Intermittent Observation Based Time Synchronization For Underwater Acoustic Sensor Networks

With the development of science and technology,people are knowing the ocean more deeply.Underwater sensor networks have attracted significant attention from both academia and industry recently.They provided a broad range of applications such as assisted navigation,ocean sampling,environmental monitoring,target recognition and communication.Most of the protocols and applications in underwater require a common time scale.Thus time synchronization become the foundation of the underwater sensor networks.While underwater sensor networks are very different from the terrestrial networks.For example,the propagation delay cannot be negligible,the capacity of battery is constrained.These factors lead to the result that time sychronization algorithms in terrestrial cannot be used in underwater sensor networks directly.Thus we need to research new energy efficient underwater sychronization algorithms to solve the challenges in underwater.On this basis,this paper puts forward an energy-efficient underwater time synchronization algorithm and underwater clock synchronization algorithm which based on intermittent observation.The second algorithm is proposed on the basis of the former algorithm.The first algorithm can be divided into two steps.On the first step,anchor nodes send packets to synchronize sensor nodes,then estimating the clock skew by using the method of least squares.While for the second step,anchor nodes send their own position and time information to synchronize sensor nodes.Sensor nodes estimate the offset through these information.The sensor nodes only need to receive packets in the whole process,therefore the energy consumption is quite low.The sensor nodes adopt the duty-cycle mode in the second algorithm.In order to acquire more accurate results and avoid frequent resynchronization,we use kalman filter based on intermittent observation to estimate the skew and offset.Then the skew and offset will be updated when we do not have any observed value.Thus the energy consumption of the time synchronization algorithm can be reduced at the same time.At the same time,we use simulation results to demonstrate energy efficiency and high precision of the algorithms.The main content and innovation of the paper is list below:(1)We propose a new energy efficient underwater time synchronization algorithm.The algorithm performs twice linear least squares method to figure out the skew and offset,which makes the estimation results more accurate.It also avoids the need for frequent resynchronization.The sensor nodes only need to receive but not send packets,which also makes the algorithm energy efficient.(2)On the basis of the first algorithm,we propose an intermittent observation time synchronization algorithm.The sensor nodes in the algorithm are in duty-cycle mode.The intermittent observation information will be transfer to kalman filter to update the skew and offset of the clock.The duty-cycle of snesor nodes and the intermittent observation scheme make the algorithm more precise and energy efficient.