A twenty-two-state extended Kalman filter for Doppler- and rangeaided underwater navigation

A twenty-two-state extended Kalman filter for Doppler- and rangeaided underwater navigation Chloe Jane Tully https://orcid.org/0009-0007-5661-7332 Tamworth, New South Wales, Australia 6 October 2026 Abstract Global navigation satellite signals are unavailable to a submerged vehicle. This paper formulates a locallevel extended Kalman filter that fuses inertial specific force and angular rate with bottom-lock and watertrack Doppler velocity, hydrostatic depth, acoustic position, magnetic heading, and scalar range to surveyed beacons or cooperating vehicles. The state comprises position, ground velocity and attitude in the northeast-down frame, gyroscope and accelerometer biases, a Doppler scale factor and lever arm, and a watercurrent vector. Acoustic and range updates are evaluated at the emission epoch rather than at packet ingest, and acoustic position innovations are subjected to a chi-square gate. Two simulations are reported. A singlevehicle survey exercises mode transitions through a bottom-lock outage. A two-vehicle, two-beacon scenario generates ranges from an independent truth trajectory; both filters, initialised 4 m from truth, finish a 40 s cruise within 0.02 m of that trajectory. The study is a controlled simulation. It is not a sea trial, and it does not establish observability of current under straight-line motion. Keywords: autonomous underwater vehicle; extended Kalman filter; Doppler velocity log; long-baseline ranging; cooperative navigation; acoustic delay