Proceedings of the 7th International Conference on Communications and Cyber Physical Engineering

Table of Contents

1. Stega Craft – The Art of Hiding Secrets

   Shirisha Reddy K., Shreejit Cheela, Vignya Durvasula, Sidhi Anish Kumar

   Abstract

   StegaCraft is an integrated system preserving traditional steganographic techniques, enabling end-to-end concealment of data within carrier media. Leveraging deep learning and LSB methods, it facilitates hiding text within text, audio within audio, and image within image, while optimizing resource utilization and ensuring data authenticity.

2. Enhancing Cloud Security Through Data Encryption and Dispersion

   K. Karpagavalli, N. Rajiya Bhanu, R. Lavanya, G. Reethika, K. Reddy Pavani

   Abstract

   In the era of widespread digital transformation and cloud adoption, ensuring robust security measures is paramount to safeguarding sensitive data. This paper explores a multifaceted approach to enhancing cloud security through the integration of data encryption and dispersion techniques. Encryption serves as a foundational layer, mitigating the risk of unauthorized access by converting data into unreadable formats without the appropriate decryption keys. However, relying solely on encryption may not be sufficient, as a single point of compromise could potentially expose the entire dataset. To address this vulnerability, data dispersion techniques are introduced, which involve breaking down the encrypted data into fragments and distributing them across various cloud servers. This dispersion not only adds an additional layer of complexity for potential attackers but also reduces the risk of a single point of failure. The combined use of encryption and dispersion not only fortifies data security but also aligns with regulatory compliance requirements. This paper delves into the technical aspects, benefits, and potential challenges associated with implementing this integrated approach, offering insights into how organizations can bolster their cloud security posture in an increasingly interconnected and data-centric landscape. We study the additive application of data encryption and dispersion ways in this in-depth analysis of cloud security. A strong barrier against unwanted access is created by using dispersion to spread pieces over several servers and encryption to make data unreadable. The entire solution addresses the changing face of cyber threats while improving security and adhering to regulations. By carefully analyzing scientific details and possible obstacles, this paper provides useful data for business looking to improve their safety record in the ever-changing world of cloud computing.

3. Online Payment Fraud Detection Using Machine Learning

   P. Swathi, M. Sravani, V. Sivani, M. Susmitha, M. Swetha

   Abstract

   Essentially, electronic bowing ID for a protected environment revolved around our main objective. We will start by utilizing the available datasets for our work with them. Following that, the client will get specific requests for the web locale to be reshaped in order to test an instructional sequence of events. In order to gather the insignificant timberlands, the assessment is applied to the overall appraisal, illuminating the mixture, and the customer provides the most recent dataset finally providing reassurance on the accuracy of the results related data. From that point on, a portion of the specified characteristics will be payable, taking into consideration the area of the impacted risk and how the reasonable model establishes an association.

4. Review on Credit Card Fraud Detection Techniques

   P. Sreesudha, P. Sanjana Reddy, T. Bhavani Goud, T. Jyoshna Reddy, T. Sri Vaishnavi

   Abstract

   The widespread adoption of credit cards as the dominant electronic payment method heightens the vulnerability to fraudulent activities, given the substantial volume of daily electronic transactions. Credit-Card fraud has incurred significant financial losses for credit card companies, and they are actively seeking optimal procedures and advanced technology to detect and mitigate instances of fraudulent credit card transactions. In this study, many ML Detection Techniques to spot credit-card fraud have been examined, highlighted, and contrasted.

5. Road Traffic Condition Monitoring Using Deep Learning

   Kumari Jelli, P. Suma Poojitha, N. Sai Puneeth Rao, Rajesh Saturi

   Abstract

   Each second, the traffic surveillance system gathers a vast amount of data related to road traffic. Observing these data points manually is a laborious undertaking that also necessitates the use of personnel. Control and monitoring can be done with Deep Learning Convolutional Neural Networks. To create training data, traffic analysis data is first obtained. A transport network is created by transforming the network into a transport application and reintroducing it using self-generated data. This transportation network can explore large areas. Moreover, it is capable of being universally implemented. Moreover, DLCNN is employed to forecast traffic conditions, including but not limited to congested traffic, light traffic, accidents, and fires, based on test samples. In conclusion, the simulations demonstrated that the performance of the proposed DLCNN was superior to that of the existing model.

6. Detection of Cardiovascular Diseases in ECG Images Using Machine Learning and Deep Learning Methods

   P. Bhavana, K. Navya Sri, Ch. Venkatesh, K. Nanda Kishore, K. Swarupa Rani

   Abstract

   The urgent global problem of cardiovascular diseases is addressed in this study, with a special emphasis on heart disorders, which continue to be the world's leading cause of mortality. Early diagnosis is essential, and because electrocardiograms (ECGs) are non-invasive and reasonably priced, they are an essential tool for heart health monitoring. This study uses sophisticated learning methods; transfer learning methods like SqueezeNet and AlexNet, along with customized Convolutional Neural Network (CNN) primarily used to increase prediction accuracy. The four major cardiac abnormalities that these techniques seek to identify are unusual heartbeat, myocardial infarction, if any history of myocardial infarction, and any normal instances. This model's remarkable performance is what makes it unique; it is attained by combining deep learning and conventional machine learning methods to extract important information. This study demonstrates how artificial intelligence can revolutionize the healthcare industry, particularly in the area of medical condition prediction via picture analysis.

7. A Comprehensive Review on Social Network Mining Using Machine Learning Algorithms

   Prabhat Kumar Tiwari, Prashant Shukla

   Abstract

   The integrated approach of utilizing social network analysis and machine learning techniques to mine social network data has become a challenging and increasingly popular domain in machine learning research, and the pace of interdisciplinary research is accelerating. It entails gathering, processing, and analysing large volumes of data from social media platforms like Twitter, Facebook, Instagram, and LinkedIn. Social network mining with machine learning aims to extract useful insights from the data so that decision-making and the user experience can be enhanced. This could involve issues such as identifying influencers, predicting user behaviour, detecting spam and bots, and understanding user sentiment in discussions. Machine learning can be used in a variety of applications for social network mining including in areas such as marketing, analysis of public opinions to social media management. However, there are also some ethical concerns around using the technology, most notably in relation to bias and privacy issues. This paper explores how the machine learning-based social network mining offers much promise, potentially bringing substantial insight into user behaviour and social trends.

8. Security, Privacy Challenges and Security Measures Using IOT Models

   Shaik Sajid, I Veera Raghava Rao, Karnati Pravallika, U. Veerajanaki

   Abstract

   As the IoT continues to revolutionize industrial automation, the deployment of sensors has become widespread, offering real-time data collection and control. The Internet of Things permeates our daily lives, safeguarding privacy within security systems has become an urgent priority. Conventional security systems, frequently reliant on intrusive monitoring methods, raise user concerns about ethics and privacy. While connectivity offers many benefits, it also creates weaknesses that malicious actors can target. This study examines the various security challenges faced by IoT-enabled sensors in automation systems, including data privacy risks, unauthorized access, and potential system manipulation. The overview of Cyber Physical Production System (CPPS), elucidating their architecture, components. The various sources of attacks targeting automation systems, which play a critical role in modern industrial environments. This research contributes to advancing understanding of Cyber Physical Production System (CPPS) and cyber security in industrial automation, guiding future efforts to secure cyber-physical systems against evolving threats and measures.

9. Smart Healthcare Monitoring Systems Using IOT

   Ponugoti Charan Sai, Lakkakula Hari Prasad, Banoth Akhila

   Abstract

   The Internet of Things has brought about significant advancements in healthcare, enabling the seamless integration of devices and systems to improve patient care and outcomes. Healthcare is shifting it is focus from solely treating illnesses to proactive health management, this means promoting preventive care, lifestyle changes, and personalized interventions to enhance overall well-being, with the rise of Internet of Things, and more data than ever before is being generated and aggregated in healthcare. This includes information from wearable devices, electronic health records, and other sources, allowing for comprehensive analysis and insights. The use of Internet of Things and data analytics in healthcare enables a more patient-centric approach, by the leveraging technology and analyzing data, healthcare providers can deliver personalized care, make informed decisions, and optimize health outcomes. This article addressed with conventional IOT models on healthcare monitoring, privacy and security and a comparison made with existing algorithms.

10. Pest Detection Methods Using Machine Learning Algorithm for Agriculture Applications

    P. Gayathri Priya, G. Venkata Hari Prasad, M. Aryan, B. Lakshmi Krishna Vamsi

    Abstract

    Apple orchards are expanding globally, with apples being a popular fruit crop in the world. The biggest challenge faced by these orchards is the threat of the apple worm, a harmful worm which destruct the apple crop. Utilizing Internet of things technology, discern gadgets equipped with machine learning algorithms canefficientlycollectdataregardingtocropandanalyzeitinreal-time, and detect anomalies such as the presence of codling moths in the apple crop. These devices can automatically capture images of traps, pre-process the images, isolate insects like cultivate worms for classification, and inform farmers if any codling moths are identified in the farm land. A low power platform is operated by Internet of things and sun cells, ensuring it can run continuously over low power networks without human intervention means an automatic technique. The hardware used in neural network working contains an Intel neural network and raspberry pi3 which is used to construct the gadget.

11. Semantic Feature-Based Image Retrieval Using Vgg19 Model: Corel Images Dataset and Elu Activation Function

    V. Sagar Reddy, Chevella Anil Kumar, G. Jaya Sheela, M. Harsha Vardhan, Afsar Hussain

    Abstract

    The affordability and accessibility of digital image sensors and internet technologies have led to the establishment of extensive picture databases for various applications. The demand for efficient picture retrieval techniques that fulfil user requirements is underscored by the proliferation of these image repositories. Considerable efforts have been invested in refining content-based, or semantic-based, image retrieval methods, aiming to overcome the semantic gap between the attributes of low-level image and human visual perception. This study introduces the development and implementation of Semantic-Based Image Retrieval (CBIR) employing the VGG19 Model and ELU (Exponential Linear Unit Activation Function) in response to the burgeoning research interest in this domain. Additionally, to stimulate further research, this investigation provides CBIR architecture overview, contemporary methods of low level feature extraction, machine learning techniques, similarity metrics, and performance evaluation metrics. Recent advancements in deep learning have yielded remarkable outcomes. The effectiveness of a classification system hinges on the feature extraction quality of an image; whereby higher-quality features correspond to improved accuracy. Despite the impressive performance of many deep learning-based systems in image classification tasks, their capacity to extract comprehensive image information remains limited due to various inherent challenges, consequently compromising overall classification accuracy. Leveraging the Corel image dataset, this study to refine semantic-based image retrieval and classification methodologies.

12. To Investigate and Analyze the Applications and Impact of Machine Learning Techniques in Enhancing Computational Processing Capabilities

    G. Archana, S. Gopalakrishna, B. Kishore, K. Haripalreddy, V. Sumathi, Pradeep Kumar

    Abstract

    Machine learning (ML) has revolutionized computational processing, driving advancements across various industries. This study investigates the applications and impact of ML techniques in enhancing computational capabilities, focusing on supervised learning, unsupervised learning, reinforcement learning, and deep learning. Specialized hardware such as GPUs, TPUs, FPGAs, and ASICs plays a crucial role in optimizing ML tasks, improving efficiency, speed, and performance. By leveraging these advanced technologies, ML enables real-time data analysis, intelligent decision-making, and the handling of complex data structures. This consonance between ML and specialized hardware facilitates innovations in healthcare, finance, autonomous systems, and consumer technology. Challenges such as energy efficiency and scalability persist, necessitating ongoing research. Future directions include the exploration of quantum computing and the development of more energy-efficient hardware.

13. IoT Enabled Multi-purpose A9G Based Tracker for Enhanced Monitoring Applications

    Rambabu Kambhampati, Sravan K. Vittapu, Ravichand Sankuru, Chennadi Thirumala, Gail Hemakanth, Ganagam Umeshchandra, Rambabu Kambhampati

    Abstract

    This paper presents an A9G-based monitoring and spying module that makes use of the module’s capabilities to provide covert surveillance and precise location tracking. To produce a small and flexible tracking device, the hardware design integrates the A9G module with antennas, power supply circuits, sensors, and auxiliary interfaces. In order to guarantee dependable functioning and data privacy, the firmware/software development is concentrated on integrating GPS/GNSS tracking algorithms, communication protocols (GSM, GPRS, Bluetooth), and security measures. Applications like asset tracking, vehicle monitoring, and personal surveillance are served by the tracking module’s real-time monitoring, remote configuration, and secure communication features. Transparency in tracking operations, adherence to legislation, and privacy-preserving design principles all address ethical and legal concerns. The proposed architecture provides a suitable solution in terms of user needs in tracking and spying applications with ethical standards.

14. Machine Learning Algorithms for Power Quality Improvement in Distributed Generation System

    Ch. Shravani, P. Rajesh Kumar, M. Rajitha, R. N. Bhargavi, G. SreeLakshmi

    Abstract

    The integration of distributed generation systems (DGS) presents challenges related to maintaining high power quality in electrical grids. Traditional methods for power quality improvement often fall short in addressing the dynamic and decentralized nature of DGS. Machine learning (ML) algorithms offer a promising solution by leveraging data-driven approaches to optimize power quality. This paper presents an examination of machine learning algorithms for power quality improvement in DGS. Various Techniques within the domain of machine learning, including SVM, ANN, and decision trees, and reinforcement learning are discussed in the context of load forecasting, fault detection, harmonic mitigation, optimal control of energy storage systems, and smart grid coordination. Additionally, the paper highlights the potential benefits of ML algorithms in enhancing grid reliability, stability, and efficiency while accommodating the increasing penetration of renewable energy sources. Case studies and future research directions are also discussed to illustrate the practical application and ongoing advancements in this field. Overall, machine learning emerges as a promising tool for addressing power quality challenges in distributed generation systems, contributing to the advancement of a more robust and sustainable electrical grid.

15. To Optimize Predictive Analytics with Machine Learning Techniques

    P. Anusha, P. Pavankumar, A. Venkata Laxmi, P. Navitha, G. Rajender, S. Naga Jyothi

    Abstract

    Predictive analytics is pivotal for data-driven decision-making in diverse fields, including finance, healthcare, and urban planning. This paper investigates the optimization of predictive analytics through advanced machine learning techniques. We examine various machine learning models, including regression, classification, and neural networks, and assess their performance using a range of optimization strategies. Key methods discussed include Gradient Descent, Adam optimization, and regularization techniques such as L1 and L2. Our study demonstrates that these optimization approaches significantly enhance model accuracy and generalization. Additionally, we explore hyperparameter tuning methods, such as Grid Search, Random Search, and Bayesian Optimization, to identify optimal configurations for predictive models. The results indicate substantial improvements in predictive performance and offer practical insights for model selection and implementation. Future research directions include exploring sophisticated optimization algorithms and integrating domain-specific adaptations to further refine predictive models. This work provides a comprehensive framework for optimizing predictive analytics, contributing to more accurate and reliable data-driven decision-making.

16. A Comprehensive Study on Stock Market Forecasting Using AI and ML Techniques

    Yash Koshti, Rahul Jain, Himanshu Barhaiya, Ram Pratap Singh, Neelesh Gour

    Abstract

    This study looks at the usage of cutting-edge AI as well as ML technologies for forecasting of stock market, with an emphasis on short-term projections for notable US-listed businesses in a range of industries. Our program seeks to generate directional forecasts by utilizing historical price data, technical indicators, and sentiment analysis of news. We explore a variety of stock market analytical topics, such as risk assessment, pattern detection, and machine learning-based investment return estimates. Using historical data, the paper thoroughly investigates the Efficient Market Hypothesis and its consequences on stock price prediction. The efficacy of many methods and models, including LSTM networks, ARIMA, and GARCH, in financial prediction is assessed. We also go over issues with using technology-driven forecasting techniques, including data scarcity, overfitting, and moral dilemmas.

17. A Power Efficient Systolic Array Architecture-Based Multiplier for DSP Applications

    Bhattu Hari Prasad Nayak, Sravan K. Vittapu, Gangadi Chandra Vardhan Reddy, Dindu Akhil goud, Banuka Sanjay

    Abstract

    A multiplier is an essential part of many digital integrated circuits (ICs) used in arithmetic computations. Multipliers for modern VLSI (Very Large-Scale Integration) circuits must be quick, power-efficient, and small in size. In the quickly changing world of today and in the current technological environment, the widespread use of computers and the Internet has been facilitated by the need for strong and quick data processing. Nevertheless, using parallel computing effectively satisfies these expectations. In particular, it offers a brand-new method for utilizing Systolic Architecture on Field Programmable Gate Arrays (FPGAs) and other Reconfigurable Systems (RS) to speed up Matrix Multiplication. Systolic arrays are highly parallel computer architectures that can handle difficulties by performing operations on data pieces in parallel. This makes them ideal for jobs like matrix multiplication. Our main goal is to create a system that greatly increases data processing speed while minimizing path delay. We can accomplish this by using Xilinx Software. Because FPGAs may be used to customize and reconfigure hardware, they are a great option for optimizing systolic array topologies to meet certain computing requirements. An effective way to process data at high speeds in complicated computational settings is to leverage the capabilities of parallel computing on FPGA devices.

18. Tackling the Flexible Job Shop: A Survey on Optimization Methods for Real-World Production Scheduling

    Kilari Jyothi, R. B. Dubey

    Abstract

    Scheduling is crucial since it maximizes productivity while lowering costs, lead times, cycle times, etc. An NP-hard combinatorial optimisation problem called the Flexible Job Shop Scheduling Problem (FJSSP) has several real-world applications, including industrial facilities and cloud computing. There has been a lot of interest in finding a solution because of how complicated and significant this problem is. The purpose of this study is to assess how well different flexible job shop scheduling strategies maximize production schedules. The study’s goal is to provide recommendations for selecting the best method for real-world industrial settings and other real-time applications. The current solution techniques for the FJSP are categorized as precise algorithms, heuristics, meta-heuristics and hybrid algorithms, In this work, a recent survey on hybrid techniques in the field of FJSS is thoroughly studied.