Swarm Intelligence Optimization

Algorithms and Applications

Abhishek Kumar, Pramod Singh Rathore, Vicente Garcia Diaz, Rashmi Agrawal

• Publisher: Wiley-Scrivener
• Year: 2020
• Print ISBN: 9781119778745
• eBook ISBN: 9781119778905
• Edition: 1
• Topic: Computer Science
• Subtopic: Artificial Intelligence (AI) & Semantics
• Index: Computer Science

Resource optimization has always been a thrust area of research, and as the Internet of Things (IoT) is the most talked about topic of the current era of technology, it has become the need of the hour. Therefore, the idea behind this book was to simplify the journey of those who aspire to understand resource optimization in the IoT. To this end, included in this book are various real-time/offline applications and algorithms/case studies in the fields of engineering, computer science, information security, and cloud computing, along with the modern tools and various technologies used in systems, leaving the reader with a high level of understanding of various techniques and algorithms used in resource optimization.

1
A Fundamental Overview of Different Algorithms and Performance Optimization for Swarm Intelligence

Manju Payal¹*, Abhishek Kumar^2† and Vicente García Díaz³

¹Software Developer, Academic Hub, Ajmer, India Rajpura, Punjab, India

²Chitkara University Institute of Engineering and Technology, Chitkara University,

³Department of Computer Science, Universidad de Oviedo, Asturias, Spain

Abstract

Swarm Intelligence (SI), normally, is based on the problem-solving ability. It solves the problem using the interaction of simple information processing units. It contains some types of the terminologies which are the distribution, multiplicity, messiness, stochasticity, and randomness. The problem-solving approach is based on three terminologies which are suggested by the SI. These terminologies are the creativity, cognition capabilities, and learning. It contains some types of the methods which depend on the optimization techniques. These methods are the ABC, ACO, and PSO. Here, ABC is referred as the Artificial Bees Colony, ACO is referred as the Ant Colony Optimization, and PSO is referred as the Particle Swarm Optimization. It also depends on the scheduling optimization. It is the massive number of homogenous. These methods have grown as, of late, with a bunch of population-based algorithms, nature-driven equipped to quick, deliver least effort, and robust answers to few composite issues. Optimization is the term of the chosen best solution of the problems. It is chosen as the best solution from the set of the solutions. This solution is based on some types of features which are the highest achievable performance, cost effectiveness, and so on.

Keywords: Swarm intelligence, ant colony optimization, artificial bee colony, machine Learning, partical swarm optimization, population algorithms, agents, artificial intelligence

1.1 Introduction

SI is an essential section of the AI. Here, SI is referred as the Swarm Intelligence and AI is referred as the Artificial Intelligence. It is the bio-inspired computation [1]. Now, it has been recognized as a developing field. It was developed by the two professors. These professors are Gerardo Beni and Jing Wang. It was developed since 1989. It is the based on the cellular robotic systems. It consists of many types of algorithms. These algorithms depend on the bio-inspired

computation. Now, it has the most growing popularity because it consists many types of the SI algorithms. These algorithms consist of many types of the features such as versatility and flexibility. It consists of two most important features which are adaptability and self-learning capability [2]. This features the performance by the SI algorithms. It has identified different types of the application areas. Lately, SI has developed in prevalence with the expanding prominence quality of NP-hard issues where the discovery of a global ideals turns out to be practically inconceivable continuously situation [3]. The quantity of potential arrangements which may exist in such issues frequently will, in general, be unending. In such circumstances, finding a work capable arrangement inside time constraints gets significant. SI discovers its utility in taking care of nonlinear structure issues with real-based applications, thinking about practically all zones of sciences, designing and enterprises, from information mining to enhancement, computational insight, commercial arranging, in bioinformatics, and commercial modern applications. Some top applications contain incorporate route control, planetary motion sensing, interferometry, malignant tumor detection, micro-robot control, micro-robot control, and control [4].

There are some types of instances available in the SI which are the flock of birds, ant colonies, bacterial growth, schools of fish, and so on. It does not contain any type of the centralized control. It depends on the collection of the behavior in the nature [5].

The fundamental objective of the SI is to enhance the performance of the complex problems. It also enhances the solution of the complex problems. The incredible accomplishment of natural swarm systems has led many researchers to find out how to solve complex problems by the swarms in nature [6]. There are three types of the SI algorithms available, which provide the best solutions with the optimal issues. These algorithms are the BA, BCO, and ACO. Here, BA is referred as the Bat algorithms, BCO is referred as the Bee Colony Optimization, and ACO is referred as the Ant Colony Optimization [7].

SI, normally, is based on the problem-solving ability. It solves the problem using the interaction of simple information processing units [8]. It contains some types of the terminologies which are the distribution, multiplicity, messiness, stochasticity, and randomness. The problem-solving approach is based on three terminologies which are suggested by the SI. These terminologies are the creativity, cognition capabilities, and learning. It contains some types of the methods which depend on the optimization techniques. These methods are the ABC, ACO, and PSO. Here, ABC is referred as the Artificial Bees Colony, ACO is referred as the Ant Colony Optimization, and PSO is referred as the Particle Swarm Optimization. It also depends on the scheduling optimization [9].

It is the massive number of homogenous. These methods have grown, as of late, with a bunch of population-based algorithms, nature-driven equipped to quick, deliver least effort, and robust answers to few composite issues [10]. Optimization is the term of the chosen best solution of the problems. It is chosen as the best solution from the set of the solutions. This solution is based on some types of the features which are the highest achievable performance, cost effectiveness, and so on [11]. Finding an option with the most practical or most noteworthy feasible execution under the given requirements is by augmenting wanted factors and limiting undesired ones. In correlation, amplification implies attempting to achieve the most noteworthy or greatest outcome or result regardless of cost [12].

This term can be characterized as the joined mindset of decentralized or self-sifted through structures in normal or reproduced [13]. The inspiration begins from commonly natural structure. SI is a trademark computation since it is created by following the evolution and task behavior of basic animals and dreadful little creatures. The instance of the swarm of birds is the flock of birds. The second instance of the SI is the bee swarming. It is based on the agents that are bees. In case we are about watch a single insect or a bumble bee, we will appreciate that they are not all that sharp, yet, rather their settlements are. Multitude information can help individuals to understand complex systems, from truck controlling to military robots. A settlement can illuminate any issue, for instance, ACO calculation is utilized for finding the most limited way in the system directing issue, and Particle Swarm Intelligence is utilized in optical system improvement [14]. As an individual, the multitude might be little fakers; yet, as provinces, they respond quickly and enough to their condition. There are two kinds of social associations among swarm people, to be specific, direct communication and roundabout collaboration [15]. Direct collaborations are the undeniable cooperation through visual or sound contact, for instance, winged creatures communicate with one another with sound. Roundabout communication is known as the Stigmergy [16], where operators collaborate with the earth. A pheromone trail of ants is a case of backhanded association.

SI, an essential part in the field of AI, is continuously retrieving conspicuousness, as increasingly more high multifaceted nature issues require arrangements, which might be imperfect, yet feasible inside a sensible timeframe. For the most part motivated by natural frameworks, swarm knowledge embraces the aggregate conduct of a composed gathering of creatures, as they endeavor to endure [17]. This investigation plans to examine the overseeing thought, distinguish the potential application territories, and present a nitty gritty review of eight SI calculations [18]. The recently evolved calculations examined in the examination are the creepy crawly–based calculations and creature-based calculations in minute detail. All the more explicitly, we center around the calculations roused by ants, honey bees, fireflies, sparkle worms, bats, monkeys, lions, and wolves [19]. The motivation examinations on these calculations feature the manner in which these calculations work [20]. Variations of these calculations have been presented after the motivation examination. Explicit territories for the utilization of such calculations have likewise been featured for analysts inspired by the space. The investigation endeavors to give an underlying comprehension to the investigation of the specialized parts of the calculations and their future extension by the scholarly world and practice [20].

Moreover, SI is not just purposely utilized in multitudes of specialized gadgets. Additionally, in the plan of (advancement) calculations, swarm insight can be applied by taking motivation from multitudes of creatures. In numerous real-world optimization issues, the real target method is not known. For example, if numerous sets of 2D medical pictures, one from CT and one from MRT, must be enlisted, i.e., be adjusted so as to make their structures overlay in an important manner, the pictures must be changed to streamline a closeness metric. The genuine target work relies upon the pictures and cannot be productively streamlined by exceptionally planned calculations. So, this is a run of the mill case, where purported metaheuristics are applied, i.e., techniques that get the target work f as a black box and are looking for an information x* that upgrades f.

1.2 Methodology of SI Framework

It has no any centralized management. It is a decentralized system. Mostly, it contains two advantages which are as follows:

• 1. Agents
• 2. Self-Organization

Agents are the collection of the possible solution to a given problem. It is not centralized on the particular agents.

Adding to this are, without a doubt, the SI focal points. Multitude does not have any outside administration; however, every specialist in the multitude controls their conduct self-sufficiently. Specialist for this situation speaks to a potential answer for a given issue. In view of that, we can reason the subsequent bit of leeway, which is self-association. The insight does not concentrate on the individual operator however rises in the mult...

Table of contents

• Cover
• Table of Contents
• Title page
• Copyright
• Preface
• 1 A Fundamental Overview of Different Algorithms and Performance Optimization for Swarm Intelligence
• 2 Introduction to IoT With Swarm Intelligence
• 3 Perspectives and Foundations of Swarm Intelligence and its Application
• 4 Implication of IoT Components and Energy Management Monitoring
• 5 Distinct Algorithms for Swarm Intelligence in IoT
• 6 Swarm Intelligence for Data Management and Mining Technologies to Manage and Analyze Data in IoT
• 7 Healthcare Data Analytics Using Swarm Intelligence
• 8 Swarm Intelligence for Group Objects in Wireless Sensor Networks
• 9 Swam Intelligence–Based Resources Optimization and Analyses and Managing Data in IoT With Data Mining Technologies
• 10 Data Management and Mining Technologies to Manage and Analyze Data in IoT
• 11 Swarm Intelligence for Data Management and Mining Technologies to Manage and Analyze Data in IoT
• 12 Swarm Intelligence–Based Energy-Efficient Clustering Algorithms for WSN: Overview of Algorithms, Analysis, and Applications
• 13 Swarm Intelligence for Clustering in Wireless Sensor Networks
• 14 Swarm Intelligence for Clustering in Wi-Fi Networks
• 15 Support Vector in Healthcare Using SVM/PSO in Various Domains: A Review
• 16 IoT-Based Healthcare System to Monitor the Sensor’s Data of MWBAN
• 17 Effectiveness of Swarm Intelligence for Handling Fault-Tolerant Routing Problem in IoT
• 18 Smart Epilepsy Detection System Using Hybrid ANN-PSO Network
• Index
• End User License Agreement