A power internet of things energy-saving communication method for mass devices

Abstract

The application discloses a kind of power internet of things energy-saving communication methods for mass equipment, comprising the following steps: constructing power internet of things based on CF-NOMA edge computing;According to the signal-to-interference noise ratio model of device end, the device end data rate model is constructed;According to the device end local task processing energy consumption model and the total power consumption model of unloading task energy consumption model, the total power consumption model of system is constructed;Through the data rate model of all devices in the system and the total power consumption model of system, the objective function is constructed, and the constraint condition is constructed by combining the SIC decoding of CF-NOMA and the task queuing delay;With the maximum of objective function as optimization goal, optimize wireless resource block allocation vector, task segmentation vector, resource allocation vector, device end transmission beamforming vector, binary SIC decoding index, so that the system energy efficiency of power internet of things reaches maximum value.The application not only has wide application range, but also effectively reduces interference, reduces energy loss, and strengthens communication quality.

Description

Technical Field

[0001] This invention relates to the field of power Internet of Things (IoT) technology, and in particular to a power IoT energy-saving communication method for a large number of devices. Background Technology

[0002] The power Internet of Things (IoT) is the application of IoT in smart grids, primarily improving the informatization level of the power system by combining communication infrastructure resources with power system infrastructure resources. With the explosive growth in the number of smart grid terminal devices and the gradual expansion of the power grid, the power IoT faces significant challenges, such as information security, big data analytics, and communication quality. Because the power IoT needs to store and process massive amounts of data, its main challenge is improving network edge computing capabilities. Furthermore, the increase in application devices easily leads to communication security issues. In addition, the power IoT requires improved network resource scheduling capabilities to better achieve ubiquitous communication.

[0003] To address the challenge of massive device access, Non-orthogonal Multiple-access (NOMA) technology has emerged. Through power reuse or signature design, it allows any number of different users to occupy the same resources, such as spectrum, time, and space. Existing multi-antenna NOMA methods are mainly divided into beamformer-based NOMA (BB-NOMA) and cluster-based NOMA (CB-NOMA), employing different successive interference cancellation (SIC) strategies and beamforming designs. Specifically, BB-NOMA assigns a dedicated beamforming vector to each user and then performs SIC according to a specific decoding order to suppress interference in the remaining space. However, it suffers from drawbacks such as high SIC decoding complexity and potential impact on system performance. Compared to BB-NOMA, CB-NOMA divides users into multiple clusters based on their spatial channel correlation. By providing the same beamforming vector to each cluster and performing SIC sequentially within each cluster, it reduces system complexity and alleviates the overuse problem of SIC. However, due to the randomness of the channel, CB-NOMA has several unusable scenarios. As can be seen from the above, both beamforming-based NOMA and cluster-based NOMA are scenario-centric, and their effectiveness depends on the specific scenario, failing to adapt to the challenges of all scenarios.

[0004] In recent years, edge computing technology has emerged as a promising solution to the processing challenges of various service tasks in power IoT devices. By deploying edge servers at the network edge, redundant tasks are offloaded to these servers, reducing processing time and saving energy. Currently, research in the power IoT primarily focuses on BB-NOMA or CB-NOMA edge computing, but these approaches lack universal applicability. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings and problems of the narrow applicability of the existing technology and to provide a power Internet of Things energy-saving communication method with a wide range of applications for massive devices.

[0006] To achieve the above objectives, the technical solution of the present invention is: a power Internet of Things energy-saving communication method for massive numbers of devices, the method comprising the following steps:

[0007] S1. Construct a power Internet of Things system based on CF-NOMA edge computing;

[0008] S2. First, construct the device-side received signal model, then construct the device-side signal-to-interference-to-noise ratio model based on the device-side received signal model, and finally construct the device-side data rate model based on the device-side signal-to-interference-to-noise ratio model.

[0009] S3. First, construct the device-side local processing task data volume and unloading task data volume model based on the task segmentation model. Then, construct the device-side local task processing energy consumption model based on the device-side local processing task data volume model. Construct the device-side unloading task energy consumption model based on the device-side unloading task data volume model. Finally, construct the system's total power consumption model based on the device-side local task processing energy consumption model and the unloading task energy consumption model.

[0010] S4. Construct an objective function using the data rate model of all devices in the system and the total power consumption model of the system. Combine this with the SIC decoding of CF-NOMA and the task queuing delay to construct constraints. Under the constraints, optimize the wireless resource block allocation vector, task segmentation vector, resource allocation vector, device-side transmission beamforming vector, and binary SIC decoding index with the objective function maximization as the optimization goal, so as to maximize the system energy efficiency of the power Internet of Things and obtain the optimized wireless resource block allocation vector, task segmentation vector, resource allocation vector, device-side transmission beamforming vector, and binary SIC decoding index.

[0011] In step S1, the power IoT system based on CF-NOMA edge computing includes one antenna base station, one edge server, and... A single-antenna power IoT device;

[0012] The antenna base station and the edge server are located in the same location, and the antenna base station is respectively connected to... Each single-antenna power IoT device is connected wirelessly in sequence;

[0013] The power Internet of Things system adopts a discrete time-slot model, which divides the total optimization period into... Each time slot, the duration of each time slot Equal, continuous Individual time slots combine to form a period. , .

[0014] In step S2, the signal received by the device is modeled as follows:

[0015]

[0016] In the formula, For equipment Received signals; , The number of power Internet of Things (IoT) devices; , The number of wireless resource blocks; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device In the first In the first time slot Additive white Gaussian noise for each device To form a beamforming matrix;

[0017] When the device Interference observed after SIC operation during decoding of radio resource block signals. Represented as:

[0018]

[0019] In the formula, , For equipment SIC operation vector; For SIC binary indicators, it specifies whether it is in the device Perform SIC operations to decode the device The signal Indicates device A SiC decoding device is used before decoding its own signal. The signal Indicates device The signal is not decoded using SIC; the wireless resource block allocation strategy uses binary indicators. express, Represents wireless resource blocks In the Period allocated to equipment ,otherwise ; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of; It is an indicator function, if the event If true, then ,otherwise ; It is additive white Gaussian noise;

[0020] When the device For equipment Interference observed after SIC operation during signal decoding for:

[0021]

[0022] In the formula, In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . , This is the binary index of SIC;

[0023] The device-side signal-to-interference-noise ratio model is as follows:

[0024]

[0025]

[0026] In the formula, For equipment The signal-to-interference-to-noise ratio of decoded wireless resource block signals. For equipment Decoding equipment The signal-to-noise ratio of the signal;

[0027] The device-side data rate model is as follows:

[0028]

[0029] In the formula, For equipment The data rate that can be achieved when decoding its own signal. This represents the bandwidth of the channel.

[0030] In step S3, each task is divided into independent subtasks of equal size, with each subtask having a size of [missing information]. ,when Sub-tasks arrive at the device At that time, it can be divided into two independent parts: local processing and unloading tasks. Therefore, the first Equipment in each time slot The task segmentation model at that location is:

[0031]

[0032] In the formula, For the first Equipment in each time slot Size of subtask data used for local processing For the first Equipment in each time slot The size of the subtask data used for unloading the task. For subtasks used for local processing;

[0033] The buffer queue model for storing local processing tasks and unloading tasks on the device side is as follows:

[0034]

[0035]

[0036] In the formula, For equipment A buffer queue for storing locally processed tasks. For the first individual time slot devices The maximum amount of data that leaves the local processing queue. For equipment A buffer queue for storing unloading tasks. For the first individual time slot devices The maximum amount of data leaving the unload task queue;

[0037] The data volume model for the local processing task on the device side is as follows:

[0038]

[0039] In the formula, For equipment In the Each time slot is allocated to the CPU cycle frequency of the local processing task. To calculate the strength;

[0040] No. The computational latency model for a local processing task on a time-slot device is as follows:

[0041]

[0042] In the formula, This refers to the delay in local processing of tasks on the device.

[0043] The device-side local task processing energy consumption model is as follows:

[0044]

[0045] In the formula, For the first individual time slot devices Energy consumption for local task processing For equipment The power coefficient;

[0046] The data volume model for the device-side unloading task is as follows:

[0047]

[0048] No. The computational latency model for the offloading task at the device end of each time slot is as follows:

[0049]

[0050] In the formula, In the first Equipment in each time slot Transmission delay;

[0051] The energy consumption model for the device-side unloading task is as follows:

[0052]

[0053] In the formula, In the first Equipment in each time slot Energy consumption for task unloading.

[0054] In step S4, the objective function is to maximize the system energy efficiency in the power Internet of Things:

[0055]

[0056] In the formula, Assign vectors to radio resource blocks. , ; For task segmentation vectors, , ; Assign a vector to the resource. , ;

[0057] The constraints are:

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064] In the formula, The average data arrival rate during the movement time of the local processing task queue. ; The maximum queuing delay for tasks processed locally on the device. The average data arrival rate for the task unloading queue is the movement time. ; The maximum queuing delay for device task unloading. , This is the binary index of SIC. This represents the maximum transmission power of the antenna base station.

[0065] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0066] This invention provides an energy-saving communication method for a power Internet of Things (IoT) targeting a large number of devices. This energy-saving communication method combines CF-NOMA technology and edge computing technology, which is applicable to a variety of different scenarios without limitations. It can effectively guarantee and even improve the user experience, and achieve ultra-flexible SIC, effectively reducing interference, improving system performance, minimizing energy consumption, and achieving better energy-saving effects. Attached Figure Description

[0067] Figure 1 This is a flowchart of an energy-saving communication method for a large number of devices in the power Internet of Things according to the present invention. Detailed Implementation

[0068] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0069] See Figure 1 A power Internet of Things (IoT) energy-saving communication method for a large number of devices, the method includes the following steps:

[0070] S1. Construct a power Internet of Things system based on CF-NOMA edge computing;

[0071] S2. First, construct the device-side received signal model, then construct the device-side signal-to-interference-to-noise ratio model based on the device-side received signal model, and finally construct the device-side data rate model based on the device-side signal-to-interference-to-noise ratio model.

[0072] S3. First, construct the device-side local processing task data volume and unloading task data volume model based on the task segmentation model. Then, construct the device-side local task processing energy consumption model based on the device-side local processing task data volume model. Construct the device-side unloading task energy consumption model based on the device-side unloading task data volume model. Finally, construct the system's total power consumption model based on the device-side local task processing energy consumption model and the unloading task energy consumption model.

[0073] S4. Construct an objective function using the data rate model of all devices in the system and the total power consumption model of the system. Combine this with the SIC decoding of CF-NOMA and the task queuing delay to construct constraints. Under the constraints, optimize the wireless resource block allocation vector, task segmentation vector, resource allocation vector, device-side transmission beamforming vector, and binary SIC decoding index with the objective function maximization as the optimization goal, so as to maximize the system energy efficiency of the power Internet of Things and obtain the optimized wireless resource block allocation vector, task segmentation vector, resource allocation vector, device-side transmission beamforming vector, and binary SIC decoding index.

[0074] In step S1, the power IoT system based on CF-NOMA edge computing includes one antenna base station, one edge server, and... A single-antenna power IoT device;

[0075] The antenna base station and the edge server are located in the same location, and the antenna base station is respectively connected to... Each single-antenna power IoT device is connected wirelessly in sequence;

[0076] The power Internet of Things system adopts a discrete time-slot model, which divides the total optimization period into... Each time slot, the duration of each time slot Equal, continuous Individual time slots combine to form a period. , .

[0077] In step S2, the signal received by the device is modeled as follows:

[0078]

[0079] In the formula, For equipment Received signals; , The number of power Internet of Things (IoT) devices; , The number of wireless resource blocks; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device In the first In the first time slot Additive white Gaussian noise for each device To form a beamforming matrix;

[0080] When the device Interference observed after SIC operation during decoding of radio resource block signals. Represented as:

[0081]

[0082] In the formula, , For equipment SIC operation vector; For SIC binary indicators, it specifies whether it is in the device Perform SIC operations to decode the device The signal Indicates equipment A SiC decoding device is used before decoding its own signal. The signal Indicates equipment The signal is not decoded using SIC; the wireless resource block allocation strategy uses binary indicators. express, Represents wireless resource blocks In the Period allocated to equipment ,otherwise ; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of; It is an indicator function, if the event If true, then ,otherwise ; It is additive white Gaussian noise;

[0083] When the device For equipment Interference observed after SIC operation during signal decoding for:

[0084]

[0085] In the formula, In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . , This is the binary index of SIC;

[0086] The device-side signal-to-interference-noise ratio model is as follows:

[0087]

[0088]

[0089] In the formula, For equipment The signal-to-interference-to-noise ratio of decoded wireless resource block signals. For equipment Decoding equipment The signal-to-noise ratio of the signal;

[0090] The device-side data rate model is as follows:

[0091]

[0092] In the formula, For equipment The data rate that can be achieved when decoding its own signal. This represents the bandwidth of the channel.

[0093] In step S3, each task is divided into independent subtasks of equal size, with each subtask having a size of [missing information]. ,when Sub-tasks arrive at the device At that time, it can be divided into two independent parts: local processing and unloading tasks. Therefore, the first Equipment in each time slot The task segmentation model at that location is:

[0094]

[0095] In the formula, For the first Equipment in each time slot Size of subtask data used for local processing For the first Equipment in each time slot The size of the subtask data used for unloading the task. For subtasks used for local processing;

[0096] The buffer queue model for storing local processing tasks and unloading tasks on the device side is as follows:

[0097]

[0098]

[0099] In the formula, For equipment A buffer queue for storing locally processed tasks. For the first individual time slot devices The maximum amount of data that leaves the local processing queue. For equipment A buffer queue for storing unloading tasks. For the first individual time slot devices The maximum amount of data leaving the unload task queue;

[0100] The data volume model for the local processing task on the device side is as follows:

[0101]

[0102] In the formula, For equipment In the Each time slot is allocated to the CPU cycle frequency of the local processing task. To calculate the strength;

[0103] No. The computational latency model for a local processing task on a time-slot device is as follows:

[0104]

[0105] In the formula, This refers to the delay in local processing of tasks on the device.

[0106] The device-side local task processing energy consumption model is as follows:

[0107]

[0108] In the formula, For the first individual time slot devices Energy consumption for local task processing For equipment The power coefficient;

[0109] The data volume model for the device-side unloading task is as follows:

[0110]

[0111] No. The computational latency model for the offloading task at the device end of each time slot is as follows:

[0112]

[0113] In the formula, In the first Equipment in each time slot Transmission delay;

[0114] The energy consumption model for the device-side unloading task is as follows:

[0115]

[0116] In the formula, In the first Equipment in each time slot Energy consumption for task unloading.

[0117] In step S4, the objective function is to maximize the system energy efficiency in the power Internet of Things:

[0118]

[0119] In the formula, Assign vectors to radio resource blocks. , ; For task segmentation vectors, , ; Assign a vector to the resource. , ;

[0120] The constraints are:

[0121]

[0122]

[0123]

[0124]

[0125]

[0126]

[0127] In the formula, The average data arrival rate during the movement time of the local processing task queue. ; The maximum queuing delay for tasks processed locally on the device. The average data arrival rate for the task unloading queue is the movement time. ; The maximum queuing delay for device task unloading. , This is the binary index of SIC. This represents the maximum transmission power of the antenna base station.

[0128] The principle of this invention is explained as follows:

[0129] Cluster-free nonorthogonal multiple access (CF-NOMA) technology combines spatial-division multiple access (SDMA), BB-NOMA, and CB-NOMA to achieve ultra-flexible SIC by releasing the limitations of clusters, effectively reducing interference and improving system performance.

[0130] This invention addresses the limitation of NOMA technology's applicability in all power IoT scenarios. It introduces CF-NOMA technology, starting from SIC decoding and long-term queuing delays in power IoT devices, and proposes a power IoT energy-saving communication method combining CF-NOMA and edge computing. The objective function is modeled as a long-term stochastic optimization problem, significantly improving power IoT network capacity and massive device connectivity, reducing energy consumption, and enhancing communication quality. Mathematically, CF-NOMA provides a universal modeling method, unifying existing methods and offering more flexible transmission options, thus overcoming the shortcomings of existing methods. Therefore, using CF-NOMA in power IoT is applicable to various scenarios without limitations, offering better communication performance and reduced energy consumption. Based on this, according to user requirements for SIC decoding conditions and queuing delay constraints, we propose an energy efficiency maximization problem to jointly optimize five variables: Radio Resource Blocks (RB) allocation vector, task segmentation vector, resource allocation vector, device-side transmission beamforming vector, and binary SIC decoding index. This reduces energy consumption while maximizing communication quality, providing a dual improvement for power IoT. The constraints include SIC decoding condition constraints and binary SIC decoding index constraints to achieve CF-NOMA.

[0131] Example:

[0132] See Figure 1 A power Internet of Things (IoT) energy-saving communication method for a large number of devices, the method includes the following steps:

[0133] S1. Construct a power Internet of Things system based on CF-NOMA edge computing;

[0134] The power IoT system based on CF-NOMA edge computing includes one antenna base station, one edge server, and... A single-antenna power IoT device;

[0135] The antenna base station and the edge server are located in the same location, and the antenna base station is respectively connected to... Each single-antenna power IoT device is connected wirelessly in sequence;

[0136] The power Internet of Things system adopts a discrete time-slot model, which divides the total optimization period into... Each time slot, the duration of each time slot Equal, continuous Individual time slots combine to form a period. , ;

[0137] S2. First, construct the device-side received signal model, then construct the device-side signal-to-interference-to-noise ratio model based on the device-side received signal model, and finally construct the device-side data rate model based on the device-side signal-to-interference-to-noise ratio model.

[0138] The signal receiving model at the device end is as follows:

[0139]

[0140] In the formula, For equipment Received signals; , The number of power Internet of Things (IoT) devices; , The number of wireless resource blocks; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device; In the first In the first time slot Additive white Gaussian noise for each device Because the proposed power Internet of Things (IoT) system eliminates the concept of clusters, each device does not need to share a beamforming vector with any other device. To form a beamforming matrix;

[0141] When the device Interference observed after SIC operation during decoding of radio resource block signals. Represented as:

[0142]

[0143] In the formula, , For equipment SIC operation vector; For SIC binary indicators, it specifies whether it is in the device Perform SIC operations to decode the device The signal Indicates equipment A SiC decoding device is used before decoding its own signal. The signal to eliminate the device Interference, Indicates equipment The signal is not decoded using SIC; the wireless resource block allocation strategy uses binary indicators. express, Represents wireless resource blocks In the Period allocated to equipment ,otherwise ; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of; It is an indicator function, if the event If true, then ,otherwise ; It is additive white Gaussian noise;

[0144] When the device For equipment Interference observed after SIC operation during signal decoding for:

[0145]

[0146] In the formula, In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . , This is the binary index of SIC;

[0147] The device-side signal-to-interference-noise ratio model is as follows:

[0148]

[0149]

[0150] In the formula, For equipment The signal-to-interference-to-noise ratio of decoded wireless resource block signals. For equipment Decoding equipment The signal-to-noise ratio of the signal;

[0151] The device-side data rate model is as follows:

[0152]

[0153] In the formula, For equipment The data rate that can be achieved when decoding its own signal. The bandwidth of the channel;

[0154] S3. First, construct the device-side local processing task data volume and unloading task data volume model based on the task segmentation model. Then, construct the device-side local task processing energy consumption model based on the device-side local processing task data volume model. Construct the device-side unloading task energy consumption model based on the device-side unloading task data volume model. Finally, construct the system's total power consumption model based on the device-side local task processing energy consumption model and the unloading task energy consumption model.

[0155] Divide each task into independent subtasks of equal size, with each subtask having a size of [size missing]. ,when Sub-tasks arrive at the device At that time, it can be divided into two independent parts: local processing and unloading tasks. Therefore, the first Equipment in each time slot The task segmentation model at that location is:

[0156]

[0157] In the formula, For the first Equipment in each time slot Size of subtask data used for local processing For the first Equipment in each time slot The size of the subtask data used for unloading the task. For subtasks used for local processing;

[0158] The buffer queue model for storing local processing tasks and unloading tasks on the device side is as follows:

[0159]

[0160]

[0161] In the formula, For equipment A buffer queue for storing locally processed tasks. For the first individual time slot devices The maximum amount of data that leaves the local processing queue. For equipment A buffer queue for storing unloading tasks. For the first individual time slot devices The maximum amount of data leaving the unload task queue;

[0162] The data volume model for the local processing task on the device side is as follows:

[0163]

[0164] In the formula, For equipment In the Each time slot is the CPU cycle frequency allocated to a local processing task, measured in cycles / s; The intensity is calculated in cycles / bit;

[0165] No. The computational latency model for a local processing task on a time-slot device is as follows:

[0166]

[0167] In the formula, This refers to the delay in local processing of tasks on the device.

[0168] The device-side local task processing energy consumption model is as follows:

[0169]

[0170] In the formula, For the first individual time slot devices Energy consumption for local task processing For equipment The power coefficient;

[0171] The data volume model for the device-side unloading task is as follows:

[0172]

[0173] No. The computational latency model for the offloading task at the device end of each time slot is as follows:

[0174]

[0175] In the formula, In the first Equipment in each time slot Transmission delay;

[0176] The energy consumption model for the device-side unloading task is as follows:

[0177]

[0178] In the formula, In the first Equipment in each time slot Task unloading energy consumption;

[0179] S4. Construct an objective function using the data rate model of all devices in the system and the total power consumption model of the system. Combine the SIC decoding of CF-NOMA and the task queuing delay to construct constraints. Under the constraints, optimize the Radio Resource Blocks (RB) allocation vector, task segmentation vector, resource allocation vector, transmission beamforming vector at the device end, and binary SIC decoding index with the objective function maximization as the optimization objective, so as to maximize the system energy efficiency of the power Internet of Things and obtain the optimized radio resource block allocation vector, task segmentation vector, resource allocation vector, transmission beamforming vector at the device end, and binary SIC decoding index.

[0180] The objective function is to maximize the system energy efficiency in the power Internet of Things (IoT).

[0181]

[0182] In the formula, Assign vectors to radio resource blocks. , ; For task segmentation vectors, , ; Assign a vector to the resource. , ;

[0183] The constraints are:

[0184]

[0185]

[0186]

[0187]

[0188]

[0189]

[0190] In the formula, The average data arrival rate during the movement time of the local processing task queue. ; The maximum queuing delay for tasks processed locally on the device. The average data arrival rate for the task unloading queue is the movement time. ; The maximum queuing delay for device task unloading. , This is the binary index of SIC. This represents the maximum transmission power of the antenna base station;

[0191] The first constraint states that each device can be allocated at most one radio resource block; the second and third constraints represent queuing delay constraints; the fourth constraint states that users cannot perform SIC decoding with each other; the fifth constraint represents the SIC decoding condition; and the sixth constraint ensures the maximum transmission power of the base station. .

Claims

1. A power Internet of Things (IoT) energy-saving communication method for a large number of devices, characterized in that, The method includes the following steps: S1. Construct a power Internet of Things system based on CF-NOMA edge computing; S2. First, construct the device-side received signal model, then construct the device-side signal-to-interference-to-noise ratio model based on the device-side received signal model, and finally construct the device-side data rate model based on the device-side signal-to-interference-to-noise ratio model. S3. First, construct the device-side local processing task data volume and unloading task data volume model based on the task segmentation model. Then, construct the device-side local task processing energy consumption model based on the device-side local processing task data volume model. Construct the device-side unloading task energy consumption model based on the device-side unloading task data volume model. Finally, construct the system's total power consumption model based on the device-side local task processing energy consumption model and the unloading task energy consumption model. S4. Construct an objective function using the data rate model of all devices in the system and the total power consumption model of the system. Combine this with the SIC decoding of CF-NOMA and the task queuing delay to construct constraints. Under the constraints, optimize the wireless resource block allocation vector, task segmentation vector, resource allocation vector, device-side transmission beamforming vector, and binary SIC decoding index with the objective function maximization as the optimization goal, so as to maximize the system energy efficiency of the power Internet of Things and obtain the optimized wireless resource block allocation vector, task segmentation vector, resource allocation vector, device-side transmission beamforming vector, and binary SIC decoding index.

2. The power Internet of Things energy-saving communication method for massive numbers of devices according to claim 1, characterized in that: In step S1, the power IoT system based on CF-NOMA edge computing includes one antenna base station, one edge server, and... A single-antenna power IoT device; The antenna base station and the edge server are located in the same location, and the antenna base station is respectively connected to... Each single-antenna power IoT device is connected wirelessly in sequence; The power Internet of Things system adopts a discrete time-slot model, which divides the total optimization period into... Each time slot, the duration of each time slot Equal, continuous Individual time slots combine to form a period. , .

3. The power Internet of Things energy-saving communication method for massive numbers of devices according to claim 2, characterized in that: In step S2, the signal received by the device is modeled as follows: In the formula, For equipment Received signals; , The number of power Internet of Things (IoT) devices; , The number of wireless resource blocks; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device In the first In the first time slot Each device is formed by a dedicated transmission beam vector. Provide services In the first In the first time slot Normalized data signals of each device In the first In the first time slot Additive white Gaussian noise for each device To form a beamforming matrix; When the device Interference observed after SIC operation during decoding of radio resource block signals. Represented as: In the formula, , For equipment SIC operation vector; For SIC binary indicators, it specifies whether it is in the device Perform SIC operations to decode the device The signal Indicates equipment A SiC decoding device is used before decoding its own signal. The signal Indicates equipment The signal is not decoded using SIC. Wireless resource block allocation strategy using binary indicators express, Represents wireless resource blocks In the Period allocated to equipment ,otherwise ; In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of; It is an indicator function, if the event If true, then ,otherwise ; It is additive white Gaussian noise; When the device For equipment Interference observed after SIC operation during signal decoding for: In the formula, In the first In the first time slot The first wireless resource block to the first Channel vectors of each device for The conjugate transpose of . , This is the binary index of SIC; The device-side signal-to-interference-noise ratio model is as follows: In the formula, For equipment Decoding the signal-to-interference-to-noise ratio of wireless resource block signals. For equipment Decoding equipment The signal-to-noise ratio of the signal; The device-side data rate model is as follows: In the formula, For equipment The data rate achievable by decoding its own signal. This represents the bandwidth of the channel.

4. The power Internet of Things energy-saving communication method for massive numbers of devices according to claim 3, characterized in that: In step S3, each task is divided into independent subtasks of equal size, with each subtask having a size of [missing information]. ,when Sub-tasks arrive at the device At that time, it can be divided into two independent parts: local processing and unloading tasks. Therefore, the first Equipment in each time slot The task segmentation model at that location is: In the formula, For the first Equipment in each time slot Size of subtask data used for local processing For the first Equipment in each time slot The size of the subtask data used for unloading the task. For subtasks used for local processing; The buffer queue model for storing local processing tasks and unloading tasks on the device side is as follows: In the formula, For equipment A buffer queue for storing locally processed tasks. For the first individual time slot devices The maximum amount of data that leaves the local processing queue. For equipment A buffer queue for storing unloading tasks. For the first individual time slot devices The maximum amount of data leaving the unload task queue; The data volume model for the local processing task on the device side is as follows: In the formula, For equipment In the Each time slot is allocated to the CPU cycle frequency of the local processing task. To calculate the strength; No. The computational latency model for a local processing task on a time-slot device is as follows: In the formula, This refers to the delay in local processing of tasks on the device. The device-side local task processing energy consumption model is as follows: In the formula, For the first individual time slot devices Energy consumption for local task processing For equipment The power coefficient; The data volume model for the device-side unloading task is as follows: No. The computational latency model for the offloading task at the device end of each time slot is as follows: In the formula, In the first Equipment in each time slot Transmission delay; The energy consumption model for the device-side unloading task is as follows: In the formula, In the first Equipment in each time slot Energy consumption for task unloading.

5. A power Internet of Things energy-saving communication method for massive numbers of devices according to claim 4, characterized in that: In step S4, the objective function is to maximize the system energy efficiency in the power Internet of Things: In the formula, Assign vectors to radio resource blocks. , ; For task segmentation vectors, , ; Assign a vector to the resource. , ; The constraints are: In the formula, The average data arrival rate during the movement time of the local processing task queue. ; The maximum queuing delay for tasks processed locally on the device. The average data arrival rate for the task unloading queue is the movement time. ; The maximum queuing delay for device task unloading. , This is the binary index of SIC. This represents the maximum transmission power of the antenna base station.

Images