Hybrid person localization in ad-hoc networks using model-based control and visualization

Localization in previously unknown and GPS-denied areas is challenging, especially, when no infrastructure can be assumed. In this paper, we present a hybrid person localization approach for ad-hoc scenarios. Example applications are search and rescue operations, but also on construction sites or during maintenance tasks of large industrial plants. The outlined scheme is based on received signal strength localization in a wireless sensor network (WSN) and data fusion of inertial sensors. However, the complexity of such systems poses a challenge to non-technicians to operate them. To improve operability, the model-based design principle has been a rigorous change to systems design methodology. In this paper, we adopt the strengths of model-based methods to a heterogeneous distributed and networked system. The advantage of our approach is that the same model-based design environment is later used for control and monitoring of the network in the actual deployment. This allows easy reconfiguration of parameters during an operation. We present some localization results and highlight advantages and experiences gained from our evaluations. We conclude that model-based development on one hand and also model-based runtime control on the other hand are important aspects for future WSN deployments.