Concurrent scheduling in the Active Bat location system

This paper looks at the scalability problems inherent in the Active Bat system: an outside-in ultrasonic location system. Such systems are typically associated with higher positioning accuracies and longer tag battery lifetimes relative to comparable inside-out systems. However, they are often criticised for a lack of scalability since multiple tags must be addressed in series to prevent interference within the positioning medium. Multiple radio zones have previously been used to address this problem. We show that this approach is sub-optimal both in terms of system complexity and in terms of the location update rates that can be achieved. We propose a general solution to the scalability problem based on the computation of dynamic bounding regions within which corresponding tags are almost certainly located. These regions are used to determine when tags are sufficiently well separated to be queried concurrently without the risk of interference. We test this approach using real data, concluding that bounding regions permit more concurrency than is possible using multiple radio zones whilst at the same time reducing the complexity of the system.