Generative AI-based communication transmission method and system with scalability

Abstract

The invention provides a generative AI-based communication transmission method and system with a telescopic transmission capability, and the method comprises the steps: enabling a transmission node to reduce or increase the transmission content in a self-adaptive manner based on a plurality of factors, such as a network environment, a user demand, and a user computing power capability; when the network transmission bandwidth sufficiency is smaller than a preset standard, the transmission node only transmits part of information of the content required by the next node, and the next node generates the required content through the generative AI technology according to the part of information; and when the network transmission bandwidth sufficiency is greater than a preset standard, the transmission node transmits the complete content required by the next node. The method effectively regulates and controls the communication data volume, relieves the pressure of a communication network, and improves the service quality of services.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a communication transmission and system with scalable transmission based on generative AI. Background Art

[0002] In 2023, IMT-2030 defines six typical application scenarios for 6G based on the three major application scenarios of 5G. Among them, the future multimedia transmission services represented by the immersive communication scenario have put forward higher requirements and challenges for the amount of data transmitted during the communication process, the communication quality, and the carrying capacity of the overall network. As the next generation of network communication technology, 6G needs to provide diversified, timely, and high-throughput communication services for different businesses while addressing the rapid growth of network traffic and the rapid improvement of communication quality in the face of various business needs.

[0003] At the same time, as one of the development achievements of artificial intelligence technology in recent years, generative AI can generate a variety of corresponding texts, pictures and videos through simple semantic information or instructions. The semantic communication related technology derived from this can greatly compress the amount of transmitted data and is one of the possible evolution directions of future communication systems. Summary of the invention

[0004] In view of the defects in the prior art, the purpose of the present invention is to provide a communication transmission method and system with scalable transmission capabilities based on generative AI.

[0005] According to the present invention, a communication transmission method with scalable transmission based on generative AI is provided, comprising:

[0006] Step S1: Based on the network environment, user needs, and user computing power, the transmission node adaptively reduces or increases the transmission content;

[0007] Step S2: When the network transmission bandwidth adequacy is less than the preset standard, the transmission node only transmits part of the information of the content required by the next node, and the next node generates the required content through generative AI technology based on the part of the information;

[0008] When the network transmission bandwidth adequacy is greater than the preset standard, the transmission node transmits the complete content required by the next node.

[0009] Preferably, in step S1:

[0010] The transmission node has computing power and is equipped with a CPU chip with preset calculation examples. It performs calculations and decision-making through software algorithms, provides computing power for feature extraction and optimization of decision-making algorithms, and provides decision-making capabilities on whether to reduce or complete the transmission content.

[0011] Preferably, the transmission node has the ability to extract content features of the transmitted information, including using a neural network to identify the content, thereby segmenting and extracting key content;

[0012] The transmission node has the ability to scale content, reduce the transmitted content, transmit only part of the content, or transmit the entire content; including extracting semantic features or information through neural networks to achieve content reduction, or generating content based on semantic information through generative network DiffusionModels to achieve content expansion.

[0013] The transmission node has an intelligent decision-making mechanism, which selects the content to be transmitted based on the network environment, user conditions, and user computing power, including the use of optimization algorithms or neural networks for intelligent decision-making.

[0014] Preferably, in step S1:

[0015] For transmission nodes that do not have computing power, the sending node uses traditional encoding and decoding technology to fully encode the data and send it to the receiving end, which then decodes and recovers it.

[0016] For nodes that have computing power but require the original content to be higher than the preset standard, the sending node will encode and send most of the preset information, and the receiving end will decode and recover it.

[0017] For nodes that have computing power and have original content requirements that are lower than preset standards, the sender only encodes and sends less preset semantic information. After the receiver receives the semantic information, it generates media content through generative AI.

[0018] Preferably, in step S2:

[0019] When the node chooses to transmit part of the content needed by the next node, the next node, after receiving the part of the content, uses the generative AI capability to expand the part of the content into the complete content;

[0020] Generative AI includes Style-GAN, Diffusion Models, DALL-3 or Stable Diffusion;

[0021] If the node chooses to transmit the required complete content to the next node, the next node obtains the required content after receiving the complete content.

[0022] According to the present invention, a communication transmission system with scalable transmission based on generative AI is provided, comprising:

[0023] Module M1: Based on network environment, user needs, and user computing power, the transmission node adaptively reduces or increases the transmission content;

[0024] Module M2: When the network transmission bandwidth is less than the preset standard, the transmission node only transmits part of the information of the content required by the next node, and the next node generates the required content through generative AI technology based on the part of the information;

[0025] When the network transmission bandwidth adequacy is greater than the preset standard, the transmission node transmits the complete content required by the next node.

[0026] Preferably, in the module M1:

[0027] The transmission node has computing power and is equipped with a CPU chip with preset calculation examples. It performs calculations and decision-making through software algorithms, provides computing power for feature extraction and optimization of decision-making algorithms, and provides decision-making capabilities on whether to reduce or complete the transmission content.

[0028] Preferably, the transmission node has the ability to extract content features of the transmitted information, including using a neural network to identify the content, thereby segmenting and extracting key content;

[0029] The transmission node has the ability to scale content, reduce the transmitted content, transmit only part of the content, or transmit the entire content; including extracting semantic features or information through neural networks to achieve content reduction, or generating content based on semantic information through generative network DiffusionModels to achieve content expansion.

[0030] The transmission node has an intelligent decision-making mechanism, which selects the content to be transmitted based on the network environment, user conditions, and user computing power, including the use of optimization algorithms or neural networks for intelligent decision-making.

[0031] Preferably, in the module M1:

[0032] For transmission nodes that do not have computing power, the sending node uses traditional encoding and decoding technology to fully encode the data and send it to the receiving end, which then decodes and recovers it.

[0033] For nodes that have computing power but require the original content to be higher than the preset standard, the sending node will encode and send most of the preset information, and the receiving end will decode and recover it.

[0034] For nodes that have computing power and have original content requirements that are lower than preset standards, the sender only encodes and sends less preset semantic information. After the receiver receives the semantic information, it generates media content through generative AI.

[0035] Preferably, in the module M2:

[0036] When the node chooses to transmit part of the content needed by the next node, the next node, after receiving the part of the content, uses the generative AI capability to expand the part of the content into the complete content;

[0037] Generative AI includes Style-GAN, Diffusion Models, DALL-3 or Stable Diffusion;

[0038] If the node chooses to transmit the required complete content to the next node, the next node obtains the required content after receiving the complete content.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] Based on the breakthrough in device computing power and the rapid development of generative AI technology, the present invention designs a communication transmission method based on generative AI with scalable transmission capabilities. This method actively chooses to reduce or expand the transmission content according to factors such as the network environment, user conditions, and user computing power, thereby making full use of the computing power of the node, effectively regulating the amount of communication data, alleviating the pressure on the communication network, and improving the service quality of the business. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0042] Figure 1 The communication network structure of embodiment 3;

[0043] Figure 2 This is the communication network structure of Example 4. DETAILED DESCRIPTION

[0044] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0045] Embodiment 1:

[0046] The present invention proposes a communication transmission method with scalable transmission based on generative AI. The transmission method is based on multiple factors such as network environment, user needs, and user computing power. The transmission node adaptively reduces or increases the transmission content. For example, when the network transmission bandwidth is tight, the transmission node can only transmit part of the content required by the next node, such as semantics and information profile, and then the next node generates the required content through generative AI technology based on the partial information; or when the network transmission bandwidth is sufficient, the transmission node can transmit the complete content required by the next node to achieve high-fidelity and error-free transmission. The next node does not need to generate the required content through generative AI, thereby realizing adaptive scalable control of the transmission content, so as to improve the communication and service quality of future multimedia services and improve network resilience.

[0047] This patent aims to invent a scalable network, node and communication transmission method based on generative AI. This transmission method can actively select scalable transmission content based on factors such as network environment, user situation and user computing power, so as to make full use of the computing power of the node and effectively regulate the amount of communication data.

[0048] According to the present invention, a communication transmission method with scalable transmission based on generative AI is provided, comprising:

[0049] Step S1: Based on the network environment, user needs, and user computing power, the transmission node adaptively reduces or increases the transmission content;

[0050] Specifically, in step S1:

[0051] The transmission node has computing power and is equipped with a CPU chip with preset calculation examples. It performs calculations and decision-making through software algorithms, provides computing power for feature extraction and optimization of decision-making algorithms, and provides decision-making capabilities on whether to reduce or complete the transmission content.

[0052] Specifically, the transmission node has the ability to extract content features of the transmitted information, including using neural networks to identify content, thereby segmenting and extracting key content;

[0053] The transmission node has the ability to scale content, reduce the transmitted content, transmit only part of the content, or transmit the entire content; including extracting semantic features or information through neural networks to achieve content reduction, or generating content based on semantic information through generative network DiffusionModels to achieve content expansion.

[0054] The transmission node has an intelligent decision-making mechanism, which selects the content to be transmitted based on the network environment, user conditions, and user computing power, including the use of optimization algorithms or neural networks for intelligent decision-making.

[0055] Specifically, in step S1:

[0056] For transmission nodes that do not have computing power, the sending node uses traditional encoding and decoding technology to fully encode the data and send it to the receiving end, which then decodes and recovers it.

[0057] For nodes that have computing power but require the original content to be higher than the preset standard, the sending node will encode and send most of the preset information, and the receiving end will decode and recover it.

[0058] For nodes that have computing power and have original content requirements that are lower than preset standards, the sender only encodes and sends less preset semantic information. After the receiver receives the semantic information, it generates media content through generative AI.

[0059] Step S2: When the network transmission bandwidth adequacy is less than the preset standard, the transmission node only transmits part of the information of the content required by the next node, and the next node generates the required content through generative AI technology based on the part of the information;

[0060] When the network transmission bandwidth adequacy is greater than the preset standard, the transmission node transmits the complete content required by the next node.

[0061] Specifically, in step S2:

[0062] When the node chooses to transmit part of the content needed by the next node, the next node, after receiving the part of the content, uses the generative AI capability to expand the part of the content into the complete content;

[0063] Generative AI includes Style-GAN, Diffusion Models, DALL-3 or Stable Diffusion;

[0064] If the node chooses to transmit the required complete content to the next node, the next node obtains the required content after receiving the complete content.

[0065] Embodiment 2:

[0066] Embodiment 2 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0067] The present invention also provides a communication transmission system with scalable transmission based on generative AI. The communication transmission system with scalable transmission based on generative AI can be implemented by executing the process steps of the communication transmission method with scalable transmission based on generative AI, that is, those skilled in the art can understand the communication transmission method with scalable transmission based on generative AI as a preferred implementation of the communication transmission system with scalable transmission based on generative AI.

[0068] According to the present invention, a communication transmission system with scalable transmission based on generative AI is provided, comprising:

[0069] Module M1: Based on network environment, user needs, and user computing power, the transmission node adaptively reduces or increases the transmission content;

[0070] Specifically, in the module M1:

[0071] The transmission node has computing power and is equipped with a CPU chip with preset calculation examples. It performs calculations and decision-making through software algorithms, provides computing power for feature extraction and optimization of decision-making algorithms, and provides decision-making capabilities on whether to reduce or complete the transmission content.

[0072] Specifically, the transmission node has the ability to extract content features of the transmitted information, including using neural networks to identify content, thereby segmenting and extracting key content;

[0073] The transmission node has the ability to scale content, reduce the transmitted content, transmit only part of the content, or transmit the entire content; including extracting semantic features or information through neural networks to achieve content reduction, or generating content based on semantic information through generative network DiffusionModels to achieve content expansion.

[0074] The transmission node has an intelligent decision-making mechanism, which selects the content to be transmitted based on the network environment, user conditions, and user computing power, including the use of optimization algorithms or neural networks for intelligent decision-making.

[0075] Specifically, in the module M1:

[0076] For transmission nodes that do not have computing power, the sending node uses traditional encoding and decoding technology to fully encode the data and send it to the receiving end, which then decodes and recovers it.

[0077] For nodes that have computing power but require the original content to be higher than the preset standard, the sending node will encode and send most of the preset information, and the receiving end will decode and recover it.

[0078] For nodes that have computing power and have original content requirements that are lower than preset standards, the sender only encodes and sends less preset semantic information. After the receiver receives the semantic information, it generates media content through generative AI.

[0079] Module M2: When the network transmission bandwidth is less than the preset standard, the transmission node only transmits part of the information of the content required by the next node, and the next node generates the required content through generative AI technology based on the part of the information;

[0080] When the network transmission bandwidth adequacy is greater than the preset standard, the transmission node transmits the complete content required by the next node.

[0081] Specifically, in the module M2:

[0082] When the node chooses to transmit part of the content needed by the next node, the next node, after receiving the part of the content, uses the generative AI capability to expand the part of the content into the complete content;

[0083] Generative AI includes Style-GAN, Diffusion Models, DALL-3 or Stable Diffusion;

[0084] If the node chooses to transmit the required complete content to the next node, the next node obtains the required content after receiving the complete content.

[0085] Embodiment 3:

[0086] Embodiment 3 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0087] 1. For transmission nodes that do not have computing power, the sending node uses traditional encoding and decoding technology to fully encode the data and send it to the receiving end, which then decodes and recovers it.

[0088] 2. For nodes that have computing power but have high requirements for original content, the sending node encodes most of the information and sends it, and the receiving end decodes and recovers it.

[0089] 3. For nodes that have computing power and low demand for original content, the sender only encodes and sends less semantic information. After the receiver receives the semantic information, it generates media content through generative AI.

[0090] Based on the above objectives, the technical solution adopted by the present invention is as follows:

[0091] 1. A communication node with computing power, equipped with a CPU chip with certain calculation examples, can perform calculations and decision-making through software algorithms. It provides computing power for feature extraction, optimization of decision algorithms, etc., and provides decision-making capabilities on whether to reduce / complete transmission content.

[0092] 2. The node has the ability to extract content features of the transmitted information, such as but not limited to using neural networks to identify content, thereby segmenting and extracting key content.

[0093] 3. The node has the ability to scale content, and can reduce the transmitted content, transmit only part of the content, or transmit the entire content. It is not limited to a specific technology, for example: extracting semantic features or information through neural networks to achieve content reduction, and generating content based on semantic information through generative network Diffusion Models to achieve content expansion.

[0094] 4. The node has an intelligent decision-making mechanism, which can select the content to be transmitted according to the network environment, user conditions, user computing power, etc. It is not limited to a specific technology, for example: optimization algorithms or neural networks can be used for intelligent decision-making.

[0095] 5. When the node chooses to transmit part of the content required by the next node, the next node uses the generative AI capability to expand the part of the content into the complete content after receiving the part of the content. This patent mainly focuses on the transmission mechanism with scalability. There are many ways to implement generative AI, such as Style-GAN, Diffusion Models, DALL-3, Stable Diffusion, etc. The specific choice will not affect the overall idea of ​​the transmission mechanism.

[0096] 6. If the node chooses to transmit the required complete content to the next node, the next node can obtain the required content after receiving the complete content.

[0097] Embodiment 3:

[0098] Embodiment 3 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0099] We consider a one-to-many video media transmission scenario, where the communication network nodes determine the actual transmission content type and allocate corresponding bandwidth resources based on the user's computing power and content requirements. The communication network structure considered in this embodiment is as follows: Figure 1 As shown in the figure. In this communication network, there is a sending node with content scalability and three receiving nodes (i.e., users). User 1 is a receiving node without computing power, user 2 is a receiving node with computing power and a high demand for original content, and user 3 is a receiving node with computing power and a demand only for semantic content.

[0100] At this point, the processing flow of the entire communication network is as follows:

[0101] 1. Each user feeds back his or her computing power and content requirements to the network communication node.

[0102] 2. The communication node makes intelligent decisions on the specific transmission content and allocated bandwidth for each user based on the feedback information received.

[0103] 3. User 1 does not have computing power and cannot generate content, so the original video media content can be sent to User 1.

[0104] 4. For user 2, since he has computing power and can generate content, but has a high demand for original content, the transmitted content can be appropriately reduced, and he can choose to send most of the original content and a small amount of semantic information.

[0105] 5. For user 3, since he has computing power and only needs the semantic content of the video, he can perform semantic extraction on the video media, greatly reducing the transmission content and only sending semantic information.

[0106] 6. After receiving the information, user 1 directly decodes it to obtain its own content.

[0107] 7. After receiving the information, user 2 generates some content to fill in the missing content.

[0108] 8. After receiving the information, user 3 uses generative AI to generate a video based on the semantic content.

[0109] Embodiment 4:

[0110] Embodiment 4 is a preferred example of Embodiment 1, and is used to illustrate the present invention in more detail.

[0111] We can consider a communication scenario that requires multi-hop transmission. In this scenario, media content is received and forwarded by multiple nodes and finally reaches the user end. Taking four nodes as an example, the communication network structure considered in this embodiment is as follows: Figure 2 As shown. The media content is first transmitted from node 1 to node 2. Node 2 processes the content through the content scaling function and transmits it to node 3. Node 3 also uses the content scaling function to process and transmit it to node 4. The content scaling process is determined according to the channel status: the channel bandwidth between node 1 and node 2 is large, and the complete content is transmitted; the channel bandwidth between node 2 and node 3 is small, so node 2 reduces the transmission content and only transmits part of the content to node 3. The bandwidth between node 3 and node 4 is large, so node 3 uses generative AI to expand part of the content into complete content and transmit it to node 4.

[0112] At this point, the processing flow of the entire communication network is as follows:

[0113] 1. Node 1 transmits complete information to node 2.

[0114] 2. Node 2 reduces the content and transmits it to node 3.

[0115] 3. Node 3 uses generative AI to expand the content into complete content and sends it to node 4.

[0116] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0117] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A communication transmission method with scalable transmission capability based on generative AI, characterized in that: include: Step S1: Based on the network environment, user needs, and user computing power, the transmission node adaptively reduces or increases the transmission content; Step S2: When the network transmission bandwidth adequacy is less than the preset standard, the transmission node only transmits part of the information of the content required by the next node, and the next node generates the required content through generative AI technology based on the part of the information; When the network transmission bandwidth adequacy is greater than the preset standard, the transmission node transmits the complete content required by the next node.

2. The communication transmission method with scalable transmission capability based on generative AI according to claim 1, characterized in that: In step S1: The transmission node has computing power and is equipped with a CPU chip with preset calculation examples. It performs calculations and decision-making through software algorithms, provides computing power for feature extraction and optimization of decision-making algorithms, and provides decision-making capabilities on whether to reduce or complete the transmission content.

3. The communication transmission method with scalable transmission based on generative AI according to claim 2, characterized in that: The transmission node has the ability to extract content features of the transmitted information, including using neural networks to identify content, thereby segmenting and extracting key content; The transmission node has the ability to scale content, reduce the transmission content, and transmit only part of the content or the entire content; This includes extracting semantic features or information through neural networks to achieve content reduction, or generating content based on semantic information through generative network Diffusion Models to achieve content expansion. The transmission node has an intelligent decision-making mechanism, which selects the content to be transmitted based on the network environment, user conditions, and user computing power, including the use of optimization algorithms or neural networks for intelligent decision-making.

4. The communication transmission method with scalable transmission capability based on generative AI according to claim 1, characterized in that: In step S1: For transmission nodes that do not have computing power, the sending node uses traditional encoding and decoding technology to fully encode the data and send it to the receiving end, which then decodes and recovers it. For nodes that have computing power but require the original content to be higher than the preset standard, the sending node will encode and send most of the preset information, and the receiving end will decode and recover it. For nodes that have computing power and have original content requirements that are lower than preset standards, the sender only encodes and sends less preset semantic information. After the receiver receives the semantic information, it generates media content through generative AI.

5. The communication transmission method with scalable transmission capability based on generative AI according to claim 1, characterized in that: In step S2: When the node chooses to transmit part of the content needed by the next node, the next node, after receiving the part of the content, uses the generative AI capability to expand the part of the content into the complete content; Generative AI includes Style-GAN, Diffusion Models, DALL-3 or Stable Diffusion; If the node chooses to transmit the required complete content to the next node, the next node obtains the required content after receiving the complete content.

6. A communication transmission system with scalable transmission capability based on generative AI, characterized in that: include: Module M1: Based on network environment, user needs, and user computing power, the transmission node adaptively reduces or increases the transmission content; Module M2: When the network transmission bandwidth is less than the preset standard, the transmission node only transmits part of the information of the content required by the next node, and the next node generates the required content through generative AI technology based on the part of the information; When the network transmission bandwidth adequacy is greater than the preset standard, the transmission node transmits the complete content required by the next node.

7. The communication transmission system with scalable transmission capability based on generative AI according to claim 6, characterized in that: In the module M1: The transmission node has computing power and is equipped with a CPU chip with preset calculation examples. It performs calculations and decision-making through software algorithms, provides computing power for feature extraction and optimization of decision-making algorithms, and provides decision-making capabilities on whether to reduce or complete the transmission content.

8. The communication transmission system with scalable transmission capability based on generative AI according to claim 7, characterized in that: The transmission node has the ability to extract content features of the transmitted information, including using neural networks to identify content, thereby segmenting and extracting key content; The transmission node has the ability to scale content, reduce the transmission content, and transmit only part of the content or the entire content; This includes extracting semantic features or information through neural networks to achieve content reduction, or generating content based on semantic information through generative network Diffusion Models to achieve content expansion. The transmission node has an intelligent decision-making mechanism, which selects the content to be transmitted based on the network environment, user conditions, and user computing power, including the use of optimization algorithms or neural networks for intelligent decision-making.

9. The communication transmission system with scalable transmission capability based on generative AI according to claim 6, characterized in that: In the module M1: For transmission nodes that do not have computing power, the sending node uses traditional encoding and decoding technology to fully encode the data and send it to the receiving end, which then decodes and recovers it. For nodes that have computing power but require the original content to be higher than the preset standard, the sending node will encode and send most of the preset information, and the receiving end will decode and recover it. For nodes that have computing power and have original content requirements that are lower than preset standards, the sender only encodes and sends less preset semantic information. After the receiver receives the semantic information, it generates media content through generative AI.

10. The communication transmission system with scalable transmission capability based on generative AI according to claim 6, characterized in that: In the module M2: When the node chooses to transmit part of the content needed by the next node, the next node, after receiving the part of the content, uses the generative AI capability to expand the part of the content into the complete content; Generative AI includes Style-GAN, Diffusion Models, DALL-3 or Stable Diffusion; If the node chooses to transmit the required complete content to the next node, the next node obtains the required content after receiving the complete content.

Images