Data transmission methods, base station equipment and mobile devices

Abstract

This application discloses a data transmission method, a base station device, and a mobile device. The method includes: receiving a sent semantic understanding request; determining data to be transmitted based on the semantic understanding request and the cached knowledge base of the base station device; and sending the data to be transmitted.

Description

Technical Field

[0001] This application relates to, but is not limited to, the field of semantic communication, and in particular to a data transmission method, base station equipment, and mobile device. Background Technology

[0002] Driven by the rapid development of artificial intelligence and wireless technology, the architecture of future mobile communication networks will transform into a form of interconnected intelligence. Semantic communication can alleviate information flow redundancy and reduce communication latency by compressing data, thereby improving system efficiency.

[0003] The implementation of semantic communication requires a semantic knowledge base and a computationally capable neural network model. Since a primary prerequisite for semantic communication is ensuring that transceivers have similar background knowledge, the semantic knowledge base is the cornerstone of semantic communication. However, in communication networks, device computing and caching resources are limited, and the enormous pressure that the deployment and generation of the knowledge base places on these limited resources cannot be ignored. Summary of the Invention

[0004] In view of this, the present application provides a data transmission method and apparatus to solve the problems existing in the prior art.

[0005] The technical solution of this application embodiment is implemented as follows:

[0006] In a first aspect, embodiments of this application provide a data transmission method, including:

[0007] Receive a semantic understanding request; determine the data to be transmitted based on the semantic understanding request and the cached knowledge base of the base station device; and send the data to be transmitted.

[0008] Secondly, embodiments of this application provide a data transmission method, including:

[0009] Send a scheduling request; receive scheduling information sent in response to the scheduling request; in response to the scheduling information, send a semantic understanding request; receive the data to be transmitted corresponding to the semantic understanding request.

[0010] Thirdly, embodiments of this application provide a data transmission apparatus, the apparatus comprising:

[0011] First acquisition module; first determination module; second determination module; first adjustment module; first display module.

[0012] Fourthly, embodiments of this application provide a data transmission apparatus, the apparatus comprising:

[0013] First acquisition module; first determination module; second determination module; first adjustment module; first display module.

[0014] This application provides a base station device for data transmission, including a memory and a processor. The memory stores a computer program that can run on the processor, and the processor executes the program to implement the steps in the method described in the first aspect.

[0015] This application provides a mobile device for data transmission, including a memory and a processor. The memory stores a computer program that can run on the processor, and the processor executes the program to implement the steps in the method described in the second aspect.

[0016] This application provides an electronic device, which includes a processor, a memory, and a communication bus; wherein the communication bus is used to establish a communication connection between the processor and the memory; and the processor is used to execute a program in the memory to implement the method described in the first or second aspect above.

[0017] Correspondingly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps described in the first or second aspect above.

[0018] In this embodiment, the base station receives a semantic understanding request from a mobile device. By combining the knowledge graph corresponding to the semantic understanding request with the cache space of the base station's cached knowledge base, the base station determines the data to be transmitted to the mobile device. Finally, the base station sends the data to be transmitted to the mobile device. Thus, after receiving the semantic understanding request from the mobile device, considering that the device's computing and caching resources are limited, the base station can adaptively adjust the data to be transmitted to the mobile device based on its current cache space and the knowledge graph corresponding to the semantic understanding request. This maximizes the value of caching while reducing the time and computing resource consumption associated with caching. Attached Figure Description

[0019] In the accompanying drawings (which are not necessarily drawn to scale), similar reference numerals may describe similar parts in different views. Similar reference numerals with different letter suffixes may indicate different examples of similar parts. The drawings illustrate, by way of example and not limitation, the various embodiments discussed herein.

[0020] Figure 1A A schematic diagram illustrating the implementation flow of a data transmission method provided in an embodiment of this application;

[0021] Figure 1B This is a schematic diagram illustrating another implementation flow of a data transmission method provided in an embodiment of this application;

[0022] Figure 1C This is a schematic diagram illustrating another implementation process of a data transmission method provided in an embodiment of this application;

[0023] Figure 2 A schematic diagram illustrating the interaction process between a mobile device and a base station device in a data transmission method provided in this application embodiment;

[0024] Figure 3 This is a schematic diagram illustrating another implementation process of a data transmission method provided in an embodiment of this application;

[0025] Figure 4 A schematic diagram illustrating the implementation of the startup time window provided in this application embodiment;

[0026] Figure 5 A schematic diagram illustrating the implementation framework of the data transmission method provided in this application embodiment;

[0027] Figure 6 This is a schematic diagram illustrating another implementation process of a data transmission method provided in an embodiment of this application;

[0028] Figure 7 This is a schematic diagram illustrating another implementation process of a data transmission method provided in an embodiment of this application;

[0029] Figure 8 This is a schematic diagram illustrating another implementation process of a data transmission method provided in an embodiment of this application;

[0030] Figure 9A This is a schematic diagram of the composition structure of the data transmission device provided in the embodiments of this application;

[0031] Figure 9B This is a schematic diagram of the composition structure of the data transmission device provided in the embodiments of this application;

[0032] Figure 10 This is a schematic diagram of the hardware entity of the electronic device provided in the embodiments of this application. Detailed Implementation

[0033] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0034] In the following description, the use of suffixes such as "module," "part," or "unit" to denote elements is solely for the purpose of illustration and has no specific meaning in itself. Therefore, "module," "part," or "unit" may be used interchangeably.

[0035] Electronic devices can be implemented in various forms. For example, the electronic devices described in this application may include mobile electronic devices such as personal digital assistants (PDAs), navigation devices, and wearable devices, as well as fixed electronic devices such as digital TVs and desktop computers that can collect fingerprints.

[0036] The following description will use mobile devices or base station devices as examples. Those skilled in the art will understand that, in addition to components specifically designed for mobile purposes, the construction according to the embodiments of this application can also be applied to fixed-type electronic devices.

[0037] Based on this, embodiments of this application provide a data transmission method applicable to base station equipment. This method uses the knowledge graph corresponding to the received semantic understanding request as the main content of a cached knowledge base, and deploys it in the edge cache of the base station equipment according to the current cache space status of the cached knowledge base. This maximizes the value of the cache while reducing the time and computational resource consumption caused by caching. In embodiments of this application, this data transmission method can be executed by the processor of a computer device. Figure 1A This is a schematic diagram illustrating the implementation flow of a data transmission method provided in an embodiment of this application. This method can be implemented by the sending end, such as... Figure 1A As shown, the method includes the following steps S101 to S103:

[0038] Step S101: Receive the sent semantic understanding request.

[0039] Here, when the mobile device needs to perform semantic communication, it sends a semantic understanding request to the base station device. This semantic understanding request is used to request information needed for semantic communication from the base station device, such as the knowledge graph required for semantic communication. The base station device receives the semantic understanding request sent by the mobile device.

[0040] Step S102: Based on the semantic understanding request and the cached knowledge base of the base station device, determine the data to be transmitted.

[0041] Here, the cached knowledge base is a semantic knowledge base used by the base station equipment to cache the knowledge graph required for semantic communication. The base station equipment determines the data that needs to be transmitted to the mobile device based on the semantic understanding request sent by the mobile device and the current cache space status of its own cached knowledge base.

[0042] In some possible implementations, the data to be transmitted includes a knowledge graph required by the mobile device for semantic communication.

[0043] Step S103: Send the data to be transmitted.

[0044] Here, after determining the transmission data that needs to be sent to the mobile device, the base station equipment sends the transmission data to the mobile device.

[0045] In some possible implementations, mobile devices can determine the corresponding communication method based on the different data to be transmitted, such as semantic communication or non-semantic communication; or they can choose to remain silent and wait for the base station to send the next data to be transmitted.

[0046] In this embodiment, the base station receives a semantic understanding request from a mobile device. By combining the knowledge graph corresponding to the semantic understanding request with the cache space of the base station's cached knowledge base, the base station determines the data to be transmitted to the mobile device. Finally, the base station sends the data to be transmitted to the mobile device. Thus, after receiving the semantic understanding request from the mobile device, considering that the device's computing and caching resources are limited, the base station can adaptively adjust the data to be transmitted to the mobile device based on its current cache space and the knowledge graph corresponding to the semantic understanding request. This maximizes the value of caching while reducing the time and computing resource consumption associated with caching.

[0047] In some embodiments, after receiving a semantic understanding request, the base station device analyzes the knowledge graph and cache status of the cached knowledge base corresponding to the semantic understanding request to accurately determine the data to be transmitted. That is, step S102 above can be achieved through... Figure 1B The steps shown are to be implemented as follows:

[0048] Step S121: Determine the knowledge graph corresponding to the semantic understanding request.

[0049] Here, since semantic communication requires a corresponding knowledge graph to understand the semantics, the mobile device needs to send a semantic understanding request to the base station device to request the corresponding knowledge graph when performing semantic communication. The base station device determines the corresponding knowledge graph required by the mobile device based on the semantic understanding request sent by the mobile device.

[0050] Step S122: Based on the knowledge graph and the cached knowledge base, determine the data to be transmitted corresponding to the semantic understanding request.

[0051] Here, the cached knowledge base is used by the base station equipment to cache the knowledge graph used for semantic communication. After determining the knowledge graph required by the mobile device, the base station equipment calculates the resource consumption required to cache the knowledge graph and the current cache space status of the base station equipment's cached knowledge base to determine the data to be transmitted to the mobile device.

[0052] In some possible implementations, the data to be transmitted may include information about the current cache space status of the base station device's cached knowledge base, i.e., whether the base station device's cached knowledge base contains prompts about the knowledge graph required by the mobile device, or the knowledge graph required by the mobile device.

[0053] In this embodiment, the base station device determines whether the cached knowledge base contains the knowledge graph required by the mobile device by comparing the knowledge graph corresponding to the semantic understanding request sent by the mobile device with the cache space in the base station device's cached knowledge base. This enables targeted data transmission to the mobile device based on different cache conditions, improving communication efficiency and ensuring accurate information transmission to the mobile device.

[0054] In some embodiments, in step S122 above, the data to be transmitted can be determined in the following two ways:

[0055] Method 1: If the knowledge graph is already cached in the cached knowledge base, determine that the knowledge graph in the cached knowledge base is the data to be transmitted.

[0056] Here, if the base station's cached knowledge base already contains the knowledge graph corresponding to the semantic understanding request sent by the mobile device, it means the base station can directly send the required knowledge graph to the mobile device. Therefore, the base station directly recognizes the data carrying the knowledge graph as the data to be transmitted.

[0057] Correspondingly, after determining that the knowledge graph in the cached knowledge base is the data to be transmitted, the data to be transmitted can be sent through the following process:

[0058] If the knowledge graph is already cached in the cached knowledge base, a first response carrying the knowledge graph is sent in the first time window.

[0059] Here, the first time window is the time period that the mobile device initiates simultaneously when sending a semantic understanding request, used to receive information from the base station device. Once the base station device confirms that its cached knowledge base has cached the knowledge graph corresponding to the semantic understanding request sent by the mobile device, it will send the first response carrying the knowledge graph required by the mobile device to the mobile device within the first time window of the mobile device.

[0060] In some possible implementations, the first response includes the knowledge graph required by the mobile device and a hint that the knowledge graph is cached in the cached knowledge base of the base station device.

[0061] Method 2: First, if the knowledge graph is not cached in the cached knowledge base, search for the knowledge graph in the preset cloud knowledge base.

[0062] Here, when the base station device confirms that it does not have a cached knowledge graph corresponding to the semantic understanding request sent by the mobile device in its own cached knowledge base, the base station device will search for the knowledge graph in the preset cloud knowledge base of the core network.

[0063] In some possible implementations, the pre-built cloud knowledge base includes a semantic knowledge graph generated from historical device request data using knowledge graph construction technology; for the deployment of a new pre-built cloud knowledge base, it can also be generated by loading from the core network.

[0064] The second step is to cache the knowledge graph in the cached knowledge base in response to finding the knowledge graph.

[0065] Here, when the base station device finds the knowledge graph required by the mobile device in the cloud knowledge base, the base station device will cache the found knowledge graph in its own cached knowledge base so that the mobile device can obtain the required knowledge graph from the base station device.

[0066] The third step is to identify the knowledge graph in the updated cached knowledge base as the data to be transmitted.

[0067] Here, after the base station device caches the knowledge graph required by the mobile device from the cloud knowledge base in its own cached knowledge base, the base station device's cached knowledge base has been updated and now contains the knowledge graph. Therefore, the base station device can determine that the knowledge graph in the updated cached knowledge base is the data that the base station device needs to transmit to the mobile device.

[0068] Correspondingly, after determining that the knowledge graph in the updated cached knowledge base is the data to be transmitted, the data to be transmitted can be sent through the following process:

[0069] If the knowledge graph is not cached in the cached knowledge base, a first response carrying the knowledge graph is sent within the second time window.

[0070] Here, the second time window is initiated by the mobile device after the first time window ends, following a preset duration, to receive information from the base station device. This preset duration is set based on the time required for the base station device to cache the knowledge graph corresponding to the semantic understanding request sent by the mobile device from the cloud knowledge base. If the base station device's cached knowledge base does not contain the required knowledge graph, after caching the knowledge graph in the cloud knowledge base, the base station device will send a first response carrying the knowledge graph to the mobile device within the second time window initiated by the mobile device. The first time window begins at the moment the semantic understanding request is sent; the second time window is initiated if no first response is received within the first time window and the preset duration following the first time window.

[0071] In this embodiment, the base station device first determines whether the knowledge graph corresponding to the semantic understanding request has been cached in its own cached knowledge base. If the base station device has cached the knowledge graph required by the mobile device in its own cached knowledge base, it can directly identify the knowledge graph as the data to be transmitted and send a first response carrying the knowledge graph to the first time window initiated by the mobile device. If the base station device has not cached the knowledge graph in its cached knowledge base, the base station device first queries the cloud knowledge base for the knowledge graph. After finding it, it stores the knowledge graph in the cached knowledge base and sends a first response carrying the knowledge graph to the second time window of the mobile device. This allows the base station device to respond as quickly as possible in different cached knowledge base cache space states, and to send the knowledge graph corresponding to the received semantic understanding request to the mobile device quickly and accurately.

[0072] In some embodiments, when the mobile device initiates the second time window, it sends a semantic understanding request to the base station device again. After receiving the semantic understanding request again, the base station device sends a first response to the mobile device. This process includes:

[0073] In response to receiving the semantic understanding request again when the second time window is initiated, a first response carrying the knowledge graph in the updated cached knowledge base is sent.

[0074] Here, the mobile device does not receive the first response within the first time window, and the first time window ends; after a preset duration, it sends a semantic understanding request to the base station device again and starts a second time window; at this time, the base station device caches the knowledge graph in the cloud knowledge base and receives the semantic understanding request sent by the mobile device again; therefore, the base station device will send a first response carrying the knowledge graph to the mobile device within the second time window started by the mobile device, based on the semantic understanding request sent by the mobile device again.

[0075] In some possible implementations, if the mobile device does not receive the first response within the first time window and is in a silent state for a preset duration, and if the base station device has already cached the knowledge graph required by the mobile device in the cloud knowledge base, then the base station device will directly send the first response carrying the knowledge graph to the mobile device.

[0076] In this embodiment, if the base station device does not cache the knowledge graph required by the mobile device in its own cached knowledge base, it will query the knowledge graph from the cloud knowledge base. Upon finding the knowledge graph, it will cache it in its own cached knowledge base and send a first response carrying the knowledge graph to the second time window after the mobile device starts. If the mobile device does not receive the first response within the first time window and remains silent for a preset duration, but the base station device has already cached the knowledge graph in the cloud knowledge base, it can directly send the first response to the mobile device. This method achieves targeted caching based on the different cache space states of the base station device's cached knowledge base and ensures timely delivery of the data required by the mobile device. This effectively improves transmission efficiency while avoiding waste of cache resources.

[0077] In some embodiments, if the knowledge graph is not cached in the cached knowledge base, in response to finding the knowledge graph in the cloud knowledge base, caching the knowledge graph in the cached knowledge base by analyzing cache resource consumption can be achieved through the following steps:

[0078] The first step is to determine the cache resource consumption of the knowledge graph corresponding to the semantic understanding request in the base station equipment based on the identification information.

[0079] Here, the identification information can be identity information such as the name or serial number of the mobile device. Cache resource consumption refers to the resource consumption required by the base station device in the process of caching the knowledge graph from the cloud knowledge base, including computing costs and cache space. The base station device uses the identification information of the mobile device to determine the specific details of the request sent by the mobile device to the base station device, and determines the cache resource consumption required for the knowledge graph corresponding to the semantic understanding request sent by the mobile device based on different circumstances.

[0080] In some possible implementations, the first step above can be achieved through the following process:

[0081] First, based on the identification information, it is determined whether the mobile device is sending a semantic understanding request for the first time, and the determination result is obtained.

[0082] Here, the base station first searches for the mobile device's identification information and then checks its historical request records to see if there is a record of the mobile device making a request. If the base station already has a request record for the mobile device's identification information, it means that this is not the first time the mobile device has sent a request to the base station; otherwise, it can be determined that this is the first time the mobile device has sent a request to the base station.

[0083] Then, based on the judgment results, the cache resource consumption of the knowledge graph subgraph in the base station equipment is determined.

[0084] Here, the cache resource consumption of knowledge graph subgraphs in base station equipment can be determined in the following two ways:

[0085] Method 1:

[0086] Step A: In response to the first time the mobile device sends the semantic understanding request, determine the amount of data in the knowledge graph subgraph corresponding to the semantic understanding request.

[0087] Here, the data volume of a knowledge graph subgraph includes the cache resource consumption required during the caching process when multiple subgraphs in the knowledge graph are jointly cached. When the base station device determines, based on the judgment result, that the mobile device is sending a semantic understanding request to the base station device for the first time, the base station device will calculate and determine the data volume of multiple subgraphs in the knowledge graph corresponding to the semantic understanding request sent by the mobile device when jointly cached.

[0088] Step B involves determining the data volume of the knowledge graph subgraph as the cache resource consumption of the knowledge graph subgraph in the base station equipment.

[0089] Here, the base station device determines the amount of data in the calculated knowledge graph subgraph as the cache resource consumption required when the knowledge graph corresponding to the semantic understanding request sent by the mobile device is cached in the base station device.

[0090] Method 2:

[0091] Step C: In response to the fact that the semantic understanding request is not being sent for the first time by the mobile device, determine the historical cache resource consumption corresponding to the last time the semantic understanding request was sent by the mobile device.

[0092] Here, historical cache resource consumption refers to the cache resource consumption of the knowledge graph corresponding to the last semantic understanding request sent by the mobile device to the base station before sending this semantic understanding request. When the base station determines that this is not the first time the mobile device has sent a semantic understanding request, it will query the historical request record of the mobile device's last request to determine the historical cache resource consumption of the knowledge graph corresponding to the last semantic understanding request sent by the mobile device to the base station.

[0093] Step D: The historical cache resource consumption is determined as the cache resource consumption corresponding to the semantic understanding request sent by the mobile device this time.

[0094] Here, the base station device will determine the historical cache resource consumption of the knowledge graph corresponding to the semantic understanding request sent by the mobile device to the base station device in the last request record in the historical request record, and determine the historical cache resource consumption as the cache resource consumption of the knowledge graph corresponding to the semantic understanding request sent by the mobile device in this time.

[0095] The second step is to cache the knowledge graph in the cached knowledge base based on the cache resource consumption.

[0096] Here, the base station device obtains the cache resource consumption by considering different request scenarios of the mobile device and different methods for determining cache resource consumption. The base station device will then determine how to cache the knowledge graph corresponding to the semantic understanding request sent by the mobile device based on the specific circumstances of different cache resource consumption, so as to cache the knowledge graph required by the mobile device in the cache knowledge base of the base station device.

[0097] In this embodiment, when the knowledge graph required by a mobile device is not cached in the base station's cached knowledge base, but is found in the cloud knowledge base, the base station determines the request sent by the mobile device by searching the identification information of the mobile device that sent the semantic understanding request. Two different methods are proposed to determine the cache resource consumption required when caching the knowledge graph corresponding to the semantic understanding request sent by the mobile device in the base station. The obtained cache resource consumption information is used to determine how to cache the knowledge graph corresponding to the semantic understanding request, so that the knowledge graph required by the mobile device is cached in the base station's cached knowledge base. This approach enables targeted processing of semantic understanding requests sent by mobile devices under different circumstances, avoiding repeated calculations of cache resource consumption for the knowledge graph corresponding to the semantic understanding request when the mobile device sends a request to the base station before, thus improving resource utilization and communication efficiency.

[0098] In some embodiments, by comparing the cache resource consumption with a preset consumption threshold, different caching strategies are adopted to cache the knowledge graph in the cached knowledge base. That is, the second step above, "caching the knowledge graph in the cached knowledge base based on cache resource consumption," can be implemented in the following two ways:

[0099] Method 1: When the cache resource consumption is less than a preset consumption threshold, based on the cached knowledge graph in the cache knowledge base and the knowledge graph corresponding to the semantic understanding request, cache the knowledge graph corresponding to the semantic understanding request in the cache knowledge base.

[0100] Here, when the cache resource consumption is less than a preset consumption threshold, the dynamic programming caching strategy for the subgraph is determined as the target caching strategy; correspondingly, based on the target caching strategy, the knowledge graph subgraph is cached in the cache knowledge base of the base station device. This method includes the following steps:

[0101] Step 1: Sort the multiple knowledge graph subgraphs corresponding to the semantic understanding request according to the first preset order to obtain the sorted subgraphs.

[0102] The first preset order is the reverse order of the sub-image sizes.

[0103] Step 2: Traverse the sorted subgraphs according to the first preset order, and traverse each subgraph in the sorted subgraphs.

[0104] Step 3: Select the cached subgraph that has the same triple as the current subgraph being traversed from the cached knowledge base of the base station device.

[0105] A knowledge graph contains multiple subgraphs, and triples are the components of each subgraph in the knowledge graph. Triples may be repeated in different knowledge graphs. Cached subgraphs refer to multiple subgraphs contained in the historical cached knowledge graph in the cached knowledge base of the base station device.

[0106] Step 4: Based on the current subgraph and the cached subgraph, cache the current subgraph in the cache knowledge base of the base station device.

[0107] In some feasible ways, the subgraph in the currently traversed knowledge graph can be compared with the subgraph of the knowledge graph already cached in the base station device's cached knowledge base, so that the current subgraph is cached in the base station device's cached knowledge base.

[0108] Method 2: If the cache resource consumption exceeds the preset consumption threshold, determine whether to cache the knowledge graph in the cache knowledge base based on the cache space required by the knowledge graph and the cache capacity of the cache knowledge base.

[0109] Here, when the cache resource consumption exceeds a preset consumption threshold, the greedy cache of triples is determined as the target cache measurement; correspondingly, based on the target cache strategy, the knowledge graph subgraph is cached in the cache knowledge base of the base station device. This method includes the following steps:

[0110] Step 5: Sort the triples of multiple knowledge graph subgraphs corresponding to the semantic understanding request according to the second preset order to obtain a list of sorted triples.

[0111] The second preset order is the order of the value density of triples; a triple is a component of multiple subgraphs contained in a knowledge graph.

[0112] Step 6: According to the second preset order, traverse the sorted triplet list and iterate through each triplet in the sorted triplet list.

[0113] Step 7: Determine the cache space required for the traversed triples and the current traversed triple; wherein, the cache space includes the cache space required by all traversed triples in the sorted triple list.

[0114] Step 8: In response to the cache space being less than the cache capacity of the cache knowledge base, the current triplet being traversed is cached in the cache knowledge base; when the currently obtained cache space, that is, the cache space required by all the triplets that have been traversed in the sorted triplet list, is less than the cache capacity of the cache knowledge base of the base station device, the triplet being traversed is cached in the knowledge base.

[0115] In this embodiment, the base station device employs different caching methods for the knowledge graphs corresponding to semantic understanding requests sent by mobile devices in different scenarios. When the caching resource consumption of the knowledge graph is less than a preset value, a subgraph-based dynamic programming caching method is adopted. Multiple knowledge graph subgraphs corresponding to the semantic understanding request are sorted according to a first preset order, and the sorted subgraphs are traversed. A cached subgraph whose triples are identical to those of the traversed current subgraph is selected from the base station device's cached knowledge base. The current subgraph is then cached in the base station device's cached knowledge base. By using the subgraph-based dynamic programming caching method, the caching value is maximized while reducing the time and computational resource consumption caused by caching. To address the curse of dimensionality issue that may arise with large-scale caching when cache resource consumption exceeds a threshold, a greedy triple-based caching method is employed. This method sorts the triples of multiple knowledge graph subgraphs corresponding to a semantic understanding request according to a second preset order. It then iterates through the sorted triple list to determine the cache space required for each traversed triple and the current traversed triple. If the cache space is less than the cache capacity of the knowledge base, the current traversed triple is cached in the knowledge base. This greedy triple-based caching method shortens the deployment and computation time of large-scale knowledge bases, maximizing the performance of cached subgraphs with limited resources.

[0116] In some embodiments, the base station device also receives a scheduling request before receiving the transmitted semantic understanding request. For example... Figure 1C The above steps S101 include steps S131 and S132:

[0117] Step S131: Receive the sent scheduling request.

[0118] Here, a scheduling request is used to request the base station equipment to send scheduling information. The base station equipment receives and analyzes the scheduling request sent by the mobile device.

[0119] Step S132: In response to the scheduling request, send scheduling information.

[0120] Here, the scheduling information includes indications such as the sending location and time that the mobile device uses when sending information to the base station. Only after the base station sends the scheduling information to the mobile device can the mobile device send a request to the base station according to the indication. After receiving the scheduling request from the mobile device, the base station will send scheduling information to the mobile device according to the scheduling request.

[0121] In this embodiment, the mobile device sends a scheduling request to the base station to obtain scheduling information from the base station. Upon receiving the scheduling request, the base station sends corresponding scheduling information as needed, thereby making data transmission between the mobile device and the base station more orderly and efficient, and improving data transmission security.

[0122] In some embodiments, when the base station device has no cache space to cache the knowledge graph, a corresponding prompt message will be sent to the mobile device, that is, step S102 above further includes:

[0123] If there is no remaining cache space in the cached knowledge base to store the knowledge graph, the prompt message used to initiate non-semantic communication is determined to be the data to be transmitted.

[0124] Here, non-semantic communication includes various communication methods other than semantic communication that do not require a knowledge graph; the prompt message is used to prompt the mobile device to initiate other communication methods. If the base station device's cached knowledge base does not cache the knowledge graph corresponding to the semantic understanding request sent by the mobile device, and there is no remaining cache space to cache the knowledge graph from the cloud knowledge base, the base station device will send a prompt message to the mobile device to initiate non-semantic communication.

[0125] Correspondingly, sending the data to be transmitted includes sending the prompt message.

[0126] Here, the base station equipment completes the process of sending a prompt message to the mobile device to initiate non-semantic communication.

[0127] In this embodiment, when the base station device's cached knowledge base lacks cache space to store the knowledge graph, the data transmitted from the base station device to the mobile device will include a prompt message to initiating non-semantic communication. This alerts the mobile device that the base station device has no cache space to store the required knowledge graph, allowing it to initiate non-semantic communication promptly and avoiding prolonged waiting that could lead to missed communication opportunities and wasted resources.

[0128] In some embodiments, after the base station device caches the knowledge graph in a non-cached space and sends the corresponding prompt information to the mobile device, the base station device also needs to check its own cache space according to a preset period. This method can be implemented through the following process:

[0129] The remaining cache space of the cached knowledge base is detected according to a preset period, and the detection result is obtained; the detection result is sent in the form of signaling according to the preset period.

[0130] Here, the preset period can be set based on the base station's cache status and the knowledge graph corresponding to the semantic understanding request sent by the mobile device. After sending a prompt message to the mobile device to initiate non-semantic communication, the base station will check the remaining cache space of its own cached knowledge base according to the preset period, and send the detection results to the mobile device in the form of signaling according to the preset period, so that the mobile device can understand the cache space status of the base station in a timely manner.

[0131] In this embodiment, after the base station device prompts the mobile device that there is no cache space in its cached knowledge base to store the knowledge graph, the base station device will periodically check the cache space of its own cached knowledge base and send the detection results to the mobile device via signaling. This allows the mobile device to know in real time whether the base station device has storage space to continue storing the knowledge graph, even before non-semantic communication is initiated. This avoids the situation where the base station device sends a prompt to the mobile device, and the cache space is updated and sufficient to cache the knowledge graph, but it cannot provide the required knowledge graph to the mobile device, thus maximizing the value of caching and reducing computational resource consumption.

[0132] This application provides a data transmission method applicable to mobile devices, which involves sending a scheduling request to request the receipt of scheduling information sent by a base station device. In this application embodiment, the data transmission method can be executed by the processor of a computer device. Figure 2 This application provides a schematic diagram of the interaction process between a mobile device and a base station device in a data transmission method according to an embodiment of the present application. Figure 2 As shown, the method includes the following steps:

[0133] Step S201: The mobile device sends a scheduling request.

[0134] Here, the mobile device sends a scheduling request to the base station device. This scheduling request is used to request the base station device to send scheduling information to the mobile device.

[0135] In step S202, the base station equipment receives the dispatch request sent.

[0136] In step S203, the base station device responds to the scheduling request by sending scheduling information; wherein, the scheduling information includes indication information such as the sending location and sending time that the mobile device should rely on when sending information to the base station device. Only after the base station device sends the scheduling information to the mobile device can the mobile device send a request to the base station device according to the indication.

[0137] Step S204: The mobile device receives scheduling information in response to the scheduling request.

[0138] Here, the mobile device receives scheduling information sent by the base station based on a scheduling request. Guided by this scheduling information, the mobile device can send semantic understanding requests to the base station according to the actual semantic communication needs.

[0139] In step S205, the mobile device responds to the scheduling information by sending a semantic understanding request.

[0140] Here, after receiving and analyzing the scheduling information, the mobile device, following the instructions in the scheduling information, sends a semantic understanding request, which it needs for semantic communication, to the base station. This semantic understanding request is used to request the knowledge graph required by the mobile device during semantic communication from the base station.

[0141] Step S206: The base station device receives the sent semantic understanding request.

[0142] In step S207, the base station device determines the data to be transmitted based on the semantic understanding request and the base station device's cached knowledge base.

[0143] Here, the data to be transmitted may include prompts to the mobile device and a knowledge graph corresponding to the semantic understanding request sent by the mobile device. The base station device analyzes the semantic understanding request and, in conjunction with the current cache space of its own cached knowledge base, determines the data to be transmitted to the mobile device.

[0144] Step S208: The base station device sends the data to be transmitted.

[0145] Step S209: The mobile device receives the data to be transmitted corresponding to the semantic understanding request.

[0146] Here, after the mobile device sends a semantic understanding request to the base station device, the base station device will send the data to be transmitted to the mobile device according to the current cache space status of its own cached knowledge base. The mobile device will then start the corresponding time window to receive the data from the base station device.

[0147] In this embodiment, the mobile device sends a scheduling request to the base station to request the receipt of scheduling information from the base station. Upon receiving the scheduling information, the mobile device, following the instructions in the scheduling information, sends the knowledge graph it needs for semantic communication to the base station as a semantic understanding request. Furthermore, the mobile device initiates a corresponding time window to receive the data to be transmitted from the base station corresponding to the semantic understanding request. This method ensures orderly and efficient data transmission between the mobile device and the base station, improving data transmission security.

[0148] In some embodiments, when a mobile device receives data to be transmitted from a base station device, it may initiate a first time window or a second time window to receive the data to be transmitted. That is, step S209 described above can be implemented through the following steps:

[0149] The first step is to start a first time window with the moment the semantic understanding request is sent as the starting moment, and to receive the data to be transmitted within the first time window.

[0150] Here, the data to be transmitted includes the knowledge graph corresponding to the semantic understanding request sent by the mobile device. When the mobile device sends a semantic understanding request to the base station device, it simultaneously activates a first time window and waits within the first time window to receive the data to be transmitted from the base station device.

[0151] The second step is to respond to the fact that the data to be transmitted is not received within the first time window, and to send the semantic understanding request again after a preset time interval and start the second time window; and to receive the data to be transmitted within the second time window.

[0152] Here, if the mobile device does not receive the data to be transmitted from the base station within the first time window, i.e., it does not receive the knowledge graph from the base station, the mobile device will send a semantic understanding request to the base station again after a preset time period, and simultaneously start a second time window. When the mobile device starts the second time window, the base station will send the data to be transmitted to the mobile device within the second time window, and the mobile device will receive the data to be transmitted within the second time window.

[0153] In this embodiment, the mobile device initiates a first time window starting from the moment it sends the semantic understanding request, and receives the data to be transmitted within the first time window. If the data to be transmitted is not received within the first time window, the mobile device resends the semantic understanding request after a preset interval and initiates a second time window; the data to be transmitted is then received within the second time window. This enables the timely and accurate transmission of the knowledge graph required by the mobile device to the mobile device under different cache space states of the cached knowledge bases of various base station devices, improving communication efficiency.

[0154] In some embodiments, after the mobile device initiates a first time window or a second time window, it may initiate non-semantic communication, i.e., the above process includes:

[0155] If the data to be transmitted is not received within the first time window or the second time window, non-semantic communication is initiated.

[0156] Here, non-semantic communication includes various communication methods other than semantic communication that do not require a knowledge graph. The data to be transmitted includes a first response and a second response carrying the knowledge graph corresponding to the semantic understanding request. When the mobile device starts the first or second time window, if the mobile device does not receive the data to be transmitted from the base station device within that window, that is, it does not receive the knowledge graph corresponding to the semantic understanding request, then the mobile device will start non-semantic communication to complete the communication process as appropriate.

[0157] In some possible implementations, after a mobile device initiates non-semantic communication, it will periodically detect signaling from the base station. This signaling contains the base station's detection results regarding its own cache space status. Based on the information contained in the signaling, the mobile device determines the current cache space status of the base station and decides whether to send a semantic understanding request to the base station again, depending on the actual situation.

[0158] In this embodiment, if the mobile device does not receive the data to be transmitted from the base station device in either the first or second time window (i.e., it does not receive the first and second responses from the base station device carrying the knowledge graph corresponding to the semantic understanding request), the mobile device will directly initiate non-semantic communication. Simultaneously, the mobile device can detect signaling from the base station device at a preset period, determine the current cache space status of the base station device based on the detection results contained in the signaling, and decide whether to send the semantic understanding request to the base station device again based on the actual situation. This allows the mobile device to successfully complete the communication process through non-semantic communication even when the base station device cannot send the knowledge graph required by the mobile device, improving communication efficiency. Furthermore, the mobile device can also understand the current cache space status of the base station device in real time and determine whether to send the semantic understanding request to the base station device again to initiate semantic communication, increasing the optional communication methods during the communication process and improving communication quality.

[0159] The following describes the application of the data transmission method provided in the embodiments of this application in a real-world scenario, taking the interactive communication between a base station device and a mobile device as an example.

[0160] This application provides a schematic diagram of the implementation process of a data transmission method, which can be implemented by a base station device and a mobile device, such as... Figure 3 As shown, the method includes the following steps S301 to S304:

[0161] Step S301: The mobile device sends a scheduling request to the base station device.

[0162] Here, the scheduling request sent by the mobile device can be a physical layer signaling, such as corresponding to a specific sequence at the physical layer, or it can be a MAC layer signaling. The scheduling request is sent in the periodically configured time-frequency resources.

[0163] Step S302: The base station equipment receives a scheduling request.

[0164] In step S303, the mobile device sends a semantic understanding request to the base station device.

[0165] Here, the mobile device receives scheduling information from the base station and sends a semantic understanding request to the base station based on this information.

[0166] Step S304: The mobile device initiates a first-time window to receive feedback information from the base station device. Here, the mobile device initiates a first-time window to receive feedback information from the base station device. This feedback information includes the following three types:

[0167] Option 31: The mobile device receives the knowledge graph subgraph sent by the base station device (the base station device has already cached the knowledge graph subgraph).

[0168] Here, the first option 31 carries the knowledge graph required by the mobile device, corresponding to the first response in the above embodiment. The first option 31 indicates that the base station device has cached the knowledge graph, and the mobile device can receive the knowledge graph.

[0169] Option 32: The mobile device ends the first time window and starts the second time window to receive the knowledge graph (the knowledge graph has been cached in the cloud knowledge base).

[0170] Here, option 32 is for cases where the base station device does not cache the knowledge graph, requiring the mobile device to close the first time window and start the second window to receive the knowledge graph; wherein, the mobile device receives feedback information carrying the knowledge graph sent by the base station device in the started second time window, and this feedback information corresponds to the second response in the above embodiment.

[0171] Option 33: The mobile device initiates non-semantic communication to transmit data.

[0172] Here, option 33 means that the base station device does not have storage space to cache the semantic information subgraph, and the mobile device needs to initiate non-semantic communication to transmit data.

[0173] If the mobile device receives option 32, it needs to initiate the second window to receive semantic understanding after a period of time T. During time T, the mobile station can choose to remain silent. Time T can be based on base station configuration or a preset value. A schematic diagram of the first and second time windows is shown below. Figure 4 As shown:

[0174] After the mobile device sends a semantic understanding request, it receives the second option 32 in the first time window 401 that is initiated; the mobile device initiates the second time window 402 after a period of time T, and receives the second response in the second time window 402.

[0175] In the embodiments of this application, such as Figure 5 The diagram illustrates an edge caching deployment mechanism for a semantic knowledge base, combined with... Figure 5 The following explanation is provided:

[0176] First, the cached knowledge base of base station 52 is initialized with error code 501.

[0177] Here, the base station cache knowledge base requests a subgraph from the knowledge graph in the cloud knowledge base and stores it. In the background knowledge base deployment mechanism proposed in this application embodiment, the initialization of the cache knowledge base of the base station device can be understood as the acquisition of the cached knowledge base content, that is, the process by which the cache knowledge base of the base station device requests a subgraph from the cloud knowledge base and stores it.

[0178] Secondly, the mobile device requested a subgraph query 502.

[0179] Here, the mobile device sends a semantic understanding request containing semantic information to the base station device. The base station device receives the semantic understanding request and executes a subgraph query to quickly respond to the semantic understanding request.

[0180] Finally, the knowledge graph is selectively updated (503).

[0181] Here, the cloud knowledge base can be updated based on the records of the newly added device request subgraph, while the cached knowledge base 504 of the base station device can be selectively updated; and different caching strategies can be selected according to the resources required by the subgraph of the knowledge graph in the computing resources 53.

[0182] In the embodiments of this application, such as Figure 6 As shown ， This application provides an application flow for a knowledge graph caching mechanism, combined with... Figure 6 The steps shown are explained below:

[0183] Step S601: The base station equipment searches for mobile device identification information.

[0184] Here, before the base station equipment searches for user information, the base station equipment first sets a cache resource consumption threshold T.

[0185] Step S602: Check if there is any historical cache resource loss for the user in the base station equipment.

[0186] Here, if yes, proceed to step S604; otherwise, proceed to step S603. First, determine whether this user is requesting the knowledge graph subgraph for the first time, and whether the base station device has any historical cached resource consumption for this user. If the mobile device is not sending a request to the base station device for the first time, and the device has any previous cached resource consumption for this user, proceed to step S604; otherwise, proceed to step S603.

[0187] Step S603: Start calculating cache resource consumption.

[0188] Here, it indicates that the mobile device is sending a request to the base station device for the first time. At this time, the base station device first performs a dynamic programming caching method based on the subgraph of the knowledge graph requested by the mobile device, and calculates the total cache resource consumption. Otherwise, based on the processing time record of the last request sent by the mobile device, the historical cache resource consumption is used as the cache resource consumption for this time. After obtaining the cache resource consumption, step S604 is executed.

[0189] Step S604: Does the cache resource consumption exceed the preset consumption threshold T? If yes, proceed to step S606; otherwise, proceed to step S605.

[0190] Here, the obtained cache resource consumption is compared with the set cache resource consumption threshold T. If the cache resource consumption exceeds the cache resource consumption threshold T, then step S606 is executed; otherwise, step S605 is executed.

[0191] Step S605: Execute the subgraph-based dynamic programming caching method.

[0192] Here, it indicates that the cache resource consumption has not exceeded the cache resource consumption threshold T, and the dynamic programming cache method based on the subgraph is executed.

[0193] Step S606: Execute the greedy caching method based on triples.

[0194] Here, it indicates that when the cache resource consumption exceeds the cache resource consumption threshold T, the greedy caching method based on triples is executed.

[0195] Step S607: Stop calculating cache resource consumption.

[0196] Step S608: Store or update the user's cached resource consumption information.

[0197] The caching method in this application embodiment includes: setting a cache resource consumption threshold T. First, it is determined whether this user is requesting a knowledge graph subgraph for the first time. If so, a dynamic programming caching method based on the subgraph is performed first, and the total cache resource consumption is calculated simultaneously; otherwise, based on the previous processing time record, if the threshold T is exceeded, a greedy caching scheme is selected; if the threshold T is not exceeded, a dynamic programming caching scheme is selected, and the total cache resource consumption is calculated simultaneously.

[0198] The two caching schemes in this application embodiment are described as follows:

[0199] The first method is a subgraph-based dynamic programming caching approach: To maximize the value of the cached knowledge graph, the caching process is viewed as a knapsack problem. The limited cache size of the knowledge base is considered as a finite knapsack capacity, the multiple subgraphs within the knowledge graph are considered as items to be placed in the knapsack, the number of relationships between subgraphs is considered as the weight of the items, and the value of each subgraph is considered as the profit of the items. Unlike the knapsack problem in related technologies, the triples in the subgraphs in this embodiment are reusable. When a subgraph is stored in the cache, the number of relationships (the weight of items in the knapsack) of other requesting subgraphs with duplicate triples in that subgraph are updated.

[0200] The subgraph-based dynamic programming caching method specifically includes:

[0201] according to Figure 5 The system has the following structure: k mobile devices and m base stations. Define B = {B1, B2, ..., B...} m} represents the base station, U = {u1, u2, ..., u k} represents the device. Each base station (BS) has a limited buffer C. m To store the semantic KG used in the semantic system.

[0202] Suppose that the cloud knowledge base caches the complete KG required by the device. Here, KG is defined as a graph, represented as a ternary model, and stored as a relation table. The KG cached by the cloud knowledge base can be defined as G. total ={V(G total ),E( total The portion of KG cached in Bm is defined as G. m ={V(G m ),E(G m ),L}, which is G cached in the cloud. total Subgraphs.

[0203] Value of Subgraphs (VoG) k,t This indicates that for a given subgraph request, the value consists of two parts: an objective assessment of translatability and a subjective assessment of effectiveness. The value of a subgraph is defined as VoG.k,t =ω1·BLEU k,t +ω2·e -γkAt BLEU, a performance metric for semantic communication systems, reflects the accuracy and complexity of different message recovery methods. k This indicates different latency tolerances for different services. A t This represents the freshness of requests originating from the same device. In other words, subgraphs with low latency tolerance and high freshness have higher cache values. ω1,ω2∈[0,1] are value factors, reflecting the requirements for translatability and effectiveness in different semantic communication systems deployed in the semantic knowledge base. V(G) is the set of all semantic bases, E(G) is the set of all relations (edges) between semantic bases, L is the graph label, and G={V(G),E(G),L}.

[0204] The semantic information structure includes a mapping relationship between messages sent by the device and KG as s = f(m) → F = {F1, F2, ..., F}. K}, where f(m) is the semantic representation process generated by message m at the source. The device request set is denoted as F = {F1, F2, ..., F}. K}, where F K It is G total The subgraph, E(F1)∪E(F2)∪…∪E(F) k )=E(G total ), V(F1)∪V(F2)∪…∪V(F k )=V(G total A subgraph can be divided into multiple triples g. n A triple g n ={h n, r n ,t n}, where h n ,t n r is the semantic entity of the subgraph. n This refers to the relationships between semantic entities.

[0205] To describe the caching behavior of triples, a binary cache placement matrix X is defined. The matrix X for base station m is... m Defined as Where x i,j Represents the triple g n Whether it is stored in Bm. That is, if h n ,t n ∈V(G m )∩r n ∈E(G m If x i,j =1. Cache size C m Defined as C m=S m / L0,S m L0 is the space for caching semantic subgraphs in base station m. L0 is the average number of bytes per triple.

[0206] To maximize the value of the cache subgraph, the limited cache C... m Treating the size as a finite knapsack capacity, the requested subgraph F is... k,t Treating them as items to be put into a knapsack, the relation number |E(F) of the subgraph is... k,t Treating each subgraph as the weight of an item, the value V of each subgraph is... O G k,t It is considered as profit from goods.

[0207] Let X g Used to determine whether the requested subgraph g is cached in the BS. If all triples in the subgraph are cached in the BS, then X... g =1. This can be described as:

[0208] max∑ g∈G VoG g x g ,st∑ I ∑ J X m (i,i)<C m ,x g ∈{0,1}, g=1,2,…,kT.

[0209] In this application's knapsack problem, triples in the subgraphs are reusable. When a subgraph is stored in the cache, the relation count of other requesting subgraphs with triples that are duplicates of that subgraph is updated. That is, each time a new subgraph is considered for addition to the cache, the new and old subgraphs in the cache are merged, duplicate triples are removed, and thus the relation count is updated, i.e., the item weights in the knapsack problem are updated. Furthermore, by calculating the value of the merged subgraphs instead of simply adding values ​​in the knapsack problem, the modification to the knapsack problem in related technologies under this special case can also be reflected.

[0210] like Figure 7 As shown in the embodiments of this application, a subgraph-based dynamic programming caching process is provided, combined with... Figure 7 The steps shown are explained below:

[0211] Step S701: Initialize sequence dp[n], buffer capacity, and sub-graphet.

[0212] Where dp[n] represents the subgraph that can achieve the maximum value when the weight of the items does not exceed n, the cache capacity refers to the cache space of the cache knowledge base of the base station device, and the subgraph set refers to the knowledge graph subgraph set corresponding to the semantic understanding request sent by the mobile device.

[0213] Step S702: Begin traversing the sub-atlas.

[0214] Here, the knowledge graph subgraphs are sorted according to a first preset order, and the sorted subgraphs are traversed. In some embodiments, the first preset order is a descending order of size.

[0215] Step S703: Run the knapsack algorithm.

[0216] Step S704: Determine whether the current subgraph can be added to the knapsack.

[0217] If yes, proceed to step S705; otherwise, proceed to step S706.

[0218] Step S705: Update the knapsack capacity and the value of the remaining subgraph.

[0219] In some embodiments, run Figure 7 The purpose of the Check and Update section is to determine whether the current subgraph can be added to the knapsack each time it is considered for addition to the cache. Let weight be the traversal from the cache capacity to the size of the selected subgraph minus 1; let i be the traversal from 0 to weight. Here, weight represents the remaining capacity of the knapsack. The Check and Update steps include:

[0220] Step S7051: Merge the subgraph and the graph of dp[i] to obtain the merged subgraph.

[0221] Step S7052: Determine the size of the merged subgraph. If the size of the merged subgraph is less than or equal to the weight, proceed to step S7053; otherwise, skip to step S7055.

[0222] Step S7053: If the subgraph value of the merged image is greater than the subgraph value of the corresponding subgraph of dp[weight], proceed to step S7054; otherwise, proceed to step S7055.

[0223] Step S7054: Update the graph corresponding to dp[weight] with the merged subgraph.

[0224] Step S7055: Exit the Check And Update subprocess and return to the previous process.

[0225] Here, the relation number is updated by merging the new and old subgraphs in the cache and removing duplicate triples, thus updating the weight of the items in the knapsack problem. In some embodiments, calculating the value of the merged subgraphs instead of simply adding values ​​in the knapsack problem also reflects the modification to the knapsack problem in this special case.

[0226] Step S706: Determine whether the subgraph has been traversed.

[0227] If yes, proceed to step S708; otherwise, proceed to step S707.

[0228] Step S707: Obtain the next subgraph.

[0229] After step S707, return to step S703.

[0230] Step S708: Return dp[cache capacity].

[0231] The second approach is a greedy caching method based on triples: To address the curse of dimensionality problem that may occur with large-scale caching of subgraph-based dynamic programming caching methods, this application proposes a greedy caching method. To improve caching performance, triple weights are designed to calculate the total value of the subgraph, and a greedy algorithm is used to maximize this value.

[0232] Among them, the greedy caching method based on triples specifically includes:

[0233] Let the triple g nm Value density r(g) nm ), x n Used to determine triple g nm Is it contained in subgraph n? If so, then x n =1. This can be described as:

[0234] max∑ n∈V(G) r n x n ,st∑ I ∑ J x m (i,j)<C m ,x n ∈{0,1},n=1,2,…,|V(G t o tal )|

[0235] Among them, the value density of triples x n Used to determine whether triple n is cached in the base station, for example, if g n,m If x is contained in subgraph n, then n =1.

[0236] like Figure 8 As shown, the embodiment of this application provides a greedy caching process based on triples, combined with Figure 8 The steps shown are explained below:

[0237] Step S801: Initialize the triplet value list and cache capacity.

[0238] Here, the triple value list refers to the value list of triples in multiple knowledge graph subgraphs corresponding to a semantic understanding request, and the cache capacity refers to the cache space of the cache knowledge base of the base station device.

[0239] Step S802: Sort the triplet value density list by size.

[0240] Here, the triples of multiple knowledge graph subgraphs corresponding to the semantic understanding request are sorted according to the second preset order to obtain a sorted triple list.

[0241] Step S803: Take the first triplet in the triplet value density list.

[0242] Step S804: Determine whether the current optimal cache size is less than the cache capacity.

[0243] If yes, proceed to step S805; otherwise, proceed to step S806.

[0244] Step S805: Add the current triplet to the optimal cache.

[0245] Here, when the optimal cache size is less than the cache capacity of the cache knowledge base, the current triplet being traversed is cached in the cache knowledge base.

[0246] Step S806: Determine whether the triplet has been traversed.

[0247] If yes, proceed to step S808; otherwise, proceed to step S807.

[0248] Step S807: Obtain the triples of the next subgraph.

[0249] After step S807, return to step S803.

[0250] Step S808: Return the best cache.

[0251] In some embodiments, the cache resource consumption will exceed the cache resource consumption threshold T. To address the dimensionality curse problem that may exist in large-scale caching for subgraph-based dynamic programming caching methods, this application proposes a triple-based greedy caching method. By designing triple weights to calculate the total value of the subgraph, and using a greedy algorithm, the deployment and computation time of large-scale knowledge bases are shortened, thereby improving the performance of cached subgraphs under limited resources and maximizing its performance.

[0252] This application provides a data transmission device, which includes various modules and units included in each module. It can be implemented by a processor in a terminal; of course, it can also be implemented by specific logic circuits. In the implementation process, the processor can be a central processing unit, a microprocessor, a digital signal processor, or a field-programmable gate array, etc.

[0253] Figure 9A This is a schematic diagram of the composition structure of the data transmission device provided in the embodiments of this application, applied to base station equipment, such as... Figure 9A As shown, the data transmission device 910 includes:

[0254] The first receiving module 911 is used to receive the sent semantic understanding request;

[0255] The first determining module 912 is used to determine the data to be transmitted based on the semantic understanding request and the cached knowledge base of the base station device;

[0256] The first sending module 913 is used to send the data to be transmitted.

[0257] In some embodiments, the first determining module 912 includes:

[0258] The first determining submodule is used to determine the knowledge graph corresponding to the semantic understanding request;

[0259] The second determining submodule is used to determine the data to be transmitted corresponding to the semantic understanding request based on the knowledge graph and the cached knowledge base.

[0260] In some embodiments, the second determining submodule includes:

[0261] The first determining unit is configured to determine, when the knowledge graph is already cached in the cached knowledge base, that the knowledge graph in the cached knowledge base is the data to be transmitted;

[0262] The second determining unit is configured to search for the knowledge graph in a preset cloud knowledge base when the knowledge graph is not cached in the cached knowledge base; in response to finding the knowledge graph, cache the knowledge graph in the cached knowledge base, and determine that the knowledge graph in the cached knowledge base has been updated as the data to be transmitted.

[0263] In some embodiments, the first transmitting module 913 includes:

[0264] The first sending submodule is configured to send a first response carrying the knowledge graph in a first time window, provided that the knowledge graph has been cached in the cached knowledge base.

[0265] The second sending submodule is used to send a first response carrying the knowledge graph within a second time window when the knowledge graph is not cached in the cached knowledge base; wherein, the start time of the first time window is the time when the semantic understanding request is sent; the second time window is started when the first response is not received within the first time window and within a preset time period after the first time window.

[0266] In some embodiments, the second sending submodule is further configured to: in response to receiving the semantic understanding request that is sent again when the second time window is initiated, send a second response carrying the knowledge graph in the updated cached knowledge base.

[0267] In some embodiments, the apparatus further includes:

[0268] The second receiving module is used to receive the sent scheduling request;

[0269] The second sending module is used to send scheduling information in response to the scheduling request.

[0270] Figure 9B This is a schematic diagram of the composition structure of the data transmission device provided in the embodiments of this application, applied to mobile devices, such as... Figure 9B As shown, the data transmission device 920 includes:

[0271] The fourth sending module 921 is used to send scheduling requests;

[0272] The third receiving module 922 is used to receive scheduling information sent in response to the scheduling request;

[0273] The fifth sending module 923 is used to send a semantic understanding request in response to the scheduling information;

[0274] The fourth receiving module 924 is used to receive the data to be transmitted corresponding to the semantic understanding request.

[0275] In some embodiments, the fourth receiving module 924 includes:

[0276] The first startup submodule is used to start a first time window with the moment the semantic understanding request is sent as the starting moment, and to receive the data to be transmitted within the first time window;

[0277] The third sending submodule is used to respond to the fact that the data to be transmitted is not received within the first time window, and to send the semantic understanding request again at a preset time interval and start the second time window.

[0278] The first receiving submodule is used to receive the data to be transmitted within the second time window.

[0279] In some embodiments, the data transmission device 920 further includes:

[0280] The first startup module is used to initiate non-semantic communication in response to the failure to receive the data to be transmitted within the first time window or the second time window.

[0281] It should be noted that, in the embodiments of this application, if the above-mentioned problem discovery method is implemented as a software functional module and sold or used as an independent product, it can also be stored in a terminal-readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, or the part that contributes to the prior art, can be embodied in the form of a software product. This terminal software product is stored in a storage medium and includes several instructions to cause a terminal (which may be a personal computer or a server, etc.) to execute all or part of the methods described in the various embodiments of this application.

[0282] It should be noted that, Figure 10 A schematic diagram of the hardware entity of the electronic device provided in the embodiments of this application, such as... Figure 10 As shown, the electronic device 1000 can be a mobile device or a base station device. The hardware entity of the electronic device 1000 includes a processor 1001, a communication interface 1002, and a memory 1003. The processor 1001 typically controls the overall operation of the electronic device 1000. The communication interface 1002 enables the terminal to communicate with other terminals or servers via a network. The memory 1003 is configured to store instructions and applications executable by the processor 1001, and can also cache data to be processed or already processed by the processor 1001 and various modules in the electronic device 1000 (e.g., image data, audio data, voice communication data, and video communication data). This can be implemented using flash memory or random access memory (RAM).

[0283] Correspondingly, embodiments of this application provide a storage medium storing executable instructions for inducing the processor to execute the aforementioned problem discovery method.

[0284] The descriptions of the storage medium and device embodiments above are similar to those of the method embodiments above, and have similar beneficial effects. For technical details not disclosed in the storage medium and device embodiments of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0285] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms. The above-mentioned separated components may or may not be physically separated, and the components shown may or may not be physical units; they can be located in one place or distributed across multiple network units; some or all of the units can be selected to achieve the purpose of the solution in this embodiment according to actual needs.

[0286] Furthermore, the functional units in the various embodiments of this application can all be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in a combination of hardware and software functional units. Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable storage medium, and when executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks. Alternatively, if the integrated unit of this application is implemented as a software functional module and sold or used as an independent product, it can also be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, or the part that contributes to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause the terminal to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage media include various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks. The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A data transmission method, characterized in that, Applied to base station equipment, the method includes: Receive and send semantic understanding requests; Determine whether the knowledge graph corresponding to the semantic understanding request is cached in the cached knowledge base of the base station device; If the knowledge graph is not cached in the cached knowledge base, determine the cache resource consumption based on the semantic understanding request; Based on the aforementioned cache resource consumption, the knowledge graph in the preset cloud knowledge base is cached in the cached knowledge base; Based on the semantic understanding request and the cached knowledge base, the data to be transmitted is determined; Send the data to be transmitted.

2. The method according to claim 1, characterized in that, If the knowledge graph is already cached in the cached knowledge base, after receiving the sent semantic understanding request, the method further includes: Determine the knowledge graph corresponding to the semantic understanding request; The knowledge graph in the cached knowledge base is identified as the data to be transmitted. Send the data to be transmitted.

3. The method according to claim 1, characterized in that, Sending the data to be transmitted includes: If the knowledge graph is already cached in the cached knowledge base, a first response carrying the knowledge graph is sent in the first time window. Alternatively, if the knowledge graph is not cached in the cached knowledge base, a second response carrying the knowledge graph is sent within a second time window; wherein, the start time of the first time window is the time when the semantic understanding request is sent; and the second time window is started if the first response is not received within the first time window and within a preset time period after the first time window.

4. The method according to claim 3, characterized in that, In the case where the knowledge graph is not cached in the cached knowledge base, within the second time window, a second response carrying the knowledge graph is sent, including: In response to receiving the semantic understanding request again when the second time window is initiated, a second response carrying the knowledge graph in the updated cached knowledge base is sent.

5. The method according to claim 1, characterized in that, Before receiving the sent semantic understanding request, the method further includes: Receive the sent scheduling request; In response to the scheduling request, scheduling information is sent.

6. A data transmission method, characterized in that, Applied to mobile devices, the method includes: Send a scheduling request; Receive scheduling information sent in response to the scheduling request; In response to the scheduling information, a semantic understanding request is sent; The system receives data to be transmitted corresponding to the semantic understanding request. The data to be transmitted is obtained based on the semantic understanding request and the cached knowledge base of the base station device. The cached knowledge base includes a knowledge graph cached from a preset cloud knowledge base based on cache resource consumption. The cache resource consumption is obtained based on the semantic understanding request when the knowledge graph is not cached in the cached knowledge base.

7. The method according to claim 6, characterized in that, The step of receiving the data to be transmitted corresponding to the semantic understanding request includes: The first time window is started with the moment the semantic understanding request is sent, and the data to be transmitted is received within the first time window. If the data to be transmitted is not received within the first time window, the semantic understanding request is sent again after a preset time interval and a second time window is started; The data to be transmitted is received within the second time window.

8. The method according to claim 7, characterized in that, After responding to the situation where the data to be transmitted is not received within the first time window, and after resending the semantic understanding request at a preset interval and initiating a second time window, the method further includes: If the data to be transmitted is not received within the first time window or the second time window, non-semantic communication is initiated.

9. A base station device for data transmission, comprising a memory and a processor, the memory storing a computer program executable on the processor, the processor executing the program to implement the steps of the method according to any one of claims 1 to 5.

10. A mobile device for data transmission, comprising a memory and a processor, the memory storing a computer program executable on the processor, the processor executing the program to implement the steps of the method of any one of claims 6 to 8.

Images