CN113260064B - Access method and system of large-scale internet of things equipment - Google Patents
Access method and system of large-scale internet of things equipment Download PDFInfo
- Publication number
- CN113260064B CN113260064B CN202110682512.0A CN202110682512A CN113260064B CN 113260064 B CN113260064 B CN 113260064B CN 202110682512 A CN202110682512 A CN 202110682512A CN 113260064 B CN113260064 B CN 113260064B
- Authority
- CN
- China
- Prior art keywords
- internet
- things equipment
- time
- uploading
- things
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/50—Allocation or scheduling criteria for wireless resources
- H04W72/53—Allocation or scheduling criteria for wireless resources based on regulatory allocation policies
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y30/00—IoT infrastructure
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W48/00—Access restriction; Network selection; Access point selection
- H04W48/18—Selecting a network or a communication service
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Abstract
The invention discloses an access method of large-scale internet of things equipment, which comprises the following steps: generating a prediction model based on the behavior pattern of the uploading time point driven by the event of the Internet of things equipment; pre-judging the uploading time point of the Internet of things equipment to be accessed according to the prediction model to generate a pre-uploading time point of the Internet of things equipment; according to the pre-uploading time point of the internet of things equipment and the network service quality requirement of the internet of things equipment, uplink wireless resources are distributed to the internet of things equipment, wherein the wireless resources at least comprise one or more of wireless bearing scheduling, wireless bandwidth management and wireless power control, so that the wireless resources can be effectively distributed, the wireless resource configuration time is reduced, the network performance is improved, and network congestion, delay, resource waste and the like are avoided.
Description
Technical Field
The invention relates to the field of wireless communication, in particular to an access method and system of large-scale internet of things equipment.
Background
The internet of things is a narrow-band low-power-consumption wide area network technology for communication by using a mobile base station, uses an authorized frequency band, can be directly deployed on the existing mobile communication network, and needs an operator to provide services and support. The Internet of things is mainly used for Internet of things equipment with low power consumption, wide area range, small transmission data volume and low updating frequency, and is suitable for various outdoor scenes such as street lamps, parking, shared bicycles, logistics containers, air monitors and the like.
When the wireless access network is connected with the internet of things device, a traditional random access process is often adopted, that is, only high-speed downlink service is provided, and an event-driven uploading mode of the internet of things device reports data at a specific time interval, which is different from periodic reporting, and the event-driven uploading time point of the internet of things device is not predetermined, so that a base station of the wireless access network cannot effectively allocate resources. Particularly, when large-scale internet of things equipment is connected with a wireless access network, a base station corresponding to the wireless access network can cause network congestion and packet loss when handling short data packets of thousands of internet of things equipment and uplink data at the same time, the overall performance of the wireless access network is affected, and the requirements of the internet of things network on low delay, high reliability, high density and the like of uplink required by 6G scene indexes cannot be met.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide an access method and system for large-scale internet of things devices, which can effectively allocate radio resources, reduce radio resource allocation time and improve network performance, and avoid network congestion, delay, resource waste, etc.
In order to solve the above technical problem, a first aspect of the present invention discloses an access method for a large-scale internet of things device, where the method includes: generating a prediction model based on the behavior pattern of the uploading time point driven by the event of the Internet of things equipment; pre-judging the uploading time point of the Internet of things equipment to be accessed according to the prediction model to generate a pre-uploading time point of the Internet of things equipment; and allocating uplink wireless resources to the Internet of things equipment according to the pre-uploading time point of the Internet of things equipment and the network service quality requirement of the Internet of things equipment, wherein the wireless resources at least comprise one or more of wireless bearer scheduling, wireless bandwidth management and wireless power control.
In some embodiments, the generating a prediction model based on behavior patterns of event-driven upload time points of the internet of things device comprises: configuring the time lines of the Internet of things equipment into time sequences with the same length; recording uploading time points of the Internet of things equipment driven by the Internet of things equipment events of the time sequence, and generating a two-dimensional matrix with time sequence length and time sequence number; taking the time point beyond the time sequence driven by the event of the Internet of things equipment as a data set label for judging whether the data uploading occurs to the Internet of things equipment or not; and inputting the two-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model.
In some embodiments, the generating a two-dimensional matrix having a time series length and a time series number is followed by: acquiring a plurality of two-dimensional matrixes to generate a three-dimensional matrix with time sequence length, time sequence number and Internet of things equipment number; and inputting the three-dimensional matrix and the data set label into a long-term and short-term memory network model for training to generate a prediction model.
In some embodiments, the method further comprises modifying, by a Bayesian optimizer, length values of the time series.
In some embodiments, the method further comprises: evaluating the accuracy value of the prediction model by a Boolean objective function.
According to a second aspect of the present invention, there is provided an access system for a large-scale internet of things device, the system comprising: the prediction model is used for generating a prediction model based on a behavior mode of an uploading time point driven by an event of the Internet of things equipment; the prediction module is used for pre-judging the uploading time point of the Internet of things equipment to be accessed according to the prediction model to generate a pre-uploading time point of the Internet of things equipment; and the resource allocation module is used for allocating uplink wireless resources to the internet of things equipment according to the pre-uploading time point of the internet of things equipment and the network service quality requirement of the internet of things equipment, wherein the wireless resources at least comprise one or more of wireless bearer scheduling, wireless bandwidth management and wireless power control.
In some embodiments, a predictive model, comprises: the configuration unit is used for configuring the time lines of the Internet of things equipment into time sequences with the same length; the first parameter unit is used for recording uploading time points of the Internet of things equipment in the event driving of the Internet of things equipment of the time sequence and generating a two-dimensional matrix with time sequence length and time sequence number; the second parameter unit is used for taking the time point which is driven by the event of the Internet of things and exceeds the time sequence as a data set label for judging whether the data uploading occurs to the Internet of things equipment or not; and the training unit is used for inputting the two-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model.
In some embodiments, the predictive model further comprises: the third parameter unit is used for acquiring a plurality of two-dimensional matrixes to generate a three-dimensional matrix with time sequence length, time sequence number and Internet of things equipment number; and the training unit is also used for inputting the three-dimensional matrix and the data set label into a long-term and short-term memory network model for training to generate a prediction model.
In some embodiments, the system further comprises: and the optimization module is used for modifying the length value of the time sequence through a Bayesian optimizer.
In some embodiments, the system further comprises: and the evaluation module is used for evaluating the precision value of the prediction model through a Boolean objective function.
Compared with the prior art, the invention has the beneficial effects that:
by implementing the method, the internal behavior pattern of the uploading time point between the event-driven internet-of-things devices can be mined by using a deep learning method, and the uploading time point of the internet-of-things devices is pre-judged, so that wireless resources (wireless bearing scheduling, bandwidth management, power control and the like) are effectively allocated, the configuration time of the wireless resources is further reduced, the network performance is improved, network congestion, delay, resource waste and the like are avoided, and the low-delay, high-reliability and high-density indexes required by service scenes such as 6G vehicle networking, industrial manufacturing 5.0, telemedicine and the like are achieved.
Drawings
Fig. 1 is a schematic flow chart of a method for accessing a large-scale internet of things device according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a deep learning model according to an embodiment of the present invention;
fig. 3 is a schematic view of an access system of a large-scale internet of things device disclosed in the embodiment of the present invention;
fig. 4 is a schematic structural diagram of an access interaction device of a large-scale internet of things device according to an embodiment of the present invention.
Detailed Description
For better understanding and implementation, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "comprises," "comprising," and any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or modules is not necessarily limited to those steps or modules explicitly listed, but may include other steps or modules not expressly listed or inherent to such process, method, article, or apparatus.
The embodiment of the invention discloses an access method and an access system of large-scale internet of things equipment, which can use a deep learning method to excavate an internal behavior pattern of an uploading time point between the event-driven internet of things equipment and pre-judge the uploading time point of the internet of things equipment, thereby effectively allocating wireless resources (wireless bearing scheduling, bandwidth management, power control and the like), further reducing the configuration time of the wireless resources and improving the network performance, and avoiding network congestion, delay, resource waste and the like, so that low-delay, high-reliability and high-density indexes required by service scenes of 6G internet of vehicles, 5.0 in industrial manufacturing, telemedicine and the like are achieved.
Example one
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating an access method of a large-scale internet of things device according to an embodiment of the present invention. The access method of the large-scale internet of things device can be applied to a wireless access network system and used for accessing large-scale and large-batch internet of things devices or other devices, and the embodiment of the invention of the application system is not limited. As shown in fig. 1, the access method of the large-scale internet of things device may include the following operations:
101. and generating a prediction model based on the behavior pattern of the event-driven uploading time point of the Internet of things equipment.
Because the event-driven uploading mode of the internet of things device is that the internet of things device reports data at a specific time interval, unlike periodic reporting, the event-driven reporting mode has no predetermined uploading time point, which may result in that a wireless base station accessing the internet of things device cannot effectively allocate resources in advance. And especially, the internet of things equipment in the URLLC (Ultra Reliable Low Latency Communication) scene needs a wireless large bandwidth to meet the requirements of Ultra Low Latency and Ultra high reliability, illustratively, remote driving needs an uplink rate of >2.5Mbps, a delay of <5ms, and a reliability of >99.999%, and an intelligent transportation system needs an uplink rate of >1.1Mbps, a delay of <10ms, and a reliability of > 99.999%. Access to a large number of URLLC internet of things devices presents challenges to radio resource management configurations.
Furthermore, each base station in the wireless network needs to be connected into thousands of internet of things devices, the uplink time rule of each device is not fixed, but the inventor discovers that certain time correlation exists between the devices, and discovers that the time pattern uploaded between the internet of things devices can be effectively pre-configured for wireless resource management according to the quality requirement of the internet of things devices through multiple analyses of the inventor, and the existing wireless network is generally predicted by using a fixed algorithm, but the base station computing resources are exhausted in long-term operation, so that the inventor chooses to adopt a statistical model such as a neural network to be more suitable for reducing computing power consumption, increasing the utilization rate of network resources, improving the uplink rate and reducing delay. Illustratively, the Long Short Term Memory (Long Short Term Memory) deep learning network is selected to be more appropriate and more accurate in time linear series prediction. The improvement of the performance of the wireless access side infrastructure enables the deep learning model training process to be more efficient and reliable. Therefore, in the present embodiment, a prediction model is mainly constructed, and the specific implementation is as follows: first, the configuration tag of the equipment of the physical association accessed by each base station may be marked as (A, B, C, D, E … … a1, B1, … … AN, BN), for example, and the time line corresponding to each equipment of the physical association is marked as a time sequence of the same length, for example, (1, 2, 3, … … 20). As a preferred embodiment, the length of the time sequence may be initialized according to the duration of the data uploaded to the corresponding base station by the internet of things device, so that the trained LSTM model can perform parameter adjustment on the length of the time sequence in the training process, so as to perform more effective prediction.
After that, the time points of the event-driven uploading of the internet of things devices in the time series are recorded, and for example, when the a internet of things devices respectively upload data at the time points 1, 3, 6, 9, and 12, the notation a20= {10100100100100000000} (0 means no uploading, and 1 means uploading). If the next time point, for example, 21, exceeds the time point of the time series, the next time point is used as a data set tag for determining whether the data upload occurs in the internet of things device, and the upload activity recorded in the time line by the remaining internet of things devices B, C, D … … in the same way may be implemented in the form of a table stored below and including the time series and the data set tag.
And importing the two-dimensional matrix of each time line of the Internet of things equipment into a deep learning model LSTM as input data, and further carrying out statistical mining on a data uploading correlation mode between the base station connection equipment. As an example, the deep learning model can be implemented as a framework as shown in fig. 2, and is essentially a specific form of Recurrent Neural Network (RNN) due to the Long Short Term Memory (LSTM) model. The LSTM model solves the problem of RNN short-term memory by adding thresholds (Gates) on the basis of the RNN model, so that the recurrent neural network can really and effectively utilize long-distance time sequence information. The LSTM adds 3 logic control units, namely an Input Gate (Input Gate), an Output Gate (Output Gate), and a forgetting Gate (Forget Gate), to the basic structure of RNN, and each logic control unit is connected to a multiplication element, and controls the Input and Output of information streams and the state of a cell (Memory cell) by setting a weight value at the edge where a Memory unit of a neural network is connected to other parts.
As a preferred embodiment, after generating the two-dimensional matrix having the time-series length and the time-series number, the method includes: and acquiring a plurality of two-dimensional matrixes to generate a three-dimensional matrix with time sequence length, time sequence number and Internet of things equipment number, and inputting the three-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model. Since a plurality of the internet-of-things devices simultaneously input a two-dimensional matrix, the training data set becomes a three-dimensional matrix, wherein the X, Y, Z axes representing the three-dimensional matrix can be respectively expressed as the time series length, the time series number and the internet-of-things device number. The three-dimensional matrix is input into the deep learning model as the two-dimensional matrix so as to generate a prediction model.
As a preferred embodiment, the length of the time sequence is subjected to a hyper-parameter adjustment process, and a bayesian optimizer is used to modify the length value step by step, so as to optimize the reconstructed time sequence length.
As a preferred embodiment, the accuracy value of the prediction model is evaluated by means of a boolean objective function. After the Boolean objective function (1 is correct, 0 is wrong) is evaluated, the final prediction accuracy of the uploading time behavior of the internet of things equipment is counted, so that the LSTM deep learning model achieves the optimal solution after the super-parameter adjustment, and the optimal prediction accuracy is obtained.
102. And pre-judging the uploading time point of the Internet of things equipment to be accessed according to the prediction model to generate the pre-uploading time point of the Internet of things equipment.
For the trained prediction model, the correlation of the intrinsic behavior of the physical device is already included, the prediction time point included in the label of the data set can be directly used for judgment, the trained LSTM can perform correlation analysis on all the equipment connected with the base station, illustratively, the probability of uploading of the equipment B after 3 time points after the equipment a uploads is about 0.8, and in order to adapt to the behavior structure mode, the multi-level entropy expression, the fully random value (equipment C), the partial correlation determination value (equipment F and equipment B2) and the fully determination value (T) can be used.
103. And allocating uplink wireless resources to the Internet of things equipment according to the pre-uploading time point of the Internet of things equipment and the network service quality requirement of the Internet of things equipment, wherein the wireless resources at least comprise one or more of wireless bearer scheduling, wireless bandwidth management and wireless power control.
Specifically, the network service quality requirement for the physical device may be determined by an existing network service requirement architecture, and for different device requirements, different radio configurations are selected, and the information flow of the user is classified, so that in the network service quality requirement architecture, the radio configuration required by the internet of things device may be matched, and uplink radio resources are allocated to the radio configuration.
According to the method provided by the embodiment, the deep learning method can be used for mining the internal behavior pattern of the uploading time point between the things and the internet of things driven by events, and the uploading time point of the things and the internet of things is pre-judged, so that the wireless resources (wireless bearing scheduling, bandwidth management, power control and the like) are effectively allocated, the wireless resource configuration time is further reduced, the network performance is improved, network congestion, delay, resource waste and the like are avoided, and the low-delay, high-reliability and high-density indexes required by service scenes of 6G vehicle networking, industrial manufacturing 5.0, telemedicine and the like are achieved.
Example two
Referring to fig. 3, fig. 3 is a schematic diagram of an access system of a large-scale internet of things device according to an embodiment of the present invention. As shown in fig. 3, the access system of the large-scale internet of things device includes:
and the prediction model 1 is used for generating a prediction model based on the behavior pattern of the event-driven uploading time point of the Internet of things equipment.
And the prediction module 2 is used for pre-judging the uploading time point of the internet of things equipment to be accessed according to the prediction model to generate the pre-uploading time point of the internet of things equipment.
And the resource allocation module 3 is configured to allocate uplink radio resources to the internet of things device according to a pre-upload time point of the internet of things device and a network service quality requirement of the internet of things device, where the radio resources at least include one or more of radio bearer scheduling, radio bandwidth management, and radio power control.
Specifically, the prediction model 1 includes: the configuration unit 101 is configured to configure the time lines of the equipment of the internet of things into time sequences with the same length. A first parameter unit 102, configured to record event-driven uploading time points of the internet of things device in the time series, and generate a two-dimensional matrix having a time series length and a time series number. And a second parameter unit 103, which takes the time point beyond the time series driven by the event of the equipment in the internet of things as a data set tag for judging whether data uploading occurs to the equipment in the internet of things. And the training unit 104 is used for inputting the two-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model. The concrete implementation is as follows: first, the configuration tag of the equipment of the physical association accessed by each base station may be marked as (A, B, C, D, E … … a1, B1, … … AN, BN), for example, and the time line corresponding to each equipment of the physical association is marked as a time sequence of the same length, for example, (1, 2, 3, … … 20). As a preferred embodiment, the length of the time sequence may be initialized according to the duration of the data uploaded to the corresponding base station by the internet of things device, so that the trained LSTM model can perform parameter adjustment on the length of the time sequence in the training process, so as to perform more effective prediction.
After that, the time points of the event-driven uploading of the internet of things devices in the time series are recorded, and for example, when the a internet of things devices respectively upload data at the time points 1, 3, 6, 9, and 12, the notation a20= {10100100100100000000} (0 means no uploading, and 1 means uploading). If the next time point, for example, 21, exceeds the time point of the time series, the next time point is used as a data set tag for determining whether the data upload occurs in the internet of things device, and the upload activity recorded in the time line by the remaining internet of things devices B, C, D … … in the same way may be implemented in the form of a table stored below and including the time series and the data set tag.
And importing the two-dimensional matrix of each time line of the Internet of things equipment into a deep learning model LSTM as input data, and further carrying out statistical mining on a data uploading correlation mode between the base station connection equipment. As an example, the deep learning model can be implemented as a framework as shown in fig. 2, and is essentially a specific form of Recurrent Neural Network (RNN) due to the Long Short Term Memory (LSTM) model. The LSTM model solves the problem of RNN short-term memory by adding thresholds (Gates) on the basis of the RNN model, so that the recurrent neural network can really and effectively utilize long-distance time sequence information. The LSTM adds 3 logic control units, namely an Input Gate (Input Gate), an Output Gate (Output Gate), and a forgetting Gate (Forget Gate), to the basic structure of RNN, and each logic control unit is connected to a multiplication element, and controls the Input and Output of information streams and the state of a cell (Memory cell) by setting a weight value at the edge where a Memory unit of a neural network is connected to other parts.
As a preferred embodiment, the prediction model 1 further includes: and a third parameter unit 105, configured to acquire a plurality of two-dimensional matrices to generate a three-dimensional matrix having a time series length, a time series number, and an internet of things device number. And the training unit 104 is further used for inputting the three-dimensional matrix and the data set labels into the long-term and short-term memory network model for training to generate a prediction model. The concrete implementation is as follows: and acquiring a plurality of two-dimensional matrixes to generate a three-dimensional matrix with time sequence length, time sequence number and Internet of things equipment number, and inputting the three-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model. Since a plurality of the internet-of-things devices simultaneously input a two-dimensional matrix, the training data set becomes a three-dimensional matrix, wherein the X, Y, Z axes representing the three-dimensional matrix can be respectively expressed as the time series length, the time series number and the internet-of-things device number. The three-dimensional matrix is input into the deep learning model as the two-dimensional matrix so as to generate a prediction model.
As a preferred embodiment, the system further comprises: and the optimization module 4 is used for modifying the length value of the time sequence through a Bayesian optimizer. For the length of the time sequence, the length value can be gradually modified by using a Bayesian optimizer in the process of carrying out hyper-parameter adjustment, so that the reconstructed time sequence length is optimized.
In a preferred embodiment, the system further comprises: and the evaluation module 5 is used for evaluating the precision value of the prediction model through a Boolean objective function. After the Boolean objective function (1 is correct, 0 is wrong) is evaluated, the final prediction accuracy of the uploading time behavior of the internet of things equipment is counted, so that the LSTM deep learning model achieves the optimal solution after the super-parameter adjustment, and the optimal prediction accuracy is obtained.
According to the system provided by the embodiment, the deep learning method can be used for mining the internal behavior pattern of the uploading time point between the things and the internet of things driven by events, and the uploading time point of the things and the internet of things equipment is pre-judged, so that wireless resources (wireless bearing scheduling, bandwidth management, power control and the like) are effectively allocated, the wireless resource configuration time is further reduced, the network performance is improved, network congestion, delay, resource waste and the like are avoided, and the low-delay, high-reliability and high-density indexes required by service scenes of 6G vehicle networking, industrial manufacturing 5.0, telemedicine and the like are achieved.
EXAMPLE III
Referring to fig. 4, fig. 4 is a schematic structural diagram of an interaction device for accessing a large-scale internet of things device according to an embodiment of the present invention. The accessing interactive device of the large-scale internet of things device described in fig. 4 may be applied to a wireless access system, and the embodiment of the present invention is not limited to the application system of the accessing interactive device of the large-scale internet of things device. As shown in fig. 4, the apparatus may include:
a memory 601 in which executable program code is stored;
a processor 602 coupled to a memory 601;
the processor 602 calls the executable program code stored in the memory 601 for executing the access method of the large-scale internet of things device described in the first embodiment.
Example four
The embodiment of the invention discloses a computer-readable storage medium which stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute the access method of the large-scale internet of things device described in the first embodiment.
EXAMPLE five
An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute the method for accessing a large-scale internet of things device described in the first embodiment.
The above-described embodiments are only illustrative, and the modules described as separate components may or may not be physically separate, and the components displayed as modules may or may not be physical modules, may be located in one place, or may be distributed on a plurality of network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above detailed description of the embodiments, those skilled in the art will clearly understand that the embodiments may be implemented by software plus a necessary general hardware platform, and may also be implemented by hardware. Based on such understanding, the above technical solutions may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, where the storage medium includes a Read-Only Memory (ROM), a Random Access Memory (RAM), a Programmable Read-Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), a One-time Programmable Read-Only Memory (OTPROM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), a Compact Disc-Read-Only Memory (CD-ROM), or other disk memories, CD-ROMs, or other magnetic disks, A tape memory, or any other medium readable by a computer that can be used to carry or store data.
Finally, it should be noted that: the method and system for accessing large-scale internet of things disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, and are only used for illustrating the technical solutions of the present invention, rather than limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art; the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. An access method of large-scale internet of things equipment is characterized by comprising the following steps:
generating a prediction model based on the behavior pattern of the uploading time point driven by the event of the Internet of things equipment;
pre-judging the uploading time point of the Internet of things equipment to be accessed according to the prediction model to generate a pre-uploading time point of the Internet of things equipment;
and allocating uplink wireless resources to the Internet of things equipment according to the pre-uploading time point of the Internet of things equipment and the network service quality requirement of the Internet of things equipment, wherein the wireless resources at least comprise one or more of wireless bearer scheduling, wireless bandwidth management and wireless power control.
2. The access method of the large-scale internet of things equipment according to claim 1, wherein the generating of the prediction model based on the behavior pattern of the event-driven uploading time point of the internet of things equipment comprises:
configuring the time lines of the Internet of things equipment into time sequences with the same length;
recording uploading time points of the Internet of things equipment driven by the Internet of things equipment events of the time sequence, and generating a two-dimensional matrix with time sequence length and time sequence number;
taking the time point beyond the time sequence driven by the event of the Internet of things equipment as a data set label for judging whether the data uploading occurs to the Internet of things equipment or not;
and inputting the two-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model.
3. The access method of the large-scale internet of things equipment according to claim 1, wherein the generating of the prediction model based on the behavior pattern of the event-driven uploading time point of the internet of things equipment comprises:
configuring the time lines of the Internet of things equipment into time sequences with the same length;
recording uploading time points of the Internet of things equipment driven by the Internet of things equipment events of the time sequence, and generating a two-dimensional matrix with time sequence length and time sequence number;
taking the time point beyond the time sequence driven by the event of the Internet of things equipment as a data set label for judging whether the data uploading occurs to the Internet of things equipment or not;
acquiring a plurality of two-dimensional matrixes to generate a three-dimensional matrix with time sequence length, time sequence number and Internet of things equipment number;
and inputting the three-dimensional matrix and the data set label into a long-term and short-term memory network model for training to generate a prediction model.
4. The method for accessing the large-scale internet of things equipment according to claim 2 or 3, wherein the method further comprises the following steps:
and modifying the length value of the time sequence by a Bayesian optimizer.
5. The method for accessing the large-scale internet of things equipment according to claim 4, wherein the method further comprises:
evaluating the accuracy value of the prediction model by a Boolean objective function.
6. An access system for large-scale internet of things equipment, the system comprising:
the prediction model is used for generating a prediction model based on a behavior mode of an uploading time point driven by an event of the Internet of things equipment;
the prediction module is used for pre-judging the uploading time point of the Internet of things equipment to be accessed according to the prediction model to generate a pre-uploading time point of the Internet of things equipment;
and the resource allocation module is used for allocating uplink wireless resources to the internet of things equipment according to the pre-uploading time point of the internet of things equipment and the network service quality requirement of the internet of things equipment, wherein the wireless resources at least comprise one or more of wireless bearer scheduling, wireless bandwidth management and wireless power control.
7. The access system of large-scale internet of things equipment according to claim 6, wherein the predictive model comprises:
the configuration unit is used for configuring the time lines of the Internet of things equipment into time sequences with the same length;
the first parameter unit is used for recording uploading time points of the Internet of things equipment in the event driving of the Internet of things equipment of the time sequence and generating a two-dimensional matrix with time sequence length and time sequence number;
the second parameter unit is used for taking the time point which is driven by the event of the Internet of things and exceeds the time sequence as a data set label for judging whether the data uploading occurs to the Internet of things equipment or not;
and the training unit is used for inputting the two-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model.
8. The access system of large-scale internet of things equipment according to claim 6, wherein the predictive model comprises:
the configuration unit is used for configuring the time lines of the Internet of things equipment into time sequences with the same length;
the first parameter unit is used for recording uploading time points of the Internet of things equipment in the event driving of the Internet of things equipment of the time sequence and generating a two-dimensional matrix with time sequence length and time sequence number;
the second parameter unit is used for taking the time point which is driven by the event of the Internet of things and exceeds the time sequence as a data set label for judging whether the data uploading occurs to the Internet of things equipment or not;
the third parameter unit is used for acquiring a plurality of two-dimensional matrixes to generate a three-dimensional matrix with time sequence length, time sequence number and Internet of things equipment number;
and the training unit is used for inputting the three-dimensional matrix and the data set label into a long-short term memory network model for training to generate a prediction model.
9. The access system of large-scale internet of things equipment according to claim 7 or 8, wherein the system further comprises:
and the optimization module is used for modifying the length value of the time sequence through a Bayesian optimizer.
10. The access system of a large-scale internet of things device of claim 9, further comprising:
and the evaluation module is used for evaluating the precision value of the prediction model through a Boolean objective function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110682512.0A CN113260064B (en) | 2021-06-21 | 2021-06-21 | Access method and system of large-scale internet of things equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110682512.0A CN113260064B (en) | 2021-06-21 | 2021-06-21 | Access method and system of large-scale internet of things equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113260064A CN113260064A (en) | 2021-08-13 |
| CN113260064B true CN113260064B (en) | 2021-09-07 |
Family
ID=77188862
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110682512.0A Active CN113260064B (en) | 2021-06-21 | 2021-06-21 | Access method and system of large-scale internet of things equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113260064B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105740967A (en) * | 2016-01-20 | 2016-07-06 | 北京理工大学 | Manufacture cloud service execution time prediction method and manufacture cloud service execution time prediction device |
| WO2018020306A1 (en) * | 2016-07-29 | 2018-02-01 | Telefonaktiebolaget Lm Ericsson (Publ) | System and method for resource-aware and time-critical iot frameworks |
| CN109327910A (en) * | 2017-07-31 | 2019-02-12 | 中兴通讯股份有限公司 | A kind of physical channel resources configuration method and device |
-
2021
- 2021-06-21 CN CN202110682512.0A patent/CN113260064B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105740967A (en) * | 2016-01-20 | 2016-07-06 | 北京理工大学 | Manufacture cloud service execution time prediction method and manufacture cloud service execution time prediction device |
| WO2018020306A1 (en) * | 2016-07-29 | 2018-02-01 | Telefonaktiebolaget Lm Ericsson (Publ) | System and method for resource-aware and time-critical iot frameworks |
| CN109327910A (en) * | 2017-07-31 | 2019-02-12 | 中兴通讯股份有限公司 | A kind of physical channel resources configuration method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113260064A (en) | 2021-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11410046B2 (en) | Learning-based service migration in mobile edge computing | |
| CN110839184B (en) | Method and device for adjusting bandwidth of mobile fronthaul optical network based on flow prediction | |
| CN111132190A (en) | Base station load early warning method and device | |
| US11412317B2 (en) | Method and apparatus for allocating bandwidth based on machine learning in passive optical network | |
| CN110662245B (en) | Base station load early warning method and device based on deep learning | |
| CN111277437A (en) | Network slice resource allocation method for smart power grid | |
| CN104463351A (en) | Communication bandwidth prediction method and device based on power business requirements | |
| Li et al. | Method of resource estimation based on QoS in edge computing | |
| Rodoplu et al. | A multiscale algorithm for joint forecasting–scheduling to solve the massive access problem of IoT | |
| CN113542002B (en) | Slice scheduling method, device, equipment and storage medium of 5G wireless access network | |
| CN111107651B (en) | Method and device for scheduling wireless resources | |
| Chen et al. | Minimizing age-of-information for fog computing-supported vehicular networks with deep Q-learning | |
| CN105472631A (en) | Service data quantity and/or resource data quantity prediction method and prediction system | |
| CN111491312A (en) | Method and equipment for predicting, allocating, acquiring and training neural network of wireless resources | |
| Kim et al. | Mathematical model of operation of a cell of a mobile communication network with adaptive modulation schemes and handover of mobile users | |
| Ganjalizadeh et al. | Saving energy and spectrum in enabling URLLC services: A scalable RL solution | |
| CN114422606A (en) | Federal learning communication overhead compression method, device, equipment and medium | |
| CN113260064B (en) | Access method and system of large-scale internet of things equipment | |
| Avranas et al. | Deep reinforcement learning for resource constrained multiclass scheduling in wireless networks | |
| CN111091138A (en) | Irradiation forecast processing method and stacked generalization model training method and device | |
| CN114339967A (en) | Method and device for predicting base station traffic | |
| Zarrini et al. | Loss and delay analysis of non‐Poisson M2M traffic over LTE networks | |
| CN110996398A (en) | Wireless network resource scheduling method and device | |
| CN112491863B (en) | IP address black and gray list analysis method, server, terminal and storage medium | |
| Wang et al. | Dynamic multichannel access for 5G and beyond with fast time-varying channel |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |