Iraqi Journal of Information and communication technology

Abstract- The rapid proliferation of the Internet of Things (IoT) has significantly transformed modern control systems
by enabling real-time data acquisition, intelligent decision-making, and seamless connectivity across distributed
environments. This paper presents a systematic review of IoT-based control systems, covering studies published
between 2015 and 2025, with emphasis on architectural foundations, enabling technologies, control strategies,
and major application domains. The review paper analyses five key domains, namely industrial automation, smart
manufacturing, healthcare, smart homes and buildings, transportation, and energy and smart grids, while examining
major control and optimization approaches integrated with IoT, including data-driven, adaptive, intelligent, and
distributed control methods. It also identifies four major cross-cutting challenge categories, namely security and
privacy, interoperability and standardization, latency and reliability, and energy efficiency and sustainability, which
continue to limit large-scale deployment. In addition, the study highlights eight emerging research directions,
including edge-fog-cloud co-design, federated learning, TinyML, digital twins, zero-trust security, and safe learningbased
control. The novelty of this review lies in its unified perspective that connects IoT architecture, embedded
intelligence, control and optimization frameworks, and cross-domain applications within a single analytical structure.
By synthesizing existing literature and revealing key research gaps, this work provides a clearer foundation for
developing secure, scalable, and resilient next-generation IoT-driven control systems.

The speedy improvement in Wireless Sensor Network (WSN) technology leads to various applications
such as smart cities, industrial applications, and health care. Energy efficiency is one of the leading challenges
in WSN because the major constraints for the process of communication are the routing protocol and energy
efficiency. A new approach has been introduced to enhance network performance and quality, an improved Ant
Colony Optimisation (ACO) with Elliptic Curve Cryptography (ECC) mechanism-based Clustered Routing Protocol
for WSN. The major sections of the protocol are LEACH-based CH selection, Ant Colony Optimisation, and ECC
mechanism. The protocol offers improved results in optimal path finding and wormhole attack protection. Simulation
results indicate that the proposed scheme yields superior performance metrics in terms of Packet Delivery Ratio
(PDR), network throughput, energy consumption, and security overhead.

Identifying brute-force and dictionary-based login attempts in modern cybersecurity systems has become increasingly challenging, as advanced techniques often fail to detect large-scale intrusion attempts.The aim of this research is to determine the effectiveness of machine learning methods in identifying such attacks in an accurately and efficiently. Two classifiers SVM and GNB, are trained on authentication log data, both with and without PCA for dimensionality reduction.The experimental results indicate that SVM achieves the highest accuracy of 97.24% without PCA and 96.55% with PCA, demonstrating that SVM is robust in high-dimensional feature spaces. Conversely, GNB shows significant with PCA, with accuracy rising from 87.93% to 91.03%, highlighting the importance of feature decorrelation in probabilistic models. The key contribution of this work is the comparative study of lightweight machine learning models demonstrating that PCA improves the performance of correlationsensitive classifiers without undermining the computational efficiency. The results provide a feasible and scalable solution to real-time intrusion detection systems.

Android malware creates a growing security risk due to the increasing number of applications that work on Android platform nowadays. The need for effective detection methods has made the use of machine learning and deep learning a viable solution. This study presents a comparison between different Machine Learning (ML) and Deep Learning (DL) models including Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (BiLSTM), Gated Recurrent Unit (GRU), Support Vector Machine (SVM), K Nearest Neighbor (KNN), and Logistic Regression (LR) on two datasets Android Malware Detection (AMD) and CICMalDroid2020 to evaluate their performance using a unified architecture with the same parameters. Random Forest method is used to select features from the AMD dataset, by calculating the most significant features that have a direct impact on the process of detecting Android malware samples, thereby increasing models’ accuracy. Result for both datasets shows that DL models, especially BiLSTM, outperform other models with accuracy metric that reach 99.77% with full features, 99.88% with selected features for AMD dataset and 89.70% for CICMalDroid2020 dataset.

Microgrids linked with renewable energy systems represent an essential solution to build sustainable power distribution systems with high resilience. The proposed research contributes new findings to existing SCADA system studies by creating a dynamic power-sharing optimization algorithm that solves energy overproduction issues and enhances inter-microgrid coordination. The authors developed a new framework which uses solar PV systems coupled to batteries and diesel generators to maintain stable power output while solar conditions fluctuate. The designed SCADA system functions through MATLAB/Simulink to operate and optimize power distribution throughout four linked houses microgrids. The suggested optimization algorithm distributes surplus power generated by overproducing microgrids to deficit nodes while keeping real-time demand-supply equilibrium as the top priority. The simulation output shows increased power reliability because total power delivery to households rises by 25.1–42.4% during low-irradiance times (sunrise/sunset). The system achieves lowering dependency on generators by 30-45% through its operations of battery optimization and microgrid power exchange techniques. The research demonstrates how SCADA coordination enables better renewable energy network scalability and energy distribution equality and waste minimization within regions like Baghdad that experience varying solar resources.

Internet of things (IoT) technologies have permeated many of the devices used in daily life. Nowhere is this more evident than in smart homes, where regular home appliances are now connected to the internet to gain additional functionality, control and automation of routine tasks. However, like most technologies, convenience often comes at the price of risking user privacy. This paper aims to remedy this security risk by providing the user with a scheme to verify their identity in an anonymous manner. The solution needs to be lightweight and fast since smart home devices are resource constraint. The proposed scheme uses two stages to authenticate remote users, the first is using biometric fingerprints locally on the user’s mobile device and the second relies on user identity (username and password) that are sent to the smart home gateway over a secure VPN tunnel between the mobile device and the gateway. The security of the proposed scheme is verified using the scyther tool which finds security flaws in internet protocols. The performance of the scheme is compared with similar existing schemes in terms of computational cost, and it was found that the proposed scheme achieves a computational cost of 0.02 ms and outperforms all previous schemes in those metrics.

Brain Tumors (BTs) are serious medical conditions characterized by abnormal cellular growth in the brain. Magnetic Resonance Imaging (MRI) may be difficult and time-consuming for brain tumor identification and separation. This research automates the brain tumor classification, focusing on 4 diseases: gliomas, meningioma, and pituitary adenomas. Before employing Discrete Wavelet Transform (DWT) for tumor segmentation, MRI images should be converted to grayscale to improve speed and reduce computer complexity. PCA and Gray-Level Cooccurrence Matrix features are used to maximize feature extraction. The obtained features are classified using supervised K-Nearest Neighbours (KNN). The approach was tested using 2870 images, categorized as: 826 glioma, 822 meningioma, 827 pituitary adenoma, and 395 cancer-free images. The recommended approach used PCA and KNN to achieve 83% accuracy, however DWT and PCA together yielded 85% accuracy. The results show that the automated technique can quickly and accurately diagnose brain tumors.