We present a novel paradigm of sensor placement concerning data precision and estimation.Multiple abstract sensors are used to measure a quantity of a moving target in the scenario of a wireless sensor network.These s...We present a novel paradigm of sensor placement concerning data precision and estimation.Multiple abstract sensors are used to measure a quantity of a moving target in the scenario of a wireless sensor network.These sensors can cooperate with each other to obtain a precise estimate of the quantity in a real-time manner.We consider a problem on planning a minimum-cost scheme of sensor placement with desired data precision and resource consumption.Measured data is modeled as a Gaussian random variable with a changeable variance.A gird model is used to approximate the problem.We solve the problem with a heuristic algorithm using branch-and-bound method and tabu search.Our experiments demonstrate that the algorithm is correct in a certain tolerance,and it is also efficient and scalable.展开更多
基金Supported of Project of Fok Ying Tong Education Foundation(No.104030)Supported of Key Project of National Natural Science of Foundation of China(No.70531020)+2 种基金Supported of Project of New Century Excellent Talent(No.NCET-06-0382)Supported of Key Project of Education Ministry of China(No.306023)Supported of Project of Doctoral Education(20070247075)
文摘We present a novel paradigm of sensor placement concerning data precision and estimation.Multiple abstract sensors are used to measure a quantity of a moving target in the scenario of a wireless sensor network.These sensors can cooperate with each other to obtain a precise estimate of the quantity in a real-time manner.We consider a problem on planning a minimum-cost scheme of sensor placement with desired data precision and resource consumption.Measured data is modeled as a Gaussian random variable with a changeable variance.A gird model is used to approximate the problem.We solve the problem with a heuristic algorithm using branch-and-bound method and tabu search.Our experiments demonstrate that the algorithm is correct in a certain tolerance,and it is also efficient and scalable.
