Kalman Filtering With Intermittent Observations

Abstract

Abstract-Motivated by navigation and tracking applications within sensor networks, we consider the problem of performing Kalman filtering with intermittent observations. When data travel along unreliable communication channels in a large, wireless, multihop sensor network, the effect of communication delays and loss of information in the control loop cannot be neglected. We address this problem starting from the discrete Kalman filtering formulation, and modeling the arrival of the observation as a random process. We study the statistical convergence properties of the estimation error covariance, showing the existence of a critical value for the arrival rate of the observations, beyond which a transition to an unbounded state error covariance occurs. We also give upper and lower bounds on this expected state error covariance.

Extended Kalman filter-based mobile robot localization with intermittent measurements

“…The contents also agree with what previous literatures have actually reported in Huang and Dissayanake (2007), Hwan Hur and Hyo-Sung (2013), Jun et al (2012), Mo and Sinopoli (2008), Payeur (2008), Plarre and Bullo (2009) and Sinopoli et al (2004). Using EKF measurement innovation and Jacobian transformation of measurement matrix, whenever measurement data are partially missing, the lower and upper limits can be determined via measurement innovation error.…”

“…Now let us analyse the effect of the intermittent measurement to the robot localization. The expectation of state covariance at time k + 1 is demonstrated by Sinopoli et al (2004), Mo and Sinopoli (2008) and Kluge et al (2010) …”

“…Under the probability of measurement data arrival (Sinopoli et al, 2004), the updated state covariance P + k+1 is given by…”

“…This is due as the updated state covariance defines the amount of uncertainties existing in the system. In this perspective, Sinopoli et al (2004) proposed that even if the problem happened, the state covariance should have its upper and lower bounds and converge to its critical value. Similar results have been suggested in different expression by Huang and Dissayanake (2007) who consistently proved that the state covariance is converging to a steady state covariance.…”

“…Robot localization is also a partially observable problem (Vidal-Calleja, Andrade-Cetto, & Sanfeliu, 2004); and even though we can control the initial input, the output is still unpredictable. Hence, this problem is more difficult in a nonlinear system than in a linear system (Censi, 2008;Jun et al, 2012;Mo & Sinopoli, 2008;Plarre & Bullo, 2009;Sinopoli et al, 2004).…”