PEFT-Based Hybrid PSO for Scheduling Complex Applications in IoT
Internet of Things (IoT) is one of the buzzwords of the recent era and the most attractive field for researchers. It is defined as a system of connected physical objects which are approachable through the Internet and are capable of exchanging data using
- PDF / 434,975 Bytes
- 14 Pages / 439.37 x 666.142 pts Page_size
- 20 Downloads / 184 Views
Abstract Internet of Things (IoT) is one of the buzzwords of the recent era and the most attractive field for researchers. It is defined as a system of connected physical objects which are approachable through the Internet and are capable of exchanging data using immerse technologies such as sensors, actuators. With the continuous evolution in IoT, number of issues arises such as confined storage space as well as limited processing capabilities. These issues can be resolved by merging IoT with cloud computing, as cloud has the immeasurable storage space as well as processing ability. This combination has proved as a boon for Internet and this combination can also be used to solve workflow scheduling problem as well. Large complex applications are often represented as workflows. Workflow scheduling is one of the eminent obstacles in both IoT and cloud computing. Several approaches have been proposed for workflow scheduling such as heuristic and meta-heuristic approaches. Commonly meta-heuristics approaches include Genetic Algorithm (GA), Simulated Annealing (SA), Particle Swarm Optimization (PSO) and heuristic approaches include Critical Path on Processor (CPOP), Heterogeneous Earliest Finish Time (HEFT), and Predict Earliest Finish Time (PEFT). But, mostly these approaches fail due to increasing of tasks, unable to execute tasks within specified budget, time, cost, and many more reasons. To overcome these above mention issues, this paper presents a hybrid PSO algorithm that uses a combine approach of both heuristic and meta-heuristic techniques namely PEFT and PSO, respectively.
Keywords Cloud computing Internet of Things Workflow scheduling Predict earliest finish time Particle swarm optimization
K. Middha (&) A. Verma U.I.E.T., Panjab University, Chandigarh, India e-mail: [email protected] A. Verma e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 A. K. Luhach et al. (eds.), Smart Computational Strategies: Theoretical and Practical Aspects, https://doi.org/10.1007/978-981-13-6295-8_23
273
274
K. Middha and A. Verma
1 Introduction Cloud Computing (CC) has emerged as an irresistible platform to solve multiple scientific applications. It provides on-demand network access to various measures such as storage, server, and services that can be rapidly purveyed and released with minimum efforts [1]. It has revolutionized the traditional business models by offering multiple services over the internet with the help of virtualization. It employs “pay-per-use” model, that means users are charged according to resource consumption [2]. It uses Internet as well as remote servers to provide numerous services to the users. Cloud has three different forms namely private, public, and hybrid [3]. Each one of them provides different aspects regarding security. Public cloud is nothing but the Internet and can be accessed by anyone while private cloud is managed and owned by single company but it is more expensive than public cloud. Hybrid cloud relies both on private and public cloud and
Data Loading...
