Visalakshi PALANISAMY, Vasanthamani KANNAIAN

Energy-efficient scheduling for real-time tasks using dynamic slack reclamation

The application domains of portable embedded real-time systems range from safety-critical applications likeelectronic control units in the automotive sector to entertainment applications like digital cameras and setup boxes inthe consumer electronics sector. These systems attract users with multiple and sophisticated functionalities in a singlesystem. The developments in VLSI technology have enabled IC manufacturers to incorporate more features in a singledie, which facilitates these multipurpose systems. The increase in features and functionalities has a direct impact onpower consumption and has to be strictly accounted for in the design of these embedded systems. The possibility ofachieving low power consumption is obtained by using only effective software techniques along with the hardware powersaving methods. The dynamic power management techniques with the support of real-time operating systems (RTOSs)provided in recent microcontrollers enable better power management using software control. The scheduling of tasksin real-time systems is based on the worst-case execution time (WCET), whereas tasks based on their environmentalconditions may complete their execution before the WCET. This leads to increased laxity (slack time) and reducedutilization of the CPU. This paper proposes Fixed Window DynaClam, a dynamic reclamation algorithm, to effectivelyutilize slack time. The clock frequency of the processor is slowed down by reclaiming the available slack time to reducethe energy consumption. The Fixed Window DynaClam algorithm is implemented for nonpreemptive periodic tasks inthe controllers with dynamic voltage and frequency scaling power management features. The novelty of the algorithmis that the clock frequency is not only adjusted based on the real-time demands of the tasks but also does not overloadthe kernel of the RTOS as the complexity is on the lower side.

PDF

___

[1] Li YT, Malik S. Performance Analysis of Real-Time Embedded Software. 1st ed. New York, NY, USA: Springer Science & Business Media, 2012.
[2] Lokuciejewski P, Marwedel P. Worst-Case Execution Time Aware Compilation Techniques for Real-Time Systems. 1st ed. Amsterdam, the Netherlands: Springer Science & Business Media, 2010.
[3] Falk H, Lokuciejewski P, Theiling H. Design of a wcet-aware c compiler. In: IEEE 2006 Workshop on Embedded Systems for Real Time Multimedia; Seoul, Korea; 2006. pp. 121-126.
[4] Sandell D. Evaluating static worst case execution time analysis for a commercial real-time operating system. MSc, Mälardalen University, Västeräs, Sweden, 2004.
[5] Dalsgaard AE, Olesen MC, Toft M, Hansen RR, Larsen KG. METAMOC: Modular execution time analysis using model checking. In: 2010 International Workshop on Worst-Case Execution Time Analysis. Brussels, Belgium: DROPS; 2010. pp. 113-123.
[6] Kelter T, Borghorst H, Marwedel P. WCET-aware scheduling optimizations for multi-core real-time systems. In: 2014 Embedded Computer Systems: Architectures, Modeling, and Simulation; Agios Konstantinos, Greece; 2014. pp. 67-74.
[7] Cazorla FJ, Quin̈ones E, Vardanega T, Cucu L, Triquet B. Probabilistically analyzable real-time systems. ACM T Embed Comput S 2013; 12: 94.
[8] Aupy G, Benoit A, Dufossé F, Robert Y. Reclaiming the energy of a schedule: models and algorithms. Concurr Comp-Pract E 2013; 25: 1505-1523.
[9] Singh AK, Das A, Kumar A. Energy optimization by exploiting execution slacks in streaming applications on multiprocessor systems. In: ACM 2013 Design Automation Conference; New York, USA; 2010. p. 115.
[10] Yu H, Veeravalli B, Ha Y. Leakage-aware dynamic scheduling for real-time adaptive applications on multiprocessor systems. In: IEEE 2010 Design Automation Conference; Anaheim, CA, USA; 2010. pp. 493-498.
[11] Chen DR. Slack computation for DVS algorithms in fixed-priority real-time systems using fluid slack analysis. J Syst Architect 2011; 57: 850-865.
[12] Khan J, Bilavarn S, Belleudy C. Energy analysis of a DVFS based power strategy on ARM platforms. In: IEEE 2012 Faible Tension Faible Consommation; Paris, France; 2012. pp. 1-4.
[13] Rizvandi NB, Taheri J, Zomaya AY. Some observations on optimal frequency selection in DVFS-based energy consumption minimization. J Parallel Distr Com 2011; 71: 1154-1164.
[14] Niu L, Li W. Energy-efficient fixed-priority scheduling for real-time systems based on threshold work-demand analysis. In: IEEE 2011 International Conference on Hardware/Software Codesign and System Synthesis; Taipei, Taiwan; 2011. pp. 159-168.
[15] Ramesh P, Ramachandraiah U. Energy aware proportionate slack management scheduling for multiprocessor systems. Procedia Comput Sci 2018; 133: 855-863
[16] Mhedhbi I, Atitallah RB, Jemai A. Dynamic slack reclamation strategy for multiprocessor systems. In: IEEE 2012 Mediterranean Electrotechnical Conference; Yasmine Hammamet, Tunisia; 2012. pp. 979-984.
[17] Saha S, Ravindran B. An experimental evaluation of real-time DVFS scheduling algorithms. In: ACM 2012 International Systems and Storage Conference; New York, NY, USA; 2012. p. 7.
[18] Jejurikar R, Gupta R. Energy aware task scheduling with task synchronization for embedded real time systems. IEEE T Comp Aid D 2006; 25: 1024–1037.
[19] Vasanthamani K, Visalakshi P. Energy optimized scheduling for non-preemptive real-time systems. Turk J Elec Eng & Comp Sci 2017; 25: 3085-3096.
[20] Park S, Park J, Shin D, Wang Y, Xie Q, Pedram M, Chang N. Accurate modeling of the delay and energy overhead of dynamic voltage and frequency scaling in modern microprocessors. IEEE T Comput Aid D 2013; 32: 695-708
[21] Vasanthamani K, Visalakshi P, Shankar SR. SPARK-A tool for synthesis and analysis for scheduling of real-time tasks. Asian J Res Soc Sci Hum 2017; 7: 990-999.