Design And Implementation Of Low-power Heterogeneous Convergence TDMA Protocol For Wireless Sensor Networks

As the rapid development of Micro Electro Mechanical Systems (MEMS), Computer Technology, and Communications Technology, a new kind of network system called Wireless Sensor Networks (WSNs) emerged. Taking advantages of the low cost and energy efficiency of single sensor node, and the simplicity of deploying of the sensor networks, WSNs is widely used in many fields such as military surveillance, business intelligence applications and environmental studies, etc. This thesis takes Dunhuang digital heritage protection as background, and studies how to implement the image data collection and transmission wirelessly under the large-scale scenes, mainly focuses on low energy cost convergence transmission of the collected image information under the circumstance of large number of sensor nodes and heavy workload. As the effective image acquisition of the cultural heritage sites must ensure highly synchronization, and wiring of the field is not realistic, the only method is to adopt wireless control and wireless data upload. Some of the current wireless convergence transmission protocols of the sensor networks are content-based; they are mainly facing the applications such as micro-environment monitoring, disaster warning and other small workload applications, and cannot meet the requirement of this large data wireless collection and transmission work. Some other protocols are schedule-based; they are almost staying in the study and simulation level, rare for the application of large-workload convergence transmission. Because of the large workload and large number of sensor nodes, how to effectively avoid the failure caused by communication collisions, and keep the feather of low energy cost of sensor nodes, besides, how to effectively manage and schedule the network topology, are the research goal and difficulty of this thesis.Firstly, this thesis analyzes and reviews the domestic and international research status of MAC layer protocols on wireless convergence transmission and topology management. Since the design and method of present protocols are not fit to meet this large workload and low energy cost requirement, by adopting some good ideas, this thesis proposes a new MAC protocol based on TDM A. By collecting the workload of each node and allocating the time slice based on the workload collected for each sensor node, this protocol guarantees that every node periodically wake-up to work in the time slice, and when the time slice runs out it goes back to sleep state to save energy and keeps low power consumption as much as possible to extend the lifecycle of the node.Because of the incompletely reliable communications of the sensor nodes, and some other interference of external environment, the sending and receiving of packets may be failure, besides there may have new nodes need to deploy, this will all cause the change of the current network topology. This thesis proposes a new type of dynamic adjustment of network topology that will take two or three cycles to finish, and restore the network back to the cyclical time-scheduled data transmission working state.Finally, according to the proposed protocol design and key algorithms, there implements two different sets of applications for Sink node and Sensor node on a specific hardware platform. After large times of testing and improving work with different network parameters, there gets a prototype system that supports dozens of sensor nodes via several hops run smoothly. Moreover, by the comparative experiments against CSMA on specific hardware platforms, the results show that the proposed TDMA method has good improvements both in power consumption and data transmission efficiency, and for the case when the workload increases the advantages are more obvious.