SPINE-based application development on heterogeneous wireless body sensor networks

Abstract

Wireless sensor networks (WSNs) are a novel technology enabling new classes of applications and systems for ubiquitous and pervasive computing. In particular, WSNs for the human body, also known as Wireless Body Sensor Networks (WBSNs), will enable not only continuous, multi-purpose monitoring of people but also will support social interaction among people coming into physical contact. In these contexts, applications demand a wide range of functionalities, in terms of sensor types, processing performance, communication capabilities. Moreover the development of such applications has to deal with the issue of handling heterogeneous WBSNs since different kinds of sensor node architectures could be necessary to fulfill all the application requirements. This paper proposes an approach based on the SPINE frameworks (SPINE1.x and SPINE2) for the programming of signal processing applications on heterogeneous wireless sensor platforms. In particular, two integrable approaches based on the proposed frameworks are described that allow the development of applications for WBSNs constituted by heterogeneous sensor nodes. The approaches are exemplified through a human activity recognition system based on a WBSN composed of two types of sensor nodes, heterogeneous with respect to base software and hardware.