Holistic Design of Body Sensor Networks: A Cyber Physical Systems Approach
Body area sensor networks (BASN) are emerging cyber-physical systems that promise to improve quality of life through improved health, augmented sensing and actuation for the disabled, independent living for the elderly, and reduced healthcare costs. However, the physical nature of BASNs introduces several new challenges. The human body is a highly dynamic and unpredictable physical environment that creates constantly changing demands on sensing, actuation, and quality of service (QoS). The various locations that users visit, their choice of clothing also adds to the unpredictability of the environment. Thus, BASNs must simultaneously deal with rapid changes to both top-down application requirements and bottom-up resource availability. This is made all the more challenging by the wearable nature of BASN devices, which necessitates a vanishingly small size and therefore extremely limited hardware resources and power budget. This project seeks to develop new principles and techniques for realiable adaptive operation in highly dynamic physical environments, using miniaturized, energy-constrained devices.
This project consists of multiple parts. We are developing design principles that enable the cross-layer mutli-view process required for such systems. With the technologies developed based on these principles and techiques, we intend to build and delpoy relaible adaptive BASNs that meet various application demands as a proof of concept.
Design Principles and Techniques
The development of BASNs requires a cross-layer multi-view approach. A BASN can be looked at from multiple connected perspectives (system architecture, system behavior, and computational, control, and communication models that characterize system operation) and from multiple layers in each perspective. In addition, the cyber-physical relationships the affect system operation must be captured adequately within each perspective. We developing design principles that captures this cyber-physical relationship and enables this multi-view design process, allowing collaborative and effecient exploration of the system desgin space and potentially more optimized designs. A related project in this space is the Cyber-Physical Systems Modeling and Simulation Infrastracture for Body Sensor Networks.
Technologies for Adaptive Operation in Highy Dynamic Physical Environments
We have identified and are exploring a number of technologies (and techniques) that will enable the adaptive operation capabilities of BASNs:
- Ultra low-power adaptive and flexible circuits
- Intelligent, low-power communication (in particular asymmetric communcation between nodes and aggregators)
- Adaptive data fusion algorithms
- Adaptive cross-layer compression and signal processing techniques
Systems for Long-term Adaptive Continuous Monitoring
We are developing systems that take advantage of the adaptive capatilities of BASNs to enable continuous long-term monitoring with end-to-end quality of service (QoS) guarantees for changing application demands.
Long-Term Continuous Monitoring Applications for Longitudinal Assessment, Detection, and Prediction of Medical Conditions
We are looking at deploying systems in various applications that highlight the adaptive capabilities of our systems as proof-of-concept for our design principles and technologies. Some general tentative application areas are
- Long-term physical activity monitoring (for athletes and general wellness)
- Health state monitoring for first responders and rescue workers during rescue operations
- Independent and assisted living for the elderly
- Monitoring effects of everyday activities on health
Cross-layer compression analysis
In a wireless BSN it is important to reduce the amount of data transmitted. Longer battery lifetime and less collision within and across different BSNs can be achieved. This can be realized by compressing the data.
However, data goes through different layers of a BSN and might suffer of distortion due to compression and transmission interferences. The node senses and produces the data, then compresses and sends it to the aggregator; the aggregator decompresses and uses the data, then compresses it again and sends it to the base station. In this process the original data might be affected by noise.
The purpose of this project is to study the effects of compression on data going across a BSN.
Multi-Scale QoS for Body Sensor Networks
Wireless body sensor networks (BSNs) represent a promising solution for a growing national crisis - affordable quality healthcare for all. Instead of expensive and invasive in-patient monitoring and highcontact care, wireless sensor nodes distributed at strategic locations on the body can continuously and non-invasively monitor individuals for a variety of physiological and biokinetic markers. The data that is wirelessly transmitted through the BSN and ultimately onto existing wireless networking infrastructure can be used to follow trends for improved diagnosis and treatment or to detect events that require immediate intervention. With such a technology, individuals can receive high-quality healthcare at a dramatically lower cost while maximizing their ability to live independently as they age.
However, this promise will not be realized if BSNs cannot meet application and user requirements. Given that people’s lives are at stake in these target applications, it is essential that BSNs reliably provide high fidelity data and analysis. When applications require that an event be detected quickly for an immediate response, low latency is necessary. When a large number of BSN wearers live in close proximity (as will be the case in many continuous care retirement communities (CCRCs)), network throughput becomes an important metric. Finally, the practical issue of wearability in terms of BSN node form factor and battery life is central to non-invasiveness and, therefore, compliance, requiring high energy efficiency to enable a small battery to last for a long time.
We therefore propose to encapsulate these contributing metrics into a quality of service (QoS) standard that can be used to design and manage BSNs in a variety of applications and environments. We will also develop new system modeling and analysis techniques for design-time optimization and run-time management of QoS, including new QoS tuning techniques at different levels in the BSN system hierarchy – on-node signal processing and data management, intra-BSN networking, and extra-BSN networking. The key to the success of this project is the multi-scale approach to QoS as shwon in the above picture. Most prior work examines BSN system levels individually, preventing the search for a global optimum. The proposed system model will enable designers to explore how QoS techniques at each level impact the other levels. Finally, this multi-scale model, as well as the level-specific QoS techniques, will be validated and improved based on experimentation with physical BSN systems deployed in real BSN applications. The ultimate goal is to address the fundamental issues related to the practical use of many BSNs operating in real-world applications and environments.
Body Sensor Networks: A Holistic Approach From Silicon to Users. Proceedings of the IEEE. 100:91-106.. 2012.