Research On Nodes Mobility In Delay Tolerant Wireless Sensor Networks

The wireless sensor network (WSN) is in the bottom layer of the Internet of Things (IoTs). It gathers the sensing data from and disseminates the control instructions to the sensors. Therefore, data collection and dissemination efficiency is crucial to the performance of IoTs.Sensors in WSN are powered by batteries, the energy of which is limited. When the energy is consumed up, the sensor node dies out and energy holes come up in the network. Besides, the complexity of communication environment and the variability of wireless channel with time, lead to the intermittent connectivity of WSN, or the delay tolerant sensor network (DTSN). In recent research, a mobile element, called data mule (DM), can be dispatched to visit sensor nodes and to collect and deliver distributed sensing data via wireless communication, improving the connectivity and reliability of DTSN.In this thesis, the network architecture of DTSN and the mobility scheme of DM in individual architectures is analyzed. Based on this, we focus on the motion control of the DM, which decides the main performance of DTSN. A unified model is constructed to describe the motion control problem, and the problem can be divided into path selection, speed control and job scheduling subproblems accordingly.Firstly, path selection, the primary subproblem, is studied. As each path is optimal if every point in this path is in the border of convex hull, the convex hull-based method is presented to select an optimal path. Simulation results show that this method utilizes the communication capability of sensor nodes. Compared with other path selection strategies, the paths obtained by the proposed method are smoother with less break points. The data collection time is80-85%of that obtained by previous methods, and is about1.3times of the lower bound minimal value, demonstrating the effectiveness of the convex hull-basetd method.Secondly, motion planing problem is studied, which combines both path selection and speed control. To minimize the data delivery latency (DDL), the two-phase method is proposed. In the first phase of the method, the convex hull structure is utilized to depict a path skeleton, and in the second phase, speed control is considered to regulate the path skeleton. Simulation experiments show that the proposed method exhibits an average of10%to20%decrease in the latency of data collection. The results suggest that the data collection efficiency can be improved by the proposed method, and the DM approach is feasible and effective for data delivery in WSN-based applications.Finally, wireless communication is considered into motion planing to improve the data transfer quality and decrease error rate. The workspace of DM is discretized according to the kinetic constraints, and the Grid search algorithm is proposed to find a feasible path. Besides, the wireless channel quality is quantified and is used as heuristic cost of the sampling method. Simulation results show that the DDL obtained by Grid search algorithm is smaller, and the communication quality obtained by the sampling method is higher.