Experimental investigation into novel methods of reliable and secure on-body communications with low system overheads

Until recently the concept of wearable biosensors for purposes of medical monitoring was restricted to wired sensor applications. Recent advances in electronics and wireless communications have made the possibility of removing the wire from sensor applications a possibility. These advances have led to the development of small scale, wearable, sensing and communication platforms that can be placed on the human body creating the foundation for a Body Sensor Network (BSN). Body Sensor Networks aim to remove the restrictions that traditional wired sensors impose. The anticipation is that BSNs will permit the monitoring of physiological signals in any environment without limitation, giving Physicians the ability to monitor patients more closely and in environments that they cannot monitor today. Even with the recent advancements of electronics and wireless communications there are still many unanswered questions for practical solutions of BSNs that prevent BSNs from replacing traditional wired systems altogether. There is a great need for research into BSN architectures to set the standard for wireless sensor monitoring. In this work a development platform has been created for the investigation into the design and implementation of practical BSN solutions. The platform is used to compare BSN architectures and provide quantifiable results. From this work BSN architecture components that provide optimizations in system performance, energy, network lifetime and security are recommended. In Chapter 3 BSN network architectures employing the use of relaying of creeping waves is investigated. The investigation includes experimental analysis of various test environments. Experimentation demonstrates that the relaying of creeping waves offers considerable performance gains when compared to non-relay networks. For example, relaying is shown to increase network-lifetime by a factor of 13, decrease energy-per-bit requirements by 13 dB and provide the ability for the network to compensate for considerably wider fade margins. In Chapter 4 utilizing the randomness of the wireless channel for securing on-body communications with low overheads is considered. A low-complexity algorithm for establishing symmetric encryption keys is presented and validated. The algorithm relies on readily available RSSI measurements obtained from existing packets being sent and received in the network. The generated bit sequences from the algorithm are evaluated for matching between two communicating parties and mismatching with a malicious eavesdropper. It is shown that the algorithm produces long sequences of highly random bits that are perfectly matched between legitimate parties and highly mismatched with the eavesdropper