Search for: All records

Wireless sensing and the Internet of Things support real-time monitoring and data-driven control of the built environment, enabling more sustainable and responsive infrastructure. As buildings and physical structures tend to be large and complex, instrumenting them to support a wide range of applications often requires numerous sensors distributed over a large area. One impediment to this type of large-scale sensing is simply tracking where exactly devices are over time, as the physical infrastructure is updated and interacted with over time. Having low-cost but accurate localization for devices (instead of users) would enable scalable IoT network management, but current localization approaches do not provide a suitable tradeoff in terms of cost, energy, and accuracy for low power devices in unknown environments. 

Time-series data gathered from smart spaces hide user's personal information that may arise privacy concerns. However, these data are needed to enable desired services. In this paper, we propose a privacy preserving framework based on Generative Adversarial Networks (GAN) that supports sensor-based applications while preserving the user identity. Experiments with two datasets show that the proposed model can reduce the inference of the user's identity while inferring the occupancy with a high level of accuracy. 

The emergence of radio frequency (RF) dependent device-free indoor occupancy detection has seen slow acceptance due to its high fragility. Experimentation shows that an RF-dependent occupancy detector initially performs well in the room to be sensed. However, once the physical arrangement of objects changes in the room, the performance of the classifier degrades significantly. To address this issue, we propose BLECS, a Bluetooth-dependent indoor occupancy detection system which can adapt itself in the dynamic environment. BLECS uses a reinforcement learning approach to predict the occupancy of an indoor environment and updates its decision policy by interacting with existing IoT devices and sensors in the room. We tested this system in five different rooms for 520 hours in total, involving four occupants. Results show that, BLECS achieves 21.4% performance improvement in a dynamic environment compared to the state-of-the-art supervised learning algorithm with an average F1 score of 86.52%. This system can also predict occupancy with a maximum 89.23% F1 score in a completely unknown environment with no initial trained model.