Cognitive Dynamic Systems for Location-Aware Radios

Reliable and accurate location information is of high importance for a wealth of applications, ranging from logistics, process monitoring, and recommender systems in smart sales-floors to emergency support systems. Location information is also of tremendous value for future wireless radio networks, where it can be used to improve performance metrics such as reliability, data throughput, transmission latency, and resource-efficient use of the infrastructure. However, wireless technologies suffer from physical impairments, e.g. multipath radio propagation, signal obstructions, and from man-made interferences from competing wireless transmissions. This results in dramatic changes in the quality of communication services and in unreliable, inaccurate radio positioning. The proposed research addresses the fundamental challenge of robustly extracting high-precision location and propagation environment information from radio signals in order to create a location- aware wireless network. To cope with harsh propagation environments, such as indoors or in urban areas, the required system has to have the ability to adapt its behavior to the surrounding environment. The structure of a cognitive dynamic system (CDS) is used in the project Cognitive Dynamic Systems for Location-Aware Radios to model these needed features. A CDS mimics the capability of the visual brain to filter out clutter and focus attention on activity and relevant information. In the course of this research project, a CDS is envisioned with many interconnected layers concerned with physical as well as application-related parameters leading to optimal sensing and control of the location and environment information. The system is thus expected to be perfectly suited to handle a variety of inter-dependent location-aware services of wireless devices in a robust way and suited for robust positioning.

For future 5G radio-communication systems, designed for ultra-reliable communications with high communication-rate, the knowledge of the propagation environment and the state of user equipment (location, velocity, orientation, clock synchronization) will be indispensable especially for safety-critical applications including autonomous driving and control for indoor robotics. This so-called situational-awareness for communication can be provided by inferring the state of user equipment as well as the radio-channel and map parameters of its propagation environment. In this project, we developed methods and algorithms that can extract the relevant information from radio-channel measurements reliably. Furthermore, we developed algorithms for localization and environment mapping that can cope with such challenging strongly time-varying and cluttered estimated provided by radio-channel parameter estimators.