Channel Optimised Distributed Passive Sensor Networks

物联网认识到将配备传感器的大量物理对象互联以实现过程自动化和新的数据应用的价值。对于广泛的应用，传感器将是无线的，在没有电池或内部电源的情况下运行，从而降低了传感器的成本和复杂性。辐射射频(RF)能量转移是一种在几米范围内向电子设备提供少量电力的有吸引力的方法，并且已经被用于日益取代条形码的射频识别(RFID)系统中。然而，传感器或标签的可用功率严重有限，这限制了传感器标签的范围，从而阻碍了有效的广域操作。本项目试图通过首次考虑由多个传感器组成的完整的传感器网络和统一的分布式询问器网络来解决无源无线传感器标签的有限范围。大多数射频功率传输系统都受到现实环境中反射的限制。在这里，多个天线的使用将允许对无线信道(反射)进行特征化，并对传输信号进行优化以减轻影响，从而在需要采集数据样本时增加传感器标签处可用的射频功率，提高数字和模拟传感器标签的性能。这将实现更耗能的传感器功能和更广泛的工作区域，并具有更高的可靠性，从而打开无源传感器标签目前性能不足的一系列应用。随着工作区域的扩大，大量传感器标签同时在询问区域工作的新问题也将得到解决。应用程序模型将考虑哪些传感器需要以什么占空比运行，以确保收集到足够的数据。这可以通过考虑多个传感器的互信息以及信道信息来优化，以选择具有最有利信道的那些传感器。最后，开发的传感器标签和询问器系统将被内置到医疗保健应用的演示器中。这将指导应用模型的开发，以应对具有不同功率和占空比要求的多个不同传感器，并允许向未来跨不同学科的潜在合作伙伴(学术和工业)演示该系统。我们认为，由于无线传感的侵入性较小，传感器体积较小，没有电池，以及传感器能够以低成本随时丢弃，因此对医疗保健的潜在影响是显著的。

The Internet of Things recognises the value of interconnecting vast numbers of physical objects equipped with sensors to enable process automation and new data applications. For a wide range of applications the sensors will be wireless and operate without a battery or internal power supply, reducing the cost and complexity of the sensor. Radiative radio frequency (RF) energy transfer is an attractive method to provide small amounts power to electronic devices over a range of a few meters and is already used in radio frequency identification (RFID) systems which are increasingly replacing barcodes. However, the power available at the sensor or tag is severely limited which has limited the range of sensor tags so far preventing effective wide area operation.This project seeks to address the limited range of passive wireless sensor tags, by considering for the first time a complete sensor network comprised of multiple sensors and a unified distributed interrogator network. Most RF power delivery systems are limited by reflections which occur in realistic environments. Here use of multiple antennas will allow the wireless channel (reflections) to be characterised and the transmitted signals optimised to mitigate the effects, increasing the RF power available at the sensor tag when it is required to collect a data sample, improving the performance of both digital and analog sensor tags. This will enable both more energy intensive sensor functions and wider operating areas with greater reliability opening up a range of applications where the performance of passive sensor tags has currently insufficient. With a wider operating area, a new problem of large populations of sensor tags simultaneously operating in the interrogation region will also be addressed. An application model will consider which sensors are required to operate with what duty cycle to ensure sufficient data is collected. This can be optimised by considering the mutual information of multiple sensors and also the channel information to select those sensors with the most favourable channels. Finally, the developed sensor tags and interrogator system will be built into a demonstrator for healthcare applications. This will both guide the development of the application model to cope with multiple different sensors with different power and duty cycle requirements, and also allow the system to be demonstrated to potential future collaborators (both academic and industrial) across various disciplines. We believe the potential impact in healthcare is significant due to the less intrusive nature of wireless sensing, less bulky sensors with no battery and the ability of sensors to be readily disposable at low cost.