Neural network model-based intelligent agent converged communication architecture and method

Abstract

Disclosed in the present invention are a neural network model-based intelligent agent converged communication architecture and method. In the solution of the present invention, the intelligent agent converged communication architecture comprises at least one intelligent agent unit used for converting audio and video data, and a converged communication exchange center unit and at least one intelligent terminal unit which separately communicate with the intelligent agent unit, the intelligent agent converged communication architecture uses a distributed intelligent agent cluster to achieve group call control, and uses an intelligent agent neural network model to learn link quality control, a method and a platform corresponding to the architecture are provided, and artificial intelligence and a block chain are introduced into an IMS converged communication architecture, resolving accessing multi-protocol converged switching communications of different systems and types in large-scale terminal complex environments, and forming the capability of uninterrupted transmission of various links, thus supporting secure, reliable and high-quality communication of data of various service applications such as voice, video and conferencing.

Description

An intelligent agent fusion communication architecture and method based on neural network model Technical Field

[0001] The present invention belongs to the field of communication processing technology, and in particular relates to an intelligent agent fusion communication architecture and method based on a neural network model. Background Art

[0002] IMS is the network architecture for implementing large-scale convergence solutions in the Next Generation Network (NGN). IMS can implement VoIP services, more effectively manage network resources, user resources, and application resources, improve network intelligence, and enable users to experience a converged communication experience across various networks and using multiple terminals.

[0003] Traditional telecommunications networks use independent signaling networks to complete call establishment, routing, and control processes. The security of this signaling network ensures network security. Furthermore, transmission utilizes dedicated time-division multiplexing (TDM) lines, and users communicate through connection-oriented channels, preventing eavesdropping and attacks from other end users. The IMS network, however, connects to the Internet. Based on the IP protocol and an open network architecture, it can share a service platform for diverse services such as voice, data, and multimedia through a variety of access methods. This increases network flexibility and interoperability between terminals, allowing different operators to efficiently and quickly develop and provide a variety of services.

[0004] Because IMS is built on IP, its security requirements are much higher than those of traditional operators operating on independent networks. Regardless of whether access is mobile or fixed, IMS security issues cannot be ignored. IMS security threats primarily arise from several areas: unauthorized access to sensitive data that compromises confidentiality; unauthorized tampering with sensitive data that compromises integrity; interference with or abuse of network services leading to denial of service or reduced system availability; user or network denial of completed operations; and unauthorized access to services. These primarily involve IMS access security (3GPP TS33.203), which covers user and network authentication and protecting services between IMS terminals and the network; and IMS network security (3GPP TS33.210), which addresses service protection between nodes within the same or different operators' networks. Furthermore, IMS security threats also pose a threat to user equipment and the Universal Integrated Circuit Card / IP Multimedia Identity Module (UICC / ISIM).

[0005] Therefore, in view of the above technical problems and defects, it is urgent to design and develop an intelligent agent fusion communication architecture and method based on a neural network model.

[0006] Summary of the Invention

[0007] To overcome the shortcomings and difficulties of the above-mentioned existing technologies, the purpose of the present invention is to provide an intelligent agent converged communication architecture and method based on a neural network model, and introduce artificial intelligence and blockchain into the IMS converged communication architecture to solve the problem of multi-protocol converged switching communications of different standards and types accessed by large-scale terminals in complex environments, forming uninterrupted transmission capabilities of multiple links, and supporting secure, reliable, and high-quality communication transmission of various business application data such as voice, video, and conferencing.

[0008] The first object of the present invention is to provide an intelligent agent fusion communication architecture based on a neural network model;

[0009] The second object of the present invention is to provide an intelligent agent fusion communication method based on a neural network model;

[0010] The third object of the present invention is to provide an intelligent agent fusion communication platform based on a neural network model;

[0011] The first object of the present invention is achieved as follows: the intelligent agent fusion communication architecture includes at least one intelligent agent unit for converting audio and video data, and a fusion communication switching center unit and at least one intelligent terminal unit respectively communicating with the intelligent agent unit;

[0012] The intelligent agent fusion communication architecture adopts a distributed intelligent agent cluster to realize group call control, and learns link quality control through an intelligent agent neural network model.

[0013] Furthermore, the converged communication switching center unit broadcasts data transmission at the network transport layer in a TCP, UDP unicast or multicast manner.

[0014] Furthermore, the intelligent agent fusion communication architecture is further provided with a synchronization processing unit, which is used to synchronize and coordinate the clusters, and synchronously collect, transmit and play audio and video data streams respectively.

[0015] Furthermore, the intelligent agent fusion communication architecture is further provided with a first creation unit, and the first creation module is used to establish a multi-intelligent agent network containment consistency model of distributed intelligent agents.

[0016] Furthermore, the first creation unit further includes:

[0017] Generate an acquisition module for generating an intelligent agent and obtaining the connection address corresponding to the MECS (Multi-Access Edge Computing Server);

[0018] The first calculation module is used to calculate the delay data from the fully connected network to the intelligent agent;

[0019] The self-check generation module is used to check whether the state of its own group call is stable and generate action control instruction data corresponding to stability or instability, wherein the action control includes: establishment control, collection control, transmission control, playback control and termination control.

[0020] The second object of the present invention is achieved in that the method comprises the steps of:

[0021] Acquire audio and video data corresponding to the smart terminal respectively, and forward and process the audio and video data in real time in conjunction with the smart agent;

[0022] According to the data processed by the intelligent agent forwarding, the data is broadcasted at the network transport layer through the fusion communication exchange center cluster and using TCP, UDP unicast or multicast mode.

[0023] Furthermore, the steps of respectively acquiring audio and video data corresponding to the smart terminals and forwarding the audio and video data in real time in conjunction with the smart agent further include:

[0024] Acquire first synchronization control data corresponding to the cluster group call communication; wherein the first synchronization control data is consistency synchronization control data between distributed intelligent agents;

[0025] According to the first synchronization control data, a multi-intelligent agent network containment consistency model of distributed intelligent agents is established.

[0026] Furthermore, the establishing of a multi-intelligent agent network containment consistency model of distributed intelligent agents based on the first synchronization control data further includes:

[0027] Generate an intelligent proxy and obtain the connection address corresponding to MECS, establish an intelligent proxy TCP connection corresponding to MECS, and create a fully connected multi-intelligent proxy network;

[0028] Based on the shortest path algorithm, calculate and generate the shortest delay data to other intelligent agents in the fully connected network;

[0029] The intelligent proxy sends its own group call status and receives group call status from all neighbors, while updating the group call status in real time based on its own dynamic equations.

[0030] Check whether the group call state of the intelligent agent itself is stable, and generate action control instruction data corresponding to stability or instability; wherein, the action control includes: establishment control, collection control, transmission control, playback control and termination control.

[0031] The third object of the present invention is achieved as follows: it includes a processor, a memory and an intelligent agent fusion communication platform control program based on a neural network model; wherein the intelligent agent fusion communication platform control program based on a neural network model is executed by the processor, the intelligent agent fusion communication platform control program based on a neural network model is stored in the memory, and the intelligent agent fusion communication platform control program based on a neural network model implements the intelligent agent fusion communication method based on a neural network model.

[0032] The present invention includes at least one intelligent agent unit for converting audio and video data through the intelligent agent fusion communication architecture, as well as a fusion communication switching center unit and at least one intelligent terminal unit that communicate with the intelligent agent unit respectively; the intelligent agent fusion communication architecture adopts a distributed intelligent agent cluster to realize group call control, and learns link quality control through an intelligent agent neural network model, as well as methods and platforms corresponding to the architecture, and introduces artificial intelligence and blockchain into the IMS fusion communication architecture to solve the problem of accessing multi-protocol fusion switching communications of different standards and types in a complex environment of large-scale terminals, forming uninterrupted transmission capabilities of multiple links, and supporting safe, reliable and high-quality communication transmission of various business application data such as voice, video, and conferencing.

[0033] In other words, the solution of the present invention ensures high-quality audio and video communications through adaptive learning of link quality control using an intelligent agent neural network model. Each intelligent terminal is responsible for interaction with an intelligent agent, and the converged communication exchange center only needs to communicate with the intelligent agent, increasing the reliability of large-scale broadcast communications. The intelligent agent and the converged communication exchange center uniformly use the SIP protocol, reducing the need for multiple protocols to be connected to the converged communication exchange center. The intelligent agent integrates into the advanced IMS architecture as a user agent to obtain the required service quality and better support registration, security, billing, bearer control, roaming and other services in converged communications. The intelligent agent acts as a blockchain node, using blockchain-based authentication, authorization, and encryption to ensure system security and prevent threats such as theft and tampering. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0035] FIG1 is a flow chart of an embodiment of an intelligent agent fusion communication architecture based on a neural network model according to the present invention;

[0036] FIG2-a is a flow chart of a second embodiment of an intelligent agent fusion communication architecture based on a neural network model according to the present invention;

[0037] FIG2-b is a schematic diagram of a third process of an embodiment of an intelligent agent fusion communication architecture based on a neural network model of the present invention;

[0038] FIG3 is a schematic diagram of an intelligent agent fusion communication architecture based on a neural network model according to the present invention;

[0039] FIG4 is a flow chart of an intelligent agent fusion communication method based on a neural network model according to the present invention;

[0040] FIG5 is a schematic diagram of an intelligent agent fusion communication platform based on a neural network model according to the present invention;

[0041] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0042] In order to better understand the purpose, technical solutions and advantages of the present invention, the present invention is further described below with reference to the accompanying drawings and specific embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification.

[0043] The present invention may also be implemented or applied through other different specific examples, and the details in this specification may also be modified and changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0044] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the various components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0045] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of such features. Secondly, the technical solutions between the various embodiments can be combined with each other, but this must be based on the fact that ordinary technicians in this field can implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0046] Preferably, the present invention's intelligent agent fusion communication method based on a neural network model is applied to one or more terminals or servers. The terminal is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc.

[0047] The terminal can be a computing device such as a desktop computer, notebook, PDA, cloud server, etc. The terminal can interact with the client through a keyboard, mouse, remote control, touchpad, or voice control device.

[0048] The present invention is to realize an intelligent agent fusion communication architecture, method and platform based on a neural network model.

[0049] As shown in FIG4 , it is a flow chart of an intelligent agent fusion communication method based on a neural network model provided by an embodiment of the present invention.

[0050] In this embodiment, the intelligent agent fusion communication method based on the neural network model can be applied to terminals with display functions or fixed terminals. The terminals are not limited to personal computers, smart phones, tablet computers, desktop computers or all-in-one computers equipped with cameras, etc.

[0051] The neural network model-based intelligent agent converged communication method can also be applied to a hardware environment consisting of a terminal and a server connected to the terminal via a network. The network includes, but is not limited to, a wide area network, a metropolitan area network, or a local area network. The neural network model-based intelligent agent converged communication method of the present invention can be executed by a server, a terminal, or both.

[0052] For example, for a terminal that needs to perform intelligent proxy convergence communication based on a neural network model, the intelligent proxy convergence communication function based on a neural network model provided by the method of the present invention can be directly integrated into the terminal, or a client for implementing the method of the present invention can be installed. For another example, the method provided by the present invention can also be run on a server or other device in the form of a software development kit (SDK). The SDK provides an interface for the intelligent proxy convergence communication function based on a neural network model. The terminal or other device can implement the intelligent proxy convergence communication function based on a neural network model through the provided interface.

[0053] The present invention will be further described below with reference to the accompanying drawings.

[0054] As shown in Figures 1-5, the present invention provides an intelligent agent fusion communication architecture based on a neural network model.

[0055] The intelligent agent fusion communication architecture includes at least one intelligent agent unit for converting audio and video data, and a fusion communication switching center unit and at least one intelligent terminal unit respectively communicating with the intelligent agent unit;

[0056] The intelligent agent fusion communication architecture adopts a distributed intelligent agent cluster to realize group call control, and learns link quality control through an intelligent agent neural network model.

[0057] The converged communication switching center unit broadcasts data transmission at the network transport layer in a TCP, UDP unicast or multicast manner.

[0058] The intelligent agent fusion communication architecture is further provided with a synchronization processing unit, which is used to synchronize and coordinate the clusters, and to synchronously collect, transmit and play the voice and video data streams.

[0059] The intelligent agent fusion communication architecture is further provided with a first creation unit, and the first creation module is used to establish a multi-intelligent agent network containment consistency model of the distributed intelligent agent.

[0060] The first creation unit further includes:

[0061] Generate and obtain modules, used to generate intelligent agents and obtain the connection address corresponding to MECS;

[0062] The first calculation module is used to calculate the delay data from the fully connected network to the intelligent agent;

[0063] The self-check generation module is used to check whether the state of its own group call is stable and generate action control instruction data corresponding to stability or instability, wherein the action control includes: establishment control, collection control, transmission control, playback control and termination control.

[0064] Specifically, in an embodiment of the present invention, link quality control is learned through an intelligent agent neural network model, and complex transmission timing problems are solved based on reliable retransmission, forward correction (FEC), service priority, end-to-end rate control, etc., to ensure the stability of the transmission link, reduce delays and eliminate central bottlenecks, and reduce network costs, thereby providing guarantees for high-quality audio and video communications; as shown in Figure 2-a.

[0065] As shown in Figure 1, each smart terminal is responsible for interacting with an intelligent agent. All audio and video data are converted through the intelligent agent. The converged communication exchange center only needs to communicate with the intelligent agent. It can use TCP, UDP unicast or multicast methods to speed up broadcast data transmission at the network transport layer and increase the reliability of large-scale broadcast communications.

[0066] In terms of collaboration, according to the characteristics of cluster communication, group calling is a key business. In traditional cluster systems, group call control is implemented through a central centralized control terminal. In the present invention, a distributed intelligent agent cluster is used to implement it. This cluster must be synchronized and coordinated to ensure the synchronous collection, transmission and playback of audio and video data streams.

[0067] Distributed collaborative control includes distributed coordination technology and consistent synchronization control technology. Distributed coordination technology can be implemented using the mature Zookeeper component. Zookeeper is a distributed coordination service that manages a large number of hosts. It features reliability, scalability, and transparency. It provides services such as naming, load balancing, configuration management, cluster management, node leader election, distributed queues, distributed locks, and data registry required for coordination between distributed intelligent agents.

[0068] In order to meet the requirements of consistent synchronization control between distributed intelligent agents in cluster group call communications, the states and changes of the intelligent agents must be consistent, which is very similar to the multi-intelligent agent network system and can be solved by establishing a multi-intelligent agent network constraint consistency model of distributed intelligent agents.

[0069] Consider a network consisting of N nodes, where each node is represented as a first-order n-dimensional dynamical system. Nodes are coupled through interaction state information. The dynamic behavior of the i-th node in the network can be described by the following equation:

[0070] where x i (t)∈R n ,c is the coupling strength, Γ is the internal coupling matrix, a ij is the adjacency matrix, u i (t)∈R n is a linear feedback controller, d i τ (t) is the controller gain, s τ (t) is the desired equilibrium state of the system, and τ is the desired equilibrium state change time.

[0071] when When, through u i (t) Feedback control input, the network system can achieve the same control, L is the Lapacian matrix of the system, the diagonal matrix D = diag (d1, ..., dN ), d i is the control gain of node i.

[0072] When the system reaches a consistent equilibrium, x r (t)→s τ (t), x i (t)→x r (t), so x i (t)→s τ (t), that is, the system consistently reaches s τ (t) state, thus changing the state of the controlling node s τ (t), will make the status of all nodes in the network change synchronously.

[0073] Based on the consistency model of multi-agent network, the following consistency synchronization control algorithm between intelligent agents is proposed:

[0074] 1. In the initialization state, an intelligent agent is generated in the MECS, communicates with the center, obtains the connection address of other MECS, and establishes a TCP connection with the intelligent agents in other MECS in the network. As the connection of the intelligent agent, a fully connected multi-intelligent agent network is formed. The communication delay T of the connection is obtained by interacting with all multi-intelligent agents through TCP connection. ij .

[0075] 2. The intelligent agent calculates the shortest delay to other intelligent agents in the fully connected network according to Dijkstra's shortest path algorithm, and retains the neighbor connection with the shortest delay. Its communication delay T ij As a ij , forming an adjacency list.

[0076] 3. Through the reserved TCP connection with the neighbor, the intelligent agent i sends its own group call status x to all neighboring intelligent agents. i (t), and receive the group call status x sent by all neighbors j≠i j (t).

[0077] 4. Intelligent agent i updates the group call state x according to its own dynamic equation (1) i (t). The group call initiating intelligent agent r acts as a restraining node, and its control gain d can be set to 1. The intelligent agent r changes s according to the group call status. τ (t) Adjust the network to be consistent. After the group call state is established, the collection, transmission and playback cycle state changes are automatically performed at time intervals τ until termination.

[0078] 5. Intelligent agent i checks whether its group call status is stable (satisfying || x i (t+Δτ)-x i(t)||<∈), if it is stable, execute the corresponding action (establish, collect, transmit, play, terminate, etc.).

[0079] 6. Repeat step 3 at time intervals T.

[0080] This collaborative control algorithm allows multiple intelligent agent networks to reach consensus, and when the dynamic update clock T and the maximum network latency are significantly less than the state change time τ, intelligent agents can update their states based solely on neighbor changes, ensuring consistency across the entire network. This eliminates the need for a central coordination center. When the planned state serves as the sampling and output control state for voice, a consistent, synchronized collection, distribution, and playback network is formed, enabling group calling.

[0081] As shown in Figure 1, the intelligent agent and the converged communication exchange center use the SIP protocol uniformly, exchanging audio, video, data, and messages based on the same communication standard, reducing the need to connect multiple protocols in the converged communication exchange center;

[0082] Intelligent agents are integrated into the advanced IMS architecture as UAs, enabling various types of terminals to establish peer-to-peer IP communications and obtain the required quality of service while completing the necessary functions for the service, such as registration, security, billing, bearer control, and roaming.

[0083] As a blockchain node, the intelligent agent performs authentication protection and authorization functions based on the blockchain. All authentication security information is stored on the chain, restricting access by illegal users and ensuring the security of the system. At the same time, digital encryption technology is used to ensure the confidentiality of calls and protect them from threats such as theft and tampering.

[0084] Intelligent agents have become a key concept in computer science and artificial intelligence in recent years. They refer to computational entities that reside in a specific environment, perceive the environment, and operate autonomously to achieve a set of goals on behalf of their designers or users. Intelligent agents can perceive, learn, reason, and act, and after training based on a knowledge base, they can mimic human behavior, thus possessing intelligence. They possess the following key attributes: Autonomy: Intelligent agents can independently control their state and behavior, operating and running without human or other program intervention. Perception and responsiveness: Intelligent agents can promptly perceive and respond to changes in their environment. Agency: Intelligent agents can proactively exhibit goal-driven behavior and independently choose appropriate actions at the right time. Communication: Intelligent agents can exchange information and interact with other entities using some form of communication. Persistence: Intelligent agents operate continuously or continuously, and their state should remain consistent throughout their operation. Reasoning and planning: Intelligent agents are capable of performing relevant reasoning and intelligent computations based on learned knowledge and experience.

[0085] Blockchain technology is a new distributed infrastructure and computing paradigm that uses block chain data structures to verify and store data, distributed node consensus algorithms to generate and update data, cryptography to ensure the security of data transmission and access, and smart contracts composed of automated script codes to program and operate data. It has the following characteristics: Decentralization: Blockchain technology does not rely on additional third-party management agencies or hardware facilities, and has no central control. Beyond the self-contained blockchain itself, each node achieves self-verification, transmission, and management of information through distributed accounting and storage. Decentralization is the most prominent and essential feature of blockchain. Openness: The foundation of blockchain technology is open source. Except for the private information of the transaction parties, which is encrypted, blockchain data is open to everyone. Anyone can query blockchain data and develop related applications through public interfaces, making the entire system highly transparent. Independence: Based on consensus-based specifications and protocols (such as hashing algorithms and other mathematical algorithms), the entire blockchain system does not rely on other third parties. All nodes can automatically and securely verify and exchange data within the system without any human intervention. Security: As long as you do not control 51% of all data nodes, you cannot arbitrarily manipulate or modify network data. This makes the blockchain itself relatively secure and avoids subjective data changes. Anonymity: Unless required by law, from a technical perspective, the identity information of each block node does not need to be disclosed or verified, and information transmission can be anonymous.

[0086] To achieve the above object, the present invention further provides an intelligent agent fusion communication method based on a neural network model, as shown in FIG4 , the method comprising the following steps:

[0087] S1. Acquire audio and video data corresponding to the smart terminal respectively, and forward and process the audio and video data in real time in conjunction with the smart agent;

[0088] S2. Based on the data processed by the intelligent agent forwarding, the data is broadcasted and transmitted at the network transport layer through the converged communication exchange center cluster using TCP, UDP unicast or multicast.

[0089] The method of respectively acquiring audio and video data corresponding to the intelligent terminal and forwarding the audio and video data in real time in conjunction with the intelligent agent further includes:

[0090] S11. Acquire first synchronization control data corresponding to the cluster group call communication; wherein the first synchronization control data is consistency synchronization control data between distributed intelligent agents;

[0091] S12. Establish a multi-intelligent agent network containment consistency model of distributed intelligent agents based on the first synchronization control data.

[0092] The step of establishing a multi-intelligent agent network containment consistency model of distributed intelligent agents according to the first synchronization control data further includes:

[0093] S121. Generate an intelligent proxy and obtain the connection address corresponding to the MECS, establish an intelligent proxy TCP connection corresponding to the MECS, and create a fully connected multi-intelligent proxy network;

[0094] S122. Calculate and generate the shortest delay data to other intelligent agents in the fully connected network based on the shortest path algorithm;

[0095] S123, sending its own group call status through the intelligent proxy, and receiving the group call status sent by all neighbors, while updating the group call status in real time according to its own dynamic equation;

[0096] S124. Check whether the group call state of the intelligent agent itself is stable, and generate action control instruction data corresponding to stability or instability; wherein the action control includes: establishment control, collection control, transmission control, playback control and termination control.

[0097] That is to say, in the solution of the present invention, the specific steps of the intelligent agent fusion communication method based on the neural network model are as follows:

[0098] S00, the smart terminal is connected to the smart agent. The smart agent, as a blockchain participating node, uses the regional chain network for identity authentication. If the authentication is passed, the connection is maintained to continue the session, otherwise the connection is disconnected and the session is ended. S01, after the smart terminal is authenticated, the audio and video timing data is sent to the smart agent when a call is needed. The smart agent first puts the received audio and video timing data into the receiving queue, inputs it into the link neural network model for link quality assessment, forward correction and end-to-end rate control, and outputs it to the next level of priority and retransmission control processing. It combines reliable retransmission and service priority to put it into the corresponding sending queue cache, and then forwards the audio and video data in real time and sends it to the converged communication switching center cluster.

[0099] As shown in Figure 2-b, link quality assessment is performed based on the GNN graph neural network, forward correction is completed based on the LSTM long short-term memory network, and end-to-end rate control is implemented based on the DNN deep neural network. When the audio and video time series data of multiple intelligent agents are input into the GNN, the GNN comprehensively evaluates the link quality of each intelligent terminal, and the LSTM synchronously converts and forward predicts the audio and video time series data. The GNN evaluation results and the audio and video data after LSTM correction are input into the DNN, which performs adaptive rate adjustment on multiple terminals, overall coordination to ensure the quality of audio and video transmission, and outputs the audio and video data to the next level for retransmission and priority control.

[0100] S02. Based on the data processed by the intelligent agent forwarding, hybrid selection or forwarding distribution is performed through the fusion communication exchange center cluster, and TCP, UDP unicast or multicast is used to broadcast data transmission at the network transport layer.

[0101] The method of respectively acquiring audio and video data corresponding to the intelligent terminal and forwarding the audio and video data in real time in conjunction with the intelligent agent further includes:

[0102] S011. Acquire first synchronization control data corresponding to the cluster group call communication; wherein the first synchronization control data is consistency synchronization control data between distributed intelligent agents;

[0103] S012. Establish a multi-intelligent agent network containment consistency model of distributed intelligent agents based on the first synchronization control data.

[0104] The step of establishing a multi-intelligent agent network containment consistency model of distributed intelligent agents according to the first synchronization control data further includes:

[0105] S0121. Generate an intelligent proxy and obtain the connection address corresponding to the MECS, establish an intelligent proxy TCP connection corresponding to the MECS, and create a fully connected multi-intelligent proxy network;

[0106] S0122. Calculate and generate the shortest delay data to other intelligent agents in the fully connected network based on the shortest path algorithm;

[0107] S0123. Send its own group call status through the intelligent proxy, receive the group call status sent by all neighbors, and update the group call status in real time according to its own dynamic equation;

[0108] S0124. Check whether the group call state of the intelligent agent itself is stable, and generate action control instruction data corresponding to stability or instability; wherein, the action control includes: establishment control, collection control, transmission control, playback control and termination control.

[0109] S03, the audio and video timing data sent by the fusion communication exchange center cluster to the smart terminal is processed similarly to S1. The intelligent agent first receives it and puts it into the receiving queue, inputs it into the link neural network model for link quality assessment, forward correction and end-to-end rate control, and outputs it to the next level of priority and retransmission control processing. Combined with reliable retransmission and service priority, it is put into the corresponding sending queue cache, and then the audio and video data is forwarded and processed in real time and sent to the smart terminal.

[0110] In the embodiment of the method scheme of the present invention, the functional modules involved in the intelligent agent fusion communication method based on the neural network model have been described above in detail and will not be repeated here.

[0111] To achieve the above objectives, the present invention also provides an intelligent proxy fusion communication platform based on a neural network model, as shown in FIG5 , comprising a processor, a memory, and an intelligent proxy fusion communication platform control program based on a neural network model; wherein the processor executes the intelligent proxy fusion communication platform control program based on a neural network model, the intelligent proxy fusion communication platform control program based on a neural network model is stored in the memory, and the intelligent proxy fusion communication platform control program based on a neural network model implements the steps of the intelligent proxy fusion communication method based on a neural network model, for example:

[0112] S1. Acquire audio and video data corresponding to the smart terminal respectively, and forward and process the audio and video data in real time in conjunction with the smart agent;

[0113] S2. Based on the data processed by the intelligent agent forwarding, the data is broadcasted and transmitted at the network transport layer through the converged communication exchange center cluster using TCP, UDP unicast or multicast.

[0114] The specific details of the steps have been explained above and will not be repeated here.

[0115] In an embodiment of the present invention, the built-in processor of the intelligent agent fusion communication platform based on the neural network model can be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and a combination of various control chips. The processor utilizes various interfaces and circuits to connect various components, and executes or runs programs or units stored in the memory, as well as calls data stored in the memory, to perform various functions of the intelligent agent fusion communication based on the neural network model and process data.

[0116] The memory is used to store program codes and various data. It is installed in the intelligent agent fusion communication platform based on the neural network model and realizes high-speed and automatic access to programs or data during operation.

[0117] The memory includes read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electronically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0118] The present invention includes at least one intelligent agent unit for converting audio and video data, and a fusion communication switching center unit and at least one intelligent terminal unit that communicate with the intelligent agent unit respectively through the intelligent agent fusion communication architecture; the intelligent agent fusion communication architecture adopts a distributed intelligent agent cluster to realize group call control, and learns link quality control through an intelligent agent neural network model; and methods and platforms corresponding to the architecture, and introduces artificial intelligence and blockchain into the IMS fusion communication architecture to solve the problem of accessing multi-protocol fusion switching communications of different standards and types in a complex environment of large-scale terminals, forming uninterrupted transmission capabilities of multiple links, and supporting safe, reliable and high-quality communication transmission of various business application data such as voice, video, and conferencing.

[0119] In other words, the solution of the present invention ensures high-quality audio and video communications through adaptive learning of link quality control using an intelligent agent neural network model. Each intelligent terminal is responsible for interaction with an intelligent agent, and the converged communication exchange center only needs to communicate with the intelligent agent, increasing the reliability of large-scale broadcast communications. The intelligent agent and the converged communication exchange center uniformly use the SIP protocol, reducing the need for multiple protocols to be connected to the converged communication exchange center. The intelligent agent integrates into the advanced IMS architecture as a user agent to obtain the required service quality and better support registration, security, billing, bearer control, roaming and other services in converged communications. The intelligent agent acts as a blockchain node, using blockchain-based authentication, authorization, and encryption to ensure system security and prevent threats such as theft and tampering.

[0120] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. An intelligent agent fusion communication architecture based on a neural network model, characterized in that: The intelligent proxy fusion communication architecture includes at least one intelligent proxy unit for converting audio and video data, and a fusion communication switching center unit and at least one intelligent terminal unit that communicate with the intelligent proxy unit respectively; The intelligent proxy fusion communication architecture adopts a distributed intelligent proxy cluster to realize group call control, and learns link quality control through an intelligent proxy neural network model.

2. According to claim 1, the intelligent agent fusion communication architecture based on the neural network model is characterized in that: The converged communication switching center unit broadcasts data transmission in the network transport layer through TCP, UDP unicast or multicast.

3. According to the intelligent agent fusion communication architecture based on the neural network model according to claim 1 or 2, it is characterized in that: The intelligent proxy fusion communication architecture is also provided with a synchronization processing unit, which is used to synchronize and coordinate the clusters, and to synchronously collect, transmit and play the audio and video data streams.

4. According to claim 3, the intelligent agent fusion communication architecture based on the neural network model is characterized in that: The intelligent agent fusion communication architecture is also provided with a first creation unit, and the first creation module is used to establish a multi-intelligent agent network containment consistency model of distributed intelligent agents.

5. According to claim 4, the intelligent agent fusion communication architecture based on the neural network model is characterized in that: The first creation unit further includes: Generate an acquisition module, used to generate an intelligent proxy and obtain a connection address corresponding to the MECS; The first calculation module is used to calculate the delay data from the fully connected network to the intelligent agent; The self-check generation module is used to check whether the group call state is stable and generate action control instruction data corresponding to stability or instability, wherein the action control includes: establishment control, collection control, transmission control, playback control and termination control.

6. An intelligent agent fusion communication method based on a neural network model, characterized in that: The method comprises the following steps: Respectively obtain audio and video data corresponding to the intelligent terminal, and forward and process the audio and video data in real time in conjunction with the intelligent agent; According to the data processed by the intelligent proxy forwarding, the data is broadcasted and transmitted at the network transport layer through the fusion communication exchange center cluster and TCP, UDP unicast or multicast.

7. The intelligent agent fusion communication method based on a neural network model according to claim 6 is characterized in that: The method of respectively acquiring audio and video data corresponding to the intelligent terminal and forwarding the audio and video data in real time in combination with the intelligent agent also includes: Acquire first synchronization control data corresponding to the cluster group call communication; wherein the first synchronization control data is consistency synchronization control data between distributed intelligent agents; According to the first synchronization control data, a multi-intelligent agent network containment consistency model of distributed intelligent agents is established.

8. The intelligent agent fusion communication method based on a neural network model according to claim 7 is characterized in that: The step of establishing a multi-intelligent agent network containment consistency model of a distributed intelligent agent according to the first synchronization control data further includes: Generate an intelligent proxy and obtain the connection address corresponding to MECS, establish an intelligent proxy TCP connection corresponding to MECS, and create a fully connected multi-intelligent proxy network; According to the shortest path algorithm, calculate and generate the shortest delay data to other intelligent agents in the fully connected network; Send its own group call status through the intelligent proxy, receive the group call status sent by all neighbors, and update the group call status in real time according to its own dynamic equation; Check whether the group call state of the intelligent agent itself has reached stability, and generate action control instruction data corresponding to stability or instability; wherein the action control includes: establishment control, collection control, transmission control, playback control and termination control.

9. An intelligent agent fusion communication platform based on a neural network model, characterized in that: It includes a processor, a memory and an intelligent proxy fusion communication platform control program based on a neural network model; wherein the intelligent proxy fusion communication platform control program based on a neural network model is executed on the processor, the intelligent proxy fusion communication platform control program based on a neural network model is stored in the memory, and the intelligent proxy fusion communication platform control program based on a neural network model implements the intelligent proxy fusion communication method based on a neural network model as described in any one of claims 6 to 8.

Images