No abstract

No abstract

No abstract

Summary Nowadays, there is an emerging need for applications based on the Internet of Things (IoT). The sensor nodes present in the IoT network produce data constantly, which directly influences the durability of the network. Therefore, two major challenges while designing IoT systems are network lifetime and energy consumption. Although the ability of IoT applications is huge, there are several limitations such as energy optimization, heterogeneity of devices storage, load balancing, privacy, and security that have to be addressed. These constraints have to be optimized for improving the efficiency of the networks. Hence, the main intention of this paper is to develop the intelligent‐based cluster head selection model for accomplishing green communication in IoT. The two famous algorithms like spotted hyena optimization (SHO) and sun flower optimization (SFO) are integrated to form sun flower‐spotted hyena optimization (SF‐SHO) by utilizing the hybrid meta‐heuristic concept for the optimal cluster head selection. The most significant parameters in IoT networks like delay, distance, energy, temperature, and load are considered for deriving a multi‐objective function to offer optimal clustering. The cluster head of the model is optimally tuned based on the hybrid SF‐SHO, to solve the multi‐objective problem, thus showing the enhanced green communication performance. The proposed model is analyzed and evaluated over different approaches in terms of energy‐specific factors, and the attained results confirm the efficiency of the developed method.

The Internet of Things (IoT) system is composed of several numbers of sensor nodes and systems, which are wirelessly interlinked to the internet. Generally, big data is the storage of a huge amount of information, which causes the classification process to be very challenging. Numerous big data classification approaches are implemented, but the computational time and secure handling of information are the major problems. The aim of the study is the development of big data approach in Internet of Things (IoT) healthcare application. Hence, this paper presents the proposed Dragonfly Rider Competitive Swarm Optimization-based Deep Residual Network (DRCSO-based DRN) for big data classification in IoT. First, the IoT nodes are simulated, and the heart disease patient data are collected through sensors. The routing is done using the Multi-objective Fractional Gravitational Search Algorithm (Multi-objective FGSA). In the Base Station (BS), the big data classification is done. Here, the classification is done using MapReduce (MR) framework, which includes two phases, like mapper and the reducer phase. The input data is initially fed to the mapper phase in the map-reduce (MR) framework. In the mapper phase, feature selection is carried out based on Dragonfly Rider Optimization Algorithm (DROA) in order to select the appropriate features for further processing. The DROA is modeled through merging Dragonfly Algorithm (DA) and Rider Optimization Algorithm (ROA). In the reducer phase, the classification is performed using DRN, which is trained by the developed DRCSO algorithm. The DRCSO is modeled by incorporating DA, ROA and Competitive Swarm Optimization (CSO). In addition, the performance of the developed method is outperformed than the existing approaches such as Linguistic Fuzzy Rules with Canopy Mapreduce (LFR-CM) + Fuzzy classifier, Machine learning-dependent k-nearest neighbors (FML-KNN), MapReduce-Fuzzy Integral-dependent Ensemble Learning Model+Single hidden layer feedforward neural network (MR-FI-ELM + SLFN) and DROA-recurrent neural network (RNN) based on the accuracy, average residual energy and throughput with the value of 0.929, 0.086[Formula: see text]J and 86.585. The proposed method is used to manage and derive meaningful information from the patient’s medical records, medical examinations results and hospital records.

No abstract

No abstract

No abstract