Background
Acoustic positioning telemetry allows to collect large amounts of data on the movement of aquatic animals by use of autonomous receiver stations. Essential in this process is the conversion from raw signal detections to reliable positions. A new advancement in the domain is Yet Another Positioning Solver (YAPS), which combines the detection data on the receivers with a model of animal movement. This transparent, flexible and on-line available positioning algorithm overcomes problems related to traditional point-by-point positioning and filtering techniques. However, its performance has only been tested on data from one telemetry system, providing transmitters with stable burst interval. To investigate the performance of YAPS on different system parameters and settings, we conducted a simulation study.
Results
This paper discusses the effect of varying burst types, burst intervals, number of observations, reflectivity levels of the environment, levels of out-of-array positioning and temporal receiver resolution on positioning accuracy. We found that a receiver resolution better than 1 ms is required for accurate fine-scale positioning. The positioning accuracy of YAPS increases with decreasing burst intervals, especially when the number of observations is low, when reflectivity is high or when information out-of-array is used. However, when the burst interval is stable, large burst intervals (in the order of 1 to 2 min) can be chosen without strongly hampering the accuracy (although this results in information loss). With random burst intervals, the accuracy can be much improved if the random sequence is known.
Conclusions
As it turns out, the key to accurate positioning is the burst type. If a stable burst interval is not possible, the availability of the random sequence improves the positioning of random burst interval data significantly.
