Novel Dynamic Paradigms for Wave Sensing Inspired by Bat Biosonar

Virginia Polytechnic Institute And State University, Blacksburg VA

Investigators

Abstract

Current technical sonar and radar technology follows an approach that appears to be fundamentally different from what can be observed in bat biosonar: Whereas engineered systems rely on large numbers of emitter and receiver elements that each have rather simple characteristics, bat biosonar operates with a very small number of intricate emitters and receivers. Bats emit their ultrasonic pulses and receive the echoes using baffle shapes that can be compared to megaphones and horn antennas to a first approximation. However, these shapes are not only geometrically much more complicated than their technical peers, they can also have a unique dynamic dimension in that they change their shapes on time scales that are similar to the duration of the animals' ultrasonic pulses. At the same time, bat biosonar appears to be far superior to engineered systems in dealing with structure-rich natural environments. Hence, developing an understanding of the role that the dynamic dimension plays in bat biosonar could lead to novel dynamic sensing paradigms that could improve the performance of sonar, radar, and related technical sensing modalities. This research will reproduce and investigate dynamic features seen in biosonar the biosonar system of certain bat species in a biomimetic prototype sonar. The principal underlying hypotheses to be tested is that the deformations of these baffle shapes add a dynamic dimension to this biological sensing system that could be used to (i) enlarge the system's general coding capacity for sensory information, (ii) enhance the encoding of certain salient features, (iii) adapt the system to different sensing scenarios. If this is the case, the unusual dynamic dimension of bat biosonar could be a key factor behind the superior ability of bats to meet the sensory needs for navigation in complex natural environments based on a very parsimonious sensory input. This hypothesis will be investigated by constructing a biomimetic sonar system that will employ baffle shape for emission as well as reception that can change their shapes in synchrony with the respective diffraction processes. The biomimetic sensory system will be used to investigate the dynamic encoding of sensory information in natural biosonar sensing tasks. The results of these experiments will be analyzed using numerical simulations and information-theoretic methods to deal with the random nature of natural biosonar scenes and the resulting echo signals.