Target tracking with packet delays and losses - QoI amid latencies and missing data

In this paper, we investigate how packet delays and losses affect the quality of target tracking. Specifically, we use Bayesian information of the posterior distribution of target locations to quantify the quality of target tracking and investigate how network quality and measurement quality affect the value of Bayesian information. We show that improving measurement quality provides diminishing gain on tracking quality, while the gain from improving network quality does not diminish. We obtain the condition under which a user obtains information gain on the target location from a tracking process. We further use Bayesian information as the metric for the gateway to select the sensor for taking measurements and determine the measurement time to control the tracking quality.