Multi-Node Collaborative Detection Technology Of LoRa Signal

The Internet of Things shows a rapid development trend.With the continuous development of the Internet of Things,more and more wireless devices are applied to actual scenarios.As the Internet of Things technology widely used,LoRa is becoming more and more popular.As more and more Io T devices are based LoRa,the conflicts between data packets will increase exponentially,which will reduce the effectiveness of system transmission.The multi-node system detection technology of LoRa signals is intended to detect LoRa signals and get location of LoRa signal source devices,so as to better support the monitoring and management of the signal spectrum and related services,and to improve the transmission efficiency of the physical.To detect of LoRa signals,the current method of scanning mostly trace the channels where LoRa signals may appear,or detect the unique preamble features of LoRa.If prior information is too limited,the detection efficiency will be lower.However,in practice,the blind detection of LoRa signal is often required,so as to making the existing detection methods inefficient.The methods based on the minimum mean square error(DLMS)distributed energy detection algorithm have high performance.However,the DLMS algorithm assumes that the input data is standard and 100 percent correct,and which is often biased by the distributed node itself due to hardware acquisition and other reasons.This paper proposes a novel distributed gradient descent total least squares energy detection algorithm(ND-GDTLS),to improve the shortcomings of the DLMS algorithm.Simulation experiments show that the stability of the(ND-GDTLS)energy detection algorithm is better than the DLMS algorithm,the node error is lower,the detection probability is higher,and the performance is better.Compared with centralized,the network structure is more stable and more suitable for wireless networks.The existing positioning algorithm of LoRa signal are mostly based anchor nodes and transmission power,the fusion center uses transmit power of the signal source which is assumed known and the position information of the anchor node to obtain the position of the signal source.But to get the location of LoRa signal sources with unknown information particularly the transmit power of the signal source,such methods are difficult to fulfill the requirements.In addition,positioning of signal often requires multi-node to coordinate.Different nodes may get the detection signal by line-of-sight transmission(LOS)or non-line-of-sight transmission(NLOS),but current node classification can not distinguish during the positioning detection of existing LoRa signal positioning algorithms with unknown information.This proposed method of the multi-node LoRa signal co-localization is achieved by using the expectation-optimized AOA positioning algorithm,distinguishing between line-of-sight nodes and non-line-of-sight nodes,and removing nodes with large errors to complete LoRa signal positioning.The experimental results show that the expected maximum AOA algorithm can detect LoRa signal well.Compared with the centralized expectation maximization algorithm,the performance of the algorithm is more robust,and the positioning error is smaller.As mentioned above,the existing LoRa signal detection equipment mostly uses the existing LoRa chip to scan the LoRa signal.Therefore the implementation and verification of the LoRa signal detection algorithm described can not be achieved.So this paper proposes to set up LoRa signal detection node make use of the software radio platform,and the node use energy detection algorithms to complete local scanning and detection of LoRa signal,the node can complete local judgment and information storage.The designed single node can detect,scan and collect the actual monitoring data,which is foundation to the complete construction of a multi-node distributed LoRa signal monitoring network,and to verify the distributed detection algorithm.The experimental results show that the LoRa signal detection node based on the software radio platform can complete the detection of LoRa signal.Compared with the channel scanning,the detection efficiency and rate are improved,and the blind detection of LoRa signal is better completed.