Kalman filter based ranging and clock synchronization for ultra wide band sensor networks

Abstract

This Thesis presents the design, implementation, and validation of a Kalman filterbased range estimation technique to precisely calculate the inter-node ranges of Ultra Wide Band (UWB) modules. In addition to that the development and validation of an improved global clock synchronization framework is presented. Noise characteristics of relative time measurements of a stationary UWB anchor pair are first analyzed using an Allan deviation plot. To track the propagation of the imprecise clocks on low cost UWB transceiver platforms, Kalman filters are used in between every anchor pair. These filters track the variation of a remote anchor’s hardware clock relative to it’s own hardware clock, while estimating the time of flight between the anchor pair as a filter state. While adhering to a simple round robin transmission schedule, both inbound and outbound message timestamp data are used to update the filter. These measurements have made the time of flight observable in the chosen state space. A faster relative clock filter convergence has been achieved with the inclusion of the clock offset ratio as a measurement additional to the timestamps. Furthermore, a modified gradient clock synchronization algorithm is used to achieve global clock synchronization throughout the network. A correction term is used in the gradient clock synchronization algorithm to enforce the global clock rate to converge at the average of individual clock rates while achieving asymptotic stability in clock rate error state. Stability of the original and modified methods for time invariant hardware clocks are compared using eigenvalue tests. Experiments are conducted to evaluate synchronization and ranging accuracy of the proposed range estimation approach.