Research On Data Transmission Scheduling In Wireiess Sensor Networks

Internet of Things (IoT) is considered to be the future key information infrastructure, and rapidly become the competitive focus in recent years. The objective of IoT is to realize the wide range of interconnection and the in-depth integration of the cyber space and physical world. As the peripheral networks of IoT, wireless sensor networks can provide IoT with the means of actively sensing the physical world at large spatial and temporal scales. On the one hand, the node energy in wireless sensor networks is limited, and the data transmission is the main source of the energy consumption. in wireless sensor nodes. On the other hand, the applications of IoT have the real-time requirement, and data transmission scheduling strategy is one of the main factors that influence the transmission delay peroformance. Therefore, it is of practical significance for studying the data transmission strategy which can reduce the transmission energy consumption and the transmission delay in wireless sensor networks.In this thesis, we investigate the data transmission scheduling problem that aims to reduce the transmission energy consumption and the transmission delay in wireless sensor networks. The basic idea is as follows. First, we build a design framework of the data transmission scheduling for the data collection in wireless sensor networks. Based on this framework, we divide the network lifetime into the multiple scheduling cycles according to the sampling period of wireless sensor nodes and the transmission delay constraints, and dynamically adjust the scheduling cycles according to the current traffic loads of wireless sensor networks. Then, we further determine the transmission schedule for the network within one scheduling cycle according to the different design goals that include data aggregation, distributed source coding and collection of heterogeneous correlative data. The main contributions of this thesis are as follows.(1) For collecting data in wireless sensor networks, we build a design framework of the data transmission scheduling. This framework represents the relationship between the time constriants and the transmission scheduling in wireless sensor networks, and can transform the complex scheduling optimization problem to a network service curve construction problem with low computing cost. We propose two basic transmission scheduling algorithms for data collection:Local Uniform Rate Scheduling (LURS) algorithm and Improved Local Uniform Rate Scheduling (I-LURS) algorithm. LURS algorithm adaptively adjusts the scheduling cycles according to the current traffic loads of wireless sensor networks, and can obtain significant energy gain compared with the traditional scheduling algorithms. Based on the operations of LURS algorithm, I-LURS algorithm supplements a constrained optimization operation within one scheduling cycle, and can save more transmission energy consumption at the cost of increasing the computational complexity. (2) For collecting the aggregated data in wireless sensor networks, we propose an in-network aggregation transmission scheduling strategy. This scheduling strategy includes two coupled parts:The in-network aggregation computing method and the aggregation scheduling algorithm. The proposed in-network aggregation computing method not only optimally fuses the data obtained from the different sensor nodes but also predicts the upper neighber sensors’data which cannot be aggregated to the sink before deadlines. The proposed aggregation scheduling algorithm can aggregate as much sensing information as possible from the network to the sink within delay constraints and reduce the state estimation error significantly.(3) For collecting the data encoded by distributed source coder in wireless sensor networks, we propose a data transmission scheduling strategy based on in-network progressive coding. By using the proposed scheduling strategy, the relay nodes perform the in-network puncturing operation on the syndorm bits received from the upper neighber sensor nodes, and the distributed coding rate can be controlled level by level. Then, the decoder can determine the schedule of the additional decoding data based on the level number of the relay nodes, and receive the additional decoding data from the relay nodes instead of the remote encoder node. Therefore, the decoding waiting delay is reduced.(4) For collecting the heterogeneous correlative data in the network system with heterogeneous data sensing capability, we first propose an adaptive wake-up transmission scheduling strategy for the wireless sensor nodes. This stragety guarantees the accurate data collection while reducing the transmission energy consumption of the portable wireless sensor node. Then, we propose a video transmission scheduling strategy based on the estimated trajectory of the suspect target in the heterogeneous sensor network. This video scheduling strategy utilizes the binary sensing data collected by the wireless sensor nodes to estimate the trajectory of the suspect target, and can reasonably schedule the video sequences based on the spatial and temporal correlation of the binary sensing data and the video data. Consequently, it reduces the delay of identifying the suspect targe.In summary, we first propose a design framework of the data transmission scheduling in wireless sensor networks, and then study the data transmission scheduling problem from the viewpoints of data aggregation, distributed source coding and collection of heterogeneous correlative data. We propose a series of strategies of the data transmission scheduling for reducing the transmission energy consumption and delay. The proposed scheduling strategies can significantly improve the performance of the transmission energy consumption and delay in wireless sensor networks.