Antenna layout planning method for indoor distributed system

Abstract

The invention discloses an antenna layout planning method and device for an indoor distributed system, and the method comprises the steps: 1, inputting initial parameter information, which comprises building scene parameters, antenna parameters and signal receiver information; in the second step, a parent population is initialized, wherein the parent population comprises N distributed antenna layout schemes; 3, generating a filial generation population containing N layout planning schemes by adopting a genetic algorithm; 4, N distributed antenna layout schemes are selected from the parent population and the child population according to the selected strategy to serve as candidate populations; and step 5, selecting KE individuals from the candidate population based on a distribution prediction agent model to determine a final distributed antenna layout scheme, wherein the distribution prediction agent model is used for fitting and calculating an evaluation index function.

Description

Technical Field

[0001] The present application relates to an antenna layout planning method, and particularly to an antenna layout planning method for an indoor distributed system of a 5G communication system based on Bayesian optimization. Background Art

[0002] In recent years, with the continuous improvement of mobile communication requirements, the number of connected devices and the scale of transmitted data have shown an explosive growth. To meet the needs of users, 5G mobile communication technology has emerged, providing three major application scenarios: large-scale access connection, ultra-reliable low latency, and enhanced mobile broadband. It is reported that nearly 80% of data transmission occurs in indoor scenarios. Considering the severe attenuation of 5G high-frequency signals with distance and penetration, it is difficult for outdoor base station antennas to provide a high-quality and high-coverage communication environment for indoor users. In China, the coverage of 5G signals varies greatly in different scenarios. The signal coverage rate on outdoor main roads such as in cities can reach 90%, while the signal coverage rate in many indoor scenarios is less than 60%. Even worse, in some more enclosed environments such as underground parking lots and elevators, the signal coverage is less than 50%. Therefore, providing reliable wireless signal coverage indoors is an important step towards achieving ubiquitous 5G signal coverage.

[0003] Existing 5G network planning software mainly focuses on the planning of outdoor base station distribution networks, and few software is specifically designed for 5G network planning of indoor distribution systems, such as Ranplan. Currently, in 5G indoor distribution projects, multiple objectives often need to be considered, such as construction costs (materials used, construction period) and network quality (power loss, signal coverage). However, existing 5G network planning software is all based on single-objective optimization methods and cannot provide users with the best compromise planning scheme between different objectives.

[0004] The indoor distributed antenna system (In-Building Distributed Antenna Systems, IB-DAS) is widely deployed as a popular technology to enhance mobile communication services within buildings. As an effective and feasible method, the indoor distribution system accesses external source signals indoors and then propagates the signals through indoor ceiling antennas to provide signal coverage for indoor scenarios. In the indoor distribution system, the layout of indoor ceiling antennas is closely related to the indoor signal coverage range, indoor signal quality, and construction cost. They exhibit superior performance in achieving higher data rates, reducing interference, and providing coverage in blind spots. Different from directly relying on outdoor base stations (BS), a typical IB-DAS transmits signals from outdoor signal sources to the indoor environment through a topology consisting of devices such as indoor antennas, coaxial cables, couplers, and distributors. The antenna placement problem is one of the important challenges in designing IB-DAS, which is to determine the optimal antenna positions and the optimal number of antennas to optimize the coverage range, reduce interference, and manage costs.

[0005] However, the existing methods for optimizing the indoor antenna layout still have the following problems: (1) low algorithm efficiency, with only one result output per operation; (2) inaccurate indoor signal propagation models adopted, which are not applicable to the propagation simulation of 5G high-frequency signals, thus resulting in a large deviation between the optimization effect and the actual effect; (3) high computational cost for methods using accurate propagation models, with expensive optimization costs. Summary of the Invention

[0006] According to one aspect of the present disclosure, there is provided a method for antenna layout planning for an indoor distributed system, the method comprising: in step S101, inputting initial parameter information, where the initial parameter information includes building scenario parameters, antenna parameters, and signal receiver information; in step S201, initializing a parental population, the parental population including N distributed antenna layout schemes; in step S301, using a genetic algorithm to generate an offspring population including N layout planning schemes; in step S401, selecting N distributed antenna layout schemes from the parental population and the offspring population as a candidate population according to a selected strategy; in step S501, selecting K E individuals from the candidate population based on a distribution prediction surrogate model for determining the final distributed antenna layout scheme, where the distribution prediction surrogate model is used to fit and calculate an evaluation index function.

[0007] According to an embodiment of the present disclosure, the selected strategy includes an environment selection strategy based on decomposition.

[0008] According to an embodiment of the present disclosure, selecting N distributed antenna layout schemes from the parent population and the offspring population as the candidate population according to a selected strategy includes determining the aggregation function values corresponding to each individual in the parent population and the offspring population according to a decomposition-based environment selection strategy, and selecting N distributed antenna layout scheme individuals from the individuals in the parent population and the offspring population based on the aggregation function values corresponding to each individual in the parent population and the offspring population.

[0009] According to an embodiment of the present disclosure, the initial parameter information further includes: parameters of the decomposition-based environment selection strategy.

[0010] According to an embodiment of the present disclosure, the building scenario parameters include the length d of the building floor plan l and the width d w ; and / or the signal receiver information includes at least one of the received signal threshold θ, the signal interference threshold θ′, and the number of input users N at the user; and / or the parameters of the decomposition-based environment selection strategy include at least one of the number K of individuals used to truly evaluate the evaluation index in each algorithm iteration R and the neighbor set size T. E

[0011] According to an embodiment of the present disclosure, truly evaluating the evaluation index includes evaluating the evaluation index using a ray tracing propagation model.

[0012] According to an embodiment of the present disclosure, the method further includes: initializing the decomposition-based environment selection strategy. In a further embodiment, initializing the decomposition-based environment selection strategy includes obtaining an initialized reference point set based on at least one of K E and T The reference point set is used to calculate the aggregation function values of each population individual.

[0013] According to an embodiment of the present disclosure, the method further includes: replacing the parent population in step S201 with the candidate population, jumping back to step S201, and looping through steps S201 to S501 until a preset number of loops is reached.

[0014] According to an embodiment of the present disclosure, selecting K E individuals from the candidate population based on the distributed prediction surrogate model to determine the final distributed antenna layout scheme includes: initializing the surrogate model data set where it contains each distributed antenna layout scheme individual in the parent population and the function value associated with the evaluation index corresponding to the individual; based on the surrogate model data set Establish a distribution prediction surrogate model; select K individual distributed antenna layout scheme individuals from the candidate population based on the distribution prediction surrogate model, and calculate the evaluation index function values corresponding to the K individuals; update the surrogate model data set based on the K individuals and their corresponding evaluation index function values; and determine the final distributed antenna layout planning scheme based on the updated surrogate model data set. E individuals, and calculate the evaluation index function values corresponding to the K E individuals; update the surrogate model data set based on the K E individuals and the evaluation index function values corresponding thereto; and determine the final distributed antenna layout planning scheme based on the updated surrogate model data set.

[0015] According to an embodiment of the present disclosure, the evaluation index includes at least one of signal coverage rate, signal interference rate, and layout cost.

[0016] According to an embodiment of the present disclosure, the signal coverage rate is determined based on the received signal strength at the signal receiver and the received signal threshold θ in the signal receiver information; and / or the signal interference rate is determined based on the received signal strength at the signal receiver and the signal interference threshold θ'; and / or the received signal strength is calculated based on the ray tracing propagation model.

[0017] According to an embodiment of the present disclosure, generating an offspring population including N layout planning schemes by using a genetic algorithm includes generating offspring individuals by using a differential evolution crossover mutation operator.

[0018] According to an embodiment of the present disclosure, determining K E individuals based on the distribution prediction surrogate model includes: obtaining the predicted value and uncertainty of the function value associated with the evaluation index based on the distribution prediction surrogate model, and selecting K E individuals based on the obtained predicted value and uncertainty of the function value associated with the evaluation index.

[0019] According to an embodiment of the present disclosure, selecting K E individuals based on the obtained predicted value and uncertainty of the function value associated with the evaluation index further includes calculating the expected improvement value based on the obtained predicted value and uncertainty of the function value associated with the evaluation index, and selecting K E individuals based on the calculated expected improvement value.

[0020] According to an embodiment of the present disclosure, the expected improvement value is calculated by the following steps: removing individuals from the candidate population that are the same as or similar to the individuals whose evaluation indices have been truly evaluated in to obtain a set obtaining the individuals Q in the set through the distribution prediction surrogate model i the predicted value y(Q i ) of the evaluation function corresponding to and the uncertainty s 2 (Q i); Based on the predicted value y(Q i ) and the uncertainty s 2 (Q i ) calculate the expected improvement value EI(Q i ). i )

[0021] According to an embodiment of the present disclosure, the distribution prediction proxy model is formed based on a Gaussian proxy model

[0022] According to an embodiment of the present disclosure, the method is applicable to a 5G communication system

[0023] According to an aspect of the present disclosure, there is also provided a device for antenna layout planning of an indoor distributed system. The device includes: an input unit, a controller, and an output unit. Among them, the input unit is configured to receive the input of initial parameter information, the output unit is configured to output a final distributed antenna layout scheme, and the controller is configured to control the input unit and the output unit to execute the method described above

[0024] According to an aspect of the present disclosure, there is also provided a computer-readable storage medium storing instructions. When the instructions are executed by a computer, the computer is prompted to execute the method described above BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a schematic plan view in which a uniform spatial division, a region of d l ×d w is divided into m×n grids, and each grid is assigned a receiver R i .

[0026] Figure 2 is a hidden meta-variable representation method according to an embodiment of the present disclosure

[0027] Figure 3 illustrates an antenna planning method for a 5G indoor distributed system according to an embodiment of the present disclosure

[0028] Figure 4 illustrates a simplified antenna planning method according to an embodiment of the present disclosure DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions of the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present disclosure. Obviously, the described embodiments are some, but not all, of the embodiments of the present disclosure. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of the present disclosure without creative efforts shall fall within the protection scope of the present disclosure

[0030] In addition, the various operations, functions, or algorithms described below can be implemented or supported by one or more computer programs, each of which is formed by computer-readable program code and embodied in a computer-readable medium. The terms "application" and "program" refer to one or more computer programs, software components, instruction sets, processes, functions, objects, classes, instances, related data, or portions thereof suitable for implementation in a suitable computer-readable program code. The phrase "computer-readable program code" includes any type of computer code, including source code, object code, and executable code. The phrase "computer-readable medium" includes any type of medium that can be accessed by a computer, such as read-only memory (ROM), random access memory (RAM), hard disk drives, compact discs (CDs), digital video discs (DVDs), or any other type of memory. A "non-transitory" computer-readable medium excludes wired, wireless, optical, or other communication links that transmit transitory electrical signals or other signals. Non-transitory computer-readable media include media that can permanently store data and media that can store and later rewrite data, such as rewritable optical discs or erasable memory devices.

[0031] The terms used herein to describe the embodiments of the present application are not intended to limit and / or define the scope of the present application. For example, unless otherwise defined, technical terms or scientific terms used in the present disclosure should have the ordinary meaning understood by those of ordinary skill in the art to which the present application pertains.

[0032] In addition to the high cost of the antenna placement optimization problem, the present application also considers several other basic objectives - including coverage and interference - for optimization. Therefore, the present application formulates the antenna placement optimization problem as an expensive multi-objective optimization problem and applies a decomposition-based Bayesian optimization method to solve this problem. Bayesian optimization is a powerful tool for solving expensive optimization problems, and the decomposition-based framework is a typical method for solving multi-objective problems and maintaining the diversity of solutions.

[0033] The present disclosure uses a ray tracing model to simulate signal strength, and establishes a computationally inexpensive surrogate model to fit computationally expensive evaluation metric functions such as the signal coverage rate, signal interference rate, and construction cost of a 5G indoor distributed antenna layout; by encoding the indoor antenna layout planning scheme as a population represented by real number encoding, using crossover and mutation operators to generate offspring populations, and combining the prediction results of the surrogate model, the population is updated based on an environmental selection strategy to obtain the candidate population for each algorithm iteration; representative individuals are selected from each generation of candidate populations for true evaluation to update the surrogate model, and finally the optimal individual solution set is selected from all individuals that have been truly evaluated and decoded to obtain the planning scheme.

[0034] The objective of the present disclosure is to solve problems existing in existing indoor distributed antenna layout planning methods, such as low algorithm efficiency and high optimization cost, and to propose a 5G indoor distributed antenna layout planning method based on Bayesian optimization.

[0035] The antenna layout planning method for a 5G indoor distributed system based on Bayesian optimization proposed by the present disclosure can provide multiple optimal compromise planning schemes for decision-makers and constructors after one run, and improve problems existing in existing indoor distributed antenna layout planning methods, such as low algorithm efficiency, inaccurate signal propagation simulation, and high optimization cost, so as to better meet the actual engineering requirements.

[0036] Figure 3 An antenna planning method for a 5G indoor distributed system based on Bayesian optimization according to an embodiment of the present disclosure is shown. Although the antenna planning method herein is claimed to be applicable to 5G communication systems, it can also be applicable to 3G communication systems, 4G communication systems, etc.

[0037] Specifically, the antenna planning method according to the present disclosure includes the following steps:

[0038] S1. Input initial parameter information. In one embodiment, the initial parameter information may include building scene parameters, antenna parameters, and signal receiver information, and initialize the receiver set according to the signal receiver information; in addition, the initial parameter information may further include environment selection strategy parameters based on decomposition, and the environment selection strategy parameters based on decomposition include population size N.

[0039] S2. Randomly initialize N distributed antenna layout schemes, and these N distributed antenna layout schemes form a parent population; initialize the surrogate model data set including each individual in the parent population and the function value associated with the evaluation index corresponding to the individual, where the evaluation index includes at least one of coverage rate, signal interference rate, and layout cost, but is not limited thereto; initialize the environment selection strategy based on decomposition. In a further embodiment, the distribution prediction surrogate model is formed based on a Gaussian surrogate model.

[0040] S3. Establish a distribution prediction surrogate model based on the surrogate model data set. In a further embodiment, the distribution prediction surrogate model can be used to fit and calculate the evaluation index function, and it can give the predicted value and uncertainty of the function value associated with the evaluation index.

[0041] S4. Use a genetic algorithm to generate an offspring population containing N layout planning schemes;

[0042] S5. Select N distributed antenna layout schemes from the parent population and the offspring population as the best population according to the environment selection strategy based on decomposition.

[0043] S6. Select K individuals from the updated best population, and calculate their evaluation indicators for these K individuals; update the surrogate model dataset based on these K individuals and the evaluation indicator function values corresponding to the individuals. E individuals, and calculate their evaluation indicators for these K E individuals; update the surrogate model dataset based on these K E individuals and the evaluation indicator function values corresponding to the individuals.

[0044] S7. Jump back to step S2 and loop until a preset number of loops is reached. Finally, decode the best individual solution set in the surrogate model dataset into a distributed antenna layout planning scheme.

[0045] In a more specific embodiment, step S1 includes the following steps:

[0046] S1.1 Input building scene parameters, and the building scene parameters may include building floor plan information, such as the length d l and width d w ; Figure 1 is a schematic plan of a plane where the area of d l ×d w is divided into m×n grids through uniform spatial partitioning, and each grid is assigned a receiver R i .

[0047] S1.2 Input antenna parameters, and the antenna parameters may include at least one of the antenna transmission power, antenna carrier frequency, antenna installation height, and the upper limit U of the number of antennas used;

[0048] S1.3 Input signal receiver information, and the signal receiver information includes at least one of the received signal threshold θ, signal interference threshold θ′, and the number of input users N R in the user; initialize the user set where respectively represent the abscissa and ordinate of the user R i in the building scene, where

[0049] S1.4 Input the parameters of the decomposition-based environmental selection strategy, and the parameters of the decomposition-based environmental selection strategy include at least one of the population size N, the number K of individuals used for real evaluation in each algorithm iteration E , and the size T of the neighbor set.

[0050] In a further embodiment, the randomly initializing N distributed antenna layout schemes in step S2 includes:

[0051] S2.1.1 Randomly establish a parent individual X composed of U antennas i , X iEach antenna in contains three antenna attributes: the abscissa of the antenna the ordinate of the antenna and the antenna activation identifier used to determine whether the j-th antenna participates in the calculation of the evaluation index function where i all the antennas participating in the calculation of the evaluation index function in In each specific embodiment, such as Figure 2 shown, if then T i does not participate in the target calculation. By optimizing the number of antennas can be automatically optimized.

[0052] S2.1.2 Repeat step S2.1.1 until N parent individuals X i are generated. All X i constitute the parent population

[0053] In a further embodiment, the evaluation index described in step S2 includes at least one of signal coverage rate, signal interference rate, and layout cost. The following describes these three evaluation indexes in detail.

[0054] Before describing the evaluation index in detail, the definition of received signal strength is introduced first. Receiver R i usually receives signals from multiple antennas. To simplify the problem, the present disclosure uses the received signal strength of the receiver as an index to determine whether the receiver is covered. To calculate the evaluation index, the received signal S i at each receiver R i should be obtained first. First, for each antenna T in the antenna set j , j = 1,..., N T , first calculate the received signal strength S j of T i at R i,j . The final received signal strength S i at R i is calculated by the following formula:

[0055]

[0056] S i,jIt can be calculated through a ray-tracing propagation model. The ray-tracing propagation model mentioned here does not specify a particular ray-tracing propagation model algorithm and can be an existing or future ray-tracing propagation model. The ray-tracing propagation model is an accurate propagation model, with accurate simulation calculation results for the true signal strength, but the calculation cost is expensive, which in turn leads to an expensive calculation cost for the following evaluation metric function.

[0057] S2.2.1 Signal coverage rate

[0058] In the antenna placement optimization problem, the signal coverage rate is the main optimization goal. Denote the signal coverage rate at the receiver R i as If the received signal strength S i at the receiver R i is greater than the received signal strength threshold θ, then it is said that R i is covered, which can be obtained from the following formula:

[0059]

[0060] The signal coverage rate provided by the antenna set can be expressed by the following formula:

[0061]

[0062] where the antenna set is the antenna set participating in the calculation of the evaluation metric.

[0063] S2.2.2 Signal interference rate

[0064] The interference rate reflects how many covered receivers are affected by the interference caused by the antenna set , and the optimization goal is to minimize the interference rate. In this application, multiple homogeneous omnidirectional antennas operating at the same frequency are used. Different from some previous work that regarded the overlapping part of the antenna coverage as the interference metric, in this application, interference is regarded as: how many covered receivers are affected by interference. Denote the signal interference rate at the receiver R i as First, R i must be covered, because it is meaningless to calculate interference for uncovered receivers. Then, for each covered R i , if the difference between the received signal strength S i at R i and the received signal strength S j from other antennas T i,j is less than θ', and the number of eligible T j is greater than n' (where n' is a predefined value), then it is said that R iInterfered, which can be expressed by the following formula

[0065]

[0066] The signal interference rate caused by the antenna array can be expressed by the following formula

[0067]

[0068] S2.2.3 Layout cost

[0069] In this application, it is assumed that the indoor distributed antennas used are all the same omnidirectional antennas, excluding other wiring costs. Therefore, the layout cost brought by the antenna array is determined by the number of antennas in That is: which is determined by the number of antennas in

[0070]

[0071] where

[0072] Based on the above three evaluation indicators, the comprehensive evaluation index function of this step can be expressed as

[0073] min(f1,f2,fs),

[0074]

[0075]

[0076]

[0077] where is the antenna array participating in the calculation of the evaluation index. In f1, maximizing the signal coverage rate is transformed into minimizing the signal non-coverage rate, so that the three comprehensive evaluation indicators can be minimized simultaneously.

[0078] In a further embodiment, initializing the proxy model data set in step S2 includes the following steps:

[0079] S2.3.1 Establish an empty set

[0080] S2.3.2 Calculate the evaluation index for all individuals X in i to obtain F(X i ) = [F1(X i ),...,F m (X i)], where \(i = 1, \ldots, N\), where \(F(\cdot)\) is the value of the evaluation index function, \(N\) is the population size, and \(m\) is the number of evaluation indices.

[0081] S2.3.3 Add all \([X i , F(X i )] to the set

[0082] The initialization of the decomposition-based environmental selection strategy in step S2 includes the following steps:

[0083] S2.4.1 Uniformly sample \(N\) weight vectors, where the \(i\)-th weight vector where \(N\) is the population size. Each \(w i is associated with the corresponding parent individual \(X i .

[0084] S2.4.2 Create their neighbor index set \(B i for each weight vector \(w i , which is used to store the indices of \(T\) weight vectors closest to \(w i .

[0085] S2.4.3 Establish an empty set Store the optimal value of each evaluation index in That is:

[0086]

[0087] where \(Z j belongs to the set Finally, obtain the initialized reference point set where the reference point set stores the best value on each current evaluation index, which is used to calculate the aggregation function value of each population individual (for updating the population individual) in subsequent steps. In each specific embodiment, the aggregation function value can be calculated using the Chebyshev method. In addition, other methods (such as weighted sum, etc.) can also be selected.

[0088] In a further embodiment, the establishment of the distribution prediction surrogate model described in step S3 based on the surrogate model dataset includes the following steps:

[0089] S3.1.1 In each algorithm iteration, establish a distribution prediction surrogate model \(\sum\) based on the surrogate model dataset . Given a new individual \(X * , on each evaluation index, the distribution prediction surrogate model \(\sum\) can give the predicted value \(y(X * ) of the evaluation index function value \(F(X *), and an uncertainty (e.g., it can be the mean square error) s 2 (X * ). The predicted value y(X * ) and the uncertainty s 2 (X * ) can be expressed by the following formula:

[0090]

[0091]

[0092] where r is the covariance matrix between X * and , C is a covariance matrix of with respect to X i , 1 is a column vector with all elements equal to 1 -dimensional. μ and σ 2 need to satisfy:

[0093]

[0094] to make the best unbiased estimate.

[0095] In a further embodiment, step S4 includes the following steps:

[0096] S4.1.1 Establish an empty set P. Randomly generate a random real number rand within the interval (0, 1). If rand is less than the probability δ, then P = B i , i = 1,..., N; otherwise P = {1,..., N}. That is

[0097]

[0098] S4.1.2 Let the index r1 = i, and then randomly select two indices r2, r3 from P.

[0099] S4.1.3 The individuals pair and use the differential evolution crossover mutation operator to generate the offspring X′ i . The differential evolution crossover mutation operator here does not specify a particular differential evolution crossover mutation operator.

[0100] S4.1.4 Repeat steps S4.1.1 to S4.1.3 until all the offspring individuals X′ i , i = 1,..., N are generated. All the offspring constitute the offspring population

[0101] In a further embodiment, step S5 includes: selecting N distributed antenna layout schemes with better performance from the parent population and the offspring population as the best population based on the decomposed environmental selection strategy, where the performance can be determined based on the aggregation function value.

[0102] More specifically, step S5 may include the following steps:

[0103] S5.1.1 Obtain X′ through the distributed prediction proxy model ∑ i The predicted value y(X′ i ) of the evaluation index function, where y(X′ i ) = [y1(X′ i ),..., y k (X′ i ), k = 1,..., m, and m is the number of evaluation indicators.

[0104] S5.1.2 Update the reference point set The update step specifically includes: comparing the predicted value y k (X′ i ) on each evaluation indicator with the current reference point on each evaluation indicator , and taking the smaller value as the updated reference point value, that is, taking as the updated value.

[0105] S5.1.3 Sequentially take the index j from the set P, and calculate the aggregation function value g i of the offspring Xi te (X′ i ) and the aggregation function value g j of the parent X te (X j ). The aggregation function value can be calculated by the following formula:

[0106]

[0107]

[0108] S5.1.4 If g te (X′ i ) < g te (X j ), then replace X i with X′ j as the new parent individual, otherwise retain the parent individual.

[0109] S5.1.5 Repeat steps S5.1.3 to S5.1.4 until all indices in the set P have been operated on.

[0110] S5.1.6 Return to S5.1.1 to operate on the next offspring individual, and repeat S5.1.1 - S5.1.5 until all N offspring have been operated on. At this time, the updated parent population is the best population for this algorithm iteration.

[0111] In a further embodiment, selecting K E individuals from the best population includes selecting the K E individuals with the highest Expectation Improvement value from the best population. The Expectation Improvement value is an indicator used to evaluate the expected goodness of a solution. Specifically, if a solution has a higher Expectation Improvement value, it is more likely to perform well in the true evaluation. In other embodiments, the selected K E individuals can also be other representative individuals that can be implemented by those skilled in the art.

[0112] In a further embodiment, the calculation of the Expectation Improvement value includes the following steps:

[0113] S6.1.1 Record the minimum aggregation function value of the individuals in the best population

[0114] S6.1.2 Establish an empty set Select all the individuals in the best population that have been truly evaluated and whose Euclidean distance from the individuals in -5 is greater than 1×10 This step is a duplicate removal operation, removing the individuals in the best population that are the same as or similar to the individuals that have been truly evaluated in . The individuals after duplicate removal form the set

[0115] S6.1.3 Obtain the predicted value y(Q ) and uncertainty s i corresponding to the individual Q in i through the distribution prediction surrogate model ∑, where 2 (Q i ). For example, through the Chebyshev decomposition method, obtain the predicted value g(Q ) and uncertainty i ) of the aggregation function g 2 (Q i ) from y(Q te ) and s i ). i ) and uncertainty

[0116] S6.1.4 Calculate the individual Qi The expected improvement EI(Q i ) can be expressed by the following formula:

[0117]

[0118] where Φ(·) is the cumulative distribution function and φ(·) is the probability density function.

[0119] In a further embodiment, updating the surrogate model data set based on the K E individuals and their corresponding evaluation metric function values described in step S6 comprises the following steps:

[0120] S6.2.1 Cluster the individuals in the set into K E settlements, without specifying a particular clustering method here;

[0121] S6.2.2 Calculate the expected improvement value for each individual in each settlement (in the specific manner shown above), and select the individual Q′ i , i = 1, ..., K E for true evaluation, and calculate the evaluation metrics corresponding to these individuals to obtain F(Q′ i );

[0122] S6.2.3 Add [Q′ i , F(Q′ i )] to the surrogate model data set

[0123] According to an embodiment of the present disclosure, an apparatus may also be provided. The apparatus includes an input unit, a controller, and an output unit. Among them, the input unit is configured to receive various inputs as described above, the output unit is configured to output the final distributed antenna layout planning scheme, and the controller is configured to control the output unit and the input unit to execute each step as described above.

[0124] As a simplified embodiment, the apparatus can implement as Figure 4The method steps shown. Among them, in step S101, initial parameter information is input into the input unit, where the initial parameter information includes building scene parameters, antenna parameters, and signal receiver information; then, the controller executes the following steps. In step S201, the parent population is initialized, and the parent population includes N distributed antenna layout schemes; and the decomposition-based environmental selection strategy is initialized. In step S301, a genetic algorithm is used to generate an offspring population including N layout planning schemes. In step S401, N distributed antenna layout schemes are selected from the parent population and the offspring population according to the decomposition-based environmental selection strategy as the candidate population. In step S501, individuals that meet the set conditions are selected from the candidate population as the final distributed antenna layout scheme. This embodiment can be implemented in combination with or independently of the above embodiments. The more specific implementation manners of the steps in this embodiment can adopt the same or similar implementation manners as those described above regarding Figure 3 in the above description.

[0125] According to an embodiment of the present disclosure, a computer-readable storage medium storing instructions may also be provided, wherein, when the instructions are executed by a computer, the computer is caused to execute any one of the above methods according to the exemplary embodiments of the present disclosure. Examples of such computer-readable storage media include: read-only memory (ROM), programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD+R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD+R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or optical disk memory, hard disk drive (HDD), solid state drive (SSD), cartridge memory (such as, multimedia card, secure digital (SD) card or extreme digital (XD) card), magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid state disk, and any other device configured to store a computer program and any associated data, data files, and data structures in a non-transitory manner and to provide the computer program and any associated data, data files, and data structures to a processor or computer such that the processor or computer can execute the computer program. The instructions or computer programs in the above computer-readable storage media may run in an environment deployed in computer devices such as clients, hosts, proxy devices, servers, etc. In addition, in one example, the computer program and any associated data, data files, and data structures are distributed over a networked computer system such that the computer program and any associated data, data files, and data structures are stored, accessed, and executed in a distributed manner by one or more processors or computers.

[0126] Those skilled in the art will understand that the above illustrative embodiments are described herein and are not intended to be limiting. It should be understood that any two or more of the embodiments disclosed herein may be combined in any combination. In addition, other embodiments may be utilized and other changes may be made without departing from the spirit and scope of the subject matter presented herein. It will be readily understood that the aspects of the invention of the present disclosure as generally described herein and illustrated in the figures can be arranged, substituted, combined, separated, and designed in a variety of different configurations, all of which are contemplated herein.

[0127] Those skilled in the art will understand that the various illustrative logical blocks, modules, circuits, and steps described in this application can be implemented as hardware, software, or a combination of both. To clearly illustrate this interchangeability of hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in terms of their functional sets. Whether such a functional set is implemented as hardware or software depends on the particular application and the design constraints imposed on the overall system. Those skilled in the art can implement the described functional sets in different ways for each specific application, but such design decisions should not be construed as causing a departure from the scope of this application.

[0128] The various illustrative logical blocks, modules, and circuits described in this application can be implemented or executed with a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general-purpose processor can be a microprocessor, but in the alternative, the processor can be any conventional processor, controller, microcontroller, or state machine. The processor can also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

[0129] The steps of the methods or algorithms described in this application can be embodied directly in hardware, in software modules executed by a processor, or in a combination of both. The software modules can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, a hard disk, a removable disk, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such that the processor can read from and write to the storage medium. In the alternative, the storage medium can be integrated into the processor. The processor and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor and the storage medium can reside as discrete components in a user terminal.

[0130] In one or more exemplary designs, the functions may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions may be stored on or transmitted via a computer-readable medium as one or more instructions or code. The computer-readable medium includes both a computer storage medium and a communication medium including any medium that facilitates transfer of a computer program from one place to another. The storage medium may be any available medium that can be accessed by a general purpose or special purpose computer.

[0131] The above description is only exemplary embodiments of the present application and is not intended to limit the scope of protection of the present application. The scope of protection of the present application is determined by the appended claims.

Claims

1. A method for antenna layout planning of an indoor distributed system, characterized in that The method includes: In step one, input initial parameter information, where the initial parameter information includes building scene parameters, antenna parameters, and signal receiver information; In step two, initialize the parent population, where the parent population includes N distributed antenna layout schemes; In step three, use a genetic algorithm to generate an offspring population containing N layout planning schemes; In step four, select N distributed antenna layout schemes from the parent population and the offspring population as the candidate population according to a selected strategy; and In step five, K individuals are selected from the candidate population based on the distribution prediction surrogate model for determining the final distributed antenna layout scheme, where the distribution prediction surrogate model is used to fit and calculate the evaluation index function. E ​ 2. The method according to claim 1, wherein The selected strategy includes a decomposition-based environmental selection strategy.

3. The method according to claim 2, wherein Selecting N distributed antenna layout schemes from the parent population and the offspring population as the candidate population according to the selected strategy includes determining the aggregation function value corresponding to each individual in the parent population and the offspring population according to the decomposition-based environmental selection strategy, and selecting N distributed antenna layout scheme individuals from the individuals in the parent population and the offspring population based on the aggregation function value corresponding to each individual in the parent population and the offspring population.

4. The method according to claim 2 or 3, characterized in that, The initial parameter information further includes: parameters of the decomposition-based environmental selection strategy.

5. The method according to claim 5, wherein The building scene parameters include the length d of the building floor plan l and the width d w ; and / or The signal receiver information includes at least one of the received signal threshold θ, signal interference threshold θ at the user, and the number of input users N; and / or ′ and the number of input users N R in; and / or The parameters of the decomposition-based environmental selection strategy include at least one of the number K of individuals used to truly evaluate the evaluation metrics in each iteration of the algorithm E and the size T of the neighbor set.

6. The method according to claim 5, wherein the true evaluation of the evaluation index includes evaluating the evaluation index using a ray tracing propagation model.

7. The method according to claim 5 or 6, further comprising: Initialize the decomposition-based environmental selection strategy, Among them, the initialization of the decomposition-based environmental selection strategy includes obtaining an initialized set of reference points based on at least one of K E and T The set of reference points is used to calculate the aggregation function value of each individual in the population.

8. The method according to any one of claims 1 to 7, further includes: Use the candidate population to replace the parent population in step S201, jump back to step S201, and loop through steps S201 to S501 until a preset number of loops is reached.

9. The method according to claim 1, wherein Selecting K individuals from the candidate population based on the distribution prediction surrogate model for determining the final distributed antenna layout scheme includes: E individuals for determining the final distributed antenna layout scheme includes: Initialize the proxy model dataset Among them It contains each individual of the distributed antenna layout scheme in the parental population and the function value associated with the evaluation index corresponding to the individual; Based on the surrogate model dataset Build a distribution prediction surrogate model; Select K individual distributed antenna layout schemes from the candidate population based on the distribution prediction surrogate model, and calculate the evaluation index function values corresponding to these K E individuals; E ​ Based on the said K E update the surrogate model data set based on the K individuals and their corresponding evaluation index function values; and Determine the final distributed antenna layout planning scheme based on the updated surrogate model dataset.

10. The method according to any one of claims 1 to 9, characterized in that The evaluation index includes at least one of signal coverage rate, signal interference rate, and layout cost.

11. The method according to claim 10, wherein, The signal coverage rate is determined based on the received signal strength at the signal receiver and the received signal threshold θ in the signal receiver information; and / or The signal interference rate is determined based on the received signal strength at the signal receiver and the signal interference threshold θ ′ and / or The received signal strength is calculated based on a ray tracing propagation model.

12. The method according to any one of claims 1 to 11, characterized in that, Using a genetic algorithm to generate an offspring population containing N layout planning schemes includes using a differential evolution crossover mutation operator to generate offspring individuals.

13. The method according to any one of claims 1 to 12, characterized in that, Determine K based on the distribution prediction surrogate model E The K individuals include: obtaining the predicted value and uncertainty of the function value associated with the evaluation index based on the distribution prediction surrogate model, and selecting K E individuals based on the obtained predicted value and uncertainty of the function value associated with the evaluation index.

14. The method according to claim 13, wherein Selecting K based on the predicted values and uncertainties of the function values associated with the evaluation metrics E selecting K individuals further includes calculating an expected improvement value based on the predicted values and uncertainties of the function values associated with the evaluation metrics, and selecting K E individuals based on the calculated expected improvement value.

15. The method according to claim 14, wherein The expected improvement value is calculated through the following steps: Remove from the candidate population individuals that are the same as or similar to the individuals whose evaluation metrics have been truly evaluated in to obtain a set Obtain a set through a distributed prediction surrogate model for the individual Q i in the corresponding predicted value y(Q i ) of the evaluation function and the uncertainty s 2 (Q i ); Calculate the expected improvement value EI(Q i ) based on the predicted value y(Q 2 ) and the uncertainty s i (Q i ). i ) 16. The method according to any one of claims 1 to 15, characterized in that, The distribution prediction surrogate model is formed based on a Gaussian surrogate model.

17. The method according to any one of claims 1 to 16, characterized in that, The method is applicable to a 5G communication system.

18. An apparatus for antenna layout planning of an indoor distributed system, the apparatus comprising: An input unit, a controller, and an output unit, where the input unit is used to receive the input of the initial parameter information, the output unit is used to output the final distributed antenna layout scheme, and the controller is configured to control the input unit and the output unit to execute the method according to any one of claims 1 to 17.

19. A computer-readable storage medium for storing instructions, wherein, When the instruction is executed by a computer, it causes the computer to execute the method according to any one of claims 1 to 17.

Images