Energy-efficient Distributed Systems

Submitted by webmaster on Sun, 03/01/2020 - 04:00
Course No: 
Course Type: 
Weekly Hours: 
Ivona Brandic
Ivan Lujic

Current ICT systems consume huge amounts of electricity. Only between year 2000 and 2007 the total power consumption of data centers worldwide went from 70 to 330 billion kWh, and is projected to grow over 1000 billion kWh until 2020. Thus, energy efficient management of computational resources is gaining constantly on importance. In this lecture we will discuss concepts and approaches for the measurement, modeling and energy efficient resource allocation at runtime in distributed systems.


In particular, we will discuss concepts of micro and massive data centers, methods for speculative resource allocation in data centers, and energy efficient monitoring of resources. Since energy efficiency is highly related to pricing policies of resource we will also discuss different pricing policies. To facilitate energy efficient runtime operation of massive data centers we will discuss various controller design techniques that also consider geo-temporal constraints. Further, we will present state of the art methodologies for energy efficient management of battery life in smart devices like cyber foraging or energy harvesting.


25 h attending lectures and exercises30 h lecture follow-up and preparation of home exercises19.5 h preparation written exam0.5 h exam------75 h overall
Slides will be provided in TISS during the semester.


J. M. Pierson. Large-scale Distributed Systems and Energy Efficiency: A Holistic View. Wiley, 2015.
Toni Mastelic, Ariel Oleksiak, Holger Claussen, Ivona Brandic, Jean-Marc Pierson, Athanasios V. Vasilakos: Cloud Computing: Survey on Energy Efficiency. ACM Comput. Surv. 47(2): 33:1-33:36 (2014)


<p>Exercise&nbsp;Work 50%</p>
<p>Exam 50%</p>
<p><strong>Overall grade distribution:<br></strong>S1: [88%,100%]<br>U2: [75%,88%)<br>B3: [63%,75%)<br>G4: [50%,63%)<br>N5: [0%,50%)<strong></strong></p>
<p><strong></strong>Final Exam</p>
<p>The final exam will be on 24th June, second opportunity will be in October 2020. The student can attend one or both exames. If both exames are done then only the best result counts, and only one certificate will be issued (for the exam with better resoults).</p>
<p>To sucessfuly complete the course both parts (exam and excercise) have to be positive. Practical part will be evaluated in the form of the presentations where students have to present the developed solutions at different stages (algorithm design, testbed, etc.). <br><br></p>