Quantum-enhanced computing power network scheduling method and device, computer device, and storage medium

Abstract

The present application belongs to the technical field of quantum computing, and relates to a quantum enhanced computing power network scheduling method and device, computer equipment and a storage medium. The method comprises the following steps: based on quantum enhancement, collecting multi-modal global resources; converting scheduling requirements and resource states into a Hamiltonian model that can be processed by a quantum computer and a feature space that can be analyzed by a classical AI; using quantum computing to accelerate AI, and using AI to feed back quantum computing to find a global optimal solution; according to the optimal scheduling strategy, establishing a quantum entanglement channel between multiple computing power nodes that need to work closely together; on the basis of the quantum cooperation channel, executing the optimal scheduling strategy; through real-time feedback driving model and continuous optimization of the strategy, self-iterative closed-loop optimization is performed. The scheduling efficiency bottleneck is broken through, and a global optimal solution is achieved; quantum resources are deeply integrated, and devices are converted into engines; based on quantum entanglement, an extremely low latency cooperation channel is constructed; endogenous security is strengthened, and quantum computing threats are resisted.

Description

Technical Field

[0001] This invention relates to the field of quantum computing technology, and in particular to a method, apparatus, computer equipment, and storage medium for scheduling quantum-enhanced computing networks. Background Technology

[0002] Existing computing power scheduling technologies exhibit the following core shortcomings when dealing with heterogeneous computing power pools comprised of central clouds, edge clouds, intelligent computing centers, and quantum computing processors (QPUs): low scheduling optimization efficiency; when facing high-dimensional, multi-constraint NP-hard combinatorial optimization problems, classical algorithms and pure AI models are prone to getting trapped in local optima, resulting in scheduling latency as high as minutes; quantum computing and the computing power network are disconnected, with quantum devices serving only as independent computing power units and not used to accelerate the training and inference of AI scheduling models; cross-domain node collaboration lacks strong quantum-level correlation, relying on classical networks, leading to low synchronization accuracy and high latency; insufficient security level of scheduling data, with classical encryption methods facing the potential risk of being cracked by quantum computing; and the lack of integrated orchestration of heterogeneous resources across the entire domain, with fragmented management of "computing power-algorithm-network-security" resources resulting in low utilization. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a quantum-enhanced computing power network scheduling method, employing the following technical solution, including the following steps: Based on quantum enhancement, multimodal global resources are acquired; Transform scheduling requirements and resource states into Hamiltonian models that can be processed by quantum computers and feature spaces that can be analyzed by classical AI. Accelerate AI with quantum computing, and at the same time use AI to feed back into quantum computing to find the global optimal solution; Based on the optimal scheduling strategy, quantum entanglement channels are established among multiple computing nodes that need to work closely together; Based on the quantum cooperative channel, the optimal scheduling strategy is executed; By driving continuous optimization of the model and strategy through real-time feedback, a self-iterative closed-loop optimization is achieved.

[0004] Preferably, the step of acquiring multimodal global resources based on quantum enhancement specifically includes: Collect quantum resource data based on compressed sensing; Based on quantum enhancement, feature extraction and noise suppression are performed on the quantum resource data; The quantum resource data after feature extraction and noise suppression is spatiotemporally aligned and standardized.

[0005] Preferably, the step of transforming scheduling requirements and resource states into a Hamiltonian model that can be processed by a quantum computer and a feature space that can be analyzed by classical AI specifically includes: AI-based preprocessing of scheduling constraints and target features generates a scheduling description vector. The scheduling description vector is mapped to a ground state solution problem of a quantum many-body system, and a Hamiltonian is constructed. A fusion model is performed based on the quantum-classical feature space.

[0006] Preferably, the step of using quantum computing to accelerate AI, and simultaneously using AI to feed back into quantum computing, to find the globally optimal solution specifically includes: An AI-driven preheating startup strategy is used to predict initial quantum circuit parameters; Global optimal solution is obtained based on quantum-classical hybrid iteration; The optimal quantum state obtained through quantum measurement is mapped to a specific classical scheduling strategy through a quantum-classical data interface.

[0007] Preferably, the step of establishing a quantum entanglement channel among multiple computing nodes that need to work closely together according to the optimal scheduling strategy specifically includes: Based on the optimal scheduling strategy, entanglement bursts and batch generation are performed on demand. Perform multi-channel parallel distribution and real-time entanglement purification; Construct multi-body entangled states and monitor fidelity in real time.

[0008] Preferably, the step of executing the optimal scheduling strategy based on the quantum cooperative channel specifically includes: Based on the quantum cooperative channel, scheduling tasks are split and distributed in a quantum-classical heterogeneous manner; Distributed computing and synchronization based on quantum collaborative channels; Based on quantum key distribution and quantum teleportation, intrinsically secure transmission of scheduling data is achieved.

[0009] Preferably, the step of continuously optimizing the model and strategy through real-time feedback to perform self-iterative closed-loop optimization specifically includes: Real-time collection of scheduling execution results, and performance evaluation of the scheduling execution results; Based on reinforcement learning, we optimize the collaborative parameters of the quantum-AI model. Update and iterate the dynamic scheduling knowledge base.

[0010] To address the aforementioned technical problems, this invention also provides a quantum-enhanced computing power network scheduling device, which employs the following technical solution, including: The acquisition module is used to acquire multimodal global resources based on quantum enhancement. The transformation module is used to convert scheduling requirements and resource states into Hamiltonian models that can be processed by quantum computers and feature spaces that can be analyzed by classical AI. The search module is used to accelerate AI with quantum computing, and at the same time use AI to feed back into quantum computing to find the global optimal solution. A module is established to create quantum entanglement channels among multiple computing nodes that need to work closely together, based on the optimal scheduling strategy. The execution module is used to execute the optimal scheduling strategy based on the quantum cooperative channel; The optimization module is used to drive continuous optimization of the model and strategy through real-time feedback, and to perform self-iterative closed-loop optimization.

[0011] To address the aforementioned technical problems, the present invention also provides a computer device that employs the technical solution described below, comprising a memory and a processor. The memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the aforementioned quantum-enhanced computing network scheduling method.

[0012] To address the aforementioned technical problems, the present invention also provides a computer-readable storage medium, which employs the technical solution described below. The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the aforementioned quantum-enhanced computing network scheduling method.

[0013] Compared with the prior art, the present invention has the following main advantages: (1) By introducing quantum computing to accelerate AI, by transforming the resource state into a Hamiltonian model, quantum parallelism is used to quickly approximate the global optimal solution in a high-dimensional space, significantly reducing scheduling latency; at the same time, AI is used to feed back into quantum computing, and quantum circuits are optimized through feature space analysis to form a synergistic gain of quantum-accelerated AI and AI-optimized quantum, which solves the optimization dilemma of pure classical algorithms or pure AI models. (2) Not only does it regard QPU as a scheduled resource, but it also regards it as the core of the scheduling engine and directly participates in the generation of scheduling strategies. This breaks the limitation that quantum computing is only used for backend computing, and upgrades quantum resources from a simple computing power provider to a network scheduling enabler, thereby improving the utilization value of quantum devices and the overall intelligence level of computing power networks. (3) Establish quantum entanglement channels between nodes that require close collaboration. By utilizing the action at a distance and strong correlation of quantum entanglement, instantaneous state synchronization and instruction distribution between multiple computing nodes can be achieved, significantly reducing communication latency and providing a high-quality quantum-level collaborative foundation for scenarios such as distributed training or real-time inference; (4) Quantum properties are introduced at the scheduling layer. Quantum entanglement channels inherently possess the properties of quantum key distribution or quantum-secure direct communication, realizing the intrinsic security of scheduling command transmission, fundamentally improving the resistance to quantum attacks in the integrated orchestration process of heterogeneous resources across the entire domain, and ensuring the data security of the computing power network. Attached Figure Description

[0014] To more clearly illustrate the solutions in this invention, the accompanying drawings used in the description of the embodiments of this invention will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0015] Figure 1 This is a flowchart of an embodiment of the quantum-enhanced computing power network scheduling method of the present invention; Figure 2 This is an exemplary system architecture diagram in which the quantum-enhanced computing power network scheduling method of the present invention can be applied; Figure 3 This is a schematic diagram of the structure of an embodiment of the quantum-enhanced computing power network scheduling device of the present invention; Figure 4 This is a schematic diagram of the structure of an embodiment of the computer device of the present invention. Detailed Implementation

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein in the specification is for the purpose of describing particular embodiments only and is not intended to limit the invention; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings are used to distinguish different objects and not to describe a particular order.

[0017] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0018] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0019] It should be noted that the quantum-enhanced computing network scheduling method provided in the embodiments of the present invention is generally executed by a server / terminal device, and correspondingly, the quantum-enhanced computing network scheduling device is generally set in the server / terminal device.

[0020] In the following embodiments, entanglement burst refers to the technical mechanism of generating high-fidelity entangled photon pairs in parallel batches to a group of distributed computing nodes in an extremely short time of microseconds to milliseconds, distributing them through multiple channels and purifying them in real time, and then constructing high-fidelity multi-body quantum entanglement associations between target nodes to form an instantaneous quantum cooperative channel for node synchronization, collaborative computing and secure communication.

[0021] The deep integration of quantum computing, AI, and computing power refers to the fact that quantum computing accelerates the training and inference of AI models through quantum algorithms, and AI models automatically optimize quantum circuits, qubit allocation, quantum noise suppression, and quantum error correction strategies. The two work together to drive the global resource scheduling of the computing power network, forming a mutually reinforcing and closed-loop iterative intelligent system.

[0022] Joint scheduling of computing power and entanglement refers to simultaneously optimizing the utilization rate of classical computing resources and the efficiency, fidelity, and concurrency of quantum entanglement distribution during the scheduling decision-making process of the computing power network, so as to achieve the optimal coordination between the computing power network and the quantum communication system and achieve integrated scheduling of computing power and communication.

[0023] Quantum intrinsic security refers to the physical layer's ability to ensure that scheduling instructions and core data are uneavesdroppable, unalterable, and unforgeable, based on the principles of quantum no-cloning, quantum measurement collapse, and quantum entanglement nonlocal correlation, thus providing unconditional security for computing network scheduling.

[0024] It should be understood that the number of terminal devices, networks, and servers is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be used.

[0025] Example 1 Please refer to Figure 1 A flowchart of an embodiment of the quantum-enhanced computing network scheduling method of the present invention is shown. The quantum-enhanced computing network scheduling method includes the following steps: Step S1: Based on quantum enhancement, collect multimodal global resources.

[0026] In this embodiment, the electronic devices (e.g., servers / terminal devices) on which the quantum-enhanced computing network scheduling method runs can receive quantum-enhanced computing network scheduling requests via wired or wireless connections. It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra-wideband) connections, and other currently known or future-developed wireless connection methods.

[0027] Figure 2 This is an exemplary system architecture diagram showing how the quantum-enhanced computing power network scheduling method of the present invention can be applied. For example... Figure 2 As shown, the system architecture is divided into five layers from bottom to top: the basic resource layer, the quantum-AI fusion computing layer, the fusion scheduling hub layer, the service orchestration and interface layer, and the application service layer. Each layer is interconnected through standardized interfaces, enabling bidirectional data exchange between layers and forming a closed-loop optimization system, as detailed below: The basic resource layer provides underlying hardware resource support for the system, including classical computing resources (central cloud, edge cloud, intelligent computing center, supercomputing, edge node), quantum computing resources (quantum computing processor QPU, digital quantum simulator), quantum communication resources (entanglement source, quantum repeater, quantum memory, fiber optic / free space quantum channel), and network and storage resources (OTN, computing gateway, high-speed switch, distributed storage), which is the physical basis for computing power scheduling and quantum operations.

[0028] Quantum-AI Fusion Computing Layer: Provides core computing capabilities for the system, enabling deep integration of quantum computing and AI. It includes quantum computing modules (quantum circuit generation, quantum gate operation, quantum state evolution, quantum measurement), quantum AI modules (quantum neural network QNN, quantum converter, quantum reinforcement learning, quantum kernel function), classical AI modules (deep reinforcement learning, graph neural network, scheduling prediction model, anomaly detection model), and quantum-classical data interface (realizing bidirectional conversion between quantum state data and classical data, providing data compatibility support for cross-layer interaction).

[0029] The integrated scheduling hub layer [core layer] serves as the core of the system's scheduling decision-making and control, enabling unified scheduling and collaborative management of resources across the entire domain. It includes modules for global resource perception and modeling, quantum-AI hybrid scheduling decision-making, entanglement burst management and control, quantum communication and security encryption, task decomposition and mapping execution, and state monitoring and closed-loop optimization. It is responsible for modeling scheduling problems, finding optimal solutions, triggering and controlling entanglement bursts, security encryption, and feedback optimization of scheduling results.

[0030] Service Orchestration and Interface Layer: This layer enables standardized orchestration of system resources and services, providing unified service interfaces for upper-layer applications. It includes computing power service orchestration, quantum service orchestration, AI service orchestration, security service orchestration, as well as open API interfaces, SDKs, and management platforms, supporting rapid access and personalized scheduling services for applications in multiple scenarios.

[0031] Application Service Layer: This is the application implementation layer of the system, adapting to the computing power scheduling needs of different industries and scenarios. It includes nationwide integrated computing power network scheduling services, AI large model training and inference acceleration services, quantum computing cloud services, high-security computing power private network services for government / finance / power, real-time intelligent computing services for industrial internet, integrated space-ground quantum computing power communication network services, and cross-border computing power interconnection services.

[0032] In this embodiment, step S1 may specifically include the following steps: S11, based on compressed sensing, collects quantum resource data.

[0033] Let a high-dimensional resource state signal be... In the transform domain It is sparse in the wavelet transform domain (e.g., the wavelet transform domain), that is... ,in It is the only A sparse vector with non-zero coefficients. Compressed sensing uses a vector with a transform domain... Uncorrelated observation matrix right Perform linear observations to obtain the observation vector. The system collected This refers to the compressed resource data. Among them: : Represents the original, high-dimensional resource status signal. : Indicates the dimension of the original signal. : denotes a sparse transformation basis. : Indicates a signal exist A sparse coefficient vector in the domain. : Represents the observation matrix. : Indicates the number of observations, usually . : Represents the compressed observation vector collected.

[0034] Observation vector The formula describes the process from high-dimensional signals. to low-dimensional observations The compression process. Its significance lies in designing a suitable observation matrix. Only a small amount of data can be collected. This significantly reduces the network bandwidth and storage resource consumption of the sensing process, while ensuring that the original resource state can be reconstructed with high accuracy in the future.

[0035] Step S11 aims to reconstruct complete resource state information with high probability by employing compressed sensing theory and leveraging the sparsity of resources, using a sampling rate far lower than that required by the Nyquist sampling theorem. Compressed sensing sampling matrices are deployed in the computing power gateway, quantum channel detector, and resource monitoring agent.

[0036] S12, based on quantum enhancement, feature extraction and noise suppression are performed on the quantum resource data.

[0037] The collected compressed data y is input into a quantum neural network (QNN) in the quantum-AI fusion computing layer for feature extraction and noise reduction. A variable-structure quantum neural network model is employed. This QNN encodes the classical data y into quantum states, which evolve through a variable quantum circuit (composed of multiple parameterized single-bit gates and two-bit controlled-NOT gates). The parameterized quantum states of the quantum circuit inherently possess the ability to process high-dimensional data and effectively filter noise. The output of the circuit is used to obtain a classical feature vector through quantum measurement. : in: : This represents the core feature vector obtained after extraction and noise reduction by the quantum neural network, which will be used for subsequent scheduling problem modeling. : Indicates compressed data The initial quantum state encoded. : Represents a parameterized quantum circuit, These are trainable parameters. : Indicates the measurement, This indicates the observation in the final quantum state. The expected value.

[0038] Classical eigenvectors The calculation formula describes the input-output relationship of a QNN, that is, after parameterized quantum circuitry. The evolved quantum state collapses into a classical eigenvector by measuring the expected value. Its significance lies in the fact that by utilizing the superposition and entanglement properties of quantum computing, classical data can be mapped to a high-dimensional quantum Hilbert space for processing, enabling the extraction of deep correlation features that are difficult to discover using classical methods. At the same time, the probabilistic characteristics of quantum measurement naturally have the effect of suppressing noise.

[0039] S13, perform spatiotemporal alignment and standardization on the quantum resource data after feature extraction and noise suppression.

[0040] Since the acquisition frequencies and data formats of different resources (such as CPU utilization and entanglement source states) vary, a spatiotemporal alignment fusion algorithm based on graph neural networks (GNNs) is adopted. Computing nodes are considered as nodes in a graph, and network connections between nodes are considered as edges. Resource features (including the classical feature vectors extracted in step S12) are used. Link latency and bandwidth are treated as edge attributes, while node attributes are treated as node attributes. Through the message passing mechanism of the GNN, the spatiotemporal information of neighboring nodes is automatically aggregated to generate a structured, spatiotemporally aligned global resource state graph. This serves as the input for scheduling decisions.

[0041] The purpose of step S1 is to break through the limitations of precision and dimensionality of traditional sensing technologies, and to achieve high-fidelity, low-latency data acquisition of classical computing power, quantum computing power, network status and quantum communication resources (entanglement sources, quantum channels), so as to provide a comprehensive and accurate data foundation for subsequent scheduling modeling.

[0042] Step S2 transforms the scheduling requirements and resource status into a Hamiltonian model that can be processed by a quantum computer and a feature space that can be analyzed by classical AI.

[0043] In this embodiment, step S2 may specifically include the following steps: S21. Based on AI, preprocessing of scheduling constraints and target features is performed to generate a scheduling description vector.

[0044] The system utilizes a classic AI module (deep reinforcement learning network) in the fusion scheduling hub to preprocess the global resource state graph. The AI ​​model matches user-submitted task requests (such as computational load, data volume, latency requirements, and security level) with the current resource state G, automatically identifying and extracting the core constraints (such as deadlines and cost limits) and optimization objectives (such as minimizing completion time and maximizing resource utilization) for this scheduling task. These are then quantified into standardized mathematical parameters, forming a scheduling description vector P. This process leverages AI's strength in complex pattern recognition, transforming unstructured task requirements into structured mathematical problems.

[0045] S22 maps the scheduling description vector to a ground state problem of a quantum many-body system and constructs the Hamiltonian.

[0046] The scheduling description vector P generated in step S21 is mapped to a ground state solution problem for a quantum many-body system. Referring to the mapping logic of the Quantum Approximate Optimization Algorithm (QAOA), each task-node pair to be scheduled is considered as a qubit, whose state |0... and |1 These represent no allocation and allocation, respectively. All constraints and optimization objectives of the scheduling problem are encoded into an Ising Model Hamiltonian. : .in: : Represents the problem Hamiltonian, whose eigenvalues ​​correspond to the total cost of different scheduling schemes. : Represents the index of a task or computing node. : indicates that the action is in the first position. Pauli-Z operators on qubits. : Represents the external field coefficient, which is quantified from the attributes of a single task or node. : Represents the coupling coefficient, which is quantified from the association constraints between two tasks or nodes.

[0047] Hamiltonian The computational formula maps the scheduling problem (similar to the job shop scheduling problem, JSSP) to the energy function of a quantum system. The significance lies in the fact that, according to the principles of quantum mechanics, a physical system always tends to evolve to the state with the lowest energy (the ground state). Therefore, finding the optimal solution to the scheduling problem is equivalent to finding the Hamiltonian of this quantum system. The ground state. By solving... By determining the ground state, the globally optimal scheduling scheme can be obtained.

[0048] S23, based on quantum-classical feature space, performs fusion modeling.

[0049] The quantum feature f extracted by QNN in step S12 is combined with the problem Hamiltonian constructed in step S22. To merge. Specifically, f will be used as an adjustment. Medium coefficient and The dynamic weights are used to generate a dynamic scheduling Hamiltonian that incorporates real-time resource awareness information. This enables the quantum solution process to reflect the latest resource changes in real time, achieving the integration of "sensing-modeling".

[0050] The purpose of step S2 is to transform the abstract scheduling requirements and complex resource states into a Hamiltonian model that can be processed by a quantum computer and a feature space that can be analyzed by classical AI, thereby realizing a formal description of the problem and paving the way for quantum-accelerated solutions.

[0051] Step S3: Use quantum computing to accelerate AI, and at the same time use AI to feed back into quantum computing to find the global optimal solution.

[0052] In this embodiment, step S3 may specifically include the following steps: S31 employs an AI-driven preheating startup strategy to predict initial quantum circuit parameters.

[0053] The classic QAOA algorithm requires iterative optimization starting with random parameters, resulting in slow convergence. This solution employs an AI-driven pre-launch strategy. Utilizing the deep reinforcement learning model from step 2.1, a set of optimal initial quantum circuit parameters is directly predicted based on the current scheduling description vector P. (Such as the β and γ parameters in QAOA). This is equivalent to having the AI ​​predict a parameter combination close to the optimal solution as the starting point for the quantum algorithm. Research shows that this pre-startup method can significantly improve the convergence speed and solution quality of quantum algorithms. The formula for calculating the initial parameter vector of a quantum circuit is: ,in: : Represents the initial parameter vector of the quantum circuit. : Represents a trained AI prediction model. : Represents the standardized scheduling description vector generated in step S21.

[0054] Initial parameter vector of quantum circuit The calculation formula describes the AI ​​model The function is to directly map a set of initial quantum circuit parameters from the input P. This serves as the starting point for quantum algorithms, improving convergence speed. Its significance lies in avoiding the inefficiency of quantum algorithms starting with completely random parameters for global search. By leveraging AI's prior knowledge, the search space is directly narrowed to a high-probability region near the optimal solution, thereby significantly reducing the number of evaluations and runtime of the quantum circuit.

[0055] S32 is based on a quantum-classical hybrid iteration to find the global optimal solution.

[0056] A quantum-classical hybrid computing architecture is employed. A quantum processing unit (QPU) or digital quantum simulator is responsible for executing the computation with preheating parameters. The quantum circuit, for the Hamiltonian H which incorporates dynamic information s Evolution and measurement are performed to obtain a set of candidate solutions (quantum states) and their corresponding expected energy values. <H s Then, the optimizer on the classical computer (such as gradient descent or Bayesian optimization) updates the quantum circuit parameters θ based on these measurements to minimize... <H s This iterative process of quantum state preparation and measurement—classical parameter optimization—is repeated continuously until it converges to... <H s The minimum value of > is the quantum state at which the global optimal solution to the scheduling problem is found.

[0057] S33 maps the optimal quantum state obtained through quantum measurement to a specific classical scheduling strategy through a quantum-classical data interface.

[0058] The optimal quantum state (a combination of |0> / |1> states of a series of qubits) obtained through quantum measurement is mapped to a specific classical scheduling strategy through a quantum-classical data interface. The mapping rule is the reverse process of task-node pair encoding in step S22. For example, if a qubit |1> is measured, it means that the corresponding task-node allocation scheme has been selected. All selected task-node pairs combined constitute the complete scheduling scheme. This includes task mapping, path planning, and resource allocation schemes. Finally, the classic AI module quickly verifies the scheme. Does it satisfy all hard constraints (such as resource limits) to ensure its physical feasibility?

[0059] Step S4: Based on the optimal scheduling strategy, establish quantum entanglement channels among multiple computing nodes that need to work closely together.

[0060] In this embodiment, step S4 may specifically include the following steps: S41, according to the optimal scheduling strategy, performs entanglement bursts and batch generation on demand.

[0061] Analysis of the Entanglement Burst Management and Control Module of the Fusion Scheduling Central Layer The module identifies the set of computing nodes that need to establish collaborative relationships. Then, it sends control commands to the quantum entanglement source in the basic resource layer, triggering an entanglement bursting mechanism. Within microseconds of receiving the command, the entanglement source bursts together high-fidelity entangled photon pairs in parallel. The number of pairs generated is dynamically determined based on the number of collaborative nodes and the reliability requirements of the entanglement link, ensuring a sufficient supply of entanglement resources.

[0062] S42 performs multi-channel parallel distribution and real-time entanglement purification.

[0063] The generated entangled photon pairs are simultaneously distributed to the target node through multiple pre-planned quantum channels (fiber optics or free space). During the distribution process, entanglement purification modules deployed at quantum repeaters or nodes are used to purify photon pairs that have decohered due to channel noise and environmental interference in real time. The purification process typically involves consuming some low-quality entangled pairs to extract a pair of entangled pairs with higher fidelity.

[0064] .in: : Indicates the fidelity of the entangled pairs output after purification. : Indicates the fidelity of the entangled pair of the two inputs used in the purification operation. The purification process typically requires two entangled pairs with low fidelity to produce one entangled pair with high fidelity. : Indicates the probability of a successful purification operation. This is a number less than 1, meaning that the purification process sacrifices resources to improve fidelity.

[0065] This formula describes the fundamental principle of entanglement purification. By comparing the Bell state measurements of two low-fidelity entangled pairs, remaining entangled pairs with improved fidelity can be selected. Its significance lies in the fact that this is a key technology for ensuring the reliability of long-distance quantum communication and quantum networks. After purification, the fidelity of the entangled pairs ultimately used for collaboration can be improved. The accuracy rate has been increased to over 90%, meeting the requirements for high-precision quantum collaboration.

[0066] S43, construct multi-body entangled states and monitor fidelity in real time.

[0067] Once all target nodes receive high-fidelity entangled photon pairs, these two-body entangled pairs are connected through a series of local quantum operations and classical communication (LOCC), constructing a multi-body entangled state (as shown in the diagram, Cluster State) covering all cooperating nodes. This provides the physical foundation for distributed quantum computing and multi-point collaboration. Simultaneously, the system runs Bell state detection and quantum state tomography techniques in real time to continuously monitor the quality of entangled links between nodes. If the entanglement fidelity of a link falls below a preset threshold, a retransmission or purification compensation mechanism is immediately triggered to ensure the reliability of the collaborative channel.

[0068] The function of step S4 is: the kernel, according to the optimal scheduling strategy... Instantly establishing high-fidelity quantum entanglement channels between multiple computing nodes that require close collaboration elevates classical network-level collaboration to quantum-level strongly correlated collaboration, fundamentally solving the problems of high latency and low synchronization accuracy in cross-domain collaboration.

[0069] Step S5: Execute the optimal scheduling strategy based on the quantum cooperative channel.

[0070] In this embodiment, step S5 may specifically include the following steps: S51, based on the quantum cooperative channel, performs scheduling task splitting and quantum-classical heterogeneous distribution.

[0071] The task splitting and mapping execution module executes according to the scheduling strategy. The system intelligently breaks down user-submitted complex tasks into quantum subtasks suitable for quantum computing (such as quantum chemical simulations, combinatorial optimization solutions, and specific layer calculations for large models) and classical subtasks suitable for classical computing power (such as I / O processing, data preprocessing, and result post-processing). Then, using a pre-built quantum cooperative channel as a high-speed signaling network, the quantum subtasks are distributed to designated QPU nodes, while the classical subtasks are distributed to CPU / GPU / NPU nodes via a classical network. This heterogeneous distribution ensures that various computing resources are utilized to their full potential.

[0072] S52, based on quantum collaborative channels, performs distributed computing and synchronization.

[0073] Quantum computing units and classical computing units begin executing their respective tasks in parallel. When performing distributed computations requiring strong synchronization (such as distributed quantum machine learning), each node uses the many-body quantum entangled state constructed in step four as a synchronization benchmark. Leveraging the non-local correlation properties of quantum entanglement, nodes can achieve coordination with several orders of magnitude higher accuracy than classical clock synchronization protocols (such as NTP). For example, a measurement of one entangled pair can instantaneously affect the particle state of another node; this quantum correlation can be used to trigger precise parallel computation.

[0074] S53 enables intrinsically secure transmission of scheduled data based on quantum key distribution and quantum teleportation.

[0075] This step constructs an inherent security system for the scheduling process. On one hand, using the entangled resources generated in step S4, a secure quantum key is generated between the source and target nodes via a quantum key distribution (QKD) protocol. Scheduling instructions and core business data are encrypted using this quantum key with a one-time pad classical encryption before transmission, achieving unconditional security. On the other hand, for data requiring extremely high security levels or minimal transmission delays, quantum teleportation can be used to directly transmit the quantum state of the data from one location to another using quantum entanglement, without transmitting the physical carrier of the information, fundamentally eliminating the possibility of channel eavesdropping.

[0076] Step S6 involves continuous optimization of the model and strategy through real-time feedback, performing self-iterative closed-loop optimization.

[0077] In this embodiment, step S6 may specifically include the following steps: S61 collects scheduling execution results in real time and evaluates the performance of the scheduling execution results.

[0078] Throughout the entire process of scheduling task execution, the state monitoring and closed-loop optimization module continuously collects key performance indicators, including: actual task completion time, resource utilization (CPU / QPU) of each node, actual synchronization accuracy of quantum entanglement collaboration, QKD key generation rate, and whether any security attack events have occurred. This data is then aggregated to form a performance evaluation report R for this scheduling execution.

[0079] S62, based on reinforcement learning, performs collaborative parameter optimization of quantum-AI models.

[0080] The scheduling execution report R is used as an immediate reward signal and input into the deep reinforcement learning model in step S21. The goal of this model is to maximize the long-term cumulative reward (i.e., scheduling efficiency). The model updates its network weights based on the reward R, thereby optimizing its ability to extract scheduling constraint features and the accuracy of pre-starting quantum circuit parameters. Simultaneously, based on the performance of the quantum algorithm in this scheduling, the model also provides optimization suggestions for the structure of the quantum circuit (such as the number of layers p in QAOA) or parameter update strategies, forming a positive feedback loop where AI feeds back into the quantum.

[0081] S63 updates the dynamic scheduling knowledge base and iterates the strategy.

[0082] The complete data for this scheduling—including task characteristics P, initial resource state G, and final scheduling policy—will be used. Execution performance R is structured and stored in a dynamic scheduling knowledge base. This knowledge base serves not only as historical data storage but, more importantly, as a means for offline mining and optimization of subsequent scheduling strategies. For example, by comparing different scheduling strategies and their effectiveness in similar historical scenarios, better scheduling rules or new quantum mapping methods can be identified. This knowledge will be used to update the scheduling problem model (e.g., adjusting the Hamiltonian). (coefficients), thereby enabling iterative evolution of scheduling strategies at a higher level.

[0083] The purpose of step S6 is to establish the system's self-evolution capability. The execution results and state changes during the scheduling task are fed back to the quantum algorithm and AI model in real time, driving continuous optimization of model parameters and scheduling strategies, enabling the system to dynamically adapt to the ever-changing computing network environment.

[0084] The beneficial effects of implementing this embodiment are: (1) By introducing quantum computing to accelerate AI, by transforming the resource state into a Hamiltonian model, quantum parallelism is used to quickly approximate the global optimal solution in a high-dimensional space, significantly reducing scheduling latency; at the same time, AI is used to feed back into quantum computing, and quantum circuits are optimized through feature space analysis to form a synergistic gain of quantum-accelerated AI and AI-optimized quantum, which solves the optimization dilemma of pure classical algorithms or pure AI models. (2) Not only does it regard QPU as a scheduled resource, but it also regards it as the core of the scheduling engine and directly participates in the generation of scheduling strategies. This breaks the limitation that quantum computing is only used for backend computing, and upgrades quantum resources from a simple computing power provider to a network scheduling enabler, thereby improving the utilization value of quantum devices and the overall intelligence level of computing power networks. (3) Establish quantum entanglement channels between nodes that require close collaboration. By utilizing the action at a distance and strong correlation of quantum entanglement, instantaneous state synchronization and instruction distribution between multiple computing nodes can be achieved, significantly reducing communication latency and providing a high-quality quantum-level collaborative foundation for scenarios such as distributed training or real-time inference; (4) Quantum properties are introduced at the scheduling layer. Quantum entanglement channels inherently possess the properties of quantum key distribution or quantum-secure direct communication, realizing the intrinsic security of scheduling command transmission, fundamentally improving the resistance to quantum attacks in the integrated orchestration process of heterogeneous resources across the entire domain, and ensuring the data security of the computing power network.

[0085] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0086] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0087] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0088] Example 2 Further reference Figure 3 As a response to the above Figure 1 The present invention provides an embodiment of a quantum-enhanced computing power network scheduling device, which, in accordance with the implementation of the method shown, provides an embodiment of such a device. Figure 1 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0089] like Figure 3As shown, the quantum-enhanced computing network scheduling device 70 described in this embodiment includes: a data acquisition module 71, a conversion module 72, a search module 73, an establishment module 74, an execution module 75, and an optimization module 76. Wherein: Acquisition module 71 is used to acquire multimodal global resources based on quantum enhancement; The transformation module 72 is used to transform scheduling requirements and resource states into a Hamiltonian model that can be processed by a quantum computer and a feature space that can be analyzed by classical AI. The search module 73 is used to accelerate AI with quantum computing and at the same time use AI to feed back into quantum computing to find the global optimal solution. Module 74 is established to create quantum entanglement channels among multiple computing nodes that need to work closely together, based on the optimal scheduling strategy. Execution module 75 is used to execute the optimal scheduling strategy based on the quantum cooperative channel; The optimization module 76 is used to drive continuous optimization of the model and strategy through real-time feedback, and to perform self-iterative closed-loop optimization.

[0090] The beneficial effects of implementing this embodiment are: breaking through the scheduling efficiency bottleneck and achieving the global optimal solution; deeply integrating quantum resources and turning devices into engines; building an ultra-low latency collaborative channel based on quantum entanglement; and strengthening intrinsic security to resist the threat of quantum computing.

[0091] Example 3 To address the aforementioned technical problems, embodiments of the present invention also provide a computer device. Please refer to [link / reference needed]. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0092] The aforementioned computer device 8 includes a memory 81, a processor 82, and a network interface 83 that are interconnected via a system bus. It should be noted that only the computer device 8 with components 81, 82, and 83 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described herein is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0093] The aforementioned computer devices can be desktop computers, laptops, handheld computers, and cloud servers, among other computing devices. These devices can facilitate human-computer interaction with users through keyboards, mice, remote controls, touchpads, or voice-activated devices.

[0094] The aforementioned memory 81 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the aforementioned memory 81 may be an internal storage unit of the aforementioned computer device 8, such as the hard disk or memory of the computer device 8. In other embodiments, the aforementioned memory 81 may also be an external storage device of the aforementioned computer device 8, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 8. Of course, the aforementioned memory 81 may also include both the internal storage unit and its external storage device of the aforementioned computer device 8. In this embodiment, the aforementioned memory 81 is typically used to store the operating system and various application software installed on the aforementioned computer device 8, such as computer-readable instructions for quantum-enhanced computing network scheduling methods. In addition, the aforementioned memory 81 can also be used to temporarily store various types of data that have been output or will be output.

[0095] In some embodiments, the processor 82 described above may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 82 is typically used to control the overall operation of the computer device 8. In this embodiment, the processor 82 is used to execute computer-readable instructions stored in the memory 81 or to process data, such as executing computer-readable instructions for the quantum-enhanced computing network scheduling method described above.

[0096] The network interface 83 may include a wireless network interface or a wired network interface, which is typically used to establish a communication connection between the computer device 8 and other electronic devices.

[0097] The beneficial effects of implementing this embodiment are: breaking through the scheduling efficiency bottleneck and achieving the global optimal solution; deeply integrating quantum resources and turning devices into engines; building an ultra-low latency collaborative channel based on quantum entanglement; and strengthening intrinsic security to resist the threat of quantum computing.

[0098] Example 4 The present invention also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the quantum-enhanced computing network scheduling method described above.

[0099] The beneficial effects of implementing this embodiment are: breaking through the scheduling efficiency bottleneck and achieving the global optimal solution; deeply integrating quantum resources and turning devices into engines; building an ultra-low latency collaborative channel based on quantum entanglement; and strengthening intrinsic security to resist the threat of quantum computing.

[0100] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0101] Obviously, the embodiments described above are merely some embodiments of the present invention, not all embodiments. The accompanying drawings show preferred embodiments of the present invention, but do not limit the patent scope of the present invention. The present invention can be implemented in many different forms; rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the patent protection scope of this invention.

Claims

1. A scheduling method for a quantum-enhanced computing network, characterized in that, Includes the following steps: Based on quantum enhancement, multimodal global resources are acquired; Transform scheduling requirements and resource states into Hamiltonian models that can be processed by quantum computers and feature spaces that can be analyzed by classical AI. Accelerate AI with quantum computing, and at the same time use AI to feed back into quantum computing to find the global optimal solution; Based on the optimal scheduling strategy, quantum entanglement channels are established among multiple computing nodes that need to work closely together; Based on the quantum cooperative channel, the optimal scheduling strategy is executed; Through real-time feedback, the model and strategy are continuously optimized, enabling self-iterative closed-loop optimization. The steps of using quantum computing to accelerate AI, and simultaneously using AI to feed back into quantum computing, to find the globally optimal solution specifically include: An AI-driven preheating startup strategy is used to predict initial quantum circuit parameters; Global optimal solution is obtained based on quantum-classical hybrid iteration; The optimal quantum state obtained through quantum measurement is mapped to a specific classical scheduling strategy through a quantum-classical data interface.

2. The quantum-enhanced computing power network scheduling method according to claim 1, characterized in that, The steps for acquiring multimodal global resources based on quantum enhancement specifically include: Collect quantum resource data based on compressed sensing; Based on quantum enhancement, feature extraction and noise suppression are performed on the quantum resource data; The quantum resource data after feature extraction and noise suppression is spatiotemporally aligned and standardized.

3. The quantum-enhanced computing power network scheduling method according to claim 1, characterized in that, The steps of transforming scheduling requirements and resource states into a Hamiltonian model that can be processed by a quantum computer and a feature space that can be analyzed by classical AI specifically include: AI-based preprocessing of scheduling constraints and target features generates a scheduling description vector. The scheduling description vector is mapped to a ground state problem of a quantum many-body system, and a Hamiltonian is constructed. A fusion model is performed based on the quantum-classical feature space.

4. The quantum-enhanced computing power network scheduling method according to claim 1, characterized in that, The steps for establishing quantum entanglement channels among multiple computing nodes that need to work closely together, based on the optimal scheduling strategy, specifically include: Based on the optimal scheduling strategy, entanglement bursts and batch generation are performed on demand. Perform multi-channel parallel distribution and real-time entanglement purification; Construct multi-body entangled states and monitor fidelity in real time.

5. The quantum-enhanced computing power network scheduling method according to claim 1, characterized in that, The steps for executing the optimal scheduling strategy based on the quantum cooperative channel specifically include: Based on the quantum cooperative channel, scheduling tasks are split and distributed in a quantum-classical heterogeneous manner; Distributed computing and synchronization based on quantum collaborative channels; Based on quantum key distribution and quantum teleportation, intrinsically secure transmission of scheduling data is achieved.

6. The quantum-enhanced computing power network scheduling method according to any one of claims 1 to 5, characterized in that, The steps of continuously optimizing the model and strategy through real-time feedback and performing self-iterative closed-loop optimization specifically include: Real-time collection of scheduling execution results, and performance evaluation of the scheduling execution results; Based on reinforcement learning, we optimize the collaborative parameters of the quantum-AI model. Update and iterate the dynamic scheduling knowledge base.

7. A quantum-enhanced computing power network scheduling device, characterized in that, include: The acquisition module is used to acquire multimodal global resources based on quantum enhancement. The transformation module is used to convert scheduling requirements and resource states into Hamiltonian models that can be processed by quantum computers and feature spaces that can be analyzed by classical AI. The search module is used to accelerate AI with quantum computing, and at the same time use AI to feed back into quantum computing to find the global optimal solution. A module is established to create quantum entanglement channels among multiple computing nodes that need to work closely together, based on the optimal scheduling strategy. The execution module is used to execute the optimal scheduling strategy based on the quantum cooperative channel; The optimization module is used to drive continuous optimization of the model and strategy through real-time feedback, and to perform self-iterative closed-loop optimization. The search module is further used for: An AI-driven preheating startup strategy is used to predict initial quantum circuit parameters; Global optimal solution is obtained based on quantum-classical hybrid iteration; The optimal quantum state obtained through quantum measurement is mapped to a specific classical scheduling strategy through a quantum-classical data interface.

8. A computer device, characterized in that, The system includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the quantum-enhanced computing network scheduling method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the quantum-enhanced computing network scheduling method as described in any one of claims 1 to 6.

Images