Exploiting relative measurements for pose and time calibration of wireless sensor networks

Pose calibration and time synchronization are among the fundamental calibration problems in wireless sensor networks. Previous research has reveals strong relation between the formulations of these problems. When sensors have identical orientation, the solution to localization problems recovers the position of sensors based on relative geometrical translation measurements. When sensorspsila clock drift parameters are all unity, time synchronization reconstructs all local clock systems based on the relative clock offset measurement. In a more general scenario, the orientation matrix and the clock drift coefficient factors can be viewed as the augmented coefficients in localization and synchronization problems, respectively. Therefore the parameters to be calibrated become position, orientation, offset and drift. Sensors thus no longer have the same orientation and the same clock drift parameter. In this paper we consider the above parameters that are assumed time invariant. Suppose sensors are able to measure relative bearing and range, and exchange time-stamped information. We introduce a unified formulation for such generalized calibration problems and imply a corresponding distributed algorithm in wireless sensor networks.