Research On Networking And Transmission Mechanisms Of Low Power Wide Area Network

Low Power Wide Area Networks(LPWANs)are pivotal in provisioning low-power,long-range connectivity and networking services for a myriad of Internet of Things(Io T)applications,such as smart grids,smart cites and so on,which makes LPWANs a focal point of contemporary Io T research.Among these LPWAN technologies,LoRa communication technology has garnered extensive attention from industrial and academic areas,attributed to its quintessential characteristics of low-power consumption,expansive coverage,large-scale device connectivity,and its open ecosystem.Efficient networking and transmission of low-power devices in LoRa networks has become a critical research hotspot among recent LoRa studies.These existing literature delineates that LoRa networks suffer from prevalent uplink packet collisions and inefficient downlink data dissemination in practical deployment.The former issue is caused by the absence of channel idle checking before transmission,the scale of device quantity,and long transmission airtime,which is likely to result in packet loss or network unavailable.The latter issue stems from the LoRa WAN protocol——employed as the MAC layer protocol in LoRa networks——which is beleaguered by high power consumption,long preparation duration,and considerable decoding overheads within its data dissemination protocol,thus decreasing the efficiency of data dissemination.These impediments significantly attenuate the network’s scalability,throughput,and operational longevity,thereby hindering LoRa from mass application.This thesis endeavors to address these challenges by exploring recovery solutions of collided packet under low SNR conditions,anti-collision schemes for multiple packet collisions,and energy-efficient data dissemination protocols.This thesis aims to augment the LoRa network’s throughput,scalability,and lifespan.The scholarly contributions of this research have been disseminated across well-known international conference and journals within the computer network and Io T domains,including IEEE INFOCOM,IEEE Transactions on Industrial Informatics,and IEEE UIC.To elucidate,this thesis delineates the following core research objects:Multi-Gateway Cooperative Recovery of LoRa Collided Packets: This segment delves into the challenge of packet collision recovery within LoRa networks.Existing researches leverage the unique temporal and spectral characteristics of LoRa collided packets to separate the different packets in the collision.Nonetheless,these approaches require high SNR signals for collided packets,rendering them ineffective under low SNR conditions.To mitigate this issue,this thesis introduces the Cooperative Packet Recovery(CPR)scheme,which employs an amalgamation of multiple gateways to recovery collided packets.CPR exploits the coherence of signals within a collision across disparate gateways and the incoherence of noise to enhance the temporal and spectral features of LoRa packets,thereby facilitating the recovery of packets under low SNR scenarios.CPR encompasses a sliding window mechanism for signal accumulation and preamble feature extraction,in order to detect collision locally.Besides,CPR proposes a novel packet signal alignment mechanism predicated on the features of LoRa’s frame structure for coherently accumulating the collision signals from different gateways.After that,a strategic gateway selection is proposed based on SNR characteristics extracted during collision detection,so that helpful gateways can be added in cooperative recovery.Empirical evaluations substantiate CPR’s efficacy in significantly elevating the symbol recovery rate for collided packets in low SNR environments.Channel Hopping-Based Anti-Collision Scheme for LoRa Networks: The distortion of temporal and spectral characteristics in multi-packet collisions hampers the effectiveness of extant LoRa data packet recovery efforts.Prevailing studies have endeavored to enhance anti-collision capabilities by modulating the physical layer of LoRa,specifically by altering the chirp which is the fundamental symbolic unit of LoRa data packet transmission.However,such modifications are incompatible with existing LoRa networks and incur substantial deployment and operational costs.In response,this thesis proposes CHLoRa(Channel Hopping LoRa),an innovative anti-collision mechanism that segments chirps into subchirps and disperses them across random channels via frequency hopping.Such a method introduces additional channel sequence characteristics that facilitate the segregation of collided packets during demodulation.To make CHLoRa come true,this thesis designs a novel LoRa data packet frame structure for direct hopping channel sequence synchronization,thereby obviating pre-negotiation costs.Moreover,CHLoRa’s implementation on commercial terminal devices through interrupt mechanism invocation and register modification reduces its deployment and operational expenditures.This thesis also introduces a parallel demodulation mechanism for CHLoRa,enabling collision detection,demodulation window alignment,and interference elimination in multi-packet collision scenarios.Experimental findings corroborate CHLoRa’s superior collision detection accuracy and demodulation success rates relative to existing methodologies.High-efficient Downlink Data dissemination Protocol in LoRa WAN Networks: The LoRa WAN protocol,as the prevalent MAC layer framework in LoRa networks,necessitates device polling,transmission parameter unification,and erasure code utilization to ensure transmission reliability.These requirements precipitate long setup time,escalated device energy consumption,and substantial decoding overheads.To counteract these inefficiencies,this thesis introduces DAU(Divide and Upload),a protocol that diminishes average device energy consumption through strategic device grouping for data dissemination.DAU utilizes and reconfigures the time synchronization beacon mechanism in LoRa WAN for data packet broadcasting and device grouping,thereby curtailing device dissemination preparation times.Additionally,DAU models the data dissemination process’ s temporal and energetic overheads and proposes a greedy algorithm for device grouping,optimizing energy efficiency within specified deadlines.This thesis also presents a lightweight coding scheme predicated on device NACK(Negative Acknowledgement)information retransmission to ensure data dissemination reliability while minimizing decoding overheads.Simulations and empirical result affirm DAU’s efficacy in significantly reducing dissemination preparation durations,device energy demands,and decoding burdens in comparison to extant LoRa WAN data dissemination protocols.