E-textile's Dependable Infrastructures And Its System-Level Low Power Management Scheme

In the ambient intelligent environment, garments are not only for warmth and fashion, but for "invisible" computation. Due to the rapid advances in integrated circuits technology, material science and textile engineering, researchers created fabric computers for the future pervasive computing paradigm.Electronic textiles, or e-textiles are smart fabrics integrated with sensors, actuators, computing devices, flexible interconnections and battery modules. As a pervasive computing platform, e-textiles provide ideal computing substrates for various applications such as combat casualty care, medical monitoring, sports monitoring and law enforcement.There are many specific challenges in developing practical e-textile applications. When e-textiles are tailored as a wearable garment or other applications, tear and wear are highly frequent. The factors potentially introduce some short-or open-circuit faults into the power network and communication networks embedded in the fabrics. Hence, designing dependable infrastructures for e-textiles is a critical problem. Furthermore, e-textiles are battery powered. Similar with other battery-driven systems, it is significant to implement low power management schemes to extend the system lifetime of e-textiles applications. This dissertation focuses on these two research problems. Firstly, we propose a dependable infrastructure of electric networks and a dependable communication networks for e-textiles. Secondly, based on the dependable infrastructures, a battery management scheme for enhancing battery efficiency is discussed. The major works and contribution of this dissertation are as follows:1) Reviewed the state-of-the-art in e-textiles researchPrevious works are thoroughly reviewed. The specific challenges in e-textiles design and development are analyzed. Then, following a research framework, the research milestones are established. And we also look ahead the future research points and possible venues.2) Proposed a dependable infrastructure of electric networks for e-textilesTo enhance the fault-tolerant ability of the electric network for e-textiles, wepresent a new infrastructure of the power networks for e-textiles, named as Flexible Power Network (FPN). Instead of drawing power from a fixed battery as in the conventional electric networks, the power consuming nodes in a FPN can obtain power energy from one of battery nodes with the help of the battery selectors. We also introduce the over current protectors into both the power consuming nodes and the battery nodes to protect the batteries from wasting the charge when short-circuit faults occur. Experiments are performed using prototype systems and the results show that our FPN is more dependable than some common electric networks published before in the cases of short- and open-circuit faults.3) Presented a dependable networking scheme for e-textilesWe introduce token rings to connect the nodes, instead of the single rings in the original e-textile token grid network. The topology of the new network is described. And we also discuss the media access control protocols, a wrong route algorithm and the reliable operations. The performance metrics of the new networking scheme are analytically evaluated. Simulation results show that the new e-textile communication network can improve the ability of e-textile applications to tolerate faults and provide the communicating services of less delay in the presence of faults.4) Developed a battery management algorithm for enhancing the battery efficiency of e-textile systemsFirstly, we present a physically justified iterative computing method to illustrate the discharge, recovery and charge process of Li/Li-ion batteries. The capacity response of a given load segment is derived. Then, considering the condition that a power consuming node is connected to the multiple battery channel can- didates in FPN, we propose an algorithm aiming to maximize battery efficiency globally. The concept of weighted battery fatigue degree is defined and the novel battery scheduling algorithm called predicted weighted fatigue degree least first (PWFDLF) is developed. Experimental results show that a considerable improvement in battery efficiency can be obtained by PWFDLF under various battery configurations and current profiles compared to conventional sequential, round-robin, weighted round-robin discharging policies.