Personal wearable monitor of the heart rate variability Piotr Augustyniak Institute of Automatics, AGH University of Science and Technology, Kraków, august@agh.edu.pl Abstract: The aim of the paper is to present of a prototype of wearable heart rate variability (HRV) monitor. The home care surveillance and sleep assessment system is partly embedded into a smart home infrastructure and partly worn by a supervised person. The prototype wearable device is designed to acquire and process the electrocardiogram and to send reports accordingly to a programmed schedule. The recording, processing and transmission modes are programmable, what allows the recorder to respond immediately in case of predefined thresholds excess, while the regular reporting is organized in delayed packets exchanged during a short transmission session. This approach significantly reduces the contribution to the total power consumption from the communication module. The prototype was based on the PXA-270 development kit, but due to very low power consumption (0,5 mW) the migration to a more compact system is considered. keywords: sleep monitoring, assisted living, wireless body area network, heart rate variability, 1. Introduction Wireless monitoring of basic vital parameters is currently one of the hottest research areas and prospective application fields as assisted living and home care. Current experiences show, that personalization of the devices helps in refinement of distributed diagnosis calculation. First home care implementations prove that numerous diagnostic devices restricted to clinical use so far are applicable as elements of home automation increasing the users safety. They role in the prevention or follow-up is doubtless and among other aspects brings considerable reduction of medical costs. Thanks to deep modulation of functionality achieved with the use of software and remote configuration they fit to the paradigm of personalized medicine. The aim of the project was to built a prototype of home care surveillance and sleep assessment system including a wearable heart rate variability (HRV) monitor. The whole system is a hybrid partly embedded into a smart home infrastructure and partly worn by the supervised person. Both components are personalized by the software settings and dynamically linked executable code. The prototype wearable device is designed to acquire and process the electrocardiogram and to send reports accordingly to a programmed schedule. The ECG-derived HRV is commonly recognized as easily available representation of autonomous nervous system (ANS) and clinically studied in pursuits for many neurological disorders [1-2]. Currents studies justify that the HRV analysis in sleep is particularly meaningful. Because of lack of the voluntary actions of the subject in sleep, all heart rate accelerations and decelerations may be interpreted as ANS-related. The advantage of heart rate analysis is a wide range of data acquisition methods. The tachogram is usually derived from the single lead ECG signal, but upon the necessity the pulseoximeter, the arterial pressure meter or the cardiac microphone may also be a reliable source of RR interval series. The ECG-derived tachogram is a low data rate signal (12 bits per second) what suggests the ECG processing to be performed directly by the embedded software of a wearable recorder. Thanks to the low datastreams, the transfer may be organized in sessions allowing for considerable power savings or through a seamless link when immediate system response is needed. In the interpretation of the HRV common time- and frequency domain parameters were used. This facilitates data referencing to the clinical results in both: development and operating stages. The algorithms for tachogram processing are more complicated [3-4], particularly due to the sampling normalization required for spectral analysis. In the sleep