Intelligent agent system and method based on MCP technology

Abstract

The invention relates to the technical field of computer networks, in particular to an intelligent agent system and method based on the MCP technology, and the system comprises an MCP intermediate layer protocol engine, an intelligent scheduling module, a channel management system, a security enhancement mechanism and an edge node self-learning mechanism. The method has the beneficial effects that support for various communication protocols (such as HTTP, MQTT, gRPC and the like) is realized through the MCP technology. Through the protocol adapter module, the system can automatically identify and load different protocol adapters, and parallel work of multiple protocols can be supported without manual intervention. The multi-protocol support greatly improves the flexibility of the system, so that the system can be seamlessly connected with equipment of different manufacturers and standards, and the requirements of complex and diverse application scenes are met.

Description

Technical Field

[0001] The present invention relates to the field of computer network technology, and in particular to an intelligent agent system and method based on MCP technology. Background Art

[0002] In today's complex and diverse network environment, a variety of communication protocols exist, such as HTTP, MQTT, gRPC, and WebSocket. Different application scenarios and devices may utilize different communication protocols, which poses significant challenges for traditional communication systems when handling cross-protocol data communication. For example, HTTP is widely used in web services, while MQTT is commonly used for communication between IoT devices. When data exchange between devices using these two different protocols is required, traditional systems often require complex conversion mechanisms, increasing system complexity and development costs. Furthermore, the data formats and communication rules differ significantly between different protocols, making data conversion between different protocols prone to errors and reducing the accuracy and reliability of data transmission.

[0003] Furthermore, traditional communication systems often lack intelligent scheduling mechanisms for network data transmission. They typically employ fixed transmission paths and protocols, failing to dynamically adjust to real-time network conditions. For example, when network bandwidth is low and latency is high, the system still uses bandwidth-intensive protocols for data transmission, resulting in slow data transmission and even data loss. Furthermore, traditional systems fail to fully consider resource utilization when handling large-scale data transmission, which can easily cause network congestion and impact overall system performance.

[0004] As cyberattacks continue to escalate, data transmission security is becoming increasingly important. However, traditional communication systems suffer from numerous security deficiencies. For one thing, the encryption algorithms they employ can be simplistic and easily cracked, failing to effectively protect data integrity and confidentiality. Furthermore, traditional systems lack real-time monitoring and early warning mechanisms, making it difficult to promptly detect and address cyberattacks and malicious intrusions, leading to security incidents such as data leaks and service interruptions. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent agent system and method based on MCP technology to solve the problems raised in the above background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solutions: an intelligent agent system based on MCP technology, comprising:

[0007] MCP middle-layer protocol engine: As the core module of the system, it is designed based on the principles of modern protocol abstraction and interface unification. It encapsulates the specific implementations of multiple communication protocols into a unified processing layer, supporting the encapsulation, decapsulation, and adaptation of common network communication protocols. Each protocol has an adapter module responsible for converting protocol data into a unified internal format to achieve interoperability between different protocols. The protocol adapter module has dynamic loading and switching functions and can perform data compression and encryption during the message conversion process. It adopts a protocol abstraction layer to define a unified interface standard, which is transparent to upper-layer applications, reduces compatibility issues between different protocols, and facilitates the addition of new protocols.

[0008] Intelligent Scheduling Module: This module intelligently schedules data transmission paths and selects the optimal communication protocol based on real-time network conditions, historical access behavior, and business needs. It analyzes network status by monitoring network bandwidth, latency, and packet loss rate, and dynamically selects the most suitable transmission channel based on a policy-based scheduling algorithm. It makes predictions based on historical data and access behavior, selecting the optimal communication protocol and transmission path in advance. It dynamically adjusts data transmission priorities based on different device priorities, task importance, and transmission rate requirements.

[0009] Channel Management System: Unified management and scheduling of multi-protocol data channels. Using a connection pool mechanism, a fixed-size connection pool is set for each protocol, which automatically expands or contracts based on the current load. Dynamic monitoring of the traffic and load of each channel occurs, automatically switching to a backup channel when traffic is too high. Using asynchronous I / O technology, data transmission operations do not block the main thread, improving system throughput and response speed. Built-in load balancing function automatically distributes traffic based on the load of each channel.

[0010] Security Enhancement Mechanism: A series of security enhancement mechanisms are integrated, using advanced encryption algorithms to encrypt all data during transmission, automatically selecting the appropriate encryption method based on the protocol type, and supporting end-to-end encryption. An integrated intrusion detection system (IDS) monitors network traffic and communication behavior in real time, immediately issuing alerts and taking protective measures when abnormal traffic or suspicious activity is detected. A built-in refined access control and permission management mechanism supports the role-based access control (RBAC) model, allowing administrators to flexibly configure permissions for different users and devices. Multi-factor authentication (MFA) and single sign-on (SSO) security authentication technologies are also supported.

[0011] Edge node self-learning mechanism: In edge computing scenarios, by collecting historical data from each edge node, building a behavioral model, and analyzing it using machine learning algorithms, edge nodes can dynamically adjust their behavior based on actual conditions; dynamically select the optimal routing and communication path based on the current node network status and task requirements; and dynamically adjust computing task allocation based on device computing power and network conditions.

[0012] Preferably, in the MCP middle layer protocol engine, when the protocol adapter module processes protocol data encapsulation and parsing, the complexity and redundancy of the protocol stack are reduced through preprocessing technology; each protocol adapter module is only coupled with the specific implementation related to the protocol and is completely transparent to the upper layer application.

[0013] Preferably, in the intelligent scheduling module, the policy scheduling algorithm selects the communication protocol according to different business scenarios. When the network bandwidth is low and the latency is high, a low-latency, high-compression ratio protocol is selected; when the network bandwidth is high and the stability is strong, HTTP or gRPC protocol is selected.

[0014] Preferably, in the security enhancement mechanism, the intrusion detection system IDS monitors network traffic and communication behavior in real time, and the protective measures taken include disconnecting and blocking IP.

[0015] Preferably, in the edge node self-learning mechanism, the edge node learns the best communication strategy for different devices and application scenarios, selects low-latency, high-throughput communication protocols for high-frequency tasks, and selects low-power, low-bandwidth consumption protocols for low-frequency tasks.

[0016] A method for an intelligent agent system based on MCP technology, comprising the following steps:

[0017] Data processing based on the MCP middle-layer protocol engine: The MCP middle-layer protocol engine is designed based on the principles of modern protocol abstraction and interface unification, encapsulating the specific implementations of multiple communication protocols into a unified processing layer. An adapter module is set up for each communication protocol, which converts the protocol data into a unified internal format to achieve interoperability between different protocols. During the data conversion process, preprocessing technology is used to reduce the complexity and redundancy of the protocol stack, and protocol adapters are dynamically loaded and switched according to demand. A protocol abstraction layer is used to define a unified interface standard, so that the protocol abstraction layer provides an interface that is transparent to upper-layer applications based on different business scenarios. Each protocol adapter module is only coupled with the specific implementation related to the protocol and is completely transparent to upper-layer applications.

[0018] Data transmission path and protocol selection based on the intelligent scheduling module: monitor network bandwidth, latency, packet loss rate information, and analyze network status; dynamically select the most suitable transmission channel and optimal communication protocol based on the policy scheduling algorithm, combined with real-time network conditions, historical access behavior, and business needs; when the network bandwidth is low and the latency is high, choose a low-latency, high-compression ratio protocol; for high-bandwidth, high-stability network environments, choose HTTP or gRPC protocols; if the device access pattern has a periodic pattern, predict it based on historical access data and select the optimal communication protocol and transmission path in advance; dynamically adjust the data transmission priority based on the priority of different devices, task importance, and transmission rate requirements, and prioritize the completion of critical tasks when resources are limited;

[0019] Data channel management based on the channel management system: Using connection pool technology, a fixed-size connection pool is set for each protocol, reusing existing connections to avoid resource waste caused by frequent connection establishment and closure, and automatically expanding or shrinking the channel connection pool according to the current load situation; dynamically monitoring the traffic and load of each channel, and automatically switching to a backup channel when the load of a channel is too high; using asynchronous IO technology, each data transmission operation can be performed without blocking the main thread, improving system throughput and response speed; built-in load balancing function automatically distributes traffic according to the load of each channel, avoiding system bottlenecks caused by excessive traffic on a certain channel;

[0020] Data security based on security enhancement mechanisms: Advanced encryption algorithms are used to encrypt all data during transmission, automatically selecting the appropriate encryption method based on the protocol type, and supporting end-to-end encryption. An integrated intrusion detection system (IDS) monitors network traffic and communication behavior in real time, immediately issuing alerts upon detecting abnormal traffic or suspicious activity, and taking protective measures such as disconnecting and blocking IP addresses. A built-in refined access control and permission management mechanism supports the role-based access control (RBAC) model, allowing administrators to flexibly configure permissions for different users and devices. Multi-factor authentication (MFA) and single sign-on (SSO) security authentication technologies are also supported.

[0021] Dynamic adjustment based on the self-learning mechanism of edge nodes: In edge computing scenarios, historical data of each edge node is collected, behavioral models are constructed, and analysis is performed using machine learning algorithms. As the system runs, the edge nodes learn the optimal communication strategies for different devices and application scenarios, such as selecting low-latency, high-throughput communication protocols for high-frequency tasks, and low-power, low-bandwidth consumption protocols for low-frequency tasks. Based on the current network status of the node and task requirements, the optimal route and communication path are dynamically selected, and the allocation of computing tasks is dynamically adjusted according to the device computing power and network conditions.

[0022] Preferably, in the data processing step based on the MCP intermediate layer protocol engine, when the protocol adapter module processes the protocol data encapsulation and parsing, the complexity and redundancy of the protocol stack are reduced through preprocessing technology, and the protocol adapter can be dynamically loaded and switched according to demand to ensure the flexibility and scalability of the system.

[0023] Preferably, in the data transmission path and protocol selection step based on the intelligent scheduling module, the policy scheduling algorithm selects the communication protocol according to different business scenarios, and the scheduling strategy not only depends on the real-time network status, but also combines historical data and access behavior for prediction, and selects the best communication protocol and transmission path in advance.

[0024] Preferably, in the data security assurance step based on the security enhancement mechanism, the intrusion detection system IDS monitors network traffic and communication behavior in real time, and the protective measures taken include disconnecting and blocking IP.

[0025] Preferably, in the dynamic adjustment step based on the edge node self-learning mechanism, after the edge node learns the optimal communication strategy based on the collected historical data, it can dynamically select the optimal route and communication path according to the current node's network status and task requirements, thereby achieving flexible and efficient data processing capabilities and reducing dependence on the central server.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The intelligent proxy system and method based on MCP technology proposed in the present invention implements support for multiple communication protocols (such as HTTP, MQTT, gRPC, etc.) through MCP technology. Through the protocol adapter module, the system can automatically identify and load different protocol adapters, and support the parallel operation of multiple protocols without human intervention. This multi-protocol support greatly enhances the flexibility of the system, allowing the system to seamlessly connect devices from different manufacturers and standards to meet the needs of complex and diverse application scenarios.

[0028] The intelligent scheduling module intelligently selects the optimal communication path and protocol by collecting real-time information about network status, device behavior, and task requirements. This scheduling mechanism automatically optimizes resource allocation based on network conditions and load, reducing latency and packet loss, and improving communication efficiency. When network congestion or device requirements change, the system can quickly adjust to ensure efficient and stable data transmission, making it particularly suitable for dynamic and unstable network environments.

[0029] Asynchronous I / O and connection pooling mechanisms effectively support highly concurrent data transmission. Data flows for each protocol are decoupled through asynchronous processing, avoiding blockage and resource waste, and improving the system's processing power and responsiveness. Connection pooling technology allows connections for each protocol to be reused, reducing the frequent establishment and removal of connections and improving overall system throughput. It is particularly suitable for scenarios requiring the simultaneous management of multiple concurrent connections.

[0030] The system integrates multiple security mechanisms, such as end-to-end encryption, role-based access control (RBAC), and an intrusion detection system, to ensure the confidentiality and integrity of data during transmission. By utilizing encryption algorithms such as AES and RSA, the system effectively prevents man-in-the-middle attacks and data leaks. Furthermore, through identity verification and behavioral analysis, the system ensures that only authorized users and devices have access, thus safeguarding the security of the entire system. This makes it particularly suitable for applications requiring high data security. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0032] In order to clearly and completely describe the objectives and technical solutions of the present invention and make the advantages more clearly understood, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are part of the embodiments of the present invention, not all of them, and are only used to explain the embodiments of the present invention, not to limit the embodiments of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0033] In the first embodiment, the present invention provides a technical solution: an intelligent agent system based on MCP technology, including: an MCP middle layer protocol engine, an intelligent scheduling module, a channel management system, a security enhancement mechanism, an edge node self-learning mechanism and other functional modules.

[0034] First of all, in the initial stage of the solution implementation, the MCP middle-layer protocol engine is the core module of this system, responsible for supporting the encapsulation, decapsulation and adaptation of multiple communication protocols. The design of the MCP protocol engine is based on the principles of modern protocol abstraction and interface unification, encapsulating the specific implementations of multiple communication protocols into a unified processing layer. It is used to support common network communication protocols such as HTTP, MQTT, gRPC, WebSocket, etc. The adapter module of each protocol is responsible for converting the protocol data into a unified internal format to ensure that different protocols can communicate with each other. The protocol adapter module not only handles the encapsulation and parsing of the protocol data, but also reduces the complexity and redundancy of the protocol stack through preprocessing technology. The protocol adapter can be dynamically loaded and switched according to demand to ensure the flexibility and scalability of the system.

[0035] For example, the MQTT adapter is responsible for converting MQTT messages into the standard message format of the MCP protocol engine. During this conversion process, the adapter module compresses and encrypts the data to improve transmission efficiency and security. Similarly, for the HTTP protocol, the adapter parses the HTTP header and body and generates appropriate data blocks for transmission to downstream modules.

[0036] The MCP protocol engine utilizes a protocol abstraction layer, which defines unified interface standards to avoid interdependencies and coupling between different protocols. Based on different business scenarios, the protocol abstraction layer provides an interface that is transparent to upper-layer applications. Each protocol adapter module is coupled only to the specific implementation of that protocol, remaining completely transparent to upper-layer applications. This design not only reduces compatibility issues between different protocols but also makes the integration of new protocols simpler and more efficient.

[0037] Next is the intelligent scheduling module, the core of MCP technology. It intelligently schedules data transmission paths and selects the optimal communication protocol based on real-time network conditions, historical access behavior, and business needs. This module improves the system's overall performance and responsiveness by dynamically adjusting and optimizing resource utilization.

[0038] The scheduling module first analyzes the network status by monitoring information such as network bandwidth, latency, and packet loss rate, and dynamically selects the most suitable transmission channel based on the policy scheduling algorithm. For example, if the current network bandwidth is low and the latency is high, the system can choose a low-latency, high-compression protocol, such as MQTT, to ensure fast data transmission. For high-bandwidth, high-stability network environments, the system may choose HTTP or gRPC protocols to achieve efficient transmission of large-scale data. In actual applications, scheduling strategies not only rely on real-time network status, but also combine historical data and access behavior. For example, if the access pattern of a device has a periodic pattern, the system will make predictions based on its historical access data and select the optimal communication protocol and transmission path in advance. This prediction-based scheduling strategy can significantly improve the system's response speed and reduce latency and packet loss caused by network fluctuations.

[0039] The scheduling module further enhances performance through multi-dimensional optimization. For example, the system dynamically adjusts data transmission priorities based on factors such as device priority, task importance, and transmission rate requirements. When resources are limited, the system prioritizes critical tasks and ensures the smooth transmission of important data. A key advantage of this module lies in its adaptive capabilities, which respond in real time to changes in the network environment and adjustments in business needs, ensuring optimal communication quality in a variety of complex scenarios.

[0040] In addition to the scheduling module, the channel management system is responsible for unified management and scheduling of multi-protocol data channels. The system optimizes data transmission throughput and system responsiveness through connection pooling, efficient I / O management, and resource reuse. To improve system resource utilization, the channel management system employs connection pooling technology. By setting a fixed-size connection pool for each protocol, the system can reuse existing connections, avoiding the waste of resources caused by frequent connection establishment and closure. Each channel connection pool automatically expands or contracts based on current load to ensure efficient use of network resources.

[0041] Furthermore, to prevent overloading of a particular channel, the system dynamically monitors the traffic and load of each channel and schedules traffic as needed. If a channel becomes overloaded, the system automatically switches to a backup channel, ensuring stable and efficient communication. The channel management system also utilizes asynchronous I / O technology, allowing each data transfer operation to proceed without blocking the main thread, thereby improving system throughput and responsiveness. Asynchronous I / O allows the system to continue processing other requests while transferring data, reducing latency and CPU idle time.

[0042] In addition, the system features built-in load balancing, which automatically distributes traffic based on the load of each channel, preventing system bottlenecks caused by excessive traffic on a single channel. Load balancing not only optimizes resource utilization but also ensures smooth and stable data transmission.

[0043] Secondly, security is a crucial consideration. The intelligent agent system of the present invention ensures the integrity and confidentiality of data during transmission by integrating a series of security enhancement mechanisms, while preventing potential security attacks.

[0044] The system uses advanced encryption algorithms, such as AES and RSA, to encrypt all data during transmission. During each data transmission, the appropriate encryption method is automatically selected based on the protocol type, ensuring that data cannot be eavesdropped or tampered with during transmission. Furthermore, the system supports end-to-end encryption, ensuring that all communications from the source device to the destination device are securely protected. To prevent malicious attacks and intrusions, the system also integrates an intrusion detection system (IDS) to monitor network traffic and communication behavior in real time. Upon detecting abnormal traffic or suspicious activity, the IDS module immediately issues an alert and takes appropriate protective measures, such as disconnecting the connection or blocking the IP address. This mechanism significantly enhances system security, preventing security risks such as data leaks and service interruptions. The system also incorporates a refined access control and permission management mechanism, supporting a role-based access control (RBAC) model to ensure that only authorized users and devices can access the proxy system. Administrators can flexibly configure permissions for different users and devices as needed to prevent unauthorized access or manipulation. Support for security technologies such as multi-factor authentication (MFA) and single sign-on (SSO) further enhances system security.

[0045] Secondly, the edge node self-learning mechanism. In edge computing scenarios, the operating environment of devices and nodes is often unpredictable, and network conditions and task requirements can change at any time. The intelligent agent system of the present invention uses a self-learning mechanism to ensure that edge nodes can dynamically adjust their behavior based on actual conditions, improving communication efficiency and resource utilization. The system collects historical data from each edge node, such as access frequency, network latency, and task type, to build a behavioral model and analyze it using machine learning algorithms. As the system continues to operate, the edge nodes can learn the optimal communication strategies for different devices and application scenarios. For example, for high-frequency tasks, the system may select a low-latency, high-throughput communication protocol, while for low-frequency tasks, it may select a low-power, low-bandwidth protocol. The edge node self-learning mechanism is also reflected in intelligent routing and dynamic optimization. The system can dynamically select the optimal route and communication path based on the current node's network status and task requirements. This not only improves communication stability and efficiency, but also effectively reduces network load and avoids network congestion. With the rapid development of the Internet of Things and edge computing, traditional cloud computing architectures have gradually exposed problems such as low processing efficiency and slow response speed. By distributing computing tasks to edge nodes, the intelligent proxy system enables more flexible and efficient data processing. In edge computing scenarios, the system can dynamically adjust the distribution of computing tasks based on the computing power of the devices and network conditions, improving computing efficiency and reducing dependence on central servers.

[0046] Through this series of detailed technical steps and careful design considerations, this solution builds an efficient and intelligent knowledge base for operations and maintenance assistants by integrating big data technologies, natural language processing, machine learning, vector database technology, and the RAG model. This system not only significantly improves the efficiency and quality of operations and maintenance work, but also adapts to the ever-changing technical environment and business needs, providing strong support for enterprise operations and maintenance management. The system's design fully considers the complexity and variability of operations and maintenance scenarios. Through intelligent data processing and retrieval mechanisms, it can quickly respond to operations and maintenance personnel's queries and provide timely and accurate information support.

[0047] Example 2, based on Example 1, proposes a method for an intelligent agent system based on MCP technology according to claim 5, characterized in that it includes the following steps:

[0048] Data processing based on the MCP middle-layer protocol engine: The MCP middle-layer protocol engine is designed based on the principles of modern protocol abstraction and interface unification, and the specific implementations of multiple communication protocols are encapsulated into a unified processing layer. An adapter module is set for each communication protocol, and the adapter module converts the protocol data into a unified internal format to achieve intercommunication between different protocols. During the data conversion process, preprocessing technology is used to reduce the complexity and redundancy of the protocol stack, and protocol adapters are dynamically loaded and switched according to demand. A protocol abstraction layer is adopted to define a unified interface standard so that the protocol abstraction layer provides an interface that is transparent to upper-layer applications according to different business scenarios. Each protocol adapter module is only coupled with the specific implementation related to the protocol and is completely transparent to the upper-layer application. When the protocol adapter module handles protocol data encapsulation and parsing, preprocessing technology is used to reduce the complexity and redundancy of the protocol stack, and the protocol adapter can be dynamically loaded and switched according to demand to ensure the flexibility and scalability of the system.

[0049] Data transmission path and protocol selection based on the intelligent scheduling module: monitor network bandwidth, latency, packet loss rate information, and analyze network status; dynamically select the most suitable transmission channel and optimal communication protocol based on the policy scheduling algorithm, combined with real-time network conditions, historical access behavior and business needs; when the network bandwidth is low and the latency is high, select a low-latency, high-compression ratio protocol; for high-bandwidth, high-stability network environments, select HTTP or gRPC protocols; if the device access pattern has a periodic regularity, make predictions based on historical access data and select the best communication protocol and transmission path in advance; dynamically adjust the data transmission priority based on the priority, task importance, and transmission rate requirements of different devices, and give priority to ensuring the completion of critical tasks when resources are limited; the policy scheduling algorithm selects communication protocols according to different business scenarios, and the scheduling strategy not only relies on the real-time network status, but also combines historical data and access behavior for prediction, and selects the best communication protocol and transmission path in advance.

[0050] Data channel management based on the channel management system: Using connection pool technology, a fixed-size connection pool is set for each protocol, reusing existing connections to avoid resource waste caused by frequent connection establishment and closure, and automatically expanding or shrinking the channel connection pool according to the current load situation; dynamically monitoring the traffic and load of each channel, and automatically switching to the backup channel when the load of a channel is too high; using asynchronous IO technology, each data transmission operation can be performed without blocking the main thread, improving system throughput and response speed; built-in load balancing function, automatically distributing traffic according to the load situation of each channel, avoiding system bottlenecks caused by excessive traffic in a certain channel.

[0051] Data security protection based on security enhancement mechanism: Advanced encryption algorithms are used to encrypt all data during transmission, the appropriate encryption method is automatically selected according to the protocol type, and end-to-end encryption is supported; integrated intrusion detection system IDS monitors network traffic and communication behavior in real time, and immediately issues an alarm when abnormal traffic or suspicious activities are detected, and takes protective measures such as disconnection and IP blocking; built-in refined access control and permission management mechanism, supports role-based access control RBAC model, and administrators can flexibly configure permissions for different users and devices; supports multi-factor authentication MFA and single sign-on SSO security authentication technology; the intrusion detection system IDS monitors network traffic and communication behavior in real time, and the protective measures taken include disconnection and IP blocking.

[0052] Dynamic adjustment based on the self-learning mechanism of edge nodes: In edge computing scenarios, historical data of each edge node is collected, behavioral models are constructed, and analysis is performed using machine learning algorithms. As the system runs, the edge nodes learn the optimal communication strategies for different devices and application scenarios, such as selecting low-latency, high-throughput communication protocols for high-frequency tasks, and low-power, low-bandwidth consumption protocols for low-frequency tasks. Based on the current node's network status and task requirements, the optimal routing and communication paths are dynamically selected, and the allocation of computing tasks is dynamically adjusted based on the device computing power and network conditions. After the edge node learns the optimal communication strategy based on the collected historical data, it can dynamically select the optimal routing and communication path based on the current node's network status and task requirements, achieving flexible and efficient data processing capabilities and reducing dependence on central servers.

[0053] Example 3, based on Example 2, proposes: implementing the product deployment and operation assistant function based on a large language model and RAG, which is described in detail below.

[0054] In a certain campus, a large number of different types of terminal devices are deployed, including security cameras, temperature and humidity sensors, access control controllers, automatic lighting systems, energy management gateways, and AI edge servers. Data exchange between these devices is frequent and uses different protocols. For example, cameras use RTSP or WebSocket protocols for video streaming, sensors use MQTT to report status, AI edge servers receive analysis tasks via gRPC, and energy management devices use traditional HTTP / HTTPS interfaces. Due to different device manufacturers and varying system launch times, these communication protocols lack unified standards. As a result, traditional network proxies or gateway devices are inefficient and prone to errors when handling communication flows involving multiple protocols. Some protocols are even unable to communicate with each other, severely restricting the collaborative capabilities of the entire campus' intelligent system.

[0055] In order to solve this problem, the MCP intelligent agent system proposed in the present invention is introduced. The system is deployed on the edge server of the park and serves as an intermediate communication bridge between all terminal devices in the park and the core business system, realizing unified management, dynamic scheduling and security of multi-protocol device data. After the system is started, the MCP middle-layer protocol engine will automatically identify various types of devices in the access network and load the required protocol adapters according to the preset protocol adaptation strategy. The system supports hot-swappable loading of protocol modules. For example, when the system detects that the newly launched air quality monitoring equipment uses the CoAP protocol, it will automatically load the CoAP adapter module from the local protocol repository. The protocol identification and initialization can be completed without human intervention, and the equipment can be included in the unified communication system of the park.

[0056] Secondly, during actual operation, the temperature and humidity sensors regularly report current indoor environmental status data, including temperature, humidity, and air quality index, using the MQTT protocol. Upon receiving this data stream, the MCP protocol engine immediately converts it into the system's standard intermediate protocol format using an MQTT adapter and passes it to the intelligent scheduling module. Based on the current system load, network channel status, and processing node resource usage, the intelligent scheduling module determines whether to upload the data directly to the central data center or prioritize local pre-processing by the edge node. If latency increases in the connection channel to the central data center, the scheduling module automatically lowers the priority of that channel and switches to the AI ​​analysis engine on edge node A for pre-processing. The results are then transmitted back via the gRPC interface. Meanwhile, the camera modules reach peak operation in the evening, with each camera transmitting a large number of high-definition video streams using the RTSP protocol. The MCP system standardizes and centralizes the streaming data using a WebSocket / RTSP adapter, avoiding the lag and data loss that can occur with traditional proxies due to protocol inconsistencies during high-frequency video transmission.

[0057] Secondly, the channel management system plays a key role in the scheduling process. It maintains independent connection pools for communications with different protocols, and ensures stable responses in high-concurrency scenarios through event-driven asynchronous I / O mechanisms. When the AI ​​edge server requests to process task results, the system finds that the current channel load is close to saturation based on the communication performance records of the past few hours, so it temporarily creates a backup channel and automatically guides new tasks to switch to this channel for processing through the load balancer, significantly improving the response speed. In particular, when the energy management system performs load scheduling for the entire campus, each energy-consuming terminal must complete status synchronization within one minute. The MCP proxy system uses its low-latency characteristics to pre-cache necessary instruction data and optimize the protocol channel path to ensure that all devices complete communication within a limited time and return control results.

[0058] On the other hand, the system's security module ensures data encryption and verification of the legitimacy of communication identities throughout the entire process. When a device is first put online, the system automatically generates a digital certificate and binds it to the device's unique ID. Each communication is accompanied by an identity authentication process. All communication data is encrypted using AES, and control instructions use the RSA asymmetric encryption mechanism. Even if the data is intercepted in the middle, it cannot be decrypted or tampered with. In addition, when the system detects a large amount of abnormal traffic at the campus boundary, it immediately activates the intrusion detection module, identifies the traffic as malicious scanning behavior, and automatically blocks the source address. At the same time, an alarm record is generated in the system log and a notification email is sent to the administrator.

[0059] Finally, during its long-term operation, the system continuously builds self-learning models by collecting and learning the communication behaviors of various devices. For example, if the system identifies a sharp increase in communication requests from a certain thermostat model during a certain time of day, and predicts this behavior through data modeling, it preheats the communication channel used by the device and establishes a pre-connection to ensure optimal response speed during peak hours. Similar intelligent learning strategies are also used in visitor identification systems around holidays. The system adjusts protocol routing based on historical data in advance to avoid network congestion caused by increased traffic.

[0060] Through the rigorous implementation method described above, deploying the MCP intelligent agent system proposed in this invention in this scenario not only resolves the compatibility issues of multi-protocol devices, but also significantly improves communication efficiency and security, and reduces the average system response time. Importantly, campus personnel no longer need to perform complex protocol adaptation and configuration for newly added devices. They simply connect the device to a unified agent node, and the system automatically identifies the protocol type, establishes a communication channel, and activates the optimal strategy. This system successfully bridges the last mile of data interoperability, building a flexible, intelligent, and secure basic communications platform with excellent promotional value and practicality.

[0061] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent agent system based on MCP technology, characterized by: include: MCP middle-layer protocol engine: As the core module of the system, it is designed based on the principles of modern protocol abstraction and interface unification. It encapsulates the specific implementations of multiple communication protocols into a unified processing layer, supporting the encapsulation, decapsulation, and adaptation of common network communication protocols. Each protocol has an adapter module responsible for converting protocol data into a unified internal format to achieve interoperability between different protocols. The protocol adapter module has dynamic loading and switching functions, and can perform data compression and encryption during the message conversion process. Adopting a protocol abstraction layer, defining a unified interface standard, making it transparent to upper-layer applications, reducing compatibility issues between different protocols, and facilitating the addition of new protocols; Intelligent Scheduling Module: This module intelligently schedules data transmission paths and selects the optimal communication protocol based on real-time network conditions, historical access behavior, and business needs. It analyzes network status by monitoring network bandwidth, latency, and packet loss rate, and dynamically selects the most suitable transmission channel based on a policy-based scheduling algorithm. It makes predictions based on historical data and access behavior, selecting the optimal communication protocol and transmission path in advance. It dynamically adjusts data transmission priorities based on different device priorities, task importance, and transmission rate requirements. Channel Management System: Unified management and scheduling of multi-protocol data channels. A connection pool mechanism is used to set a fixed-size connection pool for each protocol, which automatically expands or contracts based on the current load. The system dynamically monitors the traffic and load of each channel, automatically switching to a backup channel when traffic is too high. Asynchronous I / O technology is used to ensure that data transmission operations do not block the main thread, improving system throughput and response speed. Built-in load balancing function automatically distributes traffic according to the load conditions of each channel; Security Enhancement Mechanism: A series of security enhancement mechanisms are integrated, using advanced encryption algorithms to encrypt all data during transmission, automatically selecting the appropriate encryption method based on the protocol type, and supporting end-to-end encryption. An integrated intrusion detection system (IDS) monitors network traffic and communication behavior in real time, immediately issuing alerts and taking protective measures when abnormal traffic or suspicious activity is detected. A built-in refined access control and permission management mechanism supports the role-based access control (RBAC) model, allowing administrators to flexibly configure permissions for different users and devices. Multi-factor authentication (MFA) and single sign-on (SSO) security authentication technologies are also supported. Edge node self-learning mechanism: In edge computing scenarios, by collecting historical data from each edge node, building a behavioral model, and analyzing it using machine learning algorithms, edge nodes can dynamically adjust their behavior based on actual conditions; dynamically select the optimal routing and communication path based on the current node network status and task requirements; and dynamically adjust computing task allocation based on device computing power and network conditions.

2. The intelligent agent system based on MCP technology according to claim 1, characterized in that: In the MCP middle-layer protocol engine, the protocol adapter module uses preprocessing technology to reduce the complexity and redundancy of the protocol stack when processing protocol data encapsulation and parsing; Each protocol adapter module is only coupled with the specific implementation related to the protocol and is completely transparent to the upper-layer application.

3. The intelligent agent system based on MCP technology according to claim 2, characterized in that: In the intelligent scheduling module, the policy scheduling algorithm selects the communication protocol based on different business scenarios. When the network bandwidth is low and the latency is high, a low-latency, high-compression ratio protocol is selected; when the network bandwidth is high and the stability is strong, HTTP or gRPC protocols are selected.

4. The intelligent agent system based on MCP technology according to claim 3, characterized in that: In the security enhancement mechanism, the intrusion detection system (IDS) monitors network traffic and communication behavior in real time, and the protective measures taken include disconnecting and blocking IP addresses.

5. The intelligent agent system based on MCP technology according to claim 4, characterized in that: In the edge node self-learning mechanism, the edge node learns the optimal communication strategy for different devices and application scenarios, selecting low-latency, high-throughput communication protocols for high-frequency tasks, and low-power, low-bandwidth consumption protocols for low-frequency tasks.

6. A method for an intelligent agent system based on MCP technology according to claim 5, characterized in that: The following steps are involved: Data processing based on the MCP middle-layer protocol engine: Design an MCP middle-layer protocol engine based on modern protocol abstraction and interface unification principles, encapsulating the specific implementations of multiple communication protocols into a unified processing layer; set up an adapter module for each communication protocol, which converts the protocol data into a unified internal format to achieve intercommunication between different protocols; During data conversion, preprocessing technology is used to reduce protocol stack complexity and redundancy, and protocol adapters are dynamically loaded and switched based on demand. A protocol abstraction layer is used to define a unified interface standard, enabling the protocol abstraction layer to provide an interface that is transparent to upper-layer applications based on different business scenarios. Each protocol adapter module is only coupled to the specific implementation related to the protocol and is completely transparent to upper-layer applications. Data transmission path and protocol selection based on the intelligent scheduling module: monitor network bandwidth, latency, packet loss rate information, and analyze network status; dynamically select the most suitable transmission channel and optimal communication protocol based on the policy scheduling algorithm, combined with real-time network conditions, historical access behavior, and business needs; when the network bandwidth is low and the latency is high, choose a low-latency, high-compression ratio protocol; for high-bandwidth, high-stability network environments, choose HTTP or gRPC protocols; if the device access pattern has a periodic pattern, predict it based on historical access data and select the optimal communication protocol and transmission path in advance; dynamically adjust the data transmission priority based on the priority of different devices, task importance, and transmission rate requirements, and prioritize the completion of critical tasks when resources are limited; Data channel management based on the channel management system: Using connection pool technology, a fixed-size connection pool is set for each protocol, reusing existing connections to avoid resource waste caused by frequent connection establishment and closure, and automatically expanding or shrinking the channel connection pool according to the current load; dynamically monitoring the flow and load of each channel, and automatically switching to a backup channel when the load of a channel is too high; using asynchronous IO technology, each data transmission operation can be performed without blocking the main thread, improving system throughput and response speed; Built-in load balancing function automatically distributes traffic according to the load of each channel to avoid system bottleneck caused by excessive traffic in a certain channel; Data security based on security enhancement mechanisms: Advanced encryption algorithms are used to encrypt all data during transmission, automatically selecting the appropriate encryption method based on the protocol type, and supporting end-to-end encryption. An integrated intrusion detection system (IDS) monitors network traffic and communication behavior in real time, immediately issuing alerts upon detecting abnormal traffic or suspicious activity, and taking protective measures such as disconnecting and blocking IP addresses. A built-in refined access control and permission management mechanism supports the role-based access control (RBAC) model, allowing administrators to flexibly configure permissions for different users and devices. Multi-factor authentication (MFA) and single sign-on (SSO) security authentication technologies are also supported. Dynamic adjustment based on the self-learning mechanism of edge nodes: In edge computing scenarios, historical data of each edge node is collected, behavioral models are constructed, and analysis is performed using machine learning algorithms. As the system runs, the edge nodes learn the optimal communication strategies for different devices and application scenarios, such as selecting low-latency, high-throughput communication protocols for high-frequency tasks, and low-power, low-bandwidth consumption protocols for low-frequency tasks. Based on the current network status of the node and task requirements, the optimal route and communication path are dynamically selected, and the allocation of computing tasks is dynamically adjusted according to the device computing power and network conditions.

7. A method according to claim 6, characterized in that: In the data processing steps based on the MCP middle-layer protocol engine, the protocol adapter module reduces the complexity and redundancy of the protocol stack through preprocessing technology when processing protocol data encapsulation and parsing. The protocol adapter can be dynamically loaded and switched according to demand to ensure the flexibility and scalability of the system.

8. A method according to claim 7, characterized in that: In the data transmission path and protocol selection step based on the intelligent scheduling module, the policy scheduling algorithm selects the communication protocol according to different business scenarios. The scheduling strategy not only relies on the real-time network status, but also combines historical data and access behavior for prediction to select the optimal communication protocol and transmission path in advance.

9. A method according to claim 8, characterized in that: In the data security assurance step based on the security enhancement mechanism, the intrusion detection system IDS monitors network traffic and communication behavior in real time, and the protective measures taken include disconnecting and blocking IP.

10. A method according to claim 9, characterized in that: In the dynamic adjustment step based on the edge node self-learning mechanism, after the edge node learns the optimal communication strategy based on the collected historical data, it can dynamically select the optimal route and communication path according to the current node's network status and task requirements, thereby achieving flexible and efficient data processing capabilities and reducing dependence on central servers.

Images