Multi-tasking processing method based on device priority in photovoltaic 5G base station system

Abstract

The present invention discloses a multi-task processing method based on device priority in a photovoltaic 5G base station system, comprising: after obtaining the terminal layer device status information, forming a data packet with the priority information of the terminal layer device to send a task; the edge layer device receives the task and temporarily stores it in the edge buffer, and judges whether the status information data is abnormal according to the order of the terminal layer device priority. If the frequency of abnormal data of the same type within a period of time is higher than a set threshold, the task is uploaded to the cloud layer device for processing, otherwise the task enters the edge layer device for processing and is cleared in the edge buffer; the software defined network (SDN) controller allocates the server resources of the edge layer device to process the task to be processed according to the priority information in the data packet; the processing result is fed back to the terminal layer device; and the above steps are repeated until the edge buffer task is cleared. The present invention can improve the purposefulness and efficiency of system task processing and reduce the cost loss caused by failures to the system.

Claims

1. A multi-tasking processing method based on device priority in a photovoltaic 5G base station system, characterized in that: include: Step 1: After obtaining the terminal layer device status information in the photovoltaic 5G base station system, package it with the priority information of the terminal layer device into a data packet and issue the task; The priority information of the terminal layer device is calculated by an improved multi-objective genetic algorithm, specifically including: The optimization objective function is set as delay and power consumption. The chromosome gene corresponds to the priority information value, and the gene with high device importance has a low value. The Pareto rank, crowding, gene mean and variance are used for selection. If the Pareto rank and crowding are the same, the chromosome with high mean or large variance is preferred. Step 2: After receiving the task, the edge layer device in the photovoltaic 5G base station system temporarily stores the data packet in the edge buffer. It then determines whether the status information data is abnormal in order of priority of the terminal layer devices. If the frequency of abnormal data of the same type within a period of time exceeds the set threshold, the task is directly uploaded to the cloud layer device for processing. Otherwise, the task enters the queue of pending tasks of the edge layer device and is cleared from the edge buffer. Abnormal data that is lower than the manually set interval value is type 1 abnormal data, and data that is higher than the manually set interval value is type 2 abnormal data. The data is compared with the 2-4 months of historical data stored in the edge layer. If the frequency of the same type of abnormal data exceeds the threshold, it is uploaded to the cloud layer for processing. Otherwise, it is processed by the edge layer and the buffer task is cleared. Step 3: Allocate server resources of edge devices to process pending tasks based on priority information in the data packet through a software-defined network controller; The priority is divided into several levels according to the value of the priority information, and the network topology distance of the edge layer device server resources is divided into the same number of levels; The higher the priority, that is, the lower the priority information value, the task is assigned to the server resource with the closer topological distance; Step 4: The processing results of the edge layer device are directly fed back to the terminal layer device, and the cloud layer processing results are transmitted to the terminal layer device via the edge layer device; Step 5: Repeat steps 2 to 4 until the edge buffer tasks are cleared.

2. The multi-tasking processing method based on device priority in the photovoltaic 5G base station system according to claim 1 is characterized in that: Specifically, in step 2, after receiving the data packet, the edge layer device extracts the priority information in the data packet and sorts the data, and the data processing task sorting corresponds to the priority information sorting; The comparison is based on historical data stored in edge layer devices over a period of time. The values ​​within the manually set interval in the historical data are considered normal data, while those exceeding the manually set interval are considered abnormal data. Among the abnormal data, those below the manually set interval are considered first-type abnormal data, while those above the manually set interval are considered second-type abnormal data. Starting from the tasks with high priority, determine whether the status information data in the tasks is abnormal. If the data is abnormal, compare it with the historical data to obtain the frequency of occurrence of the same type of abnormal data. If the frequency of occurrence of the same type of abnormal data exceeds the set threshold, upload it to the cloud layer device for processing, otherwise it will be retained in the edge layer device for processing; the period of time refers to 2 to 4 months, and the edge layer device stores the data of the data packet in time as historical data, and deletes expired historical data in turn.

3. The multi-tasking processing method based on device priority in the photovoltaic 5G base station system according to claim 1 is characterized in that: In the third step, the allocation rule of the software-defined network controller is specifically as follows: The priorities are divided into several levels according to the magnitude of the priority information values; at the same time, the network topological distances of the edge layer device server resources are also divided into the same number of several levels; According to the principle that the higher the priority (i.e., the lower the priority information value), the closer the resource allocated to in the network, the tasks are allocated to the edge layer device server resources with corresponding network topological distances for processing according to the priorities corresponding to the tasks to be processed.

4. The multi-tasking processing method based on device priority in the photovoltaic 5G base station system according to claim 1 is characterized in that: In the first step, the priority information of the terminal layer devices is calculated by an improved multi-objective genetic algorithm, specifically as follows: The optimization objective function is set as delay and power consumption. The genes carried on the chromosome correspond to the priority information values of the terminal layer devices. Based on the importance of the terminal layer devices to the system, it is set that those with higher importance have higher priorities and lower corresponding gene values; those with lower importance have lower priorities and higher corresponding gene values; The following steps are carried out: Step 1: Parameter initialization; Step 2: Randomly generate a population; Step 3: Calculate the objective function values of the parent chromosomes; Step 4: According to the objective function values, perform Pareto ranking by the fast non-dominated method; Step 5: Calculate the crowding degree of the parent chromosomes; Step 6: Perform the selection operation according to the Pareto ranking and crowding degree; Step 7: Generate offspring chromosomes from the selected chromosomes through crossover and mutation, etc.; Step 8: Calculate the objective function values of the offspring chromosomes; Step 9: Mix the parent and offspring chromosomes and perform Pareto ranking by the fast non-dominated method; Step 10: Calculate the crowding degree of the parent and offspring chromosomes and the mean and variance of the genes on each chromosome; Step 11: Perform the selection operation according to the Pareto ranking, crowding degree, mean and variance, specifically as follows: Let the means and variances of chromosomes x1 and x2 be E1, E2 and D1, D2 respectively; for x1 and x2, if any of the following conditions is satisfied, it is said that x1 is better than x2; 1) The Pareto rank number of x1 is less than that of x2; 2) x1 and x2 are in the same Pareto rank and the crowding degree of x1 > the crowding degree of x2; 3) x1 and x2 are in the same Pareto rank and the crowding degrees are equal, then they are discriminated by E1, E2 and D1, D2 of x1 and x2, specifically as follows: E1 > E2, D1 > D2, x1 is better than x2; E1 > E2, D1 < D2, x1 is better than x2; E1 < E2, D1 > D2, x2 is better than x1; E1 < E2, D1 < D2, x2 is better than x1; E1 = E2, the one with larger variance is better; D1 = D2, the one with larger expectation is better; Step 12: Repeat the iteration to obtain the optimal solution.

5. The multi-tasking processing method based on device priority in the photovoltaic 5G base station system according to claim 1 is characterized in that: The terminal layer devices include: 5G base station main equipment, photovoltaic modules, storage batteries, busbar boxes, sensors and WiFi modules. The sensors are respectively connected to the 5G base station main equipment, photovoltaic modules, storage batteries and busbar boxes to collect their status information and priority information. The WiFi module is connected to the sensors for wireless connection with the edge layer devices.

6. The multi-tasking processing method based on device priority in the photovoltaic 5G base station system according to claim 1 is characterized in that: The edge layer device includes an edge server for providing data storage resources and computing resources; the edge layer device is connected to the sensor through a WiFi module and is wirelessly connected to the cloud layer through a 5G communication module; Among them, an edge buffer is set in the edge layer device, and the edge layer device allows tasks to enter the edge buffer directly. The edge buffer has enough space to store tasks transmitted from the terminal layer device.

7. The multi-tasking processing method based on device priority in a photovoltaic 5G base station system according to claim 1, characterized in that: The cloud layer device includes a cloud server, which provides sufficient data storage resources and computing resources.

8. The multi-tasking processing method based on device priority in the photovoltaic 5G base station system according to claim 6 is characterized in that: Only when a task enters the edge buffer, the processing steps from step 2 to step 4 are executed, otherwise the edge layer is in an idle state.

9. Photovoltaic 5G base station system, characterized in that: include: Terminal layer devices, edge layer devices and cloud layer devices, and perform task processing through the multi-tasking processing method according to claim 1.

10. The photovoltaic 5G base station system according to claim 9, characterized in that: The terminal layer equipment includes: a 5G base station main device, a photovoltaic module, a battery, a combiner box, a sensor, and a WiFi module. The sensors are respectively connected to the 5G base station main device, the photovoltaic module, the battery, and the combiner box to collect their status information and priority information. The WiFi module is connected to the sensor for wireless connection with the edge layer equipment. The edge layer devices include: edge servers for providing data storage resources and computing resources; edge layer devices are connected to sensors via WiFi modules and wirelessly connected to the cloud layer via 5G communication modules; wherein, edge layer devices are provided with edge buffers, which allow tasks to enter the edge buffers directly, and the edge buffers have sufficient space to store tasks transmitted from terminal layer devices; The cloud layer device includes: a cloud server, which is connected to the edge layer and provides data storage resources and computing resources.

Images