Bats represent the second largest mammalian order with more than 1100 described species worldwide. They show an enormous ecological diversity providing crucial ecosystem services such as insect pest control, pollination or seed dispersal. Unfortunately, direct tracking of bats in their natural environment is very limited due to their nocturnal activity, fast active flight and their small body size. Therefore, we know little about individual's behaviour outside their roosts and if so, only for short time intervals. However, tracking of individuals is essential for addressing numerous questions in the fields of ecology, behaviour, physiology and conservation. The development of an embedded communicating sensor systems will overcome major methodological constraints in high-resolution tracking of fast-moving small animals in their natural habitat. The main focus of the subproject is the analysis of habitat use and feeding ecology of the mouse-eared bat Myotis myotis.
This sub project's main goal is the provision of a flexible system software infrastructure, called ARTE (Adaptive Run-Time Environment for Resource-scarce Sensor Systems). It is the basis for the generation of habitat models for bats (TP 1) by means of distributed data stream system queries (TP 3), which are sent to a heterogeneous sensor network (TP 4), consisting of static (TP 5) and mobile (TP 7) sensor network nodes. A major challenge is the restricted availability of resources, such as energy and memory, but also ad-hoc network connections. Hence, the operating system has to fulfill several non-functional requirements. For instance, it has to be light-weight (in terms of memory usage), energy-aware (i.e., optimized for using energy-efficient algorithms as well as being predictable in terms of battery lifetimes), and needs to provide the infrastructure that is required for efficient network protocols other sub projects can arrange with. For this reason, ARTE will be designed as a highly configurable software product line, whose configuration space is generated with tooling support. This enables the creation of optimized implementation candidates that can be used for deployment. These candidates consider dynamic adaption of both application and system functions with regards to the project's requirements. Since the operating system will be optimized for distributed data base queries, it needs to provide a customized middleware layer, which manages the software part for communication between mobile and static sensor nodes. In order to enable online modifications, that is, modifications that do not require a hard reset (e.g., for effectively testing new algorithms or new software components), the operating system on the nodes needs to be dynamically reconfigurable or reprogrammable, respectively. Both the states of applications and the system need to be preserved during this update process. The research project will benefit from two ongoing research projects: SEEP and Sloth. Both projects have to be adapted to fit the required needs and will be # extended with new features. Sloth is a light-weight operating system that aims to investigate utilization of underlying hardware features. Particularly, Sloth exploits hardware interrupt subsystems to perform most of the scheduling and dispatching activities. That is, all control flows (including threads) are implemented as interrupt service routines (ISRs). As a result, Sloth outperforms comparable embedded operating systems in terms of speed (regarding context switching) and memory usage (regarding kernel and code size). Yet, Sloth still supports blocking tasks (ISRs usually have run-to-completion semantics) and mechanisms to switch between different stacks (blocking tasks need their own stacks). Sloth has already been ported to the ARM Cortex-M3 and the Infineon TriCore. As part of this research project, Sloth will also be ported to a core with even lower power consumption and physical dimensions but still meeting all of Sloth's requirements. SEEP is an energy estimation framework that follows a proactive methodology which aims to provide developers with accurate energy estimations before deployment. Sloth is prepared for being analyzable by the SEEP framework. On the one hand, this enables the detection of hidden energy hotspots in Sloth itself, and on the other hand, precise energy estimates for all library and system functions of Sloth can be retrieved. These estimates are kept in a trace database, which can be exploited for time savings in future estimation runs, because previously analyzed functions do not have to be analyzed again. This requires to create energy profiles for the target hardware components and to adapt a corresponding instruction simulator for being usable by SEEP.
In resource-limited sensor networks, an efficient acquisition of data is especially important. For this purpose, a distributed data-stream system is employed. Its query language can activate and control the following four components. Number one is a central data management that is built in cooperation with subproject 4 (network management) and that hides the heterogeneity of the sensor network. The mobile and the ground-based sensor nodes can be in different states as far as software and energy are concerned. Because of resource limitations, applications are distributed over different nodes. The second component is an abstract model of an evaluation that is easy to create and that offers much flexibility for implementation. Using models, global queries against the sensor network can be transformed in many ways and can thus be executed efficiently. The third component is a cost model for the quantitative assessment of data-stream queries, for instance regarding the energy efficiency. Different optimization goals can be imagined here, e.g. in reference to the precision in the measurement of flight trajectories. And the forth component finally consists of newly developed stream operators and of a description of the required precision, which allows a more advanced optimization to increase battery lifetime.
In the scope of the research group BATS, the project E2SO (TP 4) focuses on challenges related to the efficient communication with integrated self-organizing data management in highly dynamic sensor networks. The communication primitives to be developed primarily target the operation with limited energy and other systems resources. In particular, three aspects will be investigated in this research project: First, the sensor nodes developed in BATS will be extremely challenging in terms of energy (the system design expects a node weighting of 2g including the battery). Therefore, new communication techniques are needed for connecting the mobile nodes to the stationary ground network. Our partners from electrical engineering are about to develop mobile sensor nodes (TP 7) as well as the ground network systems (TP 5). Besides of the limited computing and storage resources, the limited connectivity of the mobile bats constitutes new challenges that can only be investigated in this interdisciplinary team of researchers. A second focus is on the temporal storage of dynamically generated information and on the efficient access to these data. The concept is to use DHT concepts for maintaining measurement data that will be processed in the network relying on data stream processing techniques (TP 3) to finally be used for answering questions in the field of biological behavioral research (TP 1). The third research aspect of the E2SO project is to provide continuous resource and topology monitoring in order to extend the network lifetime as a key metric describing the overall system performance and to early detect bottlenecks that can be mitigated using online reconfigurations of the data streams (TP 2 and TP 3).
The subproject Tracking4BATS provides tracking and localization of bats to support the other subprojects. The localization is based on received signal strength (RSS) measurements and realized by a wireless sensor network (WSN). In this context several questions are addressed in this research project: First of all, to what precision can lightweight transmitters be tracked by field strength measurements? The topology of the WSN plays a key role: Which is the best topology of sensor nodes to track the behavior of bats continuously over a large time span, covering a large area? How can directional antennas enhance the RSS based tracking? Robust localization algorithms, which are able to cope with imperfections of RSS measurements, have to be developed. Mobility models for the flying bats can improve the estimation of the trajectories. Furthermore, a communication link between bat and ground nodes has to be provided in this subproject.
The subproject Scalability4BATS addresses enhanced precision localization of bats. High-quality localization might give to biologists more insight into the behavior of bats. The main research topics in this sub project are: Signal design, phase-based distance measurements, phase-based angle measurements, and sensor fusion. The signal design covers aspects like power efficient generation of multi-tone signals, multipath robustness and positioning accuracy. The duty cycle and burst length of the transmitted signal is of interest to keep the power consumption at a decent level. Furthermore, the theoretical limits of phase-based distance measurements have to be determined. Finally, RSS- and phase-based measurements have to be fused to guarantee precise and consistent trajectory estimates. Close cooperation with the subprojects 7 and 8 is mandatory to ensure well-designed signals which allow precise localization. Though addressing different measurement principles, subprojects 5 and 6 are closely related.
In order to explore the behavior of bats, a sensor system is to be designed in the project. These sensors must be mounted on the bat to locate the bat in flight. So that it is unaffected, the sensor node must be light and very compact. In the presented subproject module integration of miniaturized wireless sensor node with positioning functionality is to be done. For the announced use on a flying bat the most important constraints are a minimum total weight (max. 2 g including battery, circuit board and antenna) and a form factor that does not restrict the bat in their natural movements. For this subproject, these two requirements are a major challenge to the design of a multi-band antenna solution that is both in their geometry greatly shortened and three dimensions conform to the body of the bat. Aerodynamics also needs to be considered here. In addition to the positioning functionality, which is realized by integration of IC designed in TP8 communication between different sensor nodes should also be possible. To maximize the life of the applied battery and to relieve the pursued energy management of the module energy efficient communication protocols should be analyzed. By staggering the work packages, the complexity of the mobile sensor node is increased by adding more features gradually to a realization of the basic functions in the first step and culminates at the project's end in a lightweight, miniaturized wireless sensor node with a localization and communication interface for use on a flying bat.
To study the behavior of bats, the development of a sensor platform with localization functionality is the target of this research group. Since the sensor nodes are mounted on the bats, in order to avoid interruption on the natural behavior of bats, the device must be very light and compact. Therefore, only integrated circuit can reach this requirement. In addition, to guarantee the endurance of the mobile device with limited battery, meet the modulation specification while achieve low power consumption character is the main design challenge. Base on those requirements, the energy-efficient transmitter circuit is developed with the synthesizer to transmit and receive the modulated signals. Furthermore, a multi-level approach to a wake-up receiver is considered in order to minimize unnecessary activation of the circuit. For the purpose of communication and localization, the aim of this project is to design and analyze the analog circuit for the transmitter and receiver of the mobile sensor node with a single chip.
The implementation of BATS requires a scientific coordinator for handling the following tasks: control and support the networking of subprojects; support for the implementation of measurement campaigns, methodic synchronization of the subprojects and with regards to contents; organization of scientific activities of the research group; organization of gender equality activities; organization of guest scientist stays; administrative tasks.