The success of wireless communication systems together with the recent advances in different technologies have enabled dense wireless sensor network deployments. In addition to the more conventional application areas, these deployments have been recently exploited for new sensing possibilities such as: device-free localization, sensorless sensing, and non-invasive vital sign monitoring. The aforementioned technologies use the ubiquitously available received signal strength measurements of the radios for inference, and most notably, the considered sensing modality does not require people to co-operate with the system or to carry any electronic device, tag or sensor. This thesis utilizes the received signal strength measurements of narrowband radios for gaining situational awareness of the surrounding environment and two distinct research problems are considered: i) device-free localization and tracking; and ii) non-invasive breathing rate monitoring. The complex nature of the radio propagation channel is accountable for multipath propagation, which in turn makes it a difficult task to relate the changes in radio signals to location or breathing rate of the person. The constraints of low-cost narrowband radios further degrade the system performance. This thesis concentrates on leveraging possibilities enabled by the measurement system in order to develop methods that cope with the undesirable effects of multipath propagation and limited received signal strength resolution. The key contributions of the thesis are: i) the use of frequency channel diversity; ii) techniques to enable long-term deployments; iii) algorithms to train the unknown model parameters; and iv) methods for enhancing low-granularity received signal strength measurements. The thesis provides novel solutions that improve the system performance and that expand use-case scenarios of received signal strength-based device-free localization and non-invasive breathing monitoring. Research of the thesis builds upon existing work and the development efforts aim to enhance estimation accuracy, as well as, improve the applicability and use-case scenarios of the technologies to enable home healthcare and ambient assisted living applications of the future. In the work, the development efforts are evaluated using real-world experiments conducted in a variety of environments. Furthermore, the effect of each development stage is analyzed and general guidelines on system design are discussed.
|Translated title of the contribution||Sisätilojen tilannekuvan estimointi kapeakaistaisten radiosignaalien avulla|
|Publication status||Published - 2017|
|MoE publication type||G5 Doctoral dissertation (article)|
- received signal strength
- device-free localization
- non-invasive breathing monitoring