Key Technologies Research Of Wireless Sensor Network For Event-driven Applications

Event-driven application is one of the core application scenarios of wireless sensor networks,which play an important role in industrial control,structural health,ecological environment and other fields.This dissertation conducts research on wireless sensor network for event-driven applications.Specifically,we focus on railway bridge structural health monitoring.The key technologies include high performance structural health monitoring sensor,miniaturized and integrated sensor,networked embedded system,integrated circuit,event behavior modeling and so on.From the perspective of networked embedded system,we conduct in-depth research.From the perspective of "event attributes" and "network traffic",the characteristics of railway bridge structural health monitoring sensor network are as follows:(i)The monitored train events are random,rare,and urgent;(ii)High traffic is injected in a short period of time when the trains passing,and go back to idle for a long time.(iii)In addition,the evaluation of bridge health requires multi-type sensors collaboration.The above features bring three technology requirements for railway bridge structural health monitoring: low-power environment sampling,low-power wireless communication,and highly reliable data transmission.To address theses challenges,our main work and contributions are as follows:(i)An innovative sensor architecture called Eco Sense is proposed.Eco Sense provides a hardware-based on-demand sensing mechanism that effectively eliminates the energy waste caused by a sensor working in sleep or standby mode.When desired events are not present,Eco Sense node is completely turned off to save energy.When desired events occur,an on-demand connection module will harvest energy from the events and reactivate the sensor.We implemented Eco Sense node,and evaluated their performance in realworld experiments.Experimental results show that the Eco Sense technique is able to immediately detect events,and offers reasonable reaction distances.(ii)We present ECOVIBE,an on-demand sensing system that automatically turns on itself when a train passes on the bridge and adaptively powers itself off after finishing all tasks.After that,it goes to inactive state with near-zero power dissipation.ECOVIBE achieves these by: Firstly,a fully passive event detection circuit is proposed to continuously detect passing trains without consuming any energy.Secondly,combining train-induced vibration energy harvesting with transistor-based load switch,a tiny energy is able to active ECOVIBE for a long time.Thirdly,a passive adaptive off control circuit is introduced to quickly switch off ECOVIBE without consuming energy.We present the prototype implementation of the proposed system using commercially available components and evaluate its performance in real-world scenarios.Our results show that ECOVIBE is effective in railway bridge health monitoring applications.(iii)Distributed congestion mitigation and medium access is presented.A queue length based contention window adjustment approach allows congested nodes having high priority to access the channel.Then,a batch transmission scheme is introduced to drain accumulated packets rapidly.Finally,a random preamble mechanism is designed,so that congested nodes have more chances to get channel when other transmission is happening.Through the experiments on Telos B nodes test-bed using Tiny OS,evaluation results show that our proposed schemes can mitigate congestion efficiently and have better performance than traditional approaches.(iv)We firstly establish models based on combinatorics theory to analyze the performance of dynamic multichannel medium access.Then,a Cross-layer Cooperative Multichannel Medium Access(CCMMA)is proposed to effectively avoid channel contentions by enabling neighboring devices to communicate on orthogonal channels.The CCMMA also introduces a routing-enhanced mechanism that enables relaying nodes to wake up intelligently if there is incoming traffic,that successfully mitigates delay and queue overflow problems caused by low-power operation of sensor nodes.The performance of CCMMA is evaluated through extensive simulations.The results show that it provides significant improvement in terms of quality of service over existing solutions.