In this paper, development and design of a miniature sensor node comprising of a six-channel electrocardiograph and a tri-axial accelerometer, gyroscope and magnetometer, altimeter, as well as other electronic components necessary for processing and storage of acquired data and for communication via USB and Bluetooth interface is described. Device has a battery fuel gauge which enables precise state of the charge tracking. Measurements of electrocardiogram on subjects while they were doing certain physical activity were performed. Measurements were taken while subjects were sitting, walking on a flat surface, climbing stairs slowly and fast and riding a stationary bike. The influence on the quality of the recorded electrocardiogram due to the intensity of subject movement was analyzed. Circuits and algorithms for data processing successfully suppress artefacts caused by specified types of physical activity. The electrodes maintain good contact with the skin with noticeable dips in signal quality when activities that require greater torso movement are performed. Accelerometer measurements were analyzed and it was found that in real-time it is possible to distinguish between sitting/lying and movement. However, it is not possible to accurately distinguish between walking and climbing stairs because of differences in individual performance of certain types of activity. To accurately differentiate between various kinds of movements it is necessary to process the collected data using neural networks in order to find personalized features of activity intensity and type for individuals and/or patients.