A Combined Approach for Real-Time Data Compression in Wireless Body Sensor Networks

Abstract

Wireless body sensor networks (WBSNs) represent an enabling technology for unobtrusive patient monitoring. Unlike wireless sensor networks (WSNs), they are characterized by relatively few and heterogeneous sensors placed in, on, or around the human body. An important issue consists in designing efficient solutions for optimizing network resource usage, such as computational capacity, energy, and bandwidth. Compression algorithms for WBSNs need to satisfy more stringent requirements than solutions for typical WSNs. In particular, to guarantee real-time monitoring of vital signals, the algorithms cannot introduce latency. Furthermore, the maximum reconstruction error is usually very small and it needs to be known in advance. In this scenario, we propose a combined compression algorithm, which satisfied the previous requirements. Results obtained by considering different biomedical signals show that a significant compression ratio can be achieved also when very small values of the maximum error are considered.