Reducing signal processing energy consumption in solar/battery powered ubiquitous networks: Joint optimization of communication algorithms/circuits, and software design/testing

无线网络不断发展，以更低的成本支持更高的数据速率和更广泛的覆盖范围。这反过来又需要大量的信号处理，这受到现有“电池寿命”限制的严重影响。与此同时，电池技术正在进步，然而，与预期需求相比，速度要慢得多。这使得电池寿命成为许多未来无线网络实现成本效益的主要瓶颈。未来十年的预测需求远远超出了当前技术的范围，需要突破才能实现这些迫切需求。传统上，电池寿命问题一直是资源有限的移动设备的一个问题，然而，在许多新兴的场景中，无线接入点也依赖于使用太阳能电池充电的电池。在加拿大这样一个人口密度低、通信基础设施有限的大国，使用太阳能供电的无线基础设施尤为重要。这种太阳能无线网络将使向许多遥远的地区提供负担得起的通信成为可能，否则，由于使用电线为远程无线节点供电的成本过高，这是不可能的。另一个影响很大的例子是在人口密度非常低的加拿大北部进行野生动物监测，以及在阿尔伯塔省的油田进行大规模基础设施监测。提出的研究将探讨几个新的方向，以提高无线系统的功率效率，为手持设备和骨干网络。这包括应用软件工程解决“最小化功耗”问题的新方法。此外，在无线节点上运行的大多数程序通常依赖于输入数据，由于底层无线网络的性质，这些数据以随机延迟的方式到达。该研究调查了这种延迟的统计行为，以提供：(i)优化性能的设计规则，以及（ii）在各种网络负载和传输可靠性下评估性能的测试技术。

Wireless networks are constantly evolving to support higher data rates and wider coverage at a lower cost. Thisin turn requires extensive signal processing which is severely affected by the existing limitations on "batterylife". Meanwhile, the battery technology is advancing, however, at a much slower rate as compared to theprojected needs. This makes the battery life the primary bottleneck in the cost effective realization of manyfuture wireless networks. The projected needs over the next decade are far beyond the reach of the currenttechnology and breakthroughs are needed to realize these urgent requirements. Traditionally, the problem ofbattery life has been an issue in the resource limited mobile units, however, there are many emerging scenariosin which the wireless access points also rely on batteries charged using solar cells. Access to suchsolar-powered wireless infrastructure is particularly important in a vast country like Canada with lowpopulation density and limited communication infrastructures. Such solar-powered wireless networks willmake it possible to provide affordable communication to many distant areas which would be impossibleotherwise due to the excessive cost in powering the remote wireless nodes using electrical wiring. Anotherexample of large impact is wild life monitoring in the Northern Canada which have very low populationdensity, and large-scale infrastructure surveillance in oil patch in Alberta. The proposed research willinvestigate several novel directions to increase the power efficiency of wireless systems, for both handheldunits and the backbone network. This includes novel methods in the application of software engineering to theproblem of "minimizing power consumption". In addition, most of the programs running on wireless nodesusually rely on input data that arrive with a random delay due to the nature of the underlying wireless network.The research investigates the statistical behavior of such delays towards providing: (i) design rules foroptimized performance, and (ii) testing techniques to evaluate the performance under various network loadsand transmission reliability.