Interval Sampling Tasks Oriented Data Collection In Wireless Sensor Networks

Wireless Sensor Networks(WSNs)are multi-hops ad hoc networks comprising of hundreds of wireless sensor nodes.Such nodes collect environment data by performing sampling tasks via sensor components.Those envirnment data is ususlly used to feed applications for data analysis and decision-making,espeically for surveillance applications.Since data sharing may exist amongst multiple sampling tasks,the amount of sampled data is decreased significantly,thus prolonging the lifetime of both nodes and the whole network.Generally,Such sampling tasks can be divided into two categories: discrete point sampling tasks and continuous interval sampling tasks.The former requires to sample data once;while the latter requires to sample data for a time interval continuously.Conventional studies usually focus on the discrete point sampling tasks.However,the continuous interval smapling tasks are emerging in more and more applications.Therefore,this thesis focuses on continuous interval sampling tasks and conducts research studies from four aspects.Firstly,we study the problem of tasks scheduling on a single node.Such a problem is proved to be NP complete.A 2-factor approximate algorithm is proposed.Experiment results verify the effectiveness of our scheduling method.Secondly,we study the problem of task allocation in a k-coverage and r-redundant wireless sensor network.Such a problem is proved to be NP hard.Similarly,a 2-factor approximate algorithm is proposed.Empirical studies verified the effectiveness of our allocation method.Thirdly,the problem of load balance for nodes in a network has been studied.Two approximate algorithms have been provided and the simulation results confirm that our solution is effective.Finally,the research studies have been applied in a military case named urban combat,and the military value of our research thus achieves verification.