Research On Reliable Transmission Method Of High Interference Wireless Link In Internet Of Things

With the explosive growth of Internet of Things(IoT)devices,wireless links in IoT are characterised by "massive connectivity,heterogeneous protocol coexistence and dynamic deployment environments".In this new development phase,wireless links are facing more challenging high interference problems.This thesis focuses on reliable transmission methods for high interference wireless links in IoT.The main contributions of this thesis are as follows:(1)Parallel decoding method for transmission collision interference.When multiple wireless devices send packets to a single receiver at the same time,a serious collision interference problem is created at the receiver,which degrades the reliability of the wireless link,especially for a massively connected LoRa network.LoRa devices connected to the same gateway suffer severe collisions if they transmit packets to the gateway with the same parameter configuration.Although existing collision decoding methods can decode most collision cases,none of them can effectively handle adjacent collisions that occur when collision symbols are adjacent on the spectrum.A fundamental limitation of existing methods is that they rely on the exact separation of the peaks in the frequency domain corresponding to different collision symbols,and thus can only solve the problem of collisions of symbols that are not adjacent in frequency.To address these issues,this thesis proposes a LoRa collision decoding method via parallel alignment,Paralign,which enables the decoding of LoRa collisions containing adjacent collisions.The key to Paralign is to align the demodulation windows with the symbols of each packet in parallel according to the boundaries of the collision packets,and then match the peaks of these parallel-aligned windows with the peaks of the interested window.To extract the interested symbols,Paralign performs spectrum filtering,including spectral intersection and exception,and power filtering in the frequency domain.For the multiple candidate peaks problem caused by side lobes of confusing symbols,Paralign selects the interested peaks by the periodic truncation method Experimental results show that Paralign is able to achieve higher network throughput in a practical setting than existing parallel decoding methods.(2)Channel hopping method for cross-technology interference.The coexistence of numerous heterogeneous wireless network protocols inevitably leads to Cross Technology Interference(CTI),which seriously jeopardizes the transmission reliability of low-power networks,such as ZigBee.Unlike existing binary methods of determining channel availability,this article observes that the correlated channels influenced by the same CTI source do not necessarily have the same channel qualities and even the opposite state,due to the uneven spectrum power density of CTI.To exploit these unexplored channel opportunities,this thesis proposes a quantitative correlation based channel hopping method for low-power wireless networks,CoHop.CoHop can capture channel availability opportunities by building a quantitative model of channel quality correlation and optimise the probing sequence using a Pareto model to achieve channel quality prediction without actual probing to reduce probing overhead.Experimental results show that CoHop is able to achieve higher packet reception rates in real-world environments compared to existing methods.(3)Channel access method for time-varying physical attitude interference.Due to the limitations of low-power wireless devices,the highly dynamic becomes a challenge for optimizing the reliability of wireless communication links of floating IoT devices,which restricts the deployment of IoT applications to the aquatic environments.Traditional detection-based wireless channel sensing methods fail due to the channel reliability fluctuations,making it impossible to accurately characterise and predict the future channel state using historical detection information.This thesis observes that the highly dynamic antenna polarization matching state caused by the changing spatial attitude of nodes is the root cause of reliability degradation in floating networks.To investigate the relationship between polarization matching state and link quality,a link quality model based on attitude awareness is proposed.Based on this model,a new channel access method for floating IoT devices,PolarTracker,is proposed.PolarTracker can track the dynamic attitude alignment state of nodes and schedule packets to aligned period for transmission to obtain better link quality.The experimental evaluation is based on LoRa technology as an example.Experimental results show that PolarTracker achieves higher packet reception rates in real-world environments compared to LoRa WAN using ALOHA,alleviating the problem of degraded wireless link reliability in floating networks.In summary,this thesis explores the high interference problem of wireless links in the new development phase of the IoT and designs a series of reliable transmission methods for high interference wireless links.The effectiveness is verified through theoretical analysis and real environment experiments.The designs of this thesis provide a theoretical basis and technical support for the future development of reliable transmission methods for wireless links in IoT.