Resource-aware load balancing system with artificial neural networks

Yıldız, Ali
As the distributed systems becomes popular, efficient load balancing systems taking better decisions must be designed. The most important reasons that necessitate load balancing in a distributed system are the heterogeneous hosts having different com- puting powers, external loads and the tasks running on different hosts but communi- cating with each other. In this thesis, a load balancing approach, called RALBANN, developed using graph partitioning and artificial neural networks (ANNs) is de- scribed. The aim of RALBANN is to integrate the successful load balancing deci- sions of graph partitioning algorithms with the efficient decision making mechanism of ANNs. The results showed that using ANNs to make efficient load balancing can be very beneficial. If trained enough, ANNs may load the balance as good as graph partitioning algorithms more efficiently.


Energy- aware task scheduling over mobile ad hoc networks
Bokar, Ali; Bozyiğit, Müslim; Department of Computer Engineering (2009)
Mobile ad hoc networks (MANETs) can be formed dynamically without the support of any existing infrastructure or any centralized administration. They consist of heterogeneous mobile nodes which are powered by batteries, move arbitrarily and are connected by wireless links. Battery energy limitation is one of the main challenges in the MANETs. Several hardware and software based techniques have been proposed in this field. Most of the previous studies have considered only the energy minimization of individual...
Scheduling approaches for parameter sweep applications in a heterogeneous distributed environment
Karaduman, Gülşah; Şener, Cevat; Alpdemir, Mahmut Nedim; Department of Computer Engineering (2010)
In this thesis, the focus is on the development of scheduling algorithms for Sim-PETEK which is a framework for parallel and distributed execution of simulations. Since it is especially designed for running parameter sweep applications in a heterogeneous distributed computational environment, multi-round and adaptive scheduling approaches are followed. Five different scheduling algorithms are designed and evaluated for scheduling purposes of Sim-PETEK. Development of these algorithms are arranged in a way t...
Functional mockup unit adaptation for HLA-compliant distributed simulation
Yılmaz, Faruk; Oğuztüzün, Mehmet Halit S.; Durak, Umut; Department of Computer Engineering (2014)
Conceptual design of systems require aggregate level simulations of the designed system in its operational setting. Thus, performance of the system and its interactions with the other entities in its environment can be evaluated. The complex and heterogeneous nature of these simulations often requires distributed execution. IEEE 1516 High Level Architecture (HLA) is a widely accepted standard architecture for distributed aggregate level simulations. Functional Mock-up Interface (FMI) is a recent standardiza...
An evaluation of aspect-oriented programming for embedded real-time systems
Kartal, Yusuf Bora; Schmidt, Şenan Ece; Department of Electrical and Electronics Engineering (2007)
Crosscutting concerns are the issues in software that cannot be modularized within a software module. In this thesis work, a detailed evaluation of the use of Aspect Oriented Programming for the implementation of crosscutting concerns in embedded real-time systems is presented. The pilot Audio Switch project implementations are first evaluated in terms of software quality attributes. Then a detailed analysis of the two implementations, according to embedded real-time performance metrics has been carried out...
Face detection in active robot vision
Önder, Murat; Halıcı, Uğur; Department of Electrical and Electronics Engineering (2004)
The main task in this thesis is to design a robot vision system with face detection and tracking capability. Hence there are two main works in the thesis: Firstly, the detection of the face on an image that is taken from the camera on the robot must be achieved. Hence this is a serious real time image processing task and time constraints are very important because of this reason. A processing rate of 1 frame/second is tried to be achieved and hence a fast face detection algorithm had to be used. The Eigenfa...
Citation Formats
A. Yıldız, “Resource-aware load balancing system with artificial neural networks,” M.S. - Master of Science, Middle East Technical University, 2006.