In the past few years, dead-reckoning (DR) has been frequently used to estimate the trajectory of marine animals at a fine temporal scale using bio-logger devices. The uncertainty on the dive trajectory estimation results from various accumulated errors: sensors, computation, animal behavior. The accuracy of the DR estimation is generally only evaluated from the GPS starting and ending positions. Indeed, the accuracy all along the trajectory is difficult to estimate due to the difficulty of the collection of precisely-known underwater positions. Here we aim at estimating the uncertainty at a fine temporal scale using an acoustic underwater geolocation system. This work focuses on how each sensor frequency and algorithms used for the DR affect trajectory uncertainty and the global power consumption of the bio-logger. We develop a USV (Unmanned surface vehicle) equipped with an acoustic localization system and RTK GPS. This system offers us a “reference” trajectory with a resolution of 1m/100m when tracking an acoustic pinger. The DR algorithms use for orientation 3-axis accelerometer, magnetometer, and gyroscope data as well as speed data from a bio-logger. The acoustic pinger and the bio-logger are attached to a diver swimming breaststroke for 1h30min up to 15m deep. Power consumption of sensors is measured during laboratory tests. We provide the relationship between the uncertainties on positions, the sensor used, their frequencies, and the power consumption for the DR algorithms. These diagrams can be helpful for scientists designing tags to define the trade-off between the accuracy of their bio-logger trajectories and the power needed and hence the duration of deployment for a given battery capacity.