Analysis Of Uplink Access Performance And Research On Terminal Parameter Configuration Methods For LoRa System Low Earth Orbit Satellite IoT Scenarios

The current research and development of ground-based Internet of Things(IoT)are progressing rapidly.However,in areas such as oceans,mountains,deserts,and other regions,there are many limitations that make it difficult to deploy and maintain ground-based IoT base stations.This hinders the research and application of data services in these areas.In contrast,satellite-based IoT systems have the advantage of wide coverage,making it easy to support network and service in areas with broad coverage.Satellite IoT effectively compensates for the shortcomings of ground-based IoT.Low Earth Orbit(LEO)satellite IoT is characterized by low power consumption,low cost,and wide coverage.However,most existing LoRa research and system models are focused on ground-based IoT.Therefore,in recent years,the adaptability analysis of LoRa in LEO satellite IoT has become a major research topic in the field.The coverage range of ground IoT is approximately ten kilometers with a relatively small number of access terminals,and the channel fading model is the Rayleigh channel model.In contrast,in low Earth orbit satellite IoT,the low Earth orbit satellites have beam coverage radii of several hundred kilometers,and there is a massive concurrent access of terminals.The channel fading model used is the shadowed-Loss channel model.The ground IoT base stations exhibit quasi-static characteristics,where the business and channel state information do not undergo significant changes.On the other hand,low Earth orbit satellites exhibit high dynamic characteristics,where the business and channel state information have strong temporal variations.Therefore,considering the differences between low Earth orbit satellite IoT and ground IoT,the main research focus of the paper includes:(1)This article analyzes the uplink access performance of LoRa signals in two scenarios:direct connection terminals and aggregation terminals in the low-earth orbit satellite Internet of Things.It derives a closed-form expression for the uplink access success probability.First,this article focuses on two LoRa IoT scenarios:direct-connected terminals and aggregation terminals in low Earth orbit satellite IoT,and establishes the terminal distribution and channel models for both scenarios.Then,assuming no collision of LoRa signal packets,the closed-form expression for the uplink access success probability of LoRa signals is derived from the perspective of the received signal strength indicator at the low Earth orbit satellite.Finally,considering the collision of multiple LoRa signal packets,the closed-form expression for the uplink access success probability of LoRa signals is derived from the perspective of the received signal strength indicator at the low Earth orbit satellite,using serial interference cancellation techniques.In addition,a mathematical method based on random geometry is applied to analyze the uplink access performance in the aggregation terminal IoT scenario.Simulation results demonstrate that the analysis in this article can guide the initial transmit power settings to meet the average access success probability requirements of low Earth orbit satellite IoT.Compared to direct-connected terminals in low Earth orbit satellite IoT,the aggregation terminals exhibit better uplink access performance for LoRa signals and have advantages,especially under heavy loads when the modeling density of aggregation cluster heads is higher.(2)To address the issues of dynamic business and channel conditions in low Earth orbit satellite IoT and the low throughput of satellite IoT under the LoRaWAN protocol,this article proposes a method for configuring the transmission parameters of LoRa-based low Earth orbit satellite IoT terminals.The method consists of three parts:satellite-side load estimation and prediction method under asynchronous access scenarios,channel estimation and prediction method,and joint optimization method for terminal transmission parameters.First,using the preamble sequences in the LoRa frame structure,the maximum likelihood detection method is employed to achieve asynchronous collision detection of data packets,obtaining the total access load at the satellite side.The load value is then used as historical data to predict the load value of the next satellite reception window using an LSTM neural network.Next,using the preamble sequences in the LoRa frame structure,compressed sensing is used to estimate the channel state information,and the channel state information value is used as historical data to predict the channel state information value at the next terminal access moment using an LSTM neural network.Finally,based on the predicted load and channel state information,a joint configuration method for terminal transmission power,spreading factor(SF)selection,and subchannel selection is designed using reinforcement learning.Simulation results demonstrate that the proposed method effectively allocates terminal parameters and achieves improved system throughput compared to traditional LoRaWAN.