Resource scheduling method for crowd spatiotemporal traffic prediction based on long short-term memory network

Abstract

The application provides a resource scheduling method for crowd spatiotemporal traffic prediction based on a long short-term memory network, and aims to solve 5G base station cache deployment and resource allocation decision and 5G base station active / inactive switching scheme. The method comprises the following steps: step 1: a decision node determines the attributes of a first date, and calls second crowd spatiotemporal traffic data of one or more base stations; step 2: the decision node sets an error rate threshold, initializes an iteration weight, and calculates one or more first crowd spatiotemporal traffics for one or more base stations in a coverage area; and step 3: the decision node determines the network resource demand of one or more base stations in the coverage area based on the one or more first crowd spatiotemporal traffics. The application aims at 5G base station cache deployment and resource allocation decision, and improves the resource efficiency of a wireless communication system.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of crowd spatiotemporal flow prediction, and particularly to a resource scheduling method for crowd spatiotemporal flow prediction based on a long short-term memory network. BACKGROUND

[0002] With the rapid development of mobile network technology, the use and demand of users for the network are ubiquitous and the data traffic is growing rapidly, which promotes the development of 5G technology and the commercialization process of 5G network. Through the intensive deployment of small cell networks and the application of mobile edge computing, the bandwidth and service quality of mobile networks can be greatly enhanced. However, how to truly improve the efficiency of a large number of small base stations and fully exert the capability of edge computing is one of the key problems for improving the service quality of 5G network. Since the number of users and the service demand of users are important factors affecting the deployment of base station cache, resource allocation and energy management, it is of great significance to improve the performance of 5G network by accurately predicting the number of users in the coverage of a base station, especially for a specific functional area covered by a certain base station, such as a scenic area, an office area, a residential area, etc. Since different areas have their specific service demand patterns, the prediction of the number of people in an area will be conducive to the deployment and allocation of network resources in the area. SUMMARY

[0003] The application aims to provide a resource scheduling method for crowd spatiotemporal flow prediction based on a long short-term memory network, which aims to solve the 5G base station cache deployment and resource allocation decision and the 5G base station active / inactive switching scheme.

[0004] The technical solution adopted by the application is based on the continuity and periodicity of crowd spatiotemporal flow in time, extracts the characteristics of crowd spatiotemporal flow, and predicts the crowd spatiotemporal flow in one or more base station coverage areas in a region. In the application, the wireless communication technology is based on the communication technology of 5G or the extended version of 5G, or the unlicensed frequency band communication technology of MultiFire. The communication terminal in the application is a terminal node based on the 5G standard or the unlicensed frequency band Internet of Things communication technology of MultiFire. The device that provides wireless data services for the terminal is called a network access point, that is, the base station of a cell. The device that provides resource allocation strategies and active / inactive switching schemes for the base station is called a decision node.

[0005] In the above application, the method comprises:

[0006] Step 1: The decision node determines the attribute of the first date and calls the second crowd spatiotemporal flow data of one or more base stations, which is the same as or contains the attribute of the first date;

[0007] Step 2: the decision node sets an error rate threshold, initializes an iteration weight, and calculates one or more first crowd spatiotemporal traffics for one or more base stations in the coverage area according to the second crowd spatiotemporal traffic data until the error rate threshold is met or the number of iterations reaches a limit;

[0008] Step 3: the decision node determines a network resource demand strategy for the one or more base stations in the coverage area based on the one or more first crowd spatiotemporal traffics.

[0009] In an embodiment, the attribute of the first date includes a type of the first date and time information of the first date, wherein the type of the first date includes a holiday and a weekday, and the time information of the first date is a day of the week and a number of hours in a day corresponding to the first date, or a day of the holiday and a number of hours in a day corresponding to the first date.

[0010] In an embodiment, the second crowd spatiotemporal traffic data contains historical N-day crowd spatiotemporal traffic data, wherein the crowd spatiotemporal traffic is represented as a two-dimensional data and is obtained based on a network geographic grid.

[0011] The first crowd spatiotemporal traffic is a crowd spatiotemporal traffic corresponding to the first date.

[0012] In an embodiment, a data size of the second crowd spatiotemporal traffic data is Nx1, i.e., the time information of the second crowd spatiotemporal traffic data corresponds to the first date one by one; and a data size of the selected second crowd spatiotemporal traffic data is NxM, wherein M>1, i.e., the time information of the second crowd spatiotemporal traffic data contains the time information of the first date.

[0013] In an embodiment, the type of the second crowd spatiotemporal traffic data is the same as the type of the first date.

[0014] In an embodiment, the second crowd spatiotemporal traffic data is obtained based on a call detail record or a geographic position generated when a terminal uses an APP.

[0015] In an embodiment, the decision node is a base station or a network centralized scheduling control unit, which is configured to provide a resource allocation strategy and an active / inactive switching scheme for the base station.

[0016] In an embodiment, the first crowd spatiotemporal traffic or the second crowd spatiotemporal traffic is based on a network geographic grid, and the network geographic grid is a grid of MxJ, which maps different base stations in a certain area to the grid, and a crowd spatiotemporal traffic in each sub-grid is a two-dimensional matrix.

[0017]

[0018] wherein, represents the crowd spatio-temporal flow in the (m, j) sub-grid at time point i in a day, which is a LxK matrix; L and K represent the division of the sub-grid, corresponding to the accuracy of the geographical position; The elements contained in represent the number of people in a geographical position in the sub-grid, which is an integer greater than or equal to 0; The sum of all elements in represents the number of people in the (m, j) sub-grid at time point i in a day.

[0019] In a second aspect, the present application provides a data transmission control system based on crowd scheduling, which comprises a base station for calculating a first crowd spatio-temporal flow and formulating a network resource demand strategy, including: a radio frequency transceiver module, a processor, a memory and a network resource demand strategy module, the radio frequency transceiver module is used for receiving wireless signals from the terminal, inputting into the processor for signal processing, and sending signals from the processor; the memory stores instructions, data cache and crowd spatio-temporal flow data that can be executed by the processor; the processor is used to call the instructions stored in the memory, and the processor and the network resource demand strategy module interact with information; the network resource demand strategy module includes a crowd spatio-temporal flow calculation submodule and a network resource demand strategy submodule, the crowd spatio-temporal flow calculation submodule obtains second crowd spatio-temporal flow data based on the memory, calculates a first crowd spatio-temporal flow, and the network resource demand strategy submodule generates a network resource demand strategy based on the first crowd spatio-temporal flow obtained by the crowd spatio-temporal flow calculation submodule, and provides it to the processor.

[0020] The beneficial technical effect of the present application is: compared with the prior art, the present application provides a resource scheduling method based on crowd spatio-temporal flow prediction of long short-term memory network for 5G base station cache deployment and resource allocation decision, which improves the resource efficiency of the wireless communication system. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0022] Figure 1 A wireless communication system schematic diagram of the resource scheduling method based on crowd spatio-temporal flow prediction of long short-term memory network provided by an embodiment of the present application is shown;

[0023] Figure 2A resource scheduling flowchart of crowd spatiotemporal traffic prediction based on long short-term memory network of one embodiment of the present application is shown.

[0024] Figure 3 A crowd spatiotemporal traffic diagram based on network grid of one embodiment of the present application is shown.

[0025] 100. wireless communication network 101. base station 110. first area 121. communication terminal

[0026] 170. antenna module 171. crowd spatiotemporal traffic calculation sub-module 172. network resource demand policy sub-module

[0027] 173. radio frequency transceiver module 174. processor 175. storage module 192. crowd network resource demand policy module DETAILED DESCRIPTION

[0028] In order to better understand the technical solutions of the present application, the embodiments of the present application are described in detail below in combination with the drawings.

[0029] It should be clear that the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0030] The terms used in the embodiments of the present application are only for the purpose of describing the specific embodiments, and are not intended to limit the present application. The singular forms "a", "an" and "the" used in the embodiments of the present application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0031] Embodiment one

[0032] Figure 1 A wireless communication system schematic diagram of providing crowd spatiotemporal traffic prediction based on long short-term memory network of one embodiment of the present application is shown.

[0033] The technical solution of the present application is based on the continuity and periodicity of the crowd space-time traffic in time, extracts the characteristics of the crowd space-time traffic, and predicts the crowd space-time traffic in one or more base station coverage areas in a region. The wireless communication technology in the present application is based on 5G or 5G evolved communication technology or extended version, or MultiFire unlicensed frequency band communication technology. The communication terminal in the present application is based on the terminal node of 5G standard or MultiFire unlicensed frequency band Internet of Things communication technology. The device providing wireless data service for the terminal is called network access point, that is, the base station of a cell. The device providing resource allocation strategy and active / inactive switching scheme for the base station is called decision node.

[0034] As shown in Figure 1 The wireless communication network 100 is an OFDM / OFDMA-based system, including a base station 101 and a communication terminal 121. The service area of the base station 101 is the first area 110, and the data transmission between the communication terminal 121 and the base station 101 is within this range. According to the embodiment of the present application, Figure 1 The module structure of the base station 101 is also given. The base station 101 has an antenna module 170 for receiving and transmitting signals. A radio frequency transceiver module 173 is connected to the antenna module 170, which is used to obtain signals from the antenna module 170, convert them into baseband signals, and deliver them to the processing module 174. The radio frequency transceiver module 173 can also convert the baseband signals from the processing module 174 into radio frequency signals and transmit them through the antenna module 170. The processing module 174 processes the baseband signals, acts according to the signals, and calls other modules of the base station 101 to perform other processing. The storage module 175 stores data.

[0035] The base station 101 also includes a crowd network resource demand strategy module 192, which includes a crowd space-time traffic calculation submodule 171 and a network resource demand strategy submodule 172. The crowd space-time traffic calculation submodule 171 obtains the second crowd space-time traffic data from the storage module 175 to calculate the first crowd space-time traffic, and the network resource demand strategy submodule 172 generates a network resource demand strategy based on the first crowd space-time traffic obtained by the crowd space-time traffic calculation submodule 171 and provides it to the processor 174.

[0036] The present application provides a resource scheduling method based on crowd space-time traffic prediction of long short-term memory network for 5G base station cache deployment and resource allocation decision, which improves the resource efficiency of the wireless communication system.

[0037] Step 1: The decision node determines the attributes of the first date, and calls the second crowd space-time traffic data of one or more base stations, which is the same as or contains the attributes of the first date;

[0038] Step 2: the decision node sets an error rate threshold, initializes an iteration weight, and calculates one or more first crowd spatiotemporal traffics for one or more base stations in the coverage area according to the second crowd spatiotemporal traffic data until the error rate threshold is met or the number of iterations is reached;

[0039] Step 3: the decision node determines a network resource demand strategy for the one or more base stations in the coverage area based on the one or more first crowd spatiotemporal traffics.

[0040] In one embodiment, in step 1 of the present application, the attribute of the first date includes the type of the first date (holiday and weekday), and the time information of the first date, wherein the time information of the first date is the day of the week and the hours of the day corresponding to the working day of the first date, or the day of the week and the hours of the day corresponding to the holiday.

[0041] In one embodiment, in step 1 of the present application, the second crowd spatiotemporal traffic data contains historical N-day crowd spatiotemporal traffic data, wherein the crowd spatiotemporal traffic is represented as a two-dimensional data and is obtained based on a network geographic grid.

[0042] In step 2 of the present application, the first crowd spatiotemporal traffic is the crowd spatiotemporal traffic corresponding to the first date.

[0043] In one embodiment, the data size of the second crowd spatiotemporal traffic data is N x 1, that is, the time information of the second crowd spatiotemporal traffic data corresponds to the first date one by one. In another embodiment, the data size of the selected second crowd spatiotemporal traffic data is NxM, wherein M>1, that is, the time information of the second crowd spatiotemporal traffic data contains the time information of the first date.

[0044] In one embodiment, the type of the second crowd spatiotemporal traffic data is the same as the type of the first date.

[0045] In one embodiment, the second crowd spatiotemporal traffic data is obtained based on call detail records. In another embodiment, the second crowd spatiotemporal traffic data is obtained based on geographic positions generated when terminals use APPs.

[0046] In one embodiment, the decision node is a base station. In another embodiment, the decision node is a network centralized scheduling control unit.

[0047] In one embodiment, the first crowd spatiotemporal traffic or the second crowd spatiotemporal traffic is based on a network geographic grid, and the network geographic grid is a grid of MxJ that maps different base stations in a certain area, and the crowd spatiotemporal traffic in each sub-grid is a two-dimensional matrix

[0048]

[0049] wherein, represents the crowd spatiotemporal flow in the (m, j) sub-grid at time point i in a day, which is an LxK matrix, and further, L and K represent the division of the sub-grid, corresponding to the accuracy of the geographical position; The element contained in the matrix represents the number of people in a geographical position in the sub-grid, which is an integer greater than or equal to 0. Further, The sum of all elements in the matrix represents the number of people in the (m, j) sub-grid at time point i in a day.

[0050] Embodiment two

[0051] Figure 2 The resource scheduling flowchart of the crowd spatiotemporal flow prediction based on the long short-term memory network according to an embodiment of the application is shown.

[0052] As shown in the figure, the resource scheduling flowchart of the crowd spatiotemporal flow prediction based on the long short-term memory network according to an embodiment of the application comprises: Figure 2

[0053] Step 211: The decision node determines the attributes of the first date, including the type of the first date (holiday and weekday) and the time information of the first date;

[0054] Step 212: The decision node calls the second crowd spatiotemporal flow data which is the same as or contains the attributes of the first date;

[0055] Step 213: The decision node initializes the iteration weight and the error rate threshold;

[0056] Step 214: The decision node calculates one or more first crowd spatiotemporal flows;

[0057] Step 215: The decision node judges whether the output result meets the error rate threshold or reaches the iteration number. If the result is yes, go to step 216; if the result is no, go to step 214;

[0058] Step 216: The decision node formulates the network resource demand strategy.

[0059] The application provides a resource scheduling method for crowd spatiotemporal flow prediction based on the long short-term memory network, aiming at the 5G base station cache deployment and resource allocation decision, which improves the resource efficiency of the wireless communication system.

[0060] Embodiment three

[0061] Figure 3 ​A network grid based crowd spatio-temporal flow diagram of one embodiment of the present application is shown.

[0062] As shown in FIG. 1, a network grid based crowd spatio-temporal flow diagram of one embodiment of the present application includes: Figure 3

[0063] Mapping different base stations in a certain area into an MxJ grid, the crowd spatio-temporal flow in each sub-grid is a two-dimensional matrix

[0064]

[0065] wherein, represents the crowd spatio-temporal flow in the (m, j)th sub-grid at time point i in a day, which is an LxK matrix, and further, L and K represent the division of the sub-grid, corresponding to the accuracy of the geographical position; The element contained therein represents the number of people at a geographical position in the sub-grid, which is an integer greater than or equal to 0. Further, The sum of all elements in represents the number of people in the (m, j)th sub-grid at time point i in a day.

[0066] The present application provides a resource scheduling method based on long short-term memory network crowd spatio-temporal flow prediction for 5G base station cache deployment and resource allocation decision, which improves the resource efficiency of the wireless communication system.

[0067] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0068] It should be understood that although the terms first, second, etc. can be used in the embodiments of the present application to describe time points, these time points should not be limited to these terms. These terms are only used to distinguish the time points from each other. For example, the first time point can also be referred to as the second time point, and similarly, the second time point can also be referred to as the first time point without departing from the scope of the embodiments of the present application.

[0069] Depending on the context, the word "if" as used herein can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detecting (a stated condition or event)" can be interpreted as "when determined" or "in response to determining" or "when detecting (a stated condition or event)" or "in response to detecting (a stated condition or event)".

[0070] ​It should be noted that the terminal involved in the embodiments of the present application can include, but is not limited to, a personal computer (PC), a personal digital assistant (PDA), a wireless handheld device, a tablet computer, a mobile phone, an MP3 player, an MP4 player, and the like.

[0071] In several embodiments provided by the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, another division mode can be adopted. For example, a plurality of units or components can be combined or integrated into another system, or some features can be omitted or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0072] In addition, each functional unit in the various embodiments of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware, or in the form of hardware plus software function unit.

[0073] The integrated unit realized in the form of software function unit can be stored in a computer readable storage medium. The software function unit stored in the storage medium includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute part of the steps of the method described in the various embodiments of the present application. The storage medium mentioned above includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0074] The above description is merely the preferred embodiments of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1.A resource scheduling method based on long short-term memory network for crowd spatio-temporal traffic prediction, characterized in that, Comprise: Step 1: the decision node determines the attribute of the first date, and calls one or more base stations of the second crowd space-time flow data, which is the same as or contains the first date attribute; The attribute of the first date includes the type of the first date and the time information of the first date, wherein the type of the first date includes holiday and weekday, and the time information of the first date is the day of the week and the hour of the day corresponding to the first date, or the day of the holiday and the hour of the day corresponding to the first date; The second crowd space-time flow data contains historical N-day crowd space-time flow data, wherein the crowd space-time flow is represented as a two-dimensional data based on network geographic grid; The first crowd space-time flow is the crowd space-time flow corresponding to the first date; The data size of the second crowd space-time flow data is N x1, that is, the time information of the second crowd space-time flow data corresponds to the first date one by one; the data size of the selected second crowd space-time flow data is NxM, wherein M >1, that is, the time information of the second crowd space-time flow data contains the time information of the first date; The type of the second crowd space-time flow data is the same as the type of the first date; The second crowd space-time flow data is obtained based on call detail records or geographic location generated based on terminal APP usage; Step 2: the decision node sets the error rate threshold, initializes the iteration weight, calculates one or more first crowd space-time flows for one or more base stations in the coverage area according to the second crowd space-time flow data, until the error rate threshold is met or the iteration number is reached; The decision node is a base station or a network centralized scheduling control unit, which is used to provide resource allocation strategy and active / inactive switching scheme for the base station; Step 3: the decision node determines the network resource demand strategy of one or more base stations in the coverage area based on one or more first crowd space-time flows. The first crowd spatiotemporal flow or the second crowd spatiotemporal flow is based on a network geographic grid, the network geographic grid is to map different base stations in the region into an MxJ grid, and the crowd spatiotemporal flow in each sub-grid is a two-dimensional matrix (1) wherein, denotes the spatio-temporal flow of people in the (m, j) sub-grid at time point i in a day, which is a LxK matrix; L and K represent the division of the sub-grid, corresponding to the accuracy of geographical location; The element contained in the matrix denotes the number of people at a geographical location in the sub-grid, which is an integer greater than or equal to 0; The sum of all elements in the matrix denotes the number of people in the (m, j) sub-grid at time point i in a day.

Images