LoRa-Based Optimized Routing Mechanism For Perception In Colligate Alure

Colligate alure construction is one of the important practices of "Xin Jijian"(new infrastructure)in China.LoRa,with its low cost,low power consumption and easy deployment,is the first choice to realize the perception service.Building LoRa multi-hop networks in underground corridors can achieve full coverage without adding gateways.There is a lack of link mechanism applicable to narrow space,so studying the link optimization mechanism based on LoRa is meaningful to realize the digitalization of colligate alure and ensure the safe operation and sustainable development of smart grid.The new urban colligate alure is a highly integrated infrastructure,and all kinds of pipelines share the same underground space,forming the status quo of differentiated services coexistence.To address this situation,this thesis proposes a link optimization mechanism for service quality.Firstly,the link is modeled as a function of SF and hop count in terms of packet delivery rate and data rate,and then the SF and hop count of the uplink are assigned to the sensors corresponding to the services according to the demand of each type of services.This mechanism divides the original single-root uplink data acquisition network into six non-intersecting subnetworks with the same root,and assigns a spreading factor(SF)to each subnetwork so that multiple SFs are transmitted in concurrent to improve the throughput.A capacity estimation algorithm is proposed to preset a reasonable capacity for each subnet according to the distribution of services and considering the transmission time balance of each subnet.Finally,the proposed tree-based clustering algorithm migrates each sensor to the corresponding subnet according to service type,data volume,and priority of the service.Simulation experiments verify the support of different demand and throughput enhancement of this mechanism in narrow spaces and make comparisons with square spaces.To address the problem of uneven energy consumption affecting the overall life cycle of the network,this thesis proposes a segmented variable hop mechanism to reduce the energy consumption of critical sensors(the closest sensor to gateway in next hop)by choosing a combination of segment hops instead of next hops.Firstly,all sensors are segmented according to the distance to the gateway,duty cycle restriction is considered,parent-child relationship model,data aggregation model and energy consumption model are established.The dilemma of finding the optimal combination with energy consumption limitation is modeled as a finite horizon Multi-armed Bandit(MAB)problem,and an overlap-basedε-greedy segment-hop selection algorithm is proposed instead of violent attempts.And then two different matching algorithms are used to establish its point-to-point connection with the parent segment for each sensor in the key segment and other segments,respectively.In the simulation experiments,two sensor deployment models are established with uniform distribution in narrow space and exponential growth distribution,and it is demonstrated that the present mechanism outperforms its counterpart in terms of energy consumption in narrow space and makes a comparison with the exponential growth model in square space.