CN115208755A - Internet of things equipment resource-friendly feature extractor deployment method and system - Google Patents

Abstract

The invention provides a resource-friendly feature extractor deployment method and system for Internet of Things equipment, belonging to the technical field of Internet of Things. After the IoT device collects data for preprocessing, select a feature extractor to extract the principal components of the data according to its current available resources feature, and then upload the extracted principal component feature data to the edge server. The invention reduces the storage resource consumption of the Internet of Things equipment and improves the switching efficiency of the feature extractors; the redundant feature extractors are removed, and the non-redundant feature extractors are nested together in the way of parameter sharing, thereby reducing the deployment of multiple feature extractors The storage resources consumed by the processor are reduced; the invalid feature extractor switching is reduced, and the switching efficiency of the feature extractor is improved.

Description

Resource-friendly feature extractor deployment method and system for IoT devices

technical field

The invention relates to the technical field of the Internet of Things, in particular to a resource-friendly feature extractor deployment method and system for the Internet of Things device.

Background technique

With the advent of the Internet of Everything era, hundreds of millions of IoT devices will be connected to the network and will generate massive amounts of data. Due to the limited resources of IoT devices, it is difficult to provide sufficient computing resources to complete all data processing tasks. Therefore, it has become mainstream to deploy feature extractors on IoT devices to first extract the features of the collected data, and then upload the extracted feature data to edge servers/cloud servers for further processing.

IoT devices have the characteristics of dynamically changing available resources. In order to ensure that IoT devices can provide uninterrupted feature extraction services when their available resources are dynamically changing, it is necessary to deploy multiple feature extractors with different capacities on IoT devices. However, the resources of IoT devices are limited, and deploying multiple feature extractors will occupy a large amount of storage resources of IoT devices.

Uploading massive amounts of data generated by IoT devices to edge/cloud servers will not only lead to greater pressure on core network bandwidth, but also make it difficult to meet users' needs for low-latency, high-performance application services. In order to reduce the amount of data uploaded by IoT devices to edge/cloud servers, and considering that IoT devices have dynamically changing and limited available resources, it has become mainstream to deploy multiple feature extractors with different capacities on IoT devices simultaneously. Existing methods propose to deploy multiple feature extractors on IoT devices, and propose to store multiple feature extractors with different capacities in a parameter-sharing manner. However, this type of method ignores that the feature extractor deployed on the IoT device contains a large number of redundant sub-feature extractors, which leads to the waste of the storage resources of the IoT device.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a resource-friendly feature extractor deployment method and system for IoT devices that avoids the waste of IoT device storage resources by removing redundant sub-feature extractors, so as to solve at least the problems in the above background technology. a technical issue.

In order to achieve the above object, the present invention has adopted the following technical solutions:

In one aspect, the present invention provides a resource-friendly feature extractor deployment method for IoT devices, including:

Call the stored dataset information and feature extractor generation algorithm to generate a feature extractor;

Using a single-capacity non-redundant multi-functional feature extractor on IoT devices to split the feature extractor into multiple non-redundant sub-feature extractors, and use these non-redundant sub-feature extractors to generate a single Capacity non-redundant multifunctional feature extractor;

Send the generated single-capacity multi-function feature extractor to the IoT device; when the IoT device collects data for preprocessing, select a feature extractor to extract the principal component features of the data according to its current available resources, and then extract the principal component features of the data. The ingredient feature data is uploaded to the edge server.

In a second aspect, the present invention provides a resource-friendly feature extractor deployment system for IoT devices, including:

The retrieval module is used to retrieve the stored data set information and the feature extractor generation algorithm to generate the feature extractor;

A generation module for splitting a feature extractor into multiple non-redundant sub-feature extractors with a single-capacity non-redundant multi-function feature extractor on IoT devices, and dividing these non-redundant sub-feature extractors with parameters Generate a single-capacity non-redundant multi-functional feature extractor in a shared way;

The extraction module is used to send the generated single-capacity multi-function feature extractor to the IoT device; when the IoT device collects data for data preprocessing, select a feature extractor to extract the principal component features of the data according to its current available resources, Then upload the extracted principal component feature data to the edge server.

In a third aspect, the present invention provides a non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, realize the above-mentioned orientation A resource-friendly feature extractor deployment method for IoT devices.

In a fourth aspect, the present invention provides a computer program product comprising a computer program for implementing the resource-friendly feature extractor for IoT devices as described above when run on one or more processors Deployment method.

In a fifth aspect, the present invention provides an electronic device, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device runs, the processor executes all The computer program stored in the memory to cause the electronic device to execute the instructions for implementing the resource-friendly feature extractor deployment method for IoT devices as described above.

The beneficial effects of the invention are as follows: the consumption of the storage resources of the Internet of Things equipment is reduced, and the switching efficiency of the feature extractor is improved; by analyzing the properties of the feature extractor and the relationship between the feature extractors of different capacities, the redundant feature extractors are removed, and the parameter The shared method nests non-redundant feature extractors together, reducing the storage resources consumed by deploying multiple feature extractors; removing redundant feature extractors, reducing a large number of invalid feature extractor switching, and improving feature extraction Switching efficiency.

Advantages of additional aspects of the present invention will be more apparent in the following description, or will be learned by practice of the present invention.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained from these drawings without any creative effort.

FIG. 1 is a schematic diagram of a resource-friendly feature extraction framework for IoT devices according to an embodiment of the present invention.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below through the accompanying drawings are exemplary and are only used to explain the present invention, but not to be construed as a limitation of the present invention.

It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

It should also be understood that terms such as those defined in the general dictionary should be understood to have meanings consistent with their meanings in the context of the prior art and, unless defined as herein, are not to be taken in an idealized or overly formal sense. explain.

It will be understood by those skilled in the art that the singular forms "a", "an", "the" and "the" as used herein can include the plural forms as well, unless expressly stated otherwise. It should be further understood that the word "comprising" used in the description of the present invention refers to the presence of stated features, integers, steps, operations, elements and/or components, but does not exclude the presence or addition of one or more other features, Integers, steps, operations, elements and/or groups thereof.

In the description of this specification, description with reference to the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples", etc., mean specific features described in connection with the embodiment or example , structure, material or feature is included in at least one embodiment or example of the present invention. Furthermore, the particular features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, those skilled in the art may combine and combine the different embodiments or examples described in this specification, as well as the features of the different embodiments or examples, without conflicting each other.

In order to facilitate the understanding of the present invention, the present invention will be further explained and described below with reference to the accompanying drawings with specific embodiments, and the specific embodiments do not constitute limitations to the embodiments of the present invention.

Those skilled in the art should understand that the accompanying drawings are only schematic diagrams of the embodiments, and the components in the accompanying drawings are not necessarily necessary to implement the present invention.

Example 1

This embodiment 1 provides a resource-friendly feature extractor deployment system for IoT devices, including:

The retrieval module is used to retrieve the stored data set information and the feature extractor generation algorithm to generate the feature extractor;

A generation module for splitting a feature extractor into multiple non-redundant sub-feature extractors with a single-capacity non-redundant multi-function feature extractor on IoT devices, and dividing these non-redundant sub-feature extractors with parameters Generate a single-capacity non-redundant multi-functional feature extractor in a shared way;

The extraction module is used to send the generated single-capacity multi-function feature extractor to the IoT device; when the IoT device collects data for data preprocessing, select a feature extractor to extract the principal component features of the data according to its current available resources, Then upload the extracted principal component feature data to the edge server.

In this embodiment 1, using the above system, a resource-friendly feature extractor deployment method for IoT devices is implemented, including:

Call the stored dataset information and feature extractor generation algorithm to generate a feature extractor;

Using a single-capacity non-redundant multi-functional feature extractor on IoT devices to split the feature extractor into multiple non-redundant sub-feature extractors, and use these non-redundant sub-feature extractors to generate a single Capacity non-redundant multifunctional feature extractor;

Send the generated single-capacity multi-function feature extractor to the IoT device; when the IoT device collects data for preprocessing, select a feature extractor to extract the principal component features of the data according to its current available resources, and then extract the principal component features of the data. The ingredient feature data is uploaded to the edge server.

Example 2

In this embodiment 2, a resource-friendly feature extractor deployment method is provided, which solves the problem that storing multiple feature extractors with different capacities in the prior art occupies a large amount of storage resources of IoT devices. As shown in FIG. 1 , the application process of the method proposed in Embodiment 2 is as follows: the edge server first uses a feature extractor generation algorithm, such as a principal component feature extractor generation algorithm (Principal Component Analysis, PCA) and data set information to generate a feature extractor E , and then run the single-capacity non-redundant multi-function feature extractor generation algorithm to generate E _multi . The edge server then sends the generated E _multi to the IoT device. After the IoT device collects the data, it first preprocesses the data, then selects a feature extractor to extract the principal component features of the data according to its current available resources, and then uploads the extracted principal component feature data to the edge server for further processing. The method described in this embodiment can be divided into an offline stage and an online stage.

Offline stage:

First, the feature extractor E is generated by using the data set information stored in the edge server and the feature extractor generation algorithm (PCA algorithm). Then, in order to adapt to the characteristics of the dynamic change of the available resources of the IoT device, the present invention uses a single-capacity non-redundant multi-functional feature extractor on the IoT device to split the feature extractor E into multiple non-redundant sub-feature extractors, And these non-redundant sub-feature extractors are used to generate a single-capacity non-redundant multi-functional feature extractor E _multi in the way of parameter sharing. Among them, see Definition 1 for the definition of redundant sub-feature extractor.

Definition 1 (redundant sub-feature extractors): For two sub-feature extractors S _Ei and S _Ej , where S _Ei ={E ₁ ,E ₂ ,...,E _i },S _Ej ={E ₁ ,E ₂ ,...,E _j }, i≠j, if the sub-feature extractors S _Ei and S _Ej have the same extraction precision, then S _Ei or S _Ej is redundant.

According to Definition 1, it can be determined in this embodiment which sub-feature extractors are redundant.

In addition, a method for removing redundant sub-feature extractors is also defined in this embodiment, see Definition 2.

Definition 2 (removal of feature extractors): For two sub-feature extractors S _Ei and S _Ej , where S _Ei ={E ₁ ,E ₂ ,...,E _i },S _Ej ={E ₁ ,E ₂ ,...,E _j }, if i<j, then remove the sub-feature extractor S _Ej .

To this end, redundant sub-feature extractors are removed according to Definition 2, and then a single-capacity multifunctional feature extractor E _multi is generated in a parameter-sharing manner. Finally, the edge server sends the E _multi to the IoT device and deploys it.

Online stage:

After the IoT device collects data, it first preprocesses the data, such as grayscale and dimension alignment. Then, the IoT device selects an appropriate sub-feature extractor according to its current available resources (CPU resources, storage resources or network bandwidth resources) to extract the principal component features of the preprocessed data. Finally, the IoT device sends the feature data of the extractor to the edge server.

To sum up, the method described in Embodiment 2 removes redundant sub-feature extractors, and deploys multiple non-redundant feature extractors on terminal devices in a parameter sharing manner to avoid waste of IoT device storage resources. By analyzing the composition of feature extractors, the method of defining and removing redundant feature extractors reduces the storage resources occupied by multiple feature extractors and improves the efficiency of switching between feature extractors. By nesting multiple non-redundant feature extractors in a parameter sharing manner, the consumption of storage resources of the terminal device is effectively reduced.

Example 3

Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, where the non-transitory computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, resource-friendly IoT devices are realized. feature extractor deployment method, which includes:

Call the stored dataset information and feature extractor generation algorithm to generate a feature extractor;

Using a single-capacity non-redundant multi-functional feature extractor on IoT devices to split the feature extractor into multiple non-redundant sub-feature extractors, and use these non-redundant sub-feature extractors to generate a single Capacity non-redundant multifunctional feature extractor;

Send the generated single-capacity multi-function feature extractor to the IoT device; when the IoT device collects data for preprocessing, select a feature extractor to extract the principal component features of the data according to its current available resources, and then extract the principal component features of the data. The ingredient feature data is uploaded to the edge server.

Example 4

Embodiment 4 of the present invention provides a computer program (product), including a computer program that, when running on one or more processors, is used to implement a resource-friendly feature extractor deployment method for IoT devices, The method includes:

Call the stored dataset information and feature extractor generation algorithm to generate a feature extractor;

Using a single-capacity non-redundant multi-functional feature extractor on IoT devices to split the feature extractor into multiple non-redundant sub-feature extractors, and use these non-redundant sub-feature extractors to generate a single Capacity non-redundant multifunctional feature extractor;

Send the generated single-capacity multi-function feature extractor to the IoT device; when the IoT device collects data for preprocessing, select a feature extractor to extract the principal component features of the data according to its current available resources, and then extract the principal component features of the data. The ingredient feature data is uploaded to the edge server.

Example 5

Embodiment 5 of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device runs, the processor executes the A computer program stored in a memory to cause an electronic device to execute instructions implementing a resource-friendly feature extractor deployment method for IoT devices, the method comprising:

Call the stored dataset information and feature extractor generation algorithm to generate a feature extractor;

Using a single-capacity non-redundant multi-functional feature extractor on IoT devices to split the feature extractor into multiple non-redundant sub-feature extractors, and use these non-redundant sub-feature extractors to generate a single Capacity non-redundant multifunctional feature extractor;

Send the generated single-capacity multi-function feature extractor to the IoT device; when the IoT device collects data for preprocessing, select a feature extractor to extract the principal component features of the data according to its current available resources, and then extract the principal component features of the data. The ingredient feature data is uploaded to the edge server.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.

These computer program instructions can also be loaded on a computer or other programmable data processing apparatus, where a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions for execution on the computer or other programmable apparatus Steps are provided for implementing the functions specified in a flow or flows of the flowcharts and/or a block or blocks of the block diagrams.

Although the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, they do not limit the scope of protection of the present invention. Those skilled in the art should understand that on the basis of the technical solutions disclosed in the present invention, those skilled in the art do not need to pay Various modifications or deformations that can be made by creative work shall be covered within the protection scope of the present invention.

Claims (5)

1. A feature extractor deployment method for resource-friendly equipment of the Internet of things is characterized by comprising the following steps:

calling stored data set information and a feature extractor generation algorithm to generate a feature extractor;

the method comprises the steps that a single-capacity non-redundant multifunctional feature extractor is utilized on the Internet of things equipment to split the feature extractor into a plurality of non-redundant sub-feature extractors, and the non-redundant sub-feature extractors are used for generating the single-capacity non-redundant multifunctional feature extractor in a parameter sharing mode;

sending the generated single-capacity multifunctional feature extractor to the Internet of things equipment; when the data collected by the Internet of things equipment is preprocessed, a feature extractor is selected according to the current available resources to extract the principal component features of the data, and then the extracted principal component feature data is uploaded to an edge server.

2. A resource-friendly feature extractor deployment system for equipment of the Internet of things is characterized by comprising:

the calling module is used for calling the stored data set information and generating a feature extractor by a feature extractor generating algorithm;

the generating module is used for splitting the feature extractor into a plurality of non-redundant sub-feature extractors by using the single-capacity non-redundant multifunctional feature extractor on the Internet of things equipment, and generating the single-capacity non-redundant multifunctional feature extractor by using the non-redundant sub-feature extractors in a parameter sharing mode;

the extraction module is used for sending the generated single-capacity multifunctional feature extractor to the Internet of things equipment; when the internet of things equipment collects data, preprocessing the data, selecting a feature extractor to extract principal component features of the data according to current available resources of the internet of things equipment, and uploading the extracted principal component feature data to an edge server.

3. A non-transitory computer-readable storage medium storing computer instructions which, when executed by a processor, implement the internet of things device resource-friendly feature extractor deployment method as recited in claim 1.

4. A computer program product, comprising a computer program which, when run on one or more processors, is configured to implement the internet of things device resource-friendly feature extractor deployment method of claim 1.

5. An electronic device, comprising: a processor, a memory, and a computer program; wherein a processor is connected to the memory, and a computer program is stored in the memory, and when the electronic device runs, the processor executes the computer program stored in the memory to make the electronic device execute the instructions to implement the internet-of-things device resource-friendly feature extractor deployment method as claimed in claim 1.

Images