Dynamic Load Balancing Based on Multi-Objective Extremal optimization

Multi-objective algorithms based on nature-inspired approach of Extremal optimization (EO) used in distributed processor load balancing have been studied in the paper. EO defines task migration aiming at processor load balancing in execution of graph-represented distributed programs. In the multi-objective EO approach, three objectives relevant to distributed processor load balancing are simultaneously controlled: the function dealing with the computational load imbalance in execution of application tasks on processors, the function concerned with the communication between tasks placed on distinct computing nodes and the function related to the task migration number. An important aspect of the proposed multiobjective approach is the method for selecting the best solutions from the Pareto set. Pareto front analysis based on compromise solution approach, lexicographic approach and hybrid approach (lexicographic + numerical threshold) has been performed in dependence on the program graph features, the executive system characteristics and the experimental setting. The algorithms are assessed by simulation experiments with macro data flow graphs of programs run in distributed systems. The experiments have shown that the multi-objective EO approach included into the load balancing algorithms visibly improves the quality of program execution.

[1]  Lothar Thiele.  Multi-Criteria Optimization , 2010, Encyclopedia of Machine Learning.

[2]  A Study on Extremal Optimization Based Load Balancing Techniques , 2017 .

[3]  Min-Rong Chen,et al.  A novel elitist multiobjective optimization algorithm: Multiobjective extremal optimization , 2008, Eur. J. Oper. Res..

[4]  Andrew Lewis,et al.  A hybrid multi-objective extremal optimisation approach for multi-objective combinatorial optimisation problems , 2010, IEEE Congress on Evolutionary Computation.

[5]  Fernando Guirado,et al.  A New Task Graph Model for Mapping Message Passing Applications , 2007, IEEE Transactions on Parallel and Distributed Systems.

[6]  Stefan Boettcher,et al.  Extremal Optimization: Methods derived from Co-Evolution , 1999, GECCO.

[7]  Bernard P. Zeigler,et al.  Hierarchical, modular discrete-event modelling in an object-oriented environment , 1987 .

[8]  K.J. Barker,et al.  An Evaluation of a Framework for the Dynamic Load Balancing of Highly Adaptive and Irregular Parallel Applications , 2003, ACM/IEEE SC 2003 Conference (SC'03).

[9]  D. Popescu,et al.  Multi‐Criteria Optimization , 2014 .

[10]  Amir Masoud Rahmani,et al.  Load-balancing algorithms in cloud computing: A survey , 2017, J. Netw. Comput. Appl..

[11]  Marek Tudruj,et al.  Improving Extremal Optimization in Load Balancing by Local Search , 2014, EvoApplications.

[12]  Marek Tudruj,et al.  Effective processor load balancing using multi-objective parallel extremal optimization , 2018, GECCO.

[13]  Francis C. M. Lau,et al.  Load balancing in parallel computers - theory and practice , 1996, The Kluwer international series in engineering and computer science.

[14]  Marek Tudruj,et al.  Extremal Optimization applied to load balancing in execution of distributed programs , 2015, Appl. Soft Comput..

[15]  Rafiqul Zaman Khan,et al.  Classification of Task Partitioning and Load Balancing Strategies in Distributed Parallel Computing Systems , 2012 .

[16]  Poonam Singh,et al.  A review of task scheduling based on meta-heuristics approach in cloud computing , 2017, Knowledge and Information Systems.

[17]  Shivali Agarwal,et al.  Comparative Analysis of Various Evolutionary Techniques of Load Balancing: A Review , 2013 .