Abstract
Cycle-slip detection is mandatory prior to estimating positions based on carrier-phase observations. Recently, inertial sensors are integrated to detect cycle-slips regardless of sizes and combinations of cycle-slips in different frequencies. However, since inertial sensors contain errors such as random noise and bias, performance of the cycle-slip detection depends on the sensor performance. In general, there is a trade-off relationship between cost and performance of sensors, and we need to select appropriate sensors to achieve required cycle-slip detection performance. Therefore, we need a standard to select appropriate sensors. This paper introduces a procedure to predict allowable sensor error limit for gyroscope and odometer aided cycle-slip detection using theoretical formulation. By using error models of sensors, an error equation is set to predict error contributions from the sensors contained in the monitoring value that is used for cycle-slip detection. Simulation data is used to evaluate derived error equation and predict the error limit of sensors achieving objective performance. As a result, allowable sensor error region is predicted to detect 1 cycle-slip as a minimum detectable cycle-slip. This approach can be extended to accelerometers, and then it can be applied not only to vehicles on land, but also to those in aerospace.
