1. A new multi-agent particle swarm algorithm based on birds accents for the 3D indoor deployment problem
- Author
-
Nejah Nasri, Sami Mnasri, Adrien van den Bossche, Thierry Val, Centre National de la Recherche Scientifique - CNRS (FRANCE), Institut National Polytechnique de Toulouse - INPT (FRANCE), Université Toulouse III - Paul Sabatier - UT3 (FRANCE), Université Toulouse - Jean Jaurès - UT2J (FRANCE), Université Toulouse 1 Capitole - UT1 (FRANCE), École Nationale d'Ingénieurs de Sfax - ENIS (TUNISIA), Université de Sfax (TUNISIA), Réseaux, Mobiles, Embarqués, Sans fil, Satellites (IRIT-RMESS), Institut de recherche en informatique de Toulouse (IRIT), Université Toulouse 1 Capitole (UT1), Université Fédérale Toulouse Midi-Pyrénées-Université Fédérale Toulouse Midi-Pyrénées-Université Toulouse - Jean Jaurès (UT2J)-Université Toulouse III - Paul Sabatier (UT3), Université Fédérale Toulouse Midi-Pyrénées-Centre National de la Recherche Scientifique (CNRS)-Institut National Polytechnique (Toulouse) (Toulouse INP), Université Fédérale Toulouse Midi-Pyrénées-Université Toulouse 1 Capitole (UT1), Université Fédérale Toulouse Midi-Pyrénées, Laboratoire d'Analyse et Commande des Systèmes (LACS), Faculté des Sciences Mathématiques, Physiques et Naturelles de Tunis (FST), Université de Tunis El Manar (UTM)-Université de Tunis El Manar (UTM), Université Toulouse - Jean Jaurès (UT2J), IUT Toulouse 2 Blagnac, and Institut National Polytechnique de Toulouse - Toulouse INP (FRANCE)
- Subjects
0209 industrial biotechnology ,Computer science ,Evolutionary algorithm ,Réseaux et télécommunications ,02 engineering and technology ,computer.software_genre ,Machine Learning ,Many-objective optimization ,[INFO.INFO-NI]Computer Science [cs]/Networking and Internet Architecture [cs.NI] ,020901 industrial engineering & automation ,Local optimum ,Architectures Matérielles ,Stress (linguistics) ,0202 electrical engineering, electronic engineering, information engineering ,Cluster Analysis ,Instrumentation ,Multi-agent ,education.field_of_study ,3D indoor deployment ,Behavior, Animal ,Applied Mathematics ,Particle swarm optimization ,Systèmes embarqués ,Computer Science Applications ,Algorithms ,Experimental validation ,Population ,Système d'exploitation ,Context (language use) ,Machine learning ,Birds ,Imaging, Three-Dimensional ,Cluster (physics) ,Animals ,Computer Simulation ,Electrical and Electronic Engineering ,education ,Ecosystem ,business.industry ,020208 electrical & electronic engineering ,DL-IoT collection networks ,Reproducibility of Results ,Informatique et langage ,Models, Theoretical ,Control and Systems Engineering ,Software deployment ,Artificial intelligence ,Vocalization, Animal ,business ,computer ,Accent based PSO ,Software - Abstract
International audience; The 3D indoor deployment of sensor nodes is a complex real world problem, proven to be NP-hard and difficult to resolve using classical methods. In this context, we propose a hybrid approach relying on a novel bird's accent-based many objective particle swarm optimization algorithm (named acMaPSO) to resolve the problem of 3D indoor deployment on the Internet of Things collection networks. The new concept of bird's accent is presented to assess the search ability of particles in their local areas. To conserve the diversity of the population during searching, particles are separated into different accent groups by their regional habitation and are classified into different categories of birds/particles in each cluster according to their common manner of singing. A particle in an accent-group can select other particles as its neighbors from its group or from other groups (which sing differently) if the selected particles have the same expertise in singing or are less experienced compared to this particle. To allow the search escaping from local optima, the most expert particles (parents) ''die'' and are regularly replaced by a novice (newborn) randomly generated ones. Moreover, the hybridization of the proposed acMaPSO algorithm with multi-agent systems is suggested. The new variant (named acMaMaPSO) takes advantage of the distribution and interactivity of particle agents. Experimental, numerical and statistical found results show the effectiveness of the two proposed variants compared to different other recent state-of-the-art of many-objective evolutionary algorithms.
- Published
- 2019
- Full Text
- View/download PDF