability and chronological relationship of such sequence data, a structure that applies a method of inputting one element of the data sequence at each timestep and inputting the output vector (hidden vector) of the hidden layer output at a specific timestep together with the immediately following element in the sequence is called a recurrent neural network structure.

[0147] Referring to Figure 12, a recurrent neural network (RNN) is a structure that inputs elements (x1(t), x2(t), ,..., xd(t)) of a data sequence at a time point t into a fully connected neural network, and then inputs the hidden vectors (z1(t-1), z2(t-1),..., zH(t-1)) of the immediately preceding time point t-1 together and applies a weighted sum and activation function. The reason for transmitting the hidden vector to the next time point in this way is because the information in the input vectors of the preceding time points is considered to be accumulated in the hidden vector of the current time point.

[0148] Figure 12 shows an example of a neural network structure in which a recurrent loop exists.

[0149] Referring to Figure 12, the recurrent neural network operates in a predetermined order of time for the input data sequence.

[0150] When the input vector (x1(t), x2(t), ,..., xd(t)) at time point 1 is input to the recurrent neural network, the hidden vector (z1(1), z2(1),..., zH(1)) is input together with the input vector (x1(2), x2(2),..., xd(2)) at time point 2, and the vector (z1(2), z2(2),..., zH(2)) of the hidden layer is determined through a weighted sum and an activation function. This process is repeatedly performed until time points 2, 3, ,,, T.

[0151] Figure 13 shows an example of the operating structure of a recurrent neural network.

[0152] Meanwhile, when multiple hidden layers are placed within a recurrent neural network, it is called a deep recurrent neural network (DRNN). Recurrent neural networks are designed to be useful for processing sequence data (e.g., natural language processing).

[0153] It is a neural network core used in a learning manner, and includes various deep learning techniques such as DNN, CNN, RNN, Restricted Boltzmann Machine (RBM), Deep Belief Network (DBN), and Deep Q-Network, and can be applied to fields such as computer vision, speech recognition, natural language processing, and speech / signal processing.

[0154] The symbols / abbreviations / terms used in this specification are as follows.

[0155] - AI: Artificial Intelligence

[0156] - ML: Machine Learning

[0157] - NN: Neural Network

[0158] - DNN: Deep Neural Network

[0159] - GNN: Graph Neural Network

[0160] -MLP: Multi-Layer Perceptron

[0161] - NCE: Noise Contrastive Estimation

[0162] Below, technical issues related to the embodiments proposed in this specification are examined with reference to FIG. 14.

[0163] Figure 14 is a diagram illustrating a level-based communication model to which the embodiment proposed in this specification can be applied.

[0164] Referring to Figure 14, the communication model can be defined at three levels (A to C).

[0165] Level A concerns how accurately symbols (technical messages) can be transmitted between a transmitter and a receiver. This can be considered when the communication model is understood from a technical perspective.

[0166] Level B concerns how accurately the symbols transmitted between a transmitter and a receiver convey meaning. This can be considered when the communication model is understood from a semantic perspective.

[0167] Level C concerns how effectively the meaning received at the destination contributes to subsequent actions. This can be considered when the communication model is understood in terms of effectiveness.

[0168] Not all Level A to Level C perspectives are considered in the design of a communication model and may vary depending on the implementation method.

[0169] For example, a communication model focused on Level A, such as a communication model based on prior art, may be considered. Another example may be a communication model that considers not only Level A but also Level B (and Level C) for supporting semantic communication. In such a communication model, the transmitter and receiver may be referred to as a semantic transmitter and a semantic receiver, and semantic noise may be additionally considered.

[0170] Below, we describe in detail the considerations related to graph neural networks.

[0171] One of the many goals of 6G communications is to enable a variety of new services that interconnect people and machines with varying levels of intelligence. This requires consideration not only of existing technical issues (e.g., Figure 14A), but also of semantic issues (e.g., Figure 14B). Semantic communication is described in detail below, using human-to-human communication as an example.

[0172] Words used to exchange information (word information) are associated with "meaning." Upon hearing a speaker's words, a listener can interpret the meaning or concept expressed by the speaker's words. Linking this to the communication model in Figure 14, to support semantic communication, the concepts associated with the message sent from the source must be correctly interpreted at the destination.

[0173] A source can generate semantic features based on raw data and its background knowledge. The raw data can be provided to the source in advance or collected by the source. The source can generate semantic features by considering the downstream tasks performed by the destination. The source can generate a semantic message containing the semantic features. The source can transmit the semantic message to the destination.

[0174] A destination can receive semantic messages from a source. The destination can acquire semantic features from the semantic messages. Based on the semantic features, the destination can acquire raw data and background knowledge about the source. Based on the raw data and background knowledge about the source, the destination can perform downstream tasks.

[0175] Meanwhile, semantic communication may not aim to reduce reconstruction errors that may occur during the destination's acquisition of raw data from semantic features, but rather to enable the destination to perform downstream tasks in accordance with the source's intent based on semantic features. In other words, the destination may aim to clearly reason about the source's intent based on semantic features. In this case, if the source's background knowledge is reflected in the destination's background knowledge, the destination can clearly infer the source's intent based on semantic features.

[0176] Thus, the semantic properties generated at the source and delivered to the destination must be generated with consideration for the downstream tasks operating at the destination, which may require a task-oriented semantic communication system. A task-oriented semantic communication system can introduce useful invariants for downstream tasks, thereby preserving task-relevant information.

[0177] FIG. 15 is a conceptual diagram illustrating semantic communication according to one embodiment of the present specification.

[0178] Referring to Fig. 15, a sender may be a source, and a receiver may be a destination. The sender may generate a message {M} based on background knowledge (Ks), a world model (Ws), an inference procedure (Is), and a message syntax (M). The world model (Ws) may be raw data. The world model (Ws) and the message syntax (M) may be collected from an external source or provided in advance by the sender. The sender may transmit the message {M} to the receiver.

[0179] The receiver can receive a message ({M}) from the sender. The receiver has background knowledge (K r ), reasoning process (I r ) and interpret the message based on the message syntax (M) (M r ) can be done, and the world model (W r ) can be obtained. The receiving end can transmit feedback (Feedback(F)) to the transmitting end.

[0180] The world model can be expressed as the following mathematical expression 1.

[0181]

[0182] In mathematical equation 1, can be the Shannon entropy of the world model, can be a probability distribution, can be a model distribution. The world model (W) can be the world model (Ws) of the transmitter or the world model (Wr) of the receiver. If the world model (W) is the world model (Ws) of the transmitter, can be the model entropy of the semantic source. If the world model (W) is the world model (Wr) of the receiver, can be the model entropy of the semantic destination.

[0183] The logical probability of a message can be expressed as the following mathematical expression 2.

[0184]

[0185] In Equation 2, x can be a message, m(x) can be a logical probability of the message, can mean "to accompany" or "to be a model." Also, can semantically mean "entails the following result" or "is a stronger condition". Here, x can be a message Wx where x is a set of its models Ws for which x is "true". The semantic entropy of the message (x) can be as follows:

[0186]

[0187] can be the semantic entropy of message(x).

[0188] Considering background knowledge, the conditional logical probability and semantic entropy of a message (x) can be expressed as in the following mathematical expressions 4 and 5.

[0189]

[0190]

[0191] In mathematical equations 4 and 5 may be background knowledge, is a world model If, can be the conditional logical probability of a message (x) considering background knowledge, can be semantic entropy taking into account background knowledge.

[0192] is the statistical probability, And, The truth table may be as shown in Table 2.

[0193]

[0194] In Table 2, it can mean the case #. Based on the truth table as in Table 2, the possible worlds can be #1, #2, and #4, which are truth assignments among cases #1 to #4. The conditional logical probability considering the background knowledge based on Table 2 can be expressed as the following mathematical equations 6 to 8.

[0195]

[0196]

[0197]

[0198] can be a conditional logical probability of A, can be a conditional logical probability of B, can be the logical probability that both A and B are true.

[0199] Due to background knowledge, logical probabilities and a priori statistical probabilities may not be identical, and in the new distribution, A and B may not be logically independent. This can be expressed as in the following mathematical equation (9).

[0200]

[0201] When background knowledge exists, the new variance of the set of models and the semantic entropy of the world model can be as follows: Equations 10 and 11.

[0202]

[0203]

[0204] In mathematical expressions 10 and 11, can be a variance of a set of models considering background knowledge, can be the logical entropy of the world model (W) taking into account background knowledge.

[0205] Here, the world model entropy of the source without considering background knowledge and the world model entropy of the source with considering background knowledge can be as follows: Equations 12 and 13.

[0206]

[0207]

[0208] Referring to Equations 12 and 13, the shared background knowledge between the source and destination can compress messages transmitted and received between the source and destination without losing information. In other words, the background knowledge enables communication between the source and destination using shorter messages, enabling the transmission and reception of as much information as possible between the source and destination.

[0209] That is, communication at the semantic level (Level A to Level C) can take into account background knowledge and thus can perform better than communication at the technical level (Level A).

[0210] That is, if the source generates or transmits semantic features considering the downstream tasks of the destination, utilizing background knowledge may be consistent with the purpose of performing semantic communication.

[0211] To implement semantic communication encompassing all of the aforementioned components, a new layer called a semantic layer can be added to the source and destination, governing the overall operation of semantic data and messages. This layer may reflect a task-oriented communication system. To facilitate communication between the semantic layers added to the source and destination, a protocol, which is a set of rules between the semantic layers, and a set of operations required to perform communication between the semantic layers may be required.

[0212] Figure 16 illustrates a transmission structure for semantic communication to which a method according to one embodiment of the present specification may be applied. Specifically, Figure 16 illustrates a transmission structure for semantic communication including semantic guidance generation and source-channel coding.

[0213] A structure such as Fig. 16 can be applied to build a semantic communication system composed of a newly definable semantic layer. In the semantic communication system structure, the source generates semantic features representing semantic information using CNN or transformer from source data to be transmitted. At the same time, the source classifies each semantic feature according to a semantic label using a semantic segmentation network to generate a semantic stream (Semantic Stream #1~3). In addition, the source assigns an importance score ( ~ ) is given. The importance score is specifically explained with reference to Figure 17 below.

[0214] Figure 17 illustrates a semantic feature transmission technique to which a method according to one embodiment of the present disclosure may be applied. Specifically, Figure 17 illustrates a semantic feature transmission technique that sets a source-channel coding scheme and resource allocation according to a downstream task.

[0215] Referring to Figure 17, the importance score can be set as follows depending on the downstream task.

[0216] If the downstream task of the destination is focused on reconstruction, the difference in importance scores between each semantic feature can be set to be small. Conversely, if the downstream task is focused on recognizing specific objects in images, the difference in importance scores between semantic features can be set to be large.

[0217] At the source, source-channel coding is performed on streams divided into feature units, and coding rate and resource allocation are performed based on the importance score of each stream. At the destination, source-channel decoding is performed on each received signal.

[0218] Figure 18 illustrates a semantic error correction technique to which a method according to one embodiment of the present specification can be applied. Specifically, Figure 18 illustrates a semantic error correction technique utilizing side information.

[0219] The destination performs an error correction process using the conditional generative adversarial network (GAN) structure of Figure 18 on the received semantic stream obtained through the above process. At this time, the side information corresponding to the conditions of the GAN structure utilized by the destination may include a semantic label map generated by the source, an error mask by channel, etc.

[0220] However, in real environments, it is difficult to generate error masks during the source-channel coding and decoding processes. Furthermore, the process by which the destination infers semantic information through features is not actually performed through a semantic label map created by the source and agreed upon between the two parties. In other words, inference of the destination's semantic information is performed through the knowledge base held by the destination. Unless otherwise specified, this knowledge base is assumed to be shared between the source and destination. In this specification, the "knowledge base" may also be referred to as a "shared knowledge base."

[0221] Therefore, in a real environment, when an error occurs in the received semantic stream, a solution cannot be provided for the case where the destination's knowledge base interprets an incorrect semantic label.

[0222] To address the aforementioned issues, diversity in wireless communications can be leveraged to provide error correction for received semantic streams. Specifically, this specification proposes the following operations [1]-[3] to address the aforementioned issues.

[0223] [1] The source creates multiple semantic representation maps for the source data.

[0224] [2] The source transmits semantic stream(s) to the destination based on the generated multiple semantic representation maps.

[0225] [3] The destination performs error correction for semantic information through fine segmentation and attribute decoding.

[0226] Below, embodiments related to the above-described [1]-[3] (e.g., semantic layer protocol and procedure related to the operations of [1]-[3]) are specifically described with reference to FIGS. 19 to 25.

[0227] FIG. 19 illustrates image samples used for learning to perform fine segmentation according to one embodiment of the present disclosure.

[0228] The source can perform fine segmentation down to the attribute unit of the semantic feature for the source data to be transmitted in order to generate multiple semantic representation maps. Fig. 19 illustrates an image sample (or data sample) that can be used for learning to perform fine segmentation of the source. Specifically, the data sample of Fig. 19 can be used for learning the network within the source for fine segmentation. Through such data sample-based learning, the source can perform segmentation down to the smallest unit. The smallest unit can be referred to as the attribute of the semantic feature that the source wants to convey.

[0229] Figure 20 illustrates a semantic stream generation process according to one embodiment of the present specification. Specifically, referring to Figure 20, a semantic stream can be generated based on a knowledge base-based fine-grained segmentation process and a semantic feature formation process.

[0230] Referring to Figure 20, a semantic representation map for a cat image (image data) is generated based on fine segmentation. This is described in detail below.

[0231] The source performs fine segmentation based on its knowledge base (knowledge graph). Specifically, the source divides the source data into attribute units (e.g., A, B, C, D, E, F, G), which are smaller units than the feature level.

[0232] A semantic representation map is formed through the above fine segmentation. The segmentation level can be determined based on the depth of the knowledge base held by the source.

[0233] The Source constructs semantic features by combining the attributes obtained through the above-mentioned fine-grained segmentation. The Source transmits a semantic stream based on the semantic representation map.

[0234] At this time, the semantic representation map can be configured in the form of a knowledge graph. That is, the semantic representation map can be configured with nodes and edges. Specifically, the nodes can include nodes based on semantic features (e.g., X to Z) and attributes (e.g., A to E).

[0235] A feature (e.g., one of X~Z) or an attribute (e.g., one of A~E) can be composed of a single node. The relationship between a feature (e.g., one of X~Z) and an attribute (e.g., one of A~E) can be composed of the above edges. Between a semantic feature and an attribute, a weight value ( ) can be formed through learning about source data. Specifically, the weight value( ) can be based on a knowledge base previously shared between the source and destination. Below, we will examine in detail cases where errors occur during the semantic feature reasoning process.

[0236] FIG. 21 illustrates decoding, semantic feature reasoning operations at a destination according to one embodiment of the present specification, and cases in which errors occur in the operations.

[0237] The destination, like the source, can perform fine-grained segmentation on the received semantic stream by utilizing its existing knowledge base, at the attribute level. However, depending on the quality of the channel through which the semantic stream passes, errors may occur in the attribute(s) obtained by the destination after attribute decoding.

[0238] Referring to Figure 21, Case 1 and Case 2 illustrate the results when the destination performs semantic reasoning on attributes containing errors based on the knowledge base.

[0239] Case 1 represents a case where reasoning is performed based on the semantic features (i.e., X, Y, Z) intended by the source. In other words, Case 1 represents a case where the error regarding the attribute has been restored and semantic feature reasoning is performed normally. The semantic feature reasoning operation for Case 1 is described in more detail below.

[0240] Specifically, referring to Case 1 of Fig. 21, an error occurred during the receiving process, and attributes corresponding to A and F were received as H and L. That is, the results of fine segmentation performed by the destination on the received semantic stream based on the knowledge base are H, B, C, D, E, L, and G. The above H, B, C, D, E, L, and G may be referred to as decoded attributes.

[0241] The destination contains the received attributes (H, B, C, D, E, L, G) and the weight information of the knowledge base it has ( ) can accurately determine / infer the intended features (X, Y, Z) from the source.

[0242] That is, the attributes H and L corresponding to the error can be recovered during the semantic feature reasoning process.

[0243] 1) In the knowledge base, there is no semantic feature to which attributeL is mapped, and there is a semantic feature Y to which attribute E is mapped. Semantic feature Y can be inferred from attributes E,L.

[0244] 2) In the knowledge base, there is a semantic feature U to which attribute H is mapped, and there is a semantic feature X to which attributes B~D are mapped. Since the number of attributes (3) associated with semantic feature X among attributes H, B, C, and D is greater than the number of attributes (1) associated with semantic feature Y, semantic feature X can be inferred from attributes H, B, C, and D.

[0245] Case 2 represents a case where reasoning is done with features (i.e., U, Y, Z) different from the semantic features (i.e., X, Y, Z) intended by the source. In other words, Case 2 represents a case where an error occurred in semantic feature reasoning.

[0246] Specifically, referring to Case 2 of Fig. 21, an error occurred during the receiving process, and attributes A and B were received as attributes H and I. That is, the results of fine segmentation performed by the destination on the received semantic stream based on the knowledge base are H, I, C, D, E, F, and G. The above H, I, C, D, E, F, and G may be referred to as decoded attributes.

[0247] An error occurs in semantic feature reasoning because there is a semantic feature U in the knowledge base to which attributes H, I, C, and D are mapped. Specifically, the number of attributes associated with semantic feature X in the knowledge base is 2, and the number of attributes associated with semantic feature U is 4. Therefore, semantic feature U can be inferred from attributes H, I, C, and D. In the above case, the destination performs reasoning with a semantic feature (U, Y, Z) that is different from the semantic feature (X, Y, Z) intended / transmitted by the source.

[0248] As described above, multiple semantic representation maps can be utilized to support robust semantic communication against unrecoverable errors. This is described in detail below.

[0249] Diversity schemes, commonly used in wireless communications, can be utilized to reduce errors in semantic features. Specifically, semantic stream transmission can be performed based on multiple semantic representation maps utilizing the diversity scheme. A diversity scheme is a method for increasing the reliability of received signals by using the same signal across two or more different communication channels when transmitting signals from a transmitter.

[0250] Figure 22 illustrates a semantic stream transmission process based on multiple semantic representation maps according to one embodiment of the present specification. Specifically, Figure 22 illustrates a process of generating multiple semantic representation maps and transmitting multiple semantic streams based on the multiple semantic representation maps.

[0251] A knowledge base can be shared between the Source and Destination. In this case, the knowledge base can be based on graph information (e.g., the Shared knowledge base graph in Figure 22).

[0252] The source can select a specific attribute (or attribute related to the occurrence of an error) among the attributes of semantic features X, Y, and Z to be transmitted that can affect semantic representation reasoning when an error occurs.

[0253] The above specific attribute may include attribute(s) (e.g., attributes A, B, E) that are specific to the semantic features (e.g., X, Y, and Z) that the source wishes to transmit.

[0254] For example, the specific attribute may include attribute(s) (e.g., attributes A, B or attributes H, I) unique to each semantic feature (e.g., semantic feature X or Y) among two or more semantic features (e.g., semantic features X, U) to which the same attributes (e.g., attributes C, D) are mapped.

[0255] The above specific attribute related to the error of semantic features X, Y, Z that the source wants to transmit can be determined as follows.

[0256] [1] Attributes A~D are mapped to semantic feature X. Attributes C~D are mapped not only to semantic feature X but also to semantic feature Y. Therefore, attributes A~B correspond to attributes unique to semantic feature X.

[0257] [2] Attributes E~F are mapped to semantic feature Y. Attribute F is mapped not only to semantic feature Y but also to semantic feature V. Therefore, attribute E corresponds to an attribute unique to semantic feature Y.

[0258] [3] The source can determine attributes A, B, and E among attributes A to F related to the above semantic features (X, Y, Z) as attributes related to error occurrence (error occurring attribute).

[0259] If an error occurs in attributes A, B, and E among attributes A to D related to semantic features X, Y, and Z, the semantic features X, Y, and Z intended / transmitted by the source may be inferred as other semantic features. In other words, an error may occur in semantic feature reasoning. The above attributes A, B, and E may be referred to as feature(s) related to the occurrence of the error.

[0260] The Source can generate multiple semantic representation maps so that the above error occurring attributes can be affected by multiple channels. The Source can form multiple semantic streams based on the multiple semantic representation maps and then transmit them to the destination.

[0261] Referring to Fig. 22, the multiple semantic representation maps can be generated based on a combination of attribute(s) obtained through fine segmentation of data from the source. At this time, when combining attributes into semantic features, a relation weight vector( ) can be utilized. For example, relation weight vector( ) is a semantic feature Contains the weight value for each of the attributes A, C, D, and G mapped to the relation weight vector ( ) is a semantic feature It contains the weight value for each of the attributes G, C, D, and E mapped to the relation weight vector ( ) is a semantic feature Contains weight values ​​for each of the attributes B, C, D, and G mapped to .

[0262] source is relation weight vector( ) can be created.

[0263] For example, relation weight vector( ) can be generated based on arbitrary values. For example, the relation weight vector( ) can be generated based on values ​​that are easy for attribute decoding (i.e., fine segmentation) at the destination.

[0264] Information about the above relation weight vector is shared between the destination and the source. That is, the source can transmit information about the above relation weight vector to the destination. The destination can utilize the received relation weight vector for attribute decoding (fine segmentation). The following describes in detail the process of generating multiple semantic representation maps.

[0265] Figure 23 illustrates a process for generating multiple semantic representation maps according to one embodiment of the present specification. Specifically, referring to Figure 23, Steps 1 to 4 may be performed to generate multiple semantic representation maps.

[0266] In Step 1, the source determines the number of multiple semantic representation maps (# of semantic representation maps) and the structure of each semantic representation map. For example, the number of multiple semantic representation maps is determined to be 3, and the number of attributes mapped to semantic features X, Y, and Z in each semantic representation map (1, 2, 3) is determined to be 4, 2, and 1, respectively. For example, in each semantic representation map, semantic feature Y( ~ The number of attribute(s) mapped to one of the attributes is determined as 2.

[0267] In Step 2, the source maps the error occurring attributes A, B, and E to multiple semantic representation maps to ensure that no errors occur in the error occurring attributes A, B, and E during the transmission process.

[0268] Specifically, the source maps error occurring attributes A, B, and E to different locations (i.e., different semantic features) in each semantic representation map. That is, the source performs the mapping so that each error occurring attribute belongs to each of the semantic features X, Y, and Z once. The mapping for each error occurring attribute is described in detail below.

[0269] [1] Attribute A is mapped to semantic feature X in semantic representation map 1. Attribute A is mapped to semantic feature Y in semantic representation map 2. Attribute A is mapped to semantic feature Z in semantic representation map 3.

[0270] [2] Attribute B is mapped to semantic feature Y in semantic representation map 1. Attribute B is mapped to semantic feature Z in semantic representation map 2. Attribute B is mapped to semantic feature X in semantic representation map 3.

[0271] [3] Attribute E is mapped to semantic feature Z in semantic representation map 1. Attribute E is mapped to semantic feature X in semantic representation map 2. Attribute E is mapped to semantic feature Y in semantic representation map 3.

[0272] Since each error occurring attribute belongs to the entire transmitted semantic stream, it can be influenced by multiple channels to obtain diversity gain.

[0273] In Step 3, the source maps the remaining attributes (i.e., C, D, F, and G) to semantic features of each semantic representation map.

[0274] In Step 4, the source designs a relation weight vector and then generates a representation map based on it. Each semantic representation map is structured differently from the knowledge base. In other words, due to Steps 2 and 3 described above, the semantic features mapped to attributes A through G of each semantic representation map are different from those of the knowledge base. Therefore, the source sets / determines a relation weight vector and generates each semantic representation map based on it.

[0275] Below, the transmission of semantic streams and semantic feature inference based on the generated multiple semantic representation maps are specifically described.

[0276] FIG. 24 illustrates a feature decoding and semantic feature inference process for a semantic stream performed based on multiple semantic representation maps according to one embodiment of the present specification.

[0277] Referring to Figure 24, the destination can receive multiple semantic streams generated based on multiple semantic representation maps from the source. The destination performs fine segmentation on the received multiple semantic streams and attribute decoding based on a relation weight vector (shared with the source).

[0278] Specifically, the destination can obtain i) decoded attributes H, C, D, G, B, F, E, ii) decoded attributes K, C, D, E, L, A, B and iii) decoded attributes I, C, D, G, J, F, A from three semantic streams.

[0279] The destination applies the received attribute obtained through the above decoding process to the knowledge base and performs a reasoning process for the semantic feature intended by the source.

[0280] Because the source transmits error-occurring attributes in a manner that ensures diversity across multiple semantic representation maps, error recovery is possible even when errors occur in the semantic stream received at the destination via the channel. In other words, semantic feature error restoring is possible during the process of performing semantic feature reasoning based on the received attributes.

[0281] When performing semantic feature inference, the destination is a majority of the received attributes (or the number of times each attribute is decoded) and the relation weight of the knowledge base (e.g., It is possible to infer the features transmitted by the source based on one or more relation weight values.

[0282] The graph corresponding to the semantic feature reasoning in Fig. 24 is a graph that expresses the decoded attributes in the knowledge graph of the destination (i.e., a graph based on the knowledge base) in the form of a hitmap (e.g., attribute B, H, J, K: received once (decoded x 1) / attribute A, E, F, G, I: received twice (decoded x 2) / attribute C, D: received three times (decoded x 3)).

[0283] When judging whether the received semantic feature is Y or V, the destination can infer the semantic feature as follows: Since attribute E is decoded more times than attribute J, the destination infers Y among (Y=E, F / V=F, J) as the semantic feature transmitted by the source.

[0284] When determining whether the received semantic feature is X or U, it may be difficult to determine with only the hitmap for the decoded attribute (i.e., the number of decodings of the attribute). Specifically, the decoding counts of attributes A and B, which are unique to semantic feature X, are 2 and 1, respectively, and the decoding counts of attributes I and H, which are unique to semantic feature U, are 2 and 1, respectively. Therefore, the semantic feature cannot be accurately inferred with only the number of decodings of the attribute. In this case, the destination can infer the semantic feature as follows.

[0285] The destination can perform reasoning on the actual semantic features using the relation weight of the knowledge graph it has. The relation weight of the knowledge base (knowledge graph) ( ) based on which the semantic feature X to which attributes A and B are mapped can be inferred as the semantic feature intended by the source.

[0286] Referring to FIG. 25 below, the semantic feature transmission and reasoning process between the source and destination described above is explained.

[0287] FIG. 25 illustrates a semantic feature transmission and reception procedure based on multiple semantic representation maps according to one embodiment of the present specification.

[0288] The source and destination share their knowledge bases (knowledge graphs) through background communication to perform semantic communication (S2510). Through this background communication, the source can obtain information about the destination's knowledge base.

[0289] The source performs fine segmentation on the data to be transmitted by attribute unit (S2520).

[0290] The source generates multiple semantic representation maps by combining the attribute(s) obtained through the above-described fine-grained segmentation (S2530). The multiple semantic representation maps are generated based on the attribute(s) and the relation weight vector. That is, the relation weight vector is used to combine the attribute(s).

[0291] The source transmits information about the relation weight vector related to the generation of the above multiple semantic representation maps to the destination (S2540).

[0292] The source converts semantic feature(s) into semantic stream(s) based on the above multiple semantic representation maps and transmits them through different channels (S2550). For example, the source transmits semantic streams based on the above multiple semantic representation maps to the destination.

[0293] The destination performs attribute decoding through fine segmentation based on the knowledge base it possesses and the relation weight vector (S2560).

[0294] The destination performs reasoning on the semantic feature transmitted by the source by utilizing the knowledge base it possesses regarding the attribute(s) acquired through decoding (S2570).

[0295] In terms of implementation, the operations according to the embodiments described above (e.g., operations for supporting semantic communication) can be processed by the devices of FIGS. 1 to 5 described above (e.g., processors (202a, 202b) of FIG. 2).

[0296] In addition, the operations (e.g., operations for supporting semantic communication) according to the above-described embodiments may be stored in a memory (e.g., 204a, 204b of FIG. 2) in the form of commands / programs (e.g., instructions, executable codes) for driving at least one processor (e.g., processors (202a, 202b) of FIG. 2).

[0297] The embodiments described below are specifically described with reference to FIGS. 26 and 27 in terms of the operation of wireless devices (e.g., the first wireless device (200a) and the second wireless device (200b) of FIG. 2). The methods described below are distinguished only for the convenience of explanation, and it goes without saying that some components of one method may be substituted for some components of another method or may be applied in combination with each other.

[0298] FIG. 26 is a flowchart illustrating a method performed by a first wireless device in a wireless communication system according to one embodiment of the present specification.

[0299] Referring to FIG. 26, a method performed by a first wireless device in a wireless communication system according to an embodiment of the present specification includes a step of dividing each of semantic features related to source data into one or more features (S2610), a step of generating a plurality of semantic representation maps based on the features of the semantic features and a relationship weight vector (S2620), a step of transmitting information about the relationship weight vector (S2630), and a step of transmitting a plurality of semantic streams based on the plurality of semantic representation maps (S2640).

[0300] Hereinafter, the first wireless device may mean a source in semantic communication, and the second wireless device may mean a destination in semantic communication.

[0301] In S2610, the first wireless device segments each semantic feature associated with the source data into one or more attributes based on a shared knowledge base. This segmentation may refer to the fine segmentation described above.

[0302] In S2620, the first wireless device generates a plurality of semantic representation maps based on attributes of the semantic features and a relation weight vector.

[0303] Each semantic representation map may include i) the semantic features and ii) one or more attributes mapped to each of the semantic features based on the relationship weight vector. For example, the plurality of semantic representation maps may be generated based on steps 1 to 4 of FIG. 23.

[0304] For example, the semantic features to which each of the attributes related to error occurring among the above attributes is mapped may be different for each semantic representation map. The following description assumes that the semantic features are X, Y, and Z, the attributes related to error occurrence are A, B, and E, and three semantic representation maps are created (see Fig. 23).

[0305] The semantic feature (X, Y, or Z) to which each of the attributes related to the occurrence of the above error is mapped may be different for each semantic representation map. For example, attribute A is mapped to semantic feature X in semantic representation map 1. Attribute A is mapped to semantic feature Y in semantic representation map 2. Attribute A is mapped to semantic feature Z in semantic representation map 3.

[0306] For example, each semantic representation map may be a knowledge graph represented based on nodes and edges. The nodes may be based on a single semantic feature or attribute. The edges may represent a relationship between a single semantic feature and an attribute.

[0307] For example, the relation weight vector may include multiple weight values. The relation weight vector is the relation weight vector described in FIG. 22 and FIG. 23. ) can be based on. Each weight can be a value representing a relationship between i) one semantic feature and ii) one or more attributes mapped to the one semantic feature.

[0308] In S2630, the first wireless device transmits information about the relationship weight vector to the second wireless device.

[0309] In S2640, the first wireless device transmits a plurality of semantic streams based on the plurality of semantic representation maps to the second wireless device.

[0310] Semantic feature reasoning can be performed based on the above multiple semantic streams.

[0311] Specifically, attributes based on the plurality of semantic streams can be decoded by the second wireless device based on the shared knowledge base. Semantic features can be determined through semantic feature reasoning based on the decoded attributes. Here, the determined semantic features can refer to semantic features related to the source data of the first wireless device inferred by the second wireless device.

[0312] In one embodiment, among the semantic features to which the decoded attributes are mapped, the semantic features may be determined based on at least one of the number of times each attribute is decoded and / or a relationship weight associated with the shared knowledge base.

[0313] For example, based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes are mapped: the determined semantic features may include a semantic feature to which an attribute with a high decoding count among the different semantic features is mapped. This will be described in detail with reference to FIG. 24 below.

[0314] Referring to FIG. 24, the semantic features to which the decoded attributes are mapped are X, U, Y, Z, V, and W. Different semantic features to which one or more identical attributes (F) are mapped are Y and V. The number of decodings of attribute E mapped to semantic feature Y is 2, and the number of decodings of attribute J mapped to semantic feature V is 1. Therefore, among the different semantic features Y and V, the semantic feature Y to which attribute E with a high number of decodings is mapped is included in the determined semantic features (semantic features determined through semantic feature inference).

[0315] For example, based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes having the same decoding count are mapped: the determined semantic features may include semantic features determined based on relationship weights related to the shared knowledge base among the different semantic features. This will be described below with reference to FIG. 24.

[0316] Referring to Fig. 24, the semantic features to which the decoded attributes are mapped are X, U, Y, Z, V, and W. Different semantic features to which one or more identical attributes (C, D) with the same number of decodings (2) are mapped are X and U.

[0317] At this time, the decoding counts of attributes A and B mapped to semantic feature X are 2 and 1, respectively, and the decoding counts of attributes I and H mapped to semantic feature U are 2 and 1, respectively. Therefore, semantic features cannot be inferred based solely on the number of decodings of each attribute. In this case, semantic features can be inferred as follows.

[0318] Relation weight value(s) associated with the above shared knowledge base(s) ) based on which the semantic feature X to which attributes A and B are mapped can be inferred as a semantic feature related to the source data of the first wireless device. In other words, among the different semantic features X and U, the semantic feature X determined based on the relation weight value related to the shared knowledge base is included in the determined semantic features (semantic features determined through semantic feature inference).

[0319] The operations based on S2610 to S2640 described above can be implemented by the device of FIG. 2. For example, the first wireless device (200a) can control one or more transceivers (206a) and / or one or more memories (204a) to perform the operations based on S2610 to S2640.

[0320] FIG. 27 is a flowchart illustrating a method performed by a second wireless device in a wireless communication system according to another embodiment of the present specification.

[0321] Referring to FIG. 27, a method performed by a second wireless device in a wireless communication system according to another embodiment of the present specification includes a step of receiving information on a relationship weight vector related to generation of a plurality of semantic representation maps (S2710), a step of receiving a plurality of semantic streams based on the plurality of semantic representation maps (S2720), a step of decoding features based on the plurality of semantic streams (S2730), and a step of determining semantic features based on the decoded features (S2740).

[0322] Hereinafter, the first wireless device may mean a source in semantic communication, and the second wireless device may mean a destination in semantic communication.

[0323] In S2710, the second wireless device receives information about a relation weight vector associated with the generation of a plurality of semantic representation maps from the first wireless device.

[0324] Based on a shared knowledge base, each semantic feature associated with the source data of the first wireless device can be segmented into one or more attributes. This segmentation may refer to the fine segmentation described above. Based on the attributes of the semantic features and the relationship weight vector, the plurality of semantic representation maps can be generated.

[0325] Each semantic representation map may include i) the semantic features and ii) one or more attributes mapped to each of the semantic features based on the relationship weight vector. For example, the plurality of semantic representation maps may be generated based on steps 1 to 4 of FIG. 23.

[0326] For example, the semantic features to which each of the attributes related to error occurring among the above attributes is mapped may be different for each semantic representation map. The following description assumes that the semantic features are X, Y, and Z, the attributes related to error occurrence are A, B, and E, and three semantic representation maps are created (see Fig. 23).

[0327] The semantic feature (X, Y, or Z) to which each of the attributes related to the occurrence of the above error is mapped may be different for each semantic representation map. For example, attribute A is mapped to semantic feature X in semantic representation map 1. Attribute A is mapped to semantic feature Y in semantic representation map 2. Attribute A is mapped to semantic feature Z in semantic representation map 3.

[0328] For example, each semantic representation map may be a knowledge graph represented based on nodes and edges. The nodes may be based on a single semantic feature or attribute. The edges may represent a relationship between a single semantic feature and an attribute.

[0329] For example, the relation weight vector may include multiple weight values. The relation weight vector is the relation weight vector described in FIG. 22 and FIG. 23. ) can be based on. Each weight can be a value representing a relationship between i) one semantic feature and ii) one or more attributes mapped to the one semantic feature.

[0330] In S2720, the second wireless device receives a plurality of semantic streams based on the plurality of semantic representation maps from the first wireless device.

[0331] In S2730, the second wireless device decodes attributes based on the plurality of semantic streams based on a shared knowledge base.

[0332] Specifically, the second wireless device can decode features based on the plurality of semantic streams through fine segmentation based on a shared knowledge base and the relationship weight vector.

[0333] In S2740, the second wireless device determines semantic features through semantic feature reasoning based on the decoded attributes. Here, the determined semantic features may refer to semantic features related to the source data of the first wireless device inferred by the second wireless device.

[0334] In one embodiment, among the semantic features to which the decoded attributes are mapped, the semantic features may be determined based on at least one of the number of times each attribute is decoded and / or a relationship weight associated with the shared knowledge base.

[0335] For example, based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes are mapped: the determined semantic features may include a semantic feature to which an attribute with a high decoding count among the different semantic features is mapped. This will be described in detail with reference to FIG. 24 below.

[0336] Referring to FIG. 24, the semantic features to which the decoded attributes are mapped are X, U, Y, Z, V, and W. Different semantic features to which one or more identical attributes (F) are mapped are Y and V. The number of decodings of attribute E mapped to semantic feature Y is 2, and the number of decodings of attribute J mapped to semantic feature V is 1. Therefore, among the different semantic features Y and V, the semantic feature Y to which attribute E with a high number of decodings is mapped is included in the determined semantic features (semantic features determined through semantic feature inference).

[0337] For example, based on the fact that the semantic features to which the decoded attributes are mapped include different semantic features to which one or more identical attributes having the same decoding count are mapped: the determined semantic features may include semantic features determined based on relationship weights related to the shared knowledge base among the different semantic features. This will be described below with reference to FIG. 24.

[0338] Referring to Fig. 24, the semantic features to which the decoded attributes are mapped are X, U, Y, Z, V, and W. Different semantic features to which one or more identical attributes (C, D) with the same number of decodings (2) are mapped are X and U.

[0339] At this time, the decoding counts of attributes A and B mapped to semantic feature X are 2 and 1, respectively, and the decoding counts of attributes I and H mapped to semantic feature U are 2 and 1, respectively. Therefore, semantic features cannot be inferred based solely on the number of decodings of each attribute. In this case, semantic features can be inferred as follows.

[0340] Relation weight value(s) associated with the above shared knowledge base(s) ) based on which the semantic feature X to which attributes A and B are mapped can be inferred as a semantic feature related to the source data of the first wireless device. In other words, among the different semantic features X and U, the semantic feature X determined based on the relation weight value related to the shared knowledge base is included in the determined semantic features (semantic features determined through semantic feature inference).

[0341] The operations based on S2710 to S2740 described above can be implemented by the device of FIG. 2. For example, the second wireless device (200b) can control one or more transceivers (206b) and / or one or more memories (204b) to perform the operations based on S2710 to S2740.

[0342] Here, the wireless communication technology implemented in the wireless devices (200a, 200b) of the present specification may include not only LTE, NR, and 6G, but also Narrowband Internet of Things for low-power communication. At this time, for example, NB-IoT technology may be an example of LPWAN (Low Power Wide Area Network) technology, and may be implemented with standards such as LTE Cat NB1 and / or LTE Cat NB2, and is not limited to the above-described names. Additionally or alternatively, the wireless communication technology implemented in the wireless devices (XXX, YYY) of the present specification may perform communication based on LTE-M technology. At this time, for example, LTE-M technology may be an example of LPWAN technology, and may be called by various names such as eMTC (enhanced Machine Type Communication). For example, LTE-M technology can be implemented by at least one of various standards such as 1) LTE CAT 0, 2) LTE Cat M1, 3) LTE Cat M2, 4) LTE non-BL (non-Bandwidth Limited), 5) LTE-MTC, 6) LTE Machine Type Communication, and / or 7) LTE M, and is not limited to the above-described names. Additionally or alternatively, the wireless communication technology implemented in the wireless device (XXX, YYY) of the present specification can include at least one of ZigBee, Bluetooth, and Low Power Wide Area Network (LPWAN) considering low-power communication, and is not limited to the above-described names. For example, ZigBee technology can create PAN (personal area networks) related to small / low-power digital communication based on various standards such as IEEE 802.15.4, and can be called by various names.

[0343] The embodiments described above are combinations of the components and features of the present disclosure in a predetermined form. Each component or feature should be considered optional unless explicitly stated otherwise. Each component or feature may be implemented without being combined with other components or features. Furthermore, it is also possible to combine some components and / or features to form an embodiment of the present disclosure. The order of operations described in the embodiments of the present disclosure may be changed. Some components or features of one embodiment may be included in another embodiment or may be replaced with corresponding components or features of another embodiment. It is self-evident that claims that do not have an explicit citation relationship in the patent claims may be combined to form an embodiment or may be incorporated as a new claim through a post-application amendment.

[0344] Embodiments according to the present specification may be implemented by various means, for example, hardware, firmware, software, or a combination thereof. In the case of hardware implementation, an embodiment of the present invention may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, etc.

[0345] When implemented via firmware or software, an embodiment of the present specification may be implemented in the form of a module, procedure, function, or the like that performs the functions or operations described above. The software code may be stored in memory and executed by a processor. The memory may be located within or external to the processor and may exchange data with the processor via various known means.

[0346] It will be apparent to those skilled in the art that this specification may be embodied in other specific forms without departing from the essential characteristics thereof. Therefore, the foregoing detailed description should not be construed in any way as limiting but rather as illustrative. The scope of this specification should be determined by a reasonable interpretation of the appended claims, and all changes within the scope of equivalents herein are intended to be included within the scope of this specification.

Claims

1. A method performed by a first wireless device in a wireless communication system, A step of dividing each of the semantic features related to the source data into one or more attributes based on a shared knowledge base; A step of generating a plurality of semantic representation maps based on attributes of the above semantic features and a relation weight vector; a step of transmitting information about the relationship weight vector to a second wireless device; and A step of transmitting a plurality of semantic streams based on the plurality of semantic representation maps to the second wireless device; comprising: Each semantic representation map comprises i) the semantic features and ii) one or more attributes mapped to each of the semantic features based on the relation weight vector, A method characterized in that the semantic features to which each attribute related to an error occurring among the above attributes is mapped are different for each semantic expression map.

2. In paragraph 1, Based on the shared knowledge base, attributes based on the plurality of semantic streams are decoded by the second wireless device, A method characterized in that semantic features are determined through semantic feature reasoning based on the above decoded attributes.

3. In paragraph 2, A method characterized in that the semantic features are determined based on at least one of the number of decodings of each attribute and / or the relationship weights related to the shared knowledge base among the semantic features to which the decoded attributes are mapped.

4. In paragraph 3, Based on the fact that the semantic features to which the above decoded attributes are mapped include different semantic features to which one or more of the same attributes are mapped: A method characterized in that the above-determined semantic features include a semantic feature to which an attribute having a high number of decodings among the different semantic features is mapped.

5. In paragraph 3, Based on the fact that the semantic features to which the above decoded attributes are mapped include different semantic features to which one or more identical attributes are mapped having the same decoding count: A method characterized in that the determined semantic features include semantic features determined based on relationship weights related to the shared knowledge base among the different semantic features.

6. In paragraph 1, Each semantic representation map is a knowledge graph expressed based on nodes and edges. The above node is based on one semantic feature or one attribute, A method characterized in that the above edge represents a relationship between one semantic feature and one attribute.

7. In paragraph 1, The above relationship weight vector includes multiple weight values, A method characterized in that each weight is a value representing a relationship between i) one semantic feature and ii) one or more attributes mapped to the one semantic feature.

8. In a first wireless device operating in a wireless communication system, One or more transmitters and receivers; one or more processors; and comprising one or more memories connected to said one or more processors and storing instructions; A first wireless device, characterized in that said instructions, based on being executed by said one or more processors, cause said one or more processors to perform all steps of a method according to any one of claims 1 to 7.

9. A device comprising one or more memories and one or more processors functionally connected to the one or more memories, A device characterized in that said one or more memories store instructions that cause said one or more processors to perform all steps of a method according to any one of claims 1 to 7, based on being executed by said one or more processors.

10. In one or more non-transitory computer-readable media storing instructions, One or more non-transitory computer-readable media characterized in that the instructions executable by one or more processors cause the one or more processors to perform all steps of a method according to any one of claims 1 to 7.

11. A method performed by a second wireless device in a wireless communication system, A step of receiving information about a relation weight vector related to generation of a plurality of semantic representation maps from a first wireless device; A step of receiving a plurality of semantic streams based on the plurality of semantic representation maps from the first wireless device; A step of decoding attributes based on the plurality of semantic streams based on a shared knowledge base; and A step of determining semantic features through semantic feature reasoning based on the above decoded attributes; Including, A method characterized in that the semantic features are determined based on at least one of the number of decodings of each attribute and / or the relationship weights related to the shared knowledge base among the semantic features to which the decoded attributes are mapped.

12. In paragraph 11, Based on the fact that the semantic features to which the above decoded attributes are mapped include different semantic features to which one or more of the same attributes are mapped: A method characterized in that the above-determined semantic features include a semantic feature to which an attribute having a high number of decodings among the different semantic features is mapped.

13. In paragraph 11, Based on the fact that the semantic features to which the above decoded attributes are mapped include different semantic features to which one or more identical attributes are mapped having the same decoding count: A method characterized in that the determined semantic features include semantic features determined based on relationship weights related to the shared knowledge base among the different semantic features.

14. In paragraph 11, Based on the shared knowledge base, each of the semantic features related to the source data of the first wireless device is divided into one or more attributes. A method characterized in that the plurality of semantic representation maps are generated based on attributes of the semantic features and the relationship weight vector.

15. In paragraph 14, Each semantic representation map includes i) semantic features associated with the source data, and ii) attributes mapped to each of the semantic features associated with the source data based on the relationship weight vector. A method characterized in that the semantic features to which each attribute related to an error occurring among the above attributes is mapped are different for each semantic expression map.

16. In a second wireless device operating in a wireless communication system, One or more transmitters and receivers; one or more processors; and comprising one or more memories connected to said one or more processors and storing instructions; A second wireless device, characterized in that said instructions, based on being executed by said one or more processors, cause said one or more processors to perform all steps of the method according to claim 15.

Images