Communication apparatus, communication method, and program

Abstract

This communication apparatus comprises: a communication unit that acquires a first reception packet of a first protocol for use in communication with another communication apparatus; an AI control unit that inputs, to a generative AI, a first input that includes the first reception packet, a prompt for inquiring the definition of the first reception packet, and at least a portion of a written specification defined by a first version of the first protocol; and a comparison unit that, on the basis of a response from the generative AI to the first input, determines whether or not the first reception packet is defined in the first version. Accordingly, the communication apparatus can detect the version of a protocol used in communication by a communication partner.

Description

Communication device, communication method, and program

[0001] The present disclosure relates to a communication device, a communication method, and a program.

[0002] In recent years, various devices, such as residential equipment, energy generation and storage equipment, IoT (Internet of Things) home appliances, and FA (Factory Automation) equipment, have communication capabilities and communicate with PCs, smartphones, tablets, servers, and the like on the Internet using communication protocols such as IEC-14543-4-3 (ECHONET (registered trademark) Lite) of the international communication middleware standard (OSI (Open Systems Interconnection) Layer 5-7), IEC-60802 (CC-Link (registered trademark) IE TSN (Time-Sensitive Networking)) of the international communication standard (OSI Layer 1-3), and IETF RFC 1122 (TCP / IP) of the international communication standard (OSI Layer 1-7). Examples of such residential equipment include floor heating, smart meters, intercoms, electronic locks, cameras, storm shutters, shutters, and blinds. Such energy generation and storage equipment includes electric vehicle chargers, hot water supply systems, storage batteries, fuel cells, solar power generators, etc. IoT home appliances include cleaning robots, healthcare devices, microwave ovens, smart speakers, televisions, air conditioners, lighting fixtures, etc.

[0003] Patent Document 1 discloses a protocol processing device that learns the characteristics of protocol data exchanged between systems, effectively utilizes the waiting time for reception from the other system, predicts the next protocol data to be received, and prepares the next protocol data to be sent to the other system based on the prediction results.

[0004] Japanese Patent Application Publication No. 04-117839

[0005] In the past, communication devices have been unable to ascertain the version of the protocol used by a communication partner. The present disclosure has been made in consideration of such circumstances, and provides a communication device, a communication method, and a program that can detect the version of the protocol used by a communication partner.

[0006] This disclosure has been made to solve the above-mentioned problems, and one aspect of the present disclosure is a communication device comprising: a communication unit that acquires a first received packet of a first protocol used for communication with another communication device; an AI control unit that inputs a first input to a generation AI (artificial intelligence), the first input including the first received packet, a prompt inquiring about the definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol; and a matching unit that determines whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

[0007] Another aspect of the present disclosure is the above-mentioned communication device, wherein at least a portion of the specification is a portion of the specification that relates to the first received packet and the prompt.

[0008] Another aspect of the present disclosure is the above-mentioned communication device, wherein, when the matching unit determines that the first received packet is not defined in the first version, the AI ​​control unit obtains a specification defined by a second version of the first protocol that is different from the first version from a site that publishes the specification defined by the first protocol, and inputs a second input to the generation AI that includes the first received packet, the prompt, and at least a portion of the specification defined by the second version, and the matching unit determines whether the first received packet is defined in the second version based on a response from the generation AI to the second input.

[0009] Another aspect of the present disclosure is the above-mentioned communication device, wherein the AI ​​control unit stores the specification defined by the vectorized second version in a vector database.

[0010] Another aspect of the present disclosure is the above-mentioned communication device, comprising an interface unit that generates a response to a request to the communication device using the first protocol, and a confirmation unit that generates a hypothetical request to the communication device that is defined in a version that the matching unit determines as the version in which the first received packet is defined, and causes the interface unit to generate a response to the hypothetical request.

[0011] Another aspect of the present disclosure is the above-mentioned communication device, which is equipped with an interface unit that generates a response to a request to the communication device using the first protocol, and the AI ​​control unit inputs a third input to the generation AI, the third input including a prompt inquiring about a recommended value for the response and at least a portion of a document defining an appropriate range of values ​​in the first protocol for the version in which the first received packet is defined and the version determined by the matching unit.

[0012] Another aspect of the present disclosure is the above-mentioned communication device, wherein the communication unit acquires a second received packet of a second protocol different from the first protocol and used for communication with the other communication device, the AI ​​control unit inputs a fourth input to the generation AI, the fourth input including the second received packet, a prompt inquiring about a definition of the second received packet, and at least a portion of a specification defined by a third version of the second protocol, and the matching unit determines whether the second received packet is defined in the third version based on a response from the generation AI to the fourth input.

[0013] Another aspect of the present disclosure is the above-described communication device, wherein the second protocol is a security protocol.

[0014] Another aspect of the present disclosure is the above-mentioned communication device, comprising the generation AI.

[0015] Another aspect of the present disclosure is the above-mentioned communication device, including a vector database that stores vectorized specifications that define any version of the first protocol.

[0016] Another aspect of the present disclosure is a communication method having a first step of acquiring a first received packet of a first protocol used for communication with another communication device; a second step of inputting a first input to a generation AI (Artificial Intelligence), the first input including the first received packet, a prompt inquiring about a definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol; and a third step of determining whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

[0017] Another aspect of the present disclosure is a program for causing a computer to function as a communication unit that acquires a first received packet of a first protocol used for communication with another communication device, an AI control unit that inputs a first input to a generation AI (Artificial Intelligence) that includes the first received packet, a prompt inquiring about the definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol, and a matching unit that determines whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

[0018] The communication device, communication method, and program disclosed herein have the advantage of being able to detect the version of the protocol used by the communication partner for communication.

[0019] FIG. 1 is a schematic block diagram showing the configuration of a communication system 10 according to a first embodiment of the present disclosure. FIG. 2 is a flowchart illustrating a first operation example of an edge device 200 in the embodiment. FIG. 3 is a flowchart illustrating a second operation example of an edge device 200 in the embodiment. FIG. 4 is a flowchart illustrating a third operation example of an edge device 200 in the embodiment. FIG. 5 is a schematic block diagram showing the configuration of a communication system 10 according to a second embodiment of the present disclosure. FIG. 6 is a schematic block diagram showing the configuration of a communication system 10 according to a third embodiment of the present disclosure. FIG. 7 is a schematic block diagram showing the configuration of a communication system 10 according to a fourth embodiment of the present disclosure. FIG. 8 is a schematic block diagram showing the configuration of a communication system 10 according to a fifth embodiment of the present disclosure. FIG. 9 is an explanatory diagram illustrating the hardware configuration of each device according to each embodiment.

[0020] First Embodiment Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings. FIG. 1 is a schematic block diagram showing the configuration of a communication system 10 according to a first embodiment of the present disclosure. The communication system 10 includes a cloud service 100, a plurality of edge devices 200, a middleware providing server 300, an LLM (Large Language Model) 400, and a vector DB (Date Base) 500. The cloud service 100 is communicatively connected to the plurality of edge devices 200 via a network such as the Internet or an intranet. The plurality of edge devices 200 are communicatively connected to the middleware providing server 300, the LLM 400, and the vector DB 500 via a network such as the Internet or an intranet. Note that these networks may be the same or different.

[0021] The cloud service 100 is, for example, a service built on the cloud or a server built on the cloud, which collects information from the edge device 200 and provides services to the edge device 200. The cloud service 100 may be realized by one or more computers reading and executing a program. The edge device 200 is, for example, a home appliance, an energy generation / storage facility, an IoT home appliance, an FA device, etc., which provides information to the cloud service 100 and receives services provided by the cloud service 100. Note that at least a portion of the configuration of the edge device 200 may be realized by one or more computers reading and executing a program.

[0022] The edge device 200 uses the LLM 400, which has artificial intelligence (AI) capabilities, to detect the version of the first protocol used for communication by the cloud service 100. In this embodiment, the first protocol is IEC-14543-4-3 (ECHONET (registered trademark) Lite), an international communication middleware standard (OSI Layer 5-7), but other protocols may be used. The other protocols may be protocols of any layer of the OSI reference model (Open Systems Interconnection Reference Model). Hereinafter, the software for the first protocol will also be referred to as middleware. The standard defining the first protocol will also be referred to as middle standard A. Furthermore, the software for the first protocol may be drivers or applications for some or all of the protocols of any layer. The standard defining the first protocol may also be referred to as protocol standard A.

[0023] The edge device 200 includes a device unit 210, an I / F (interface) unit 220, a control unit 230, and a communication unit 240. The device unit 210 realizes the main function of the edge device 200. For example, if the edge device 200 is a washing machine, the main function of the edge device 200 is to wash clothes. Alternatively, if the edge device 200 is a washing machine, the main function of the edge device 200 is to wash clothes. The I / F unit 220 is an interface between the control unit 230 and the device unit 210, allowing the control unit 230 to control the device unit 210 and obtain information indicating the state of the device unit 210. The I / F unit 220 may be, for example, a serial bus such as an Inter-Integrated Circuit (I2C) or a Serial Peripheral Interface (SPI), or a parallel bus such as a 16-bit bus.

[0024] The control unit 230 controls the device unit 210 in response to a request from the cloud service 100 and acquires information indicating the status of the device unit 210. The control unit 230 detects the version of the first protocol used for communication by the cloud service 100 using the LLM 400, which has an artificial intelligence function. Furthermore, the control unit 230 may acquire software for the detected version of the protocol from the middleware providing server 300 and install it in the edge device 200.

[0025] The control unit 230 includes a matching unit 231, an AI control unit 232, a confirmation unit 233, and an accumulation reference unit 234. The AI ​​control unit 232 inputs a first input to the LLM 400 (generating AI (Artificial Intelligence)), the first input including a first received packet (frame, message) of a first protocol used for communication with the cloud service 100 (another communication device), a prompt inquiring about a definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol. The first version may be the version of the software of the first protocol installed in the edge device 200, or may be another version.

[0026] At least a portion of the specification may be a portion of the specification that is related to the first received packet and the prompt. The prompt inquiring about the definition of the first received packet may be a prompt inquiring about the definition of the header or the definition of the payload of the first received packet. The AI ​​control unit 232 may search the vector DB 500 using the first received packet and the prompt to acquire at least a portion of the specification. That is, the AI ​​control unit 232 may perform Retrieval-Augmented Generation (RAG) using the LLM 400 and the vector DB 500. The number of first received packets may be one or more.

[0027] The collation unit 231 determines whether the first received packet is defined in the first version based on the response from the LLM 400 to the first input. The confirmation unit 233 generates a provisional request for the edge device 200, which is defined in the version determined by the collation unit 231 to be the version in which the first received packet is defined, and causes the interface unit 220 to generate a response to the provisional request. If the response is normal, the confirmation unit 233 may determine the version determined by the collation unit 231 as the version in which the first received packet is defined. The storage reference unit 234 stores the version of the first protocol supported by the software installed on the edge device 200, the first received packet, etc.

[0028] The communication unit 240 communicates with the cloud service 100, the middleware providing server 300, the LLM 400, and the vector DB 500. For example, the communication unit 240 receives a received packet from the cloud service 100 to obtain the first received packet.

[0029] The middleware providing server 300 is a website that publishes a specification that defines a first protocol used for communication between the cloud service 100 and the edge device 200. The middleware providing server 300 may also publish software (middleware) for the first protocol used for communication between the cloud service 100 and the edge device 200. Furthermore, the format of the specification that defines the first protocol is not limited to a document format such as text format or PDF format, and may be any format including, for example, HTML format, tables, images, and the like.

[0030] The LLM 400 includes artificial intelligence with intelligent functions such as neural network learning and inference, and an operating environment for the artificial intelligence. The LLM 400 may be built on a cloud. The LLM 400 may perform its intelligent functions using a processor that performs tensor processing, such as an NPU (Neural Network Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose Computing on Graphics Processing Units), an LPU (Language Processing Unit), or a TPU (Tensor Processing Unit), and a trained model stored in memory. Furthermore, the LLM 400 may access a trained model via a processor that executes Tensor processing and an API such as DirectML to execute intelligent functions. The cloud on which the LLM 400 is built may be different from the cloud on which the cloud service 100 and the vector DB 500 are built. The LLM 400 is composed of a trained model unit 402 and a model control unit 401. The LLM 400 is a model and its operating environment configured to output a received packet of a first protocol received by the edge device 200 from the cloud service 100, a prompt inquiring about the definition of the received packet, and a definition of the received packet that corresponds to at least a portion of a specification defined by a certain version of the first protocol. When the LLM 400 receives a received packet of a first protocol received by the edge device 200 from the cloud service 100, a prompt inquiring about the definition of the received packet, and at least a portion of the specification document defining a certain version of the first protocol from the edge device 200, the LLM 400 outputs a definition of the received packet based on the input and the learned model described below.

[0031] The trained model unit 402 includes information about the model, which will be described later. The trained model is a model that has been trained in advance. The trained model may also include model parameters, which are information that defines the behavior of the model, such as constraints, weighting variables, and evaluation functions. The trained model unit 402 may also store the trained model in memory, or access the trained model via an API and execute the NPU. Furthermore, the trained model unit 402 may be formed by completing training using the NPU.

[0032] The models include, for example, NN (Neural Network), CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), VAE (Variational Autoencoder), GAN (Generative Adversarial Networks), Diffusion model, Transformer, LLM (Large Language Model), VLM (Visual Language Model), and BERT (Bidirectional Encoder Representations from The model may be a model called Transformers, GPT (Generative Pre-trained Transformer), or CLIP (Contrastive Language Image Pre-training). Note that the above-mentioned models are not exclusive, and for example, LLM, VLM, BERT, and GPT are included in Transformers. Also, for example, Transformers are included in NN. Furthermore, the learning algorithm and model may be a combination of multiple types. Models also include what are called multimodal models that are learned by combining multiple different types of data.

[0033] When the model control unit 401 acquires a received packet of the first protocol received by the edge device 200 from the cloud service 100, a prompt inquiring about the definition of the received packet, and at least a part of the specification defining a certain version of the first protocol, it inputs these to the trained model unit 402, and when the trained model unit 402 generates a definition of the received packet corresponding to the input, it outputs the definition to the source of the inquiry. In other words, when the model control unit 401 acquires the input, it uses the trained model unit 402 to generate and output a definition of the received packet corresponding to the input.

[0034] The trained model and other information used by the artificial intelligence unit may be prepared in advance, or may be obtained via a network as needed.

[0035] Vector DB 500 stores vectorized specifications that define a first protocol. Vector DB 500 includes vector DBs 501, 502, 503, and 504. Vector DB 501 stores vectorized specifications that define a first version of the first protocol. Vector DB 502 stores vectorized specifications that define a second version of the first protocol. Vector DB 503 stores vectorized specifications that define a third version of the first protocol. Vector DB 504 stores vectorized specifications that define a fourth version of the first protocol.

[0036] FIG. 2 is a flowchart illustrating a first operation example of the edge device 200 according to this embodiment. The first operation example is an operation example related to the initial processing of the edge device 200. First, the storage reference unit 234 saves the version number P of the middleware standard A corresponding to the middleware installed on the edge device 200 (step Sa1). Next, the AI ​​control unit 232 acquires X, one of the version numbers of the officially deployed middleware standard A, from the middleware provision server 300 (step Sa2). Note that the AI ​​control unit 232 may acquire the URL (Uniform Resource Locator) of the middleware provision server 300 by an Internet search. Furthermore, the AI ​​control unit 232 may download a web page published by the middleware provision server 300 based on the acquired URL of the middleware provision server 300, chunk and vectorize it, store it in the vector DB 500, and acquire X, one of the version numbers of the officially deployed middleware standard A, by RAG using the vector DB 500 and the LLM 400. For example, the AI ​​control unit 232 inputs to the LLM 400 a prompt such as "Have ECHONET (registered trademark) Lite version 1.0 released already?" and the search results of the web pages published by the middleware providing server 300 and stored in the vector DB 500 at the prompt, and obtains a response to the input (e.g., "Yes, the ECHONET (registered trademark) Lite Specification version 1.0 has been released.").

[0037] Next, the AI ​​control unit 232 determines whether the acquired version number X matches the version number P saved in step Sa1 (step Sa3). If it determines that they do not match (step Sa3—NO), the AI ​​control unit 232 returns to step Sa2 and acquires a different version number X from the previous time. If it determines that they match (step Sa3—YES), the AI ​​control unit 232 acquires the specification defined by the version number P of middleware standard A from the middleware providing server 300 and saves it in the vector DB 500 (step Sa4). Note that chunking and vectorization for saving the specification in the vector DB 500 may be performed by the AI ​​control unit 232, the vector DB 500, or another device. The AI ​​control unit 232 may also delete the web page stored in the vector DB 500 in step Sa2.

[0038] 3 is a flowchart illustrating a second operation example of the edge device 200 according to this embodiment. The second operation example is an operation example when the edge device 200 communicates with the cloud service 100, or an operation example when the edge device 200 receives a packet from the cloud service 100 that the edge device 200 cannot interpret.

[0039] First, the communication unit 240 receives multiple received packets (first received packets) from the cloud service 100 and stores them in the storage reference unit 234 (step Sb1). Next, the AI ​​control unit 232 extracts the definition of the received packets stored in step Sb1 using the vector DB 500 of version P of the middleware standard A installed on the edge device 200 (step Sb2). For example, the AI ​​control unit 232 searches the vector DB 500 of version P of the middleware standard A using the received packet and a prompt inquiring about the definition of the received packet, inputs a first input including the received packet, the prompt, and the search result to the LLM 400, and obtains a response to the first input, thereby extracting the definition of the received packet. For example, the prompt is, "Please tell the request payload data of the request frame format each byte defined by the request specification as mentioned sequence.", and the response is, "1. EHD1 (1 byte): 10, 2. EHD2 (1 byte): 81, 3. TID (2 bytes): 23 08, 4. SEOJ (3 bytes): 05 ff 01, 5. DEOJ (3 bytes): 02 6b 01, 6. ESV (1 byte): 64."

[0040] The collation unit 231 determines whether the definition extracted in step Sb2 is present in version P of the middle standard A (step Sb3). In step Sb3, the collation unit 231 determines whether the values ​​of each item in the response, such as "EHD1 (1 Byte): 10," "EHD2 (1 Byte): 81," and "TID (2 Bytes): 2308," are defined in version P of the middle standard A. If all the item values ​​are defined, the collation unit 231 may determine that the definition is present. If any of the item values ​​are not defined, the collation unit 231 may determine that the definition is absent. The AI ​​control unit 232 may determine whether each item value is defined using the vector DB 500 of version P of the middle standard A, similar to the extraction of the definition in step Sb2. The prompt used in this determination may be, for example, "Please tell the definition of ESV (1 Byte): 64" when determining whether the item value "ESV (1 Byte): 64" is defined.

[0041] If it is determined in step Sb3 that a definition exists (step Sb3-YES), the collating unit 231 determines version T of the middle standard A, which is the version in which the received packet is defined, as version P (step Sb4).

[0042] Furthermore, if it is determined in step Sb3 that there is no definition (no definition) (step Sb3-NO), the collation unit 231 causes the AI ​​control unit 232 to obtain the latest version number of the middle standard A, which is set as X (step Sb5). The AI ​​control unit 232 may obtain the latest version number using a method similar to that used to obtain X, one of the version numbers of the middle standard A, in step Sa2 of FIG. 2. The prompt used in this case may be, for example, "What is the latest version of the ECHONET (registered trademark) Lite?"

[0043] Next, the AI ​​control unit 232 downloads the specification defined by version X (second version) of middle standard A from the middleware providing server 300 (Sb6). Next, the AI ​​control unit 232 chunks and vectorizes the specification defined by version X of middle standard A and stores it in the vector DB 502 (step Sb7). Note that this chunking and vectorization may be performed by the AI ​​control unit 232, the vector DB 500, or another device.

[0044] Next, the AI ​​control unit 232 uses the vector DB 502 of version X of the middle standard A to extract the definition of the received packet saved in step Sb1 (step Sb8). This step Sb8 is the same process as step Sb2, except that the vector DB 502 to be used is used. Next, the matching unit 231 determines whether the definition extracted in step Sb2 is present in version X of the middle standard A (step Sb9). This step Sb9 is the same process as step Sb3, except that the vector DB 502 to be used is used.

[0045] If it is determined in step Sb9 that a definition exists (step Sb9-YES), the collating unit 231 determines that version T of the middle standard A, which is the version in which the received packet is defined, is version X (step Sb11).

[0046] If it is determined in step Sb9 that there is no definition (NO in step Sb9), the collation unit 231 decrements the value of X by the value Down (step Sb10) and returns to step Sb6. The value Down may be a predetermined value, or may be a value determined so that the value of X is the number of the version in which middle standard A is released.

[0047] 4 is a flowchart illustrating a third operation example of the edge device 200 in this embodiment. The third operation example is an operation example for confirming the version T determined by the collation unit 231. First, the confirmation unit 233 obtains the version number T determined by the collation unit 231 (step Sc1). Next, the confirmation unit 233 downloads the middleware of version T of middle standard A from the middleware providing server 300 (step Sc2). Next, the confirmation unit 233 installs the middleware downloaded in step Sc2 (step Sc3). Note that the confirmation unit 233 retains the state before installation so that the state before installation (the version of the middleware before installation) can be restored.

[0048] The confirmation unit 233 inputs a request to the device unit 210 based on the tentative received packet to the I / F unit 220 (step Sc5). Here, the request to the device unit 210 based on the tentative received packet is a request using a received packet defined in version T of the middle standard A. The request to the device unit 210 may be, for example, a request to turn the power on or off, a request for a response indicating whether an abnormal state has occurred, or a request for measured instantaneous power consumption. Next, the AI ​​control unit 232 obtains a response to the request input to the I / F unit 220 and determines whether the response is normal (step Sc6).

[0049] The determination in step Sc6 may be made based on whether the response is defined in version T of middle standard A. Whether the response is defined in version T of middle standard A may be determined based on a response obtained by inputting the response, a prompt inquiring whether the response is defined in version T of middle standard A, and at least a portion of the specification defined by version T of middle standard A into the LLM 400. If the response is normal (step Sc6—YES), the confirmation unit 233 determines whether confirmation of N types of responses has been completed (step Sc7). The value of N may be a predetermined value.

[0050] If it is determined in step Sc7 that confirmation of N types of responses has been completed (step Sc7-YES), the confirmation unit 233 determines that there is no hallucination in LLM400 or that the possibility is low, and confirms version T of the middle standard A (step Sc8).If confirmation of N types of responses has not been completed in step Sc7 (step Sc7-NO), the confirmation unit 233 returns to step Sc5 and performs the process.

[0051] Also, if it is determined in step Sc6 that the response is not normal (step Sc6-NO), the confirmation unit 233 determines that there is a risk of hallucination in LLM400, uninstalls the middleware installed in step Sc4, and returns to the state before installation (step Sc9).

[0052] Second Embodiment In a communication system 10 according to a second embodiment, a document defining appropriate ranges of values ​​for each version of middleware is stored in a vector DB 500. The edge device 200 references the document defining the appropriate ranges to obtain recommended values ​​for responses to requests to the device unit 210 based on received packets.

[0053] 5 is a schematic block diagram showing the configuration of a communication system 10 according to a second embodiment of the present disclosure. The communication system 10 of FIG. 5 is similar to the communication system 10 of FIG. 1, but differs in that the vector DB 500 includes vector DBs 511, 512, 513, and 514, and the control unit 230 includes an AI control unit 232a and a confirmation unit 233a. The following description will focus on the differences between the communication system 10 of this embodiment and the first embodiment, and other details will be omitted. The vector DBs 511, 512, 513, and 514 store documents defining appropriate ranges of values ​​for the first, second, third, and fourth versions of the middle standard A, respectively.

[0054] The AI ​​control unit 232a is similar to the AI ​​control unit 232, but inputs a third input to the LLM 400, which includes a prompt inquiring about a recommended value for a response to a request to the device unit 210 based on the received packet, and at least a portion of a document defining the appropriate range of values ​​in the first protocol for the version in which the first received packet is defined and the version determined by the comparison unit 231.

[0055] The confirmation unit 233a is similar to the confirmation unit 233, but further acquires a response to the third input of the LLM 400, and notifies the user of the edge device 200 of the recommended value acquired as the response.

[0056] In a communication system 10 according to a third embodiment, a vector DB 500 stores specifications defining protocols other than the first protocol and documents defining appropriate ranges. The edge device 200 performs version detection and the like for protocols other than the first protocol in the same manner as for the first protocol.

[0057] Fig. 6 is a schematic block diagram showing the configuration of a communication system 10 according to a third embodiment of the present disclosure. The communication system 10 in Fig. 6 is similar to the communication system 10 in Fig. 5, except that the vector DB 500 includes vector DBs 521, 531, 541, and 551, and the control unit 230 includes a matching unit 231b, an AI control unit 232b, a confirmation unit 233b, and an accumulation reference unit 234b instead of the matching unit 231, the AI ​​control unit 232a, the confirmation unit 233a, and the accumulation reference unit 234.

[0058] Vector DB 521 stores a specification that defines a certain version of the second protocol. Vector DB 531 stores a document that defines an appropriate range of values ​​for a certain version of the second protocol. Vector DB 541 stores a specification that defines a certain version of the third protocol. Vector DB 551 stores a document that defines an appropriate range of values ​​for a certain version of the third protocol.

[0059] The matching unit 231b, AI control unit 232b, confirmation unit 233b, and storage reference unit 234b perform the same processing for the second and third protocols as the matching unit 231, AI control unit 232a, confirmation unit 233a, and storage reference unit 234 perform for the first protocol.

[0060] In a communication system 10 according to a fourth embodiment, a vector DB 500 stores a specification that defines a security protocol that encrypts packets and manages keys for encryption and decryption. The edge device 200 detects the version of the security protocol in the same manner as the first protocol.

[0061] Fig. 7 is a schematic block diagram showing the configuration of a communication system 10 according to a fourth embodiment of the present disclosure. The communication system 10 in Fig. 7 is similar to the communication system 10 in Fig. 6, but differs in that the vector DB 500 includes vector DBs 561, 562, 571, 581, and 551, and the control unit 230 includes a matching unit 231c, an AI control unit 232c, a confirmation unit 233c, and an accumulation reference unit 234c instead of the matching unit 231b, the AI ​​control unit 232b, the confirmation unit 233b, and the accumulation reference unit 234b, and includes LLMs 403 and 404.

[0062] The collation unit 231c, AI control unit 232c, confirmation unit 233c, and storage reference unit 234c perform the same processing for the security protocol as the collation unit 231b, AI control unit 232b, confirmation unit 233b, and storage reference unit 234b perform for the first to third protocols. The AI ​​control unit 232c manages which of the LLMs 400, 403, and 404 to use, and can process multiple packets or multiple protocols in parallel. The LLMs 403 and 404 may use the same model as the LLM 400, or different models.

[0063] Fifth Embodiment In a communication system 10 according to a fifth embodiment, a control unit 230 includes an LLM unit 235 and a vector DB 236 instead of the LLM 400 and the vector DB 500, respectively.

[0064] Fig. 8 is a schematic block diagram showing the configuration of a communication system 10 according to a fifth embodiment of the present disclosure. The communication system 10 in Fig. 8 is similar to the communication system 10 in Fig. 1, but differs in that it does not include the LLM 400 and the vector DB 500, and the control unit 230 includes an LLM unit 235 and a vector DB 236.

[0065] The LLM unit 235 is similar to the LLM 400, but is provided in the control unit 230 and provides services to the AI ​​control unit 232 of the control unit 230. The vector DB 236 is similar to the vector DB 500, but is provided in the control unit 230 and provides services to the AI ​​control unit 232 of the control unit 230.

[0066] FIG. 9 is an explanatory diagram illustrating the hardware configuration of each device according to each embodiment. The devices are a cloud service 100, an edge device 200, a middleware providing server 300, an LLM 400, and a vector DB 500. Each device includes an input / output module I, a storage module M, and a control module P. The input / output module I includes some or all of the following: a communication module H11, a connection module H12, a pointing device H21, a keyboard H22, a display H23, a button H3, a microphone H41, a speaker H42, a camera H51, and a sensor H52. The storage module M includes a drive H7. The storage module M may further include some or all of a memory H8. The control module P includes a memory H8 and a processor H9. These hardware components are connected to each other via a bus (Bus) for communication and receive power from a power supply H6.

[0067] The connection module H12 is a digital input / output port such as a USB (Universal Serial Bus). The pointing device H21, keyboard H22, and display H23 may be touch panels. The sensor H52 is an acceleration sensor, a gyro sensor, a GPS receiving module, a proximity sensor, or the like. The power supply H6 is a power supply unit that supplies the electricity necessary to operate each device. The power supply H6 may be a battery. The drive H7 is an auxiliary storage medium such as a hard disk drive or a solid-state drive. The drive H7 may be a non-volatile memory such as an EEPROM, flash memory, or SOM (Selector Only Memory), or may be a magneto-optical disk drive or a flexible disk drive. Furthermore, the drive H7 is not limited to being built into each device, but may also be an external storage device connected to a connector of the connection module H12. The memory H8 is a main storage medium such as a random access memory or a high bandwidth memory (HBM). The memory H8 may be a cache memory. The memory H8 stores instructions when the instructions are executed by one or more processors H9. The processor H9 is a central processing unit (CPU). The processor H9 may be a microprocessing unit (MPU), a graphics processing unit (GPU), a general-purpose computing on graphics processing unit (GPGPU), a tensor processing unit (TPU), a language processing unit (LPU), or a neural network processing unit (NPU). The processor H9 reads out programs and various data from the drive H7 via the memory H8 and performs calculations to execute instructions stored in one or more memories H8.

[0068] The input / output module I is used in the cloud service 100, the edge device 200, the middleware providing server 300, the LLM 400, the vector DB 500, etc. The control module P is used to implement each part of the cloud service 100, the edge device 200, the middleware providing server 300, the LLM 400, and the vector DB 500. Note that in this specification and the like, the terms cloud service 100, the edge device 200, the middleware providing server 300, the LLM 400, and the vector DB 500 may be replaced with the term control module P.

[0069] The present disclosure may be embodied as follows: (1) One embodiment of the present disclosure is a communication device including: a communication unit that acquires a first received packet of a first protocol used for communication with another communication device; an AI control unit that inputs a first input to a generation AI (Artificial Intelligence), the first input including the first received packet, a prompt inquiring about a definition of the first received packet, and at least a part of a specification defined by a first version of the first protocol; and a comparison unit that determines whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

[0070] (2) Another embodiment of this disclosure is a communication device described in (1), wherein at least a portion of the specification is a portion of the specification that is related to the first received packet and the prompt.

[0071] (3) Another embodiment of this disclosure is a communication device described in (1) or (2), wherein, when the matching unit determines that the first received packet is not defined in the first version, the AI ​​control unit obtains a specification defined by a second version of the first protocol that is different from the first version from a site that publishes the specification defining the first protocol, and inputs a second input to the generation AI that includes the first received packet, the prompt, and at least a portion of the specification defined by the second version, and the matching unit determines whether the first received packet is defined in the second version based on a response from the generation AI to the second input.

[0072] (4) Another embodiment of the present disclosure is a communication device according to any one of (1) to (3), wherein the AI ​​control unit stores the specification defined by the vectorized second version in a vector database.

[0073] (5) Another embodiment of the present disclosure is a communication device according to any one of (1) to (4), comprising: an interface unit that generates a response to a request to the communication device according to the first protocol; and a confirmation unit that generates a hypothetical request to the communication device that is defined in a version determined by the matching unit as the version in which the first received packet is defined, and causes the interface unit to generate a response to the hypothetical request.

[0074] (6) Another embodiment of the present disclosure is a communication device described in any one of (1) to (5), comprising an interface unit that generates a response to a request to the communication device via the first protocol, and the AI ​​control unit inputs a third input to the generation AI, the third input including a prompt inquiring about a recommended value for the response and at least a portion of a document defining an appropriate range of values ​​in the first protocol for the version in which the first received packet is defined and the version determined by the matching unit.

[0075] (7) Another embodiment of this disclosure is a communication device described in any one of (1) to (6), wherein the communication unit acquires a second received packet of a second protocol different from the first protocol and used for communication with the other communication device, the AI ​​control unit inputs a fourth input to the generation AI, the fourth input including the second received packet, a prompt inquiring about a definition of the second received packet, and at least a portion of a specification defined by a third version of the second protocol, and the matching unit determines whether the second received packet is defined in the third version based on a response from the generation AI to the fourth input.

[0076] (8) Another embodiment of the present disclosure is the communication device according to any one of (1) to (7), wherein the second protocol is a security protocol.

[0077] (9) Another embodiment of the present disclosure is a communication device according to any one of (1) to (8), comprising the generation AI.

[0078] (10) Another embodiment of the present disclosure is a communication device according to any one of (1) to (9), including a vector database that stores vectorized specifications that define any version of the first protocol.

[0079] (11) Another embodiment of the present disclosure is a communication method including: a first step of acquiring a first received packet of a first protocol used for communication with another communication device; a second step of inputting a first input to a generation AI (Artificial Intelligence), the first input including the first received packet, a prompt inquiring about a definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol; and a third step of determining whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

[0080] (12) Another embodiment of the present disclosure is a program for causing a computer to function as: a communication unit that acquires a first received packet of a first protocol used for communication with another communication device; an AI control unit that inputs a first input to a generation AI (Artificial Intelligence), the first input including the first received packet, a prompt inquiring about a definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol; and a matching unit that determines whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

[0081] 1, 5, 6, 7, and 8, a program for realizing the functions of the cloud service 100, edge device 200, middleware providing server 300, LLM 400, and vector DB 500 may be recorded on a computer-readable recording medium, and the program recorded on this recording medium may be read into a computer system and executed to realize the cloud service 100, edge device 200, middleware providing server 300, LLM 400, and vector DB 500. Note that the term "computer system" here includes hardware such as an OS and peripheral devices.

[0082] Furthermore, "computer system" also includes the homepage provision environment (or display environment) if the WWW system is used. Furthermore, "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into computer systems. Furthermore, "computer-readable recording medium" also includes devices that dynamically store programs for a short period of time, such as communication lines used when transmitting programs over networks like the Internet or communication lines like telephone lines, and devices that store programs for a fixed period of time, such as volatile memory within the computer systems that serve as servers or clients in such cases. Furthermore, the above-mentioned programs may be those that implement part of the aforementioned functions, or may be those that can realize the aforementioned functions in combination with programs already stored in the computer system.

[0083] The embodiments of this disclosure have been described in detail above with reference to the drawings, but the specific configuration is not limited to this embodiment, and design changes and the like are also included within the scope that does not deviate from the gist of this disclosure.

[0084] 10 Communication system 100 Cloud service 200 Edge device 210 Device unit 220 I / F unit 230 Control unit 231, 231b, 231c Collation unit 232, 232a, 232b, 232c AI control unit 233, 233a, 233b, 233c Confirmation unit 234, 234b, 234c Storage reference unit 235 LLM unit 236 Vector DB 240 Communication unit 300 Middleware providing server 400, 403, 404 LLM 500, 501, 502, 503, 504, 511, 512, 513, 514, 521, 531, 541, 551, 561, 562, 571, 581 Vector DB

Claims

1. A communications device comprising: a communications unit that acquires a first received packet of a first protocol used for communications with other communications devices; an AI control unit that inputs a first input to a generation AI (Artificial Intelligence), the first input including the first received packet, a prompt inquiring about the definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol; and a comparison unit that determines whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

2. The communication device according to claim 1, wherein at least a portion of the specification is a portion of the specification that is related to the first received packet and the prompt.

3. The communication device described in claim 1, wherein, when the matching unit determines that the first received packet is not defined in the first version, the AI ​​control unit obtains a specification defined by a second version of the first protocol that is different from the first version from a site that publishes the specification defining the first protocol, and inputs a second input to the generation AI that includes the first received packet, the prompt, and at least a portion of the specification defined by the second version, and the matching unit determines whether the first received packet is defined in the second version based on a response from the generation AI to the second input.

4. The communication device according to claim 3, wherein the AI ​​control unit stores the vectorized specifications defined by the second version in a vector database.

5. A communication device as described in claim 1, comprising: an interface unit that generates a response to a request to the communication device using the first protocol; and a confirmation unit that generates a provisional request to the communication device defined in a version determined by the matching unit to be the version in which the first received packet is defined, and causes the interface unit to generate a response to the provisional request.

6. A communication device as described in claim 1, comprising an interface unit that generates a response to a request made to the communication device via the first protocol, wherein the AI ​​control unit inputs to the generation AI a third input including a prompt inquiring about a recommended value for the response and at least a portion of a document defining an appropriate range of values ​​in the first protocol for the version in which the first received packet is defined and the version determined by the matching unit.

7. The communication device described in claim 1, wherein the communication unit acquires a second received packet of a second protocol different from the first protocol and used for communication with the other communication device; the AI ​​control unit inputs a fourth input to the generation AI, the fourth input including the second received packet, a prompt inquiring about a definition of the second received packet, and at least a portion of a specification defined by a third version of the second protocol; and the matching unit determines whether the second received packet is defined in the third version based on a response from the generation AI to the fourth input.

8. The communication device of claim 7, wherein the second protocol is a security protocol.

9. A communication device according to any one of claims 1 to 8, comprising the generating AI.

10. A communication device according to any one of claims 1 to 8, comprising a vector database for storing vectorized specifications defining any version of said first protocol.

11. A communication method comprising: a first step of acquiring a first received packet of a first protocol used for communication with another communication device; a second step of inputting a first input to a generation AI (Artificial Intelligence), the first input including the first received packet, a prompt inquiring about the definition of the first received packet, and at least a portion of a specification defined by a first version of the first protocol; and a third step of determining whether the first received packet is defined in the first version based on a response from the generation AI to the first input.

12. A program for causing a computer to function as: a communication unit that acquires a first received packet of a first protocol used for communication with other communication devices; an AI control unit that inputs a first input to a generation AI (Artificial Intelligence) that includes the first received packet, a prompt inquiring about the definition of the first received packet, and at least a portion of the specification defined by a first version of the first protocol; and a matching unit that determines whether the first received packet is defined in the first version based on the response from the generation AI to the first input.

Images