Performance Evaluation of 6LoWPAN Based Networks for Ubiquitous Health Monitoring System

摘要

Machine-to-Machine (M2M) networks have become very useful in a wide range of applications such as ubiquitous healthcare (u-healthcare) monitoring, air quality monitoring, insect monitoring, etc. Such applications require providing Internet connectivity to wireless sensor networks for the purpose of data collection and gathering. IPv6 over low power wireless personal area network (6LoWPAN) protocol specifications support the Internet-of-Things (IoTs), through an adaptation layer that provides efficient header compression. The 6LoWPAN middleware constitutes a crucial enabling technology for M2M networking. An optimum deployment of a 6LoWPAN-based system requires an accurate evaluation of the 6LoWPAN network's performance. Among various metrics, throughput and Packet Error Rate (PER) are of paramount importance, especially for time-critical and delay sensitive applications. In U-Health System where real-time ECG monitoring requires high throughput with reduced PER, an intolerable level of latency could result into lethal effects on patients. Thus, the performance of Wireless Sensor Network (WSN) used in U-Health should conform strictly to the timeliness requirement of health monitoring applications. In this study, we consider a system architecture where a 6LoWPAN node acts as a proximity body sensor to form a Wireless Body Area Network (WBAN) for a given patient. The objective of the study aims at assessing the scalability range of 6LoWPAN based WBANs and at estimating whether the performance bounds of such networks can serve efficiently u-healthcare applications. We evaluate the network performance by simulating different scenarios using Cooja (a Contiki-Network Simulator). Scenarios of a Constant bit rate (CBR) traffic and a Variable-Bit-Rate traffic are simulated, emulating real life biomedical data (e.g., ECG, temperature, etc.). The CBR-based scenario is conceived to determine the verge of the 6LoWPAN network in acute cases of U-Health system. Through simulation we demonstrate the viability of the performance bounds of the 6LoWPAN-based WBAN for u-healthcare applications.