SmartLoCore: A Concept for an Adaptive Power-Aware Localization Processor

Abstract

During the last decades, a great shift to processing on mobile and ubiquitous devices has been taken place. With this movement also the integration of different sensors, especially inertial sensors, has gained strong interest. One approach is to combine inertial sensors with other localization systems. However, this needs real-time processing capabilities and low power consumption, which is usually a problem when performing this task on the main processor with the operating system on top. In this paper, we present an adaptive localization processor called SmartLoCore including an intuitive development flow that allows hardware/software co-design trade-offs during design-time. These options include running the algorithm in software, adding new additional instructions to the core instruction-set and offloading complex mathematics onto dedicated hardware. We show the applicability with a complex localization application example and highlight the benefits of the proposed concept. We conclude that the SmartLoCore concept shows a very promising potential for a large variety of mobile systems.