Safety Critical Embedded Systems Of Energy Consumption Measurement Analysis And Real-time Scheduling

The rapid developments of embedded systems play a significant role in nationeconomic and deeply promote the informatization and intellectualization of humansociety. Embedded systems often work in the harsh and complex environment whichmake it, especially the battered-powered systems, face serious energy consumptionconstraints. Furthermore, with the further evolvement of Internet and the spring up ofthe Internet of Things, embedded systems are increasingly threatened by security issues.How to design energy-efficient and high secure embedded systems becomes be one ofthe great challenges that we never faced before.The measurement and analysis of embedded system is surely necessary for thedesign of low power consumption systems. To make up the defects of traditionalmeasurement for embedded software, a method based on LabVIEW from the field ofprofessional observation and control is presented to measure energy consumption ofsoftware accurately. This method directly acquires the voltage signal and current signalof the target device in which the specified software is running. All those raw data isdiscrete which is processed approximately to evaluate energy consumption. Finally, ourmethods are proved to be correct and valid by real comparing measurements.To deeply investigate the relationship between the security and energyconsumption, it is of critical importance to the characteristics of the energy consumptionof the cryptographic algorithms. In order to obtain the accurate energy and performancedata of typical security algorithms, the physical test platform is set up based on theproposed method. By that platform, the energy consumption data of several types ofciphers are acquired, mainly including: symmetric cryptosystem, asymmetriccryptographic algorithm, hash algorithm, and corresponding security parameters.According to the measured data, we propose a multidimensional analysis frameworkand reveal security algorithms’ energy consumption characteristics from the angles ofpower, speed, and unit energy cost respectively. The related mathematical models arealso built up to summarize the explained energy consumption characteristics. High efficient scheduling for security-and energy-aware real-time tasks can ofcourse help to improve the system’s energy efficient and quality of security dramatically.Considering real-time embedded systems under heterogeneously distributedenvironments, we devise a security-and energy-aware task scheduling algorithm WeightTuning for Schedule Length and Energy Consumption or WTSE. WTSE that select theoptimize computing node by tuning the weights of schedule length and energyconsumption gives adequate consideration for several factors including networkcommunication, security service, system energy and task real-time. Furthermore, tolimit the security service consume too much energy and processing time, a mechanismof flexible utilization of task slack time is set up. Compared to most related schedulingalgorithms, WTSE can not only enhance the security and energy efficiency effectively,but also keep the high performance of job schedule length and job speedup.