Gaussian Process Models for Censored Sensor Readings

Sensor data models are key components of the design and testing of sensor network applications. In addition to their utility in validation of applications and network services, they provide a theoretical basis for the design of algorithms for efficient sampling, compression and exfiltration of the sensor data. In this paper we introduce a novel modeling technique for constructing probabilistic models for censored sensor readings. The proposed technique is an extension of the Gaussian process regression and applies to continuous valued readings subject to censoring. We treat the censored variable as a mixture of binary and a normal random variable. The Gaussian process framework provides a natural way of integrating information from both types of observations to estimate the parameters of the underlying random process. We illustrate the performance of the proposed technique in modeling wireless propagation between nodes of a wireless sensor network. The model can capture the anisotropic nature of the propagation characteristics and utilizes the implicit information from the packet reception failures.