Mobile Sensor Network Navigation Using Gaussian Processes With Truncated Observations

Xu, Yunfei; Choi, Jongeun; Oh, Songhwai

doi:10.1109/tro.2011.2162766

Cited by 96 publications

(74 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The true values for the parameters used in simulating the Gaussian process are given by (β, σ 2 f , σ s , σ t ) = (20, 10,2,8). Notice that the mean function is assumed to be an unknown random variable, i.e., the dimension of the regression coefficient β is 1.…”

Section: Simulation Resultsmentioning

confidence: 99%

“…To overcome this drawback while maintaining the Bayesian framework, we propose to use subsets of all observations y 1:t . However, instead of using truncated local observations only as in [10], Bayesian inference will be drawn based on two sets of observations: First, a set of local observations near target pointsỹ which will improve the quality of the prediction, and a second cumulative set of observationsȳ which will minimize the uncertainty in the estimated parameters. Taken together, they improve the quality of prediction as the number of observations increases.…”

Section: A a Scalable Bayesian Prediction Algorithmmentioning

confidence: 99%

“…In [10], the authors analyzed the conditions under which near-optimal prediction can be achieved using only truncated observations. This motivates the usage of sparse Gaussian process proposed in [11].…”

mentioning

confidence: 99%

See 2 more Smart Citations

Sequential Bayesian Prediction and Adaptive Sampling Algorithms for Mobile Sensor Networks

Choi

Dass

et al. 2012

IEEE Trans. Automat. Contr.

View full text Add to dashboard Cite

In this paper, we formulate a fully Bayesian approach for spatio-temporal Gaussian process regression such that multifactorial effects of observations, measurement noise and prior distributions are all correctly incorporated in the predictive distribution. Using discrete prior probabilities and compactly supported kernels, we provide a way to design sequential Bayesian prediction algorithms in which exact predictive distributions can be computed in constant time as the number of observations increases.For a special case, a distributed implementation of sequential Bayesian prediction algorithms has been proposed for mobile sensor networks. An adaptive sampling strategy for mobile sensors, using the maximum a posteriori (MAP) estimation, has been proposed to minimize the prediction error variances.Simulation results illustrate the practical usefulness of the proposed theoretically-correct algorithms.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

Section: A a Scalable Bayesian Prediction Algorithmmentioning

confidence: 99%

See 1 more Smart Citation

Sequential Bayesian Prediction and Adaptive Sampling Algorithms for Mobile Sensor Networks

Choi

Dass

et al. 2012

IEEE Trans. Automat. Contr.

View full text Add to dashboard Cite

show abstract

“…Sensor measurements can be corrupted by noises from the sensor circuitry and environment [Xu et al 2011]. The reading at time-space coordinates ( p, t), denoted by R( p, t), is given by R( p, t) = Z( p, t) + W, where W is a zero-mean Gaussian noise with variance of σ 2 w .…”

Section: Physical Fieldmentioning

confidence: 99%

Spatiotemporal Aquatic Field Reconstruction Using Robotic Sensor Swarm

Wang

Tan

Xing

et al. 2012

2012 IEEE 33rd Real-Time Systems Symposium

View full text Add to dashboard Cite

Monitoring important aquatic processes like harmful algal blooms is of increasing interest to public health, ecosystem sustainability, marine biology, and aquaculture industry. This article presents a novel approach to spatiotemporal aquatic field reconstruction using inexpensive, low-power mobile sensing platforms called robotic fish. Robotic fish networks are a typical example of cyber-physical systems where the design of cyber components (sensing, communication, and information processing) must account for inherent physical dynamics of the robots and the aquatic environment. Our approach features a rendezvous-based mobility control scheme where robotic fish collaborate in the form of a swarm to sense the aquatic environment in a series of carefully chosen rendezvous regions. We design a novel feedback control algorithm that maintains the desirable level of wireless connectivity for a sensor swarm in the presence of significant environment and system dynamics. Information-theoretic analysis is used to guide the selection of rendezvous regions so that the spatiotemporal field reconstruction accuracy is maximized subject to the limited sensor mobility. The effectiveness of our approach is validated via implementation on sensor hardware and extensive simulations based on real data traces of water surface temperature field and on-water ZigBee wireless communication.

show abstract

“…Regression analysis for Gaussian processes requires growing computational complexity since the size of the covariance matrix increases as the number of observations increases. This problem in the context of the mobile sensor networks has been tackled in different directions [15], [16]. Computational complexity of a full Bayesian prediction algorithm for spatio-temporal Gaussian processes with unknown covariance functions grows in a prohibitively fast rate as the observation number increases due to the MCMC method.…”

Section: Introductionmentioning

confidence: 99%

Bayesian prediction and adaptive sampling algorithms for mobile sensor networks

Choi

Dass

et al. 2011

Proceedings of the 2011 American Control Conference

View full text Add to dashboard Cite

Abstract-In this paper, we formulate a full Bayesian approach for spatio-temporal Gaussian process regression under practical conditions such as measurement noise and unknown hyperparmeters (particularly, the bandwidths). Thus, multifactorial effects of observations, measurement noise and prior distributions of hyperparameters are all correctly incorporated in the computed predictive distribution. Using discrete prior probabilities and compactly supported kernels, we provide a way to design sequential Bayesian prediction algorithms that can be computed (without using the Gibbs sampler) in constant time as the number of observations increases. Both centralized and distributed sequential Bayesian prediction algorithms have been proposed for mobile sensor networks. An adaptive sampling strategy for mobile sensors, using the maximum a posteriori (MAP) estimation, has been proposed to minimize the prediction error variances. Simulation results illustrate the effectiveness of the proposed algorithms.

show abstract

Mobile Sensor Network Navigation Using Gaussian Processes With Truncated Observations

Cited by 96 publications

References 20 publications

Sequential Bayesian Prediction and Adaptive Sampling Algorithms for Mobile Sensor Networks

Sequential Bayesian Prediction and Adaptive Sampling Algorithms for Mobile Sensor Networks

Spatiotemporal Aquatic Field Reconstruction Using Robotic Sensor Swarm

Bayesian prediction and adaptive sampling algorithms for mobile sensor networks

Contact Info

Product

Resources

About