Quantum Computing-Based Network Interference Optimization Method, Apparatus, Device, and Medium

Abstract

The present invention relates to a network interference optimization method, device, equipment and medium based on quantum computing. The method includes: constructing an undirected graph corresponding to the network to be optimized; generating a plurality of initial global solutions based on the undirected graph, where the initial global solution is a multi-bit binary number; respectively performing multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimized solution corresponding to the initial global solution, and calculating the weight sum of each first optimized solution; selecting the first optimized solution with the largest weight sum as the optimal solution of the network to be optimized; and determining the network unit corresponding to the bit representing the selected vertex in the optimal solution according to the corresponding relationship between the undirected graph and the network to be optimized. The present invention requires fewer qubits and is easy to implement, effectively simplifies the execution complexity, and improves the calculation accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of quantum computing, and particularly to a method, apparatus, device and medium for optimizing network interference based on quantum computing. Background Art

[0002] Networks play an important role in all walks of life in today's society. For example, in the field of mobile communication, the mobile communication network composed of base stations provides indispensable communication services for people's life and work; for another example, the wireless sensor network (WSN) applied in fields such as environmental monitoring, medical detection, and industrial automation also plays an important role in data transmission and communication. In the present invention, the base stations constituting the mobile communication network, the wireless modules constituting the WSN, etc. are collectively referred to as network units. For a network, the number of network units is closely related to the networking cost, management complexity, and operation stability; the importance of each network unit in the overall network is not exactly the same; for some networks, for the sake of safe operation, the signal coverage range of each network unit will have an overlapping area with the signal coverage range of other network units, so in specific applications, only some network units need to be turned on to meet the requirements; in addition, there may be interference between network units. For example, in the field of mobile communication, interference will occur between base stations in the same frequency band, and the interference between network units will affect the performance of the overall network. Therefore, for such a network, one purpose of network interference optimization is to determine the network units with the least interference, the largest number that can be turned on simultaneously, and the greatest importance.

[0003] Currently, the common optimization method is to use the graph theory method in mathematics to model the given network. The network units are represented as vertices in an undirected graph, and their attributes such as importance, coverage range, and capacity are represented by the assigned weights. The interference between network units is reflected as the connection between vertices. Therefore, the optimization problem of determining the network units with the least interference, the largest number that can be turned on simultaneously, and the greatest importance from the given network can be modeled as the maximum weighted independent set (MWIS) problem, whose purpose is to find a set formed by pairwise non-adjacent vertices and the sum of the weights of the vertices in the set is the largest. The commonly used solution method is to find the optimal solution of the maximum weighted independent set through the exhaustive method. However, as the network scale increases, the complexity of the interference relationship makes the difficulty of the exhaustive algorithm increase exponentially, so that the limited classical computing power is difficult to meet the requirements such as timeliness in actual applications.

[0004] At present, quantum algorithms based on the principles of quantum mechanics are gradually applied to various industries. A quantum algorithm is a new type of algorithm with quantum bits (Qubits) as the smallest computing unit. It makes full use of the characteristics of interference, superposition, and entanglement in the quantum system and can quickly and efficiently solve some classical algorithm problems whose complexity increases exponentially with the problem scale. However, in the noisy intermediate-scale quantum (NISQ) era, due to the limited number of qubits provided by quantum computers, it is difficult to solve the realistic network interference optimization problems with a large number of vertices and complex network structures. Summary of the Invention

[0005] In view of the technical problems existing in the prior art, the present invention proposes a network interference optimization method, device, equipment, and medium based on quantum computing, which can accurately solve the network interference optimization problem using a small number of qubits.

[0006] To solve the above technical problems, according to one aspect of the present invention, a network interference optimization method based on quantum computing is provided, including the following steps:

[0007] Construct an undirected graph corresponding to the network to be optimized. The network to be optimized includes multiple network units. The undirected graph includes vertices corresponding to each network unit in the network to be optimized. Each vertex includes the weight of the corresponding network unit, and the connection between vertices is used to represent the existence of interference between network units;

[0008] Generate multiple initial global solutions based on the undirected graph. The initial global solution is a multi-bit binary number, and each bit of the binary number corresponds to the corresponding vertex in the undirected graph. The bit value "1" in the binary number represents that the vertex corresponding to the bit is selected, and the bit value "0" in the binary number represents that the vertex corresponding to the bit is not selected;

[0009] Perform multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimized solution corresponding to each initial global solution, and calculate the sum of weights of each first optimized solution;

[0010] Select the first optimized solution with the largest sum of weights as the optimal solution of the network to be optimized;

[0011] According to the correspondence between the undirected graph and the network to be optimized, determine the network units corresponding to the bits representing the selected vertices in the optimal solution.

[0012] Optionally, the step of performing multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimized solution corresponding to each initial global solution includes:

[0013] Extract a preset number of bits and their values from the current global solution to form an optimization subset. Among them, when constructing the first optimization subset, the current global solution is the corresponding initial global solution. After each construction of the optimization subset, the current global solution is the latest second optimization solution obtained by optimizing based on the previous optimization subset.

[0014] Optimize each optimization subset through a quantum circuit with adjustable parameters to obtain a local optimal solution.

[0015] After obtaining the local optimal solution each time, use the local optimal solution to replace the corresponding bit value in the current global solution to obtain the latest second optimization solution, and calculate the weight sum of the latest second optimization solution.

[0016] Among them, after obtaining the latest second optimization solution each time, check whether the preset requirements for optimizing stop are met.

[0017] In response to meeting the preset requirements for optimizing stop, use the latest second optimization solution with the largest weight sum as the first optimization solution obtained by optimizing the corresponding initial global solution; in response to not meeting the preset requirements for optimizing stop, continue to construct optimization subsets for optimization.

[0018] Optionally, the step of extracting a preset number of bits and their values from the current global solution to form an optimization subset includes: determining the preset number based on a decay function.

[0019] Optionally, the decay function is a linear function, and the independent variable of the linear function represents the number of bits determined when constructing the previous optimization subset, and the linear coefficient of the linear function is a value less than 1.

[0020] Optionally, the step of extracting a preset number of bits and their values from the current global solution to form an optimization subset includes:

[0021] Execute the process of traversing the current global solution for a preset number of times. Extract a bit and its value from each process of traversing the current global solution as subset elements, and update the current global solution based on the bit value. Among them, the updated current global solution is used as the current global solution in the next process of traversing the current global solution, and the preset number is the same as the number of bits in the optimization subset.

[0022] Optionally, the step of traversing the current global solution process includes:

[0023] Traverse each bit in the current global solution, and respectively use each bit as the target bit.

[0024] Flip the bit value of the target bit to obtain the first target current global solution after flipping.

[0025] Calculate the current global solution original weight sum and the first target current global solution weight sum before the bit value of the target bit is flipped, and calculate the difference between the two weight sums;

[0026] After traversing all the bits of the current global solution, determine the bit with the largest weight sum difference as the subset bit, and determine the bit value after flipping the bit as the bit value of the subset bit;

[0027] Update the bit value of the corresponding bit in the current global solution based on the bit value of the subset bit to obtain a new current global solution.

[0028] Optionally, the step of calculating the first target current global solution weight sum includes:

[0029] Obtain the adjacent bit values of the target bit;

[0030] When the original bit value of the target bit and any one of the adjacent bit values are not all 1 and the first bit value obtained after flipping the target bit and the adjacent bit values are not all 1, when the first bit value is 1, add the weight of the vertex corresponding to the target bit to the current global solution original weight sum to obtain the first target current global solution weight sum; when the first bit value is 0, subtract the weight of the vertex corresponding to the target bit from the current global solution original weight sum to obtain the first target current global solution weight sum;

[0031] When the original bit value of the target bit and the adjacent bit values are all 1, calculate the sum of the weight of the vertex corresponding to the target bit and the weight of the vertex corresponding to the adjacent bit to obtain the first weight sum;

[0032] Calculate the product of the first weight sum and the first penalty coefficient as the first adjusted weight sum;

[0033] Add the first adjusted weight sum to the current global solution original weight sum to obtain the first target current global solution weight sum;

[0034] When the first bit value obtained after flipping the target bit and the adjacent bit values are all 1, calculate the sum of the weight of the vertex corresponding to the target bit and the weight of the vertex corresponding to the adjacent bit to obtain the first weight sum;

[0035] Calculate the product of the first weight sum and the second penalty coefficient as the second adjusted weight sum;

[0036] Subtract the second adjusted weight sum from the current global solution original weight sum to obtain the first target current global solution weight sum.

[0037] Optionally, the step of optimizing the optimization subset through a quantum circuit with adjustable parameters to obtain a locally optimal solution includes:

[0038] Construct a Hamiltonian with interference constraints based on the optimized subset, where the interference constraints include minimizing the interference between network units corresponding to the optimized subset, and where the bits in the optimized subset correspond to the applied qubits;

[0039] Adjust the parameter values in the quantum circuit to reduce the expected value of the Hamiltonian until convergence; and

[0040] Measure the quantum state when the expected value of the Hamiltonian converges, and determine the quantum state with the highest probability in the measurement result as the local optimal solution of the optimized subset.

[0041] Optionally, the step of generating multiple initial global solutions based on the undirected graph includes:

[0042] Determine the number of bits of the initial global solution based on the number of vertices of the undirected graph;

[0043] Randomly select a vertex and its adjacent vertices of the undirected graph;

[0044] Take each selected vertex as a target vertex, set the bit value of the bit corresponding to the target vertex to "1", and set the bit values of other bits to "0" to obtain an initial global solution corresponding to the target vertex.

[0045] According to another aspect of the present invention, the present invention also provides a network interference optimization device based on quantum computing, including:

[0046] An undirected graph construction module configured to construct an undirected graph corresponding to the network to be optimized, where the network to be optimized includes multiple network units, the undirected graph includes vertices corresponding to each network unit in the network to be optimized, each vertex includes the weight of the corresponding network unit, and the connection lines between the vertices are used to represent the existence of interference between network units;

[0047] A parameter acquisition module configured to generate multiple initial global solutions based on the undirected graph, where the initial global solutions are multi-bit binary numbers, each bit of the binary number corresponds to a corresponding vertex in the undirected graph, the bit value "1" of the binary number represents that the vertex corresponding to the bit is selected, and the bit value "0" in the binary number represents that the vertex corresponding to the bit is not selected;

[0048] An optimization module configured to perform multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimization solution corresponding to each initial global solution, and calculate the sum of the weights of each first optimization solution; and

[0049] The optimal solution acquisition module is configured to select the first optimization solution with the largest sum of weights as the optimal solution of the network to be optimized; and determine the network unit corresponding to the bit where the representative vertex in the optimal solution is selected according to the correspondence between the undirected graph and the network to be optimized.

[0050] According to another aspect of the present invention, the present invention also provides an electronic device, including a processor and a memory. Computer instructions are stored in the memory, and when the processor runs the computer instructions, the foregoing network interference optimization method based on quantum computing is executed.

[0051] According to another aspect of the present invention, the present invention also provides a computer-readable storage medium. The computer-readable storage medium stores computer instructions, and when the computer instructions are run by a processor, the foregoing network interference optimization method based on quantum computing is executed.

[0052] Based on the distributed quantum computing method, the present invention can accurately solve the network interference optimization problem with a small number of qubits. The number of qubits required is small, which can well adapt to the NISQ era, is easy to implement in physical experiments, effectively simplifies the execution complexity of implementation in a real quantum computer, and improves the calculation accuracy; compared with the quantum computing method of global encoding, the quantum circuit used in a single iteration of the present invention requires fewer qubits and has a shallower quantum circuit depth, thus significantly reducing the computing resources. Especially for solving large-scale network interference optimization problems, the effect is more significant. The present invention has a wide range of application fields, such as network communication, task scheduling, resource allocation, social network analysis, supply chain, and logistics optimization, etc., and effectively solves a variety of practical problems. Description of the Drawings

[0053] Next, the preferred embodiments of the present invention will be further described in detail with reference to the drawings, where:

[0054] Figure 1 is a flowchart of a network interference optimization method based on quantum computing according to an embodiment of the present invention;

[0055] Figure 2 is a flowchart of a method for generating multiple initial global solutions according to an embodiment of the present invention;

[0056] Figure 3 is a flowchart of a method for locally optimizing a target initial global solution according to an embodiment of the present invention;

[0057] Figure 4 is a flowchart of a method for extracting a preset number of bits and bit values from a target initial global solution to form an optimization subset according to an embodiment of the present invention;

[0058] Figure 5Flowchart of a method for calculating the weight sum of the first target current global solution according to an embodiment of the present invention;

[0059] Figure 6 Principle block diagram of a network interference optimization device based on quantum computing according to an embodiment of the present invention;

[0060] Figure 7 Schematic diagram of an undirected graph of a mobile network according to Application Example 1 of the present invention;

[0061] Figure 8 Schematic diagram of an undirected graph of a mobile network showing the optimal solution according to Application Example 1 of the present invention;

[0062] Figure 9 Schematic diagram of a quantum circuit acting on 5 qubits according to an embodiment of the present invention;

[0063] Figure 10 Schematic diagram of an undirected graph of a sensor network showing the optimal solution according to Application Example 2 of the present invention;

[0064] Figure 11 Schematic diagram of a quantum circuit acting on 4 qubits according to an embodiment of the present invention;

[0065] Figure 12 Structural principle block diagram of an electronic device according to an embodiment of the present invention. Detailed implementation manners

[0066] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0067] In the following detailed description, reference may be made to the accompanying drawings that form a part hereof, and in which are shown by way of illustration specific embodiments in which the application may be practiced. In the drawings, similar reference numerals describe substantially similar components in different views. The various specific embodiments of the present application have been described in sufficient detail below to enable those of ordinary skill in the art with relevant knowledge and technology to implement the technical solutions of the present application. It should be understood that other embodiments may be utilized or structural, logical or electrical changes may be made to the embodiments of the present application.

[0068] Efficient and accurate execution of quantum algorithms relies on system support centered around quantum computers. The current quantum systems are in the noisy intermediate-scale quantum (NISQ) era. Therefore, although there are various quantum-algorithm-based processing solutions for component optimization problems, when faced with the demands of large-scale quantum systems, although theoretically feasible, they are limited by the actual situation of the current development of quantum computers and are difficult to apply to practical scenarios. Additionally, the "first", "second", etc. in the technical feature names of the present invention do not indicate ranking but are only used to distinguish technical features with the same name.

[0069] Figure 1 is a flowchart of a network interference optimization method based on quantum computing according to an embodiment of the present invention. The method includes:

[0070] Step S1, construct an undirected graph corresponding to the network to be optimized. The network to be optimized includes multiple network units. The undirected graph includes vertices corresponding to each network unit in the network to be optimized. Each vertex includes the weight of the corresponding network unit, and the connection lines between vertices are used to represent the existence of interference between network units.

[0071] Step S2, generate multiple initial global solutions based on the undirected graph. Among them, the initial global solution is a multi-bit binary number, and the bits of the binary number correspond one-to-one with the vertices of the undirected graph. The bit value "1" in the binary number represents that the vertex corresponding to the bit is selected, and the bit value "0" in the binary number represents that the vertex corresponding to the bit is not selected.

[0072] Step S3, take an initial global solution as the target initial global solution.

[0073] Step S4, perform multiple local optimizations on the target initial global solution through a quantum circuit with adjustable parameters to obtain a first optimized solution corresponding to the initial global solution.

[0074] Step S5, calculate the weight sum of the first optimized solution.

[0075] Step S6, determine whether there is still an initial global solution that has not been optimized. If so, return to Step S3. If not, execute Step S7.

[0076] Step S7, sort the weight sums of all first optimized solutions and select the first optimized solution with the largest weight sum as the optimal solution.

[0077] Step S8, based on the correspondence between the undirected graph and the network to be optimized, determine the network units corresponding to the bits representing the selected vertices in the optimal solution.

[0078] Among them, the data of the network to be optimized includes the number of network units in the network, the weights assigned to each network unit based on attributes such as its importance, coverage, capacity, etc., and two network units with interference. In step S1, each network unit in the network to be optimized is used as a vertex. Based on the interference relationship between the network units in the network to be optimized, for two network units with an interference relationship, a connection is used to connect the vertices corresponding to these two network units. The weight of the vertex is the weight of the corresponding network unit.

[0079] In step S2, the initial global solution is a multi-bit binary number. The bits of the binary number correspond one-to-one with the vertices of the undirected graph. The bit values "1" and "0" of each bit represent that the vertex corresponding to the bit is selected and not selected, respectively. In one embodiment, refer to Figure 2 , Figure 2 which is a flowchart of a method for generating multiple initial global solutions according to an embodiment of the present invention. The method includes the following steps:

[0080] Step S21, determine the number of bits of the initial global solution based on the number of vertices of the undirected graph.

[0081] Step S22, randomly select a vertex of the undirected graph.

[0082] Step S23, obtain an initial global solution based on the vertex. Specifically, using the vertex as the target vertex, set the bit value of the bit corresponding to the target vertex to "1", and set the bit values of other bits to "0" to obtain an initial global solution corresponding to the target vertex.

[0083] Step S24, find all adjacent vertices of the vertex according to the connection relationship.

[0084] Step S25, obtain an initial global solution based on each adjacent vertex. Specifically, using each adjacent vertex as the target vertex, set the bit value of the bit corresponding to the target vertex to "1", and set the bit values of other bits to "0" to obtain an initial global solution corresponding to the adjacent vertex.

[0085] It can be known that step S22 and step S24 can be one step, that is, after randomly selecting a vertex of the undirected graph, find all adjacent vertices of the vertex according to the connection relationship, so as to obtain multiple determined vertices, and then generate initial global solutions corresponding to each vertex. In the initial global solution, the bit value corresponding to the vertex is "1", and the bit values of other bits are "0".

[0086] According to the foregoing method, multiple initial global solutions are obtained with a vertex and its adjacent vertices, which increases the diversity of solutions and the possible optimization space.

[0087] In step S4, when performing multiple local optimizations on the target initial global solution through a quantum circuit with adjustable parameters to obtain a first optimized solution corresponding to the initial global solution, refer to Figure 3 , Figure 3 is a flowchart of a method for locally optimizing a target initial global solution according to an embodiment of the present invention. In this embodiment, the method for locally optimizing a target initial global solution includes the following steps:

[0088] Step S41, extracting a preset number of bits and bit values from the target initial global solution to form an optimization subset.

[0089] Step S42, optimizing the optimization subset through a quantum circuit with adjustable parameters to obtain a local optimal solution.

[0090] Step S43, replacing the corresponding bit values in the target initial global solution with the local optimal solution to obtain the latest second optimized solution.

[0091] Step S44, calculating the weight sum of the latest second optimized solution.

[0092] Step S45, checking whether the preset requirements for stopping optimization are met. If the preset requirements for stopping optimization are met, execute step S46. If the preset requirements for stopping optimization are not met, return to step S41.

[0093] Step S46, sorting the weight sums of the latest second optimized solutions in descending order.

[0094] Step S47, taking the latest second optimized solution with the largest weight sum as the first optimized solution obtained by optimizing the target initial global solution, and ending the processing flow.

[0095] In the foregoing process, when extracting a preset number of bits from the target initial global solution, the preset number can be a fixed value, that is, the number of bits in each constructed optimization subset is fixed and equal. Additionally, the preset number can also be a value that changes according to a certain rule, that is, the number of bits in each constructed optimization subset is not equal. In one embodiment, the preset number is a value determined based on a decay function. That is, the number of bits in each constructed optimization subset decreases sequentially. In a more specific embodiment, the decay function is a linear function, the independent variable of which is the number of bits in the optimization subset determined in the previous construction of the optimization subset, the linear coefficient is a value less than 1, and the initial value of the preset number is a first threshold, and the first threshold is not greater than the number of qubits supported by the quantum computing device. For example, when determining the specific value of the preset number each time, it is calculated according to the following expression 1-1:

[0096] n_sub = n_sub • decay_rate (1 - 1)

[0097] Among them, n_sub on the left side of the equal sign is the value of the preset quantity to be obtained this time, and n_sub on the right side of the equal sign is the value of the preset quantity applied last time. When calculating the number of bits in the first optimized subset, n_sub on the right side of the equal sign is the initial value of the preset quantity, which is the maximum number of qubits supported by the quantum computing device. decay_rate on the right side of the equal sign is a decimal less than 1. Therefore, the value of the preset quantity to be obtained this time is less than the value of the preset quantity applied last time.

[0098] Of course, the linear function can also be in the form shown in Expression 1 - 2 below

[0099] f(n) = a - bn (1 - 2)

[0100] Among them, a is the initial value, b is the decay rate, and n is the number of iterations. Through this function, in each iteration, the number of bits in the optimized subset decreases according to a fixed ratio or quantity.

[0101] The aforementioned linear decay function is simple in form and easy to implement.

[0102] It should be noted that the decay function can also be implemented by other functions, such as the exponential decay function shown in Expression 1 - 3

[0103] f(n) = a • b n (1 - 3)

[0104] Among them, a is the initial value, b is a positive number less than 1, and n is the number of iterations, indicating that the number of bits in the optimized subset decreases exponentially after each iteration. The exponential decay function simulates a process of rapid decrease and then gradually tending to be stable.

[0105] Another example is the logarithmic decay function shown in Expression 1 - 4

[0106] f(n) = a • log(b • n), (1 - 4)

[0107] Among them, a and b are constants, and n is the number of iterations. As the number of iterations increases, the rate of decrease in the number of bits in the optimized subset gradually slows down.

[0108] Another example is the reciprocal decay function (f(n) = a / n), the polynomial decay function, and the custom decay function, etc. Those of ordinary skill in the art can choose any decay function according to the scale of the network to be optimized, application habits, etc. to determine the number of bits in the optimized subset.

[0109] In this embodiment, the preset requirement for determining whether to stop optimization is that the preset quantity is not greater than a second threshold, where the second threshold is, for example, a positive integer such as 3. That is, when the number of bits in the optimization subset is not greater than 3, this process ends.

[0110] When performing multiple local optimizations on the global solution based on the optimization subset in this embodiment, the global solution is optimized in the order from a large range to a small range, thereby improving the optimization efficiency.

[0111] See Figure 4 , Figure 4 is a flowchart of a method for extracting a preset number of bits and bit values from a target initial global solution to form an optimization subset according to an embodiment of the present invention. The method includes the following steps:

[0112] Step S410, determine each bit and its bit value of the target initial global solution x i For example, the target initial global solution is represented as x i = (y0,y1,y k ,…,y n-1 ). Here, i represents the serial number of the target initial global solution, k represents the bit serial number, y k represents the k-th bit value, "1" means that the vertex of the undirected graph corresponding to this bit is selected, and "0" means that the vertex of the undirected graph corresponding to this bit is not selected.

[0113] Step S411, calculate the original weight sum Ws_o of the target initial global solution x i .

[0114] Step S412, set k = 0 to determine the target bit.

[0115] Step S413, obtain the k-th bit value.

[0116] Step S414, change the k-th bit value in the target initial global solution x i to obtain the first target current global solution x i_k . For example, when the original bit value is "1", change it to "0", and when the original is "0", change it to "1".

[0117] Step S415, calculate the weight sum Ws_ak of the first target current global solution x i_k .

[0118] Step S416, calculate the difference △Ws_k between the two weight sums.

[0119] Where, △Ws_k =|Ws_o - Ws_ak|.

[0120] Step S417, determine whether k is less than n - 1. If k is not less than n - 1, but equal to n - 1, then execute Step S418. If k is less than n - 1, then in Step S4111, set k = k + 1, and return to Step S413.

[0121] Step S418, sort all the differences in weight sums △Ws_0, △Ws_1, △Ws_k... △Ws_(n - 1) in descending order.

[0122] Step S419, determine the differences in weight sums of the preset number ranked at the front, and obtain the corresponding bit numbers.

[0123] Step S4110, extract the corresponding bits and their bit values from the target initial global solution x i to construct an optimized subset.

[0124] Through Figure 4 the process shown, determine the bits corresponding to the undirected graph vertices that have the greatest impact on the weight sum from the initial global solution, so as to be able to preferentially optimize the vertices with the greatest global impact, improving the optimization efficiency.

[0125] See Figure 5 , Figure 5 is a flowchart of a method for calculating the weight sum of the first target current global solution according to an embodiment of the present invention. Calculating the weight sum of the first target current global solution in Step S415 includes the following steps:

[0126] Step S4150, obtain the adjacent bit values of the target bit.

[0127] Step S4151, determine whether all the original bit values of the target bit before flipping and any one of the adjacent bit values are all "1". If so, execute Step S4160. If not, execute Step S4152.

[0128] Step S4152, determine whether all the first bit values obtained after flipping the target bit and the adjacent bit values are all "1". If so, execute Step S4170. If not, execute Step S4153.

[0129] Step S4153, determine whether the first bit value obtained after flipping the target bit is "0", that is, whether it is flipped from "1" to "0". If so, execute Step S4154. If not, it means it is flipped from "0" to "1", and execute Step S4155.

[0130] Step S4154, subtract the weight of the vertex corresponding to the target bit from the original weight sum of the current global solution to obtain the weight sum of the first target current global solution, and end the process.

[0131] Step S4155, increase the weight of the vertex corresponding to the target bit in the current global solution original weight sum to obtain the first target current global solution weight sum, and end the process.

[0132] Step S4160, calculate the sum of the weight of the vertex corresponding to the target bit and the weight of the vertex corresponding to the adjacent bit to obtain the first weight sum.

[0133] Step S4161, calculate the product of the first weight sum and the first penalty coefficient as the first adjusted weight sum.

[0134] Step S4162, increase the first adjusted weight sum in the current global solution original weight sum to obtain the first target current global solution weight sum, and end the process.

[0135] Step S4170, calculate the sum of the weight of the vertex corresponding to the target bit and the weight of the vertex corresponding to the adjacent bit to obtain the first weight sum.

[0136] Step S4171, calculate the product of the first weight sum and the first penalty coefficient as the first adjusted weight sum.

[0137] Step S4172, subtract the first adjusted weight sum from the current global solution original weight sum to obtain the first target current global solution weight sum, and end the process.

[0138] Through the method for calculating the weight sum in this embodiment, when flipping the bit value, an appropriate weight sum can be calculated according to various situations before and after flipping, thereby increasing the overall calculation accuracy.

[0139] See Figure 3 , in step S42, the quantum circuit with adjustable parameters is, for example, a quantum circuit implementing the quantum approximate optimization algorithm (QAOA), a quantum circuit implementing the variational quantum eigenvalue solving algorithm (VQE), or a quantum circuit implementing the quantum neural network (QNN). Those skilled in the art can construct any quantum circuit according to their usage habits, which will not be elaborated here.

[0140] Among them, when optimizing the target optimization subset through a quantum circuit with adjustable parameters to obtain the target local optimal solution, first, a Hamiltonian with interference constraint conditions is constructed based on the target optimization subset. The interference constraint conditions include minimizing the interference between network units corresponding to the target optimization subset. Among them, the bits in the target optimization subset correspond to the applied qubits. Then, the parameters of the used quantum circuit are adjusted. When adjusting the parameters, the Hamiltonian is used as the loss function, and the parameters in the quantum circuit are adjusted in a way that reduces the expected value of the Hamiltonian until the Hamiltonian converges. When the expected value of the Hamiltonian converges, the quantum state at the time of convergence of the expected value of the Hamiltonian is measured, and the quantum state with the highest probability in the measurement result is determined as the target local optimal solution corresponding to the target optimization subset.

[0141] On the other hand, the present invention also provides a network interference optimization device based on quantum computing. Refer to Figure 6 , Figure 6 which is a schematic block diagram of a network interference optimization device based on quantum computing according to an embodiment of the present invention. The network interference optimization device 10 in the present invention includes an undirected graph construction module 11, a parameter acquisition module 12, an optimization module 13, and an optimal solution acquisition module 14. Among them, the undirected graph construction module 11 constructs an undirected graph corresponding to the network to be optimized. The network to be optimized includes multiple network units. The undirected graph includes vertices corresponding to each network unit in the network to be optimized. Each vertex includes the weight of the corresponding network unit. The connection lines between the vertices are used to represent that there is interference between the network units. The parameter acquisition module 12 generates multiple initial global solutions based on the undirected graph. Among them, the initial global solution is a multi-bit binary number. The bits of the binary number correspond one-to-one with the vertices of the undirected graph. The "1" and "0" of the binary number represent that the vertex corresponding to the bit is selected and not selected, respectively. The optimization module 13 performs multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimization solution corresponding to the initial global solution, and calculates the sum of the weights of each first optimization solution. The optimal solution acquisition module 14 selects the first optimization solution with the largest sum of weights as the optimal solution of the network to be optimized; and based on the correspondence between the undirected graph and the network to be optimized, determines the network unit corresponding to the bit representing the selected vertex in the optimal solution.

[0142] In one embodiment, the network interference optimization device based on quantum computing includes a classical computing device and a quantum computing device. For example, when the optimization module 13 performs multiple local optimizations on a target initial global solution through a quantum circuit with adjustable parameters, the quantum computing device runs the quantum circuit, measures the quantum state after the evolution of the quantum circuit, sends the measurement result to the classical computing device, and the classical computing device obtains the optimized optimal solution according to the measurement result and determines the corresponding network unit. The quantum computing device can be a quantum simulator or various types of real quantum machines.

[0143] Application Example 1

[0144] See Figure 7 , Figure 7 is a schematic diagram of an undirected graph based on a mobile network according to Application Example 1 of the present invention. The mobile network includes 15 base stations, and each vertex in the undirected graph corresponds to a base station. In the mobile network, based on attributes such as its importance, coverage area, and capacity, each base station has a corresponding weight. After quantization, the obtained weight values are shown as the numbers in the vertices in. When there is interference between two base stations, it is represented by a connection line between the corresponding vertices. The purpose of this application example is to determine the base station combination with the largest weighted sum and the smallest interference from the mobile network.

[0145] In the application of this embodiment, let v represent the network unit (base station) in the network, and the subscript j represents the serial number of the base station. Then v j represents the j-th base station in the base station set V, j = 0, 2,..., n - 1. The weight of the j-th base station v j is denoted as w j . In the following description, for the convenience of description, the undirected graph is corresponded to the network one by one, and the vertex set in the undirected graph is also denoted as V, and the vertex representing the base station is also denoted as v j , and the weight of the vertex is also the weight of the base station, denoted as w j . See Figure 7 , and the meaning of the numbers in the vertices is "vertex serial number - weight".

[0146] According to the method provided by the present invention, multiple initial global solutions are first constructed based on Figure 7 the undirected graph shown. For example, since the number of vertices in the undirected graph is 15, the number of bits for determining the initial global solution is 15, and the serial numbers of the bits correspond one by one to the vertex serial numbers. Then randomly select a vertex in the undirected graph and its adjacent vertices. For example, in Figure 7In [the above], when vertex v4 is selected, its adjacent vertices v2 and v14 are obtained according to the connection relationship. Then, each selected vertex is used as a target vertex, and the bit value of the bit corresponding to the target vertex is set to "1", and the bit values of other bits are set to "0" to obtain an initial global solution corresponding to the target vertex. The initial global solution for vertex v4 is:

[0147] x1 = [0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0].

[0148] That is, the bit value of the 5th bit is set to "1", and the others are "0".

[0149] The initial global solution for vertex v2 is:

[0150] x2 = [0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0].

[0151] That is, the bit value of the 3rd bit is set to "1", and the others are "0".

[0152] The initial global solution for vertex v14 is:

[0153] x3 = [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1].

[0154] That is, the bit value of the 15th bit is set to "1", and the others are "0".

[0155] After obtaining multiple initial global solutions, local optimization is performed respectively based on each initial global solution, and the optimized global solution is called the first optimized solution. Then, the weight sum corresponding to each optimized solution is calculated, and the first optimized solution with the largest weight sum is selected as the optimal solution of the network to be optimized. Based on the correspondence between vertices and base stations in the undirected graph, the base station corresponding to the bit representing the selected vertex in the optimal solution.

[0156] In this embodiment, when performing local optimization based on each initial global solution, according to the number of qubits 8 supported by the used quantum computing device, the number of bits in the first optimized subset is 8, and the number of bits in the second and third optimized subsets is determined by the adopted attenuation function to be 7, and the number of bits in the fourth to sixth optimized subsets is 7, and the optimization is stopped when the number of bits is equal to 4.

[0157] When performing local optimization based on the first initial global solution (such as x1), 8 bits and their bit values are extracted from the target initial global solution to form the first optimized subset. For example:

[0158] x 11 = [8, 5, 11, 0, 13, 9, 12, 3].

[0159] Each of these numbers is the bit number in the initial global solution and also the vertex number in the undirected graph. After optimizing the first optimization subset, the first local optimal solution X is obtained. 11 :

[0160] X 11 = [1, 1, 1, 1, 0, 0, 0, 0].

[0161] Using the first local optimal solution X 11 to replace the values in the original initial global solution to obtain the first first-optimized solution X1 1 :

[0162] X1 1 = [1, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 0, 0, 0].

[0163] Then calculate the first first weight sum as W1 1 = 334.0. In the first first-optimized solution X1 1 extract 7 bits and their bit values to form the second optimization subset x 12 :

[0164] x 12 = [13, 9, 0, 3, 12, 11, 1].

[0165] After optimizing the second optimization subset, the second local optimal solution X is obtained 12 :

[0166] X 12 = [1, 0, 0, 1, 0, 1, 0].

[0167] Using the second local optimal solution X 12 to replace the values in the first first-optimized solution X1 1 to obtain the second first-optimized solution X1 2 :

[0168] X1 2 = [0, 0, 0, 1, 1, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0].

[0169] The second first weight sum W1 2 = 317.0. And so on, until the optimization subset only includes the bit values of 4 bits. In this embodiment, the complete data is shown in Table 1 below:

[0170] Table 1:

[0171]

[0172] Select the first optimal solution corresponding to the maximum value 344.0 from the foregoing multiple first weight sums:

[0173] [1. 0. 0. 0. 1. 1. 0. 0. 1. 0. 0. 1. 0. 0. 0].

[0174] Similarly, similar data is obtained based on the second initial global solution x2.

[0175] Among them, x2 = [0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0].

[0176] Select the first optimal solution corresponding to the maximum value 350.0 from the multiple second weight sums among them:

[0177] [1. 0. 0. 0. 1. 1. 1. 0. 1. 0. 0. 0. 1. 0. 0].

[0178] Similar data is obtained based on the third initial global solution x3.

[0179] Among them, x3 = [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1].

[0180] Select the first optimal solution corresponding to the maximum value 350.0 from the multiple second weight sums among them:

[0181] [1. 0. 0. 0. 1. 1. 1. 0. 1. 0. 0. 0. 1. 0. 0].

[0182] Sorting the weight sums of the foregoing three first optimal solutions gives the maximum weight sum of 350.0. Therefore, the corresponding first optimal solution [1. 0. 0. 0. 1. 1. 1. 0. 1. 0. 0. 0. 1. 0. 0] is the optimal solution of this application embodiment. According to the bit number with a value of 1, determine the vertices with the same number in the undirected graph, such as Figure 8 shown, Figure 8 is a schematic diagram of an undirected graph of a mobile network showing the optimal solution according to Application Embodiment 1 of the present invention. The vertices marked with arrows in the figure are the vertices obtained according to the optimal solution. The vertex numbers are the same as the base station numbers. Therefore, the 0th, 4th, 5th, 6th, 8th, and 12th base stations with the least interference and the largest weight sum are determined from the current network.

[0183] Among them, when optimizing the optimization subset, the quantum circuit used when optimizing the first optimization subset acts on 8 quantum bits, and the quantum circuit used when optimizing the second optimization subset acts on 7 quantum bits, and so on. The number of quantum bits in the quantum circuit corresponds to the number of bits in the optimization subset. Each quantum circuit includes an initialization module, a problem Hamiltonian module, and a measurement module in the order of evolution. The initialization module includes a Hadamard gate, and a mixed superposition state is obtained by executing the Hadamard gate on the zero-state quantum bit. The problem Hamiltonian module, for example, uses a quantum circuit of the QAOA algorithm, or other variational quantum circuits, such as VQE, QNN, etc. The measurement module measures the quantum state of each quantum bit after the evolution of the problem Hamiltonian module.

[0184] In order to determine the Hamiltonian module of the problem in this application embodiment, the problem of this application embodiment is first abstracted into the objective function and constraints shown in the following expression 2-1:

[0185] (2-1)

[0186] The vertex set in the undirected graph is represented as V, and the interference situation corresponding to the network unit in the undirected graph is represented by the adjacency matrix ɛ. Each element in the matrix represents the interference between two vertices. In one embodiment, when there is interference between the two vertices, the element value is 1, and when there is no interference between the two vertices, the element value is 0. Corresponding to the interference optimization problem of the network of the present invention, it is modeled as a problem of solving the maximum weighted independent set. For a known network, its corresponding undirected graph is known, that is, its adjacency matrix ɛ and vertex v j and its weight w j All are known quantities.

[0187] In expression 2-1 Represents any vertex v j . Represents the value of the adjacency matrix element. According to the conditions that should be met above, when there is interference between two vertices, the value of the adjacency matrix element is 1. When it is substituted into the inequality, the inequality does not hold. Therefore, the two vertices are not elements in the maximum weighted independent set. When there is no interference between the two vertices, the value of the adjacency matrix element is 0. When it is substituted into the inequality, the inequality holds. Then the two vertices can be used as candidate elements of the maximum weighted independent set, and then the weighted sum of the candidate elements is calculated. The candidate elements with the largest weighted sum are combined to form the optimal solution of the maximum weighted independent set, thereby determining the corresponding base station in the network.

[0188] Based on the objective function and constraint conditions in the foregoing Expression 2-1, the Hamiltonian is constructed as shown in the following Expression 2-2:

[0189] (2-2)

[0190] Among them, Ho represents the target Hamiltonian mapped by the objective function; is the Pauli Z operator acting on the j-th qubit, and w j is the vertex weight corresponding to the j-th qubit. When the value of the weighted sum of all vertices is the largest, the value of the target Hamiltonian H O is the smallest. H C represents the constraint Hamiltonian; and are the Pauli Z operators acting on the k-th and the l -th qubits respectively; w k and w l are the vertex weights corresponding to the k-th qubit and the l -th qubit respectively, is the interference value between the vertex corresponding to the k-th qubit and the vertex corresponding to the l -th qubit. When there is interference between the two vertices ; when there is no interference between the two vertices . When and only when the selected vertices are not connected to each other, the value of the constraint Hamiltonian H C is the smallest. ρ is the specific gravity coefficient of the constraint Hamiltonian , which is used to express the importance degree of interference relative to the weighted sum.

[0191] Each optimization subset in the embodiments of this application also satisfies Expression 2-1, and the Hamiltonian based on the optimization subset also conforms to Expression 2-2.

[0192] Refer to Figure 9 , Figure 9 is a schematic diagram of a quantum circuit acting on 5 qubits according to an embodiment of the present invention. In the order of evolution, the quantum circuit sequentially includes an initialization module 21, a problem Hamiltonian module 22, and a measurement module 23. Among them, the problem Hamiltonian module 22 includes 3 layers of identical quantum circuit units 201, which include a plurality of RZ quantum gates and RX quantum gates, and each quantum gate is a parameterized quantum gate. During the parameter adjustment process, the parameter values in each quantum gate are changed to gradually reduce the Hamiltonian expectation value obtained based on the measured quantum state until convergence. When the calculated Hamiltonian expectation value no longer decreases, it is considered that the Hamiltonian expectation value has converged, and the parameter adjustment is completed.

[0193] Figure 9The structure of the quantum circuit shown is only an example, and those of ordinary skill in the art can design or apply any existing quantum circuit by themselves.

[0194] Application Example 2

[0195] As Figure 10 shown, Figure 10 is a schematic diagram of an undirected graph of a sensor network showing the optimal solution according to Application Example 2 of the present invention. The left vertices in the figure are the vertices determined to be selected according to the optimal solution, and the right side are the unselected vertices. The density of the undirected graph is 0.2. The numbers in the vertices represent weights. There are 35 sensors in the sensor network in this application example, and each sensor corresponds to Figure 10 one of the vertices. Applying the foregoing method, the first optimized subset includes 5 bits, that is, corresponding to 5 vertices, and is optimized using the quantum circuit of 5 qubits shown in Figure 9 , and the second optimized subset includes 4 bits, that is, corresponding to 4 vertices, and is optimized using the quantum circuit of 4 qubits shown in Figure 11 . See Figure 11 , Figure 11 is a schematic diagram of a quantum circuit acting on 4 qubits according to an embodiment of the present invention. Figure 11 The quantum circuit diagram in Figure 10 successively includes an initialization module 21, a problem Hamiltonian module 22, and a measurement module 23. Among them, the problem Hamiltonian module 22 includes 3 layers of the same quantum circuit units 201, and the RZ gates and RX gates therein are parameterized quantum gates. The optimal solution finally obtained in this embodiment is as shown in

[0196] The maximum weight sum is 737. After verification, it is consistent with the optimal solution and the maximum weight sum calculated by the classical solver.

[0197] In another embodiment, for an undirected graph constructed based on 100 network units, using an 8-qubit (MAX_QUBIT) quantum computer and applying 3 layers of quantum circuits, the optimal solution and its maximum weight sum can be efficiently solved. After verification, it is consistent with the optimal solution and the maximum weight sum calculated by the classical solver.

[0198] On the other hand, the embodiment of the present invention also provides an electronic device. See Figure 12 , Figure 12 is a block diagram of the structural principle of an electronic device according to an embodiment of the present invention. As Figure 12As shown, the electronic device includes a processor and a memory. Computer instructions are stored in the memory, and when the processor runs the computer instructions, it executes the network interference optimization method based on quantum computing provided by the present invention.

[0199] Specifically, the processor 601 may include a central processing unit (CPU) or a graphics processing unit (GPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. The memory 602 may include a memory for data or instructions. For example, the memory 602 may be at least one of the following: a hard disk drive (HDD), a read-only memory (ROM), a random access memory (RAM), a floppy disk drive, a flash memory, an optical disc, a magneto-optical disc, a magnetic tape, a universal serial bus (USB) drive, or other physical / tangible memory storage devices. Also, the memory 602 includes a removable or non-removable (or fixed) medium. Additionally, the memory 602 may be inside or outside the integrated gateway disaster recovery device. The memory 602 may be a non-volatile solid-state memory. In other words, generally, the memory 602 includes a tangible (non-transitory) computer-readable storage medium (such as a memory device) encoded with executable instructions, and when the executable instructions stored therein are executed by the processor 601 (such as by one or more processors), the network interference optimization method based on quantum computing in the embodiments of the present invention can be implemented.

[0200] In one example, Figure 12 The shown electronic device may further include a communication interface 603 and a bus 610. Among them, the processor 601, the memory 602, and the communication interface 603 are connected through the bus 610 to complete communication with each other. The communication interface 603 is mainly used to implement communication between various modules, devices, units, and / or devices in the electronic device.

[0201] The bus 610 includes hardware, software, or both, and can couple the components of the online data flow metering device to each other. For example, the bus can include at least one of the following: Accelerated Graphics Port (AGP) or other graphics buses, Extended Industry Standard Architecture (EISA) bus, Front Side Bus (FSB), HyperTransport (HT) interconnect, Industry Standard Architecture (ISA) bus, InfiniBand interconnect, Low Pin Count (LPC) bus, Memory bus, MicroChannel Architecture (MCA) bus, Peripheral Component Interconnect (PCI) bus, PCI-Express (PCI-X) bus, Serial Advanced Technology Attachment (SATA) bus, Video Electronics Standards Association Local (VLB) bus, or other suitable buses. The bus 610 can include one or more buses. Although embodiments of the present invention describe or illustrate specific buses, embodiments of the present invention can contemplate any suitable bus or interconnect method.

[0202] In another aspect, embodiments of the present invention further provide a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the foregoing quantum computing-based network interference optimization method is implemented. The computer-readable storage medium is, for example, a classical computer-readable storage medium, such as a Read-Only Memory (ROM), Random Access Memory (RAM), disk storage medium device, optical storage medium device, flash memory device, electrical, optical, or other physical / tangible memory storage device. It can also be a storage medium for storing quantum information and readable by a quantum computer, such as a Quantum Random Access Memory (QRAM). QRAM can be regarded as the quantum version of RAM in a classical computer. Through QRAM, a quantum superposition state of information can be created. Compared with RAM that needs to read data one by one, data in superposition can be read at superposed addresses. QRAM can be implemented in physical ways such as optical, semiconductor quantum dots, superconducting circuits, ion traps, etc.

[0203] The flowcharts and / or block diagrams of the methods and systems of embodiments of the present invention have been described above by way of example, and the related aspects have been described. It should be understood that each block or combination of blocks in the flowchart and / or block diagram can be implemented by computer program instructions, or by dedicated hardware that performs the specified function or action, or by a combination of dedicated hardware and computer instructions. When implemented in hardware, it can be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), appropriate firmware, a plug-in, a functional card, etc.; when implemented in software, it is a program or code segment used to perform the required tasks. The program or code segment can be stored in a memory, or transmitted through a data signal carried in a carrier wave on a transmission medium or a communication link. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0204] The above embodiments are only for illustrating the present invention and are not intended to limit the present invention. Those of ordinary skill in the relevant technical field can still make various changes and modifications without departing from the scope of the present invention. Therefore, all equivalent technical solutions should also fall within the scope of the disclosure of the present invention.

Claims

1. A network interference optimization method based on quantum computing, characterized in that Including: Construct an undirected graph corresponding to the network to be optimized, where the network to be optimized includes multiple network units, the undirected graph includes vertices corresponding to each network unit in the network to be optimized, each vertex includes the weight of the corresponding network unit, and the connection lines between vertices are used to represent the interference between network units; Generate multiple initial global solutions based on the undirected graph, where the initial global solutions are multi-bit binary numbers, each bit of the binary number corresponds to a corresponding vertex in the undirected graph, and the bit value "1" in the binary number represents that the vertex corresponding to the bit is selected, and the bit value "0" in the binary number represents that the vertex corresponding to the bit is not selected; Perform multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimized solution corresponding to each initial global solution, and calculate the weight sum of each first optimized solution; Select the first optimized solution with the largest weight sum as the optimal solution of the network to be optimized; According to the corresponding relationship between the undirected graph and the network to be optimized, determine the network units corresponding to the bits representing vertex selection in the optimal solution.

2. The network interference optimization method based on quantum computing according to claim 1, wherein The step of performing multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimized solution corresponding to each initial global solution includes: Extract a preset number of bits and bit values from the current global solution to form an optimization subset. When constructing the first optimization subset, the current global solution is the corresponding initial global solution. After each construction of the optimization subset, the current global solution is the latest second optimized solution obtained after optimizing based on the previous optimization subset; Optimize each optimization subset through a quantum circuit with adjustable parameters to obtain a local optimal solution; After obtaining the local optimal solution each time, replace the corresponding bit value in the current global solution with the local optimal solution to obtain the latest second optimized solution, and calculate the weight sum of the latest second optimized solution; Among them, after obtaining the latest second optimized solution each time, check whether the preset requirements for stopping optimization are met; In response to meeting the preset requirements for stopping optimization, use the latest second optimized solution with the largest weight sum as the first optimized solution obtained by optimizing the corresponding initial global solution; in response to not meeting the preset requirements for stopping optimization, continue to construct optimization subsets for optimization.

3. The network interference optimization method based on quantum computing according to claim 2, wherein The step of extracting a preset number of bits and bit values from the current global solution to form an optimization subset includes: determining the preset number based on an attenuation function.

4. The network interference optimization method based on quantum computing according to claim 3, wherein The attenuation function is a linear function, and the independent variable of the linear function represents the number of bits determined when constructing the previous optimization subset, and the linear coefficient of the linear function is a value less than 1.

5. The network interference optimization method based on quantum computing according to claim 2, characterized in that The step of extracting a preset number of bits and bit values from the current global solution to form an optimization subset includes: Execute the process of traversing the current global solution for a preset number of times, extract a bit and bit value from each process of traversing the current global solution as subset elements, and update the current global solution based on the bit value. The updated current global solution is used as the current global solution in the next process of traversing the current global solution, and the preset number is the same as the number of bits in the optimization subset.

6. The network interference optimization method based on quantum computing according to claim 5, wherein The step of traversing the current global solution process includes: Traverse each bit in the current global solution, and respectively take each bit as the target bit; Flip the bit value of the target bit to obtain the first target current global solution after flipping; Calculate the original weight sum of the current global solution before the bit value of the target bit is flipped and the weight sum of the first target current global solution, and calculate the difference between the two weight sums; After traversing all the bits of the current global solution, determine the bit with the largest weight sum difference as the subset bit, and determine the bit value after flipping the bit as the bit value of the subset bit; Update the bit value of the corresponding bit in the current global solution based on the bit value of the subset bit to obtain the updated current global solution.

7. The network interference optimization method based on quantum computing according to claim 6, wherein The step of calculating the weight sum of the first target current global solution includes: Obtain the adjacent bit values of the target bit; When the original bit value of the target bit and any adjacent bit value are not all "1" and the first bit value obtained after flipping the target bit and the adjacent bit value are not all "1", when the first bit value is "1", add the weight of the vertex corresponding to the target bit to the original weight sum of the current global solution to obtain the weight sum of the first target current global solution; when the first bit value is "0", subtract the weight of the vertex corresponding to the target bit from the original weight sum of the current global solution to obtain the weight sum of the first target current global solution; When the original bit value of the target bit and the adjacent bit values are all "1", calculate the sum of the weights of the vertex corresponding to the target bit and the weights of the adjacent vertices to obtain the first weight sum; Calculate the product of the first weight sum and the first penalty coefficient as the first adjusted weight sum; Add the first adjusted weight sum to the original weight sum of the current global solution to obtain the weight sum of the first target current global solution; When the first bit value obtained after flipping the target bit and the adjacent bit values are all "1", calculate the sum of the weights of the vertex corresponding to the target bit and the weights of the adjacent vertices to obtain the first weight sum; Calculate the product of the first weight sum and the second penalty coefficient as the second adjusted weight sum; Subtract the second adjusted weight sum from the original weight sum of the current global solution to obtain the weight sum of the first target current global solution.

8. The network interference optimization method based on quantum computing according to claim 2, wherein The step of optimizing the optimization subset through a quantum circuit with adjustable parameters to obtain a local optimal solution includes: Construct a Hamiltonian with interference constraint conditions based on the optimization subset, where the interference constraint conditions include minimizing the interference between network units corresponding to the optimization subset, and where the bits in the optimization subset correspond to the applied qubits; Adjust the parameter values in the quantum circuit so that the expected value of the Hamiltonian decreases until convergence; Measure the quantum state when the expected value of the Hamiltonian converges, and determine the quantum state with the highest probability in the measurement result as the local optimal solution of the optimization subset.

9. The network interference optimization method based on quantum computing according to claim 1, characterized in that The step of generating multiple initial global solutions based on the undirected graph includes: Determine the number of bits of the initial global solution based on the number of vertices of the undirected graph; Randomly select a vertex and its adjacent vertices of the undirected graph; Each selected vertex is used as a target vertex, and the bit value of the bit corresponding to the target vertex is set to "1", and the bit values of other bits are set to "0" to obtain an initial global solution corresponding to the target vertex.

10. A network interference optimization device based on quantum computing, characterized in that, Including: An undirected graph construction module configured to construct an undirected graph corresponding to the network to be optimized, the network to be optimized including a plurality of network units, the undirected graph including vertices corresponding to the respective network units in the network to be optimized, each vertex including the weight of the corresponding network unit, and the connections between the vertices being used to represent interference between network units; A parameter acquisition module configured to generate a plurality of initial global solutions based on the undirected graph, wherein the initial global solutions are multi-bit binary numbers, each bit of the binary number corresponding to a corresponding vertex in the undirected graph, and the bit value "1" of the binary number representing that the vertex corresponding to the bit is selected, and the bit value "0" in the binary number representing that the vertex corresponding to the bit is not selected; An optimization module configured to perform multiple local optimizations on each initial global solution through a quantum circuit with adjustable parameters to obtain a first optimization solution corresponding to each initial global solution, and calculate the sum of weights of each first optimization solution; An optimal solution acquisition module configured to select the first optimization solution with the largest sum of weights as the optimal solution of the network to be optimized; and determine the network unit corresponding to the bit representing that the vertex is selected in the optimal solution according to the correspondence between the undirected graph and the network to be optimized.

11. An electronic device, comprising a processor and a memory, characterized in that, Computer instructions are stored in the memory, and when the processor runs the computer instructions, it executes the network interference optimization method based on quantum computing as described in any one of claims 1-9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and when the computer instructions are run by the processor, it executes the network interference optimization method based on quantum computing as described in any one of claims 1-9.

Images