8333 research outputs found
Sort by
Fault-Tolerant Workflow Scheduling for Microservices in Cloud Environments Under Time and Cost Constraints
No full textThe main idea is a new way of making the scheduling of workflows practical, taking into account the intricate tradeoffs between efficiency, reliability and adherence to time constraints in cloud environments and this research explores its applications based on microservices. This study uses the greedy scheduling for microservices (GSMS) algorithm which runs with container over VM in a two layer resource structure to allocate resources greedily and adjust them dynamically. The principal objective for GSMS is to ensure that tasks are performed at the lowest possible cost while remaining within given performance and reliability requirements. Besides theoretical frameworks, this paper also shows the application of GSMS in Amazon Web Services (AWS) leveraging Amazon Elastic Kubernetes Service (EKS) for container orchestration. Real world datasets from Google and Alibaba were used to simulate workload scenarios in order to verify how effective the algorithm is. During evaluation, the performance is compared against other scheduling algorithms on the basis of different metrics using various performance indicators such as cost optimization, adherence to deadlines, and fault tolerance. By exposing those nuanced tradeoffs on this studies region, this investigation tackles know-how in cloud computing and micro-services control inside the clouds
Kubernetes Security Best Practices
No full textThe security and adaptation that comes along with Kubernetes, the systems used to build containerized apps and microservices, is what this research examines. However, as more and more industries adopt Kubernetes, it also comes with additional complexities and opens up whole new areas of risk to security, from the vulnerabilities in container images to misconfigured policies. Adding to the complexity is the fact that many organizations use Kubernetes in conjunction with popular cloud platforms like AWS, which requires them to effectively address two types of security challenges the cloud-native and the Kuber-native. Related works identify important threats, show how microservices increase complexity, and propose new tools or techniques to improve Kubernetes security. The objective of this work is to evaluate the security of Kubernetes systems as well as existing and future security approaches. Thematic analysis revealed the key security issues such who are deploying these systems, what are the top insecure practices such as RBAC misconfigurations, container image vulnerabilities, network security, complexity of microservices, continuous monitoring, and zero-trust-based policy frameworks. The research indicated considerable shortcomings in existing security approaches, although work is being done to remedy issues surrounding vulnerabilities created by misconfigurations and imperfect network policies. Besides theoretical aspects, practical implementation was done on the live environment of Kubernetes with the actual approach and practices for implementing security on deploy. Important implementation steps were addressed, like configuring RBAC, securing container images, applying network policies, and configuring monitoring with Prometheus. The work provides insights into Kubernetes security, which is valuable for organizations looking to enhance the security of their environments. Organizations can be better equipped with a more secure, scalable and resilient containerized infrastructure by remediating the identified weaknesses and implementing strong security measures
Enhancing Resource Management in Cloud with Blockchain for Transparent Recording of Scaling Events
No full textCloud computing is a growing field where many services are used by businesses due to its features such as scalability and ease of access to resources on demand. Since the cloud providers manage the resource allocation, transparency in it would be a requirement. Bringing in a decentralized approach would make the system more transparent to the cloud users. This study proposes to build a system that would store the scaling information and the logs that help to identify the scaling requirement to Hyperledger Fabric. To perform this, external monitoring tools are used along with AWS Services such as EC2, SQS, DynamoDB and designed to build a well-performing system under various workloads. Various experiments were conducted to identify the issues in the system and identify the design that could process all the requests successfully. The system’s performance outperformed the previous research and showed improved latency and throughput in the benchmarking test performed using Hyperledger Caliper
Monitoring the Security Vulnerabilities in CI/CD Pipeline Using DevSecOps Security Testing Tools
No full textWith the increasing use of Development & Operations (DevOps) in the Software Development Life Cycle, the security is an important concern which needs to be taken into consideration. So, “Static Application Security Testing” (SAST) and “Dynamic Application Security Testing” (DAST) have been integrated in CI/CD Pipelines to check the vulnerabilities of a software at the build and deployment stage. However, the current tools were not efficient enough to detect the vulnerabilities in real-time, so in this research the main motive is to overcome those gaps by advancing the security tool to check vulnerabilities and attacks in a software. By integrating new methods like “Interactive Application Security Testing” (IAST) it provides a real-time vulnerability detection by monitoring the internal behaviour of applications. In this research, the demonstration of improved results of security testing tools are showcased. Where “Snyk” (SAST) tool identified 82 vulnerabilities in more effective and efficient way than the previous work, on the other hand “Stackhawk” (DAST) tool delivered decent results, lastly the integration of “Datadog” (IAST) was successfully done, as it does the work of both SAST & DAST in one tool
Securing Cloud Data: Developing a File Storage System on AWS S3 for Enhanced Security
No full textProtecting sensitive information is one of the great challenges that an organization and individuals face in this era of cloud computing. This project focuses on designing and implementing a secure, scalable, and cost-effective file storage system using AWS S3 along with its integrated services: KMS, IAM, and VPC. The advanced encryption, network isolation, and strict access controls make the system assure confidentiality, integrity, and accessibility of data. Important characteristics involve server-side encryption by the help of AWS KMS, versioning in place, and the least privileged approach adopted in IAM policies. Both functional and security tests result in 98% of file operations with average 200 ms of response times and 100% of data integrity. Leveraging the cloud infrastructure of AWS, as well as automating key processes, the project demonstrated the possibility of using a mix of cloud-native tools with custom development to address significant data security challenges. It thus provides a blueprint for organizations looking to store sensitive information securely in a cloud environment
Enhancing Security in IOT Home Automation Systems
No full textThe integration of IoT technologies into home automation has transformed modern living, offering greater convenience, efficiency, and centralized control over household systems. However, the interconnected nature of these devices introduces significant security vulnerabilities, exposing users to risks such as data breaches, unauthorized access, and system disruptions. These challenges are compounded by the resource limitations of IoT devices, which often struggle to support advanced security mechanisms. This research focuses on identifying critical security threats in IoT-enabled smart homes and proposes a comprehensive framework to address these vulnerabilities. Guided by Zero Trust Architecture principles, the study incorporates advanced encryption, multi-factor authentication protocols, and real-time anomaly detection models to protect devices against evolving cyber threats. A user-centric approach is also emphasized, enhancing homeowner awareness and encouraging secure practices to mitigate risks effectively. The methodology combines theoretical analysis with empirical testing in a simulated smart home environment, allowing for the evaluation of security solutions under real-world conditions. Results demonstrate that adaptive security measures can effectively address the evolving threat landscape without compromising device functionality. By bridging the gap between innovation and security, this research offers practical insights into enhancing the resilience of IoT-enabled home automation systems. The findings contribute to developing smarter, safer homes, ensuring that advanced technologies can be adopted confidently and securely
Optimizing IoT Secure Data Processing and Using Azure Edge Computing Solutions
No full textThis research investigates an Azure IoT Edge-enabled solution that could collect telemetry data, store, and analyze it in real-time. The study focuses on designing an integrated architecture with the integration of Azure IoT Hub, Logic Apps, and Azure Monitor for seamless processing of data to address emerging demands for effective IoT pipelines. IoT Edge devices were provisioned, and the modules configured to start ingesting and processing data. The implementation entails a variety of workflows relating to the analytics of telemetry. Evaluations emphasize key metrics, such as message latency, data throughput, and system reliability to ensure there is a functional IoT pipeline where measurable improvements could be identified in data handling despite issues in connectivity and module configuration at the beginning. These results confirm the scalability and reliability of the proposed solution but also point out certain areas for further improvement, such as enhanced configuration automation. The work done here provides a good base for future projects that might deal with advanced analytics or scaling up multi-cloud support and boasts considerable academic and commercial potential
Analysis of automated zero trust AWS home network for confidentiality and authentication issues
No full textIn a home network environment, this research investigates the issues of maintaining confidentiality and ensuring authentication issues that involve unauthorized access, data leakage, and insufficient segmentation of the network which are common. Moreover, this research also addresses traditional approaches like Virtual Private Network-based solutions which often fail to effectively address the evolving challenges of securing home networks, particularly when it comes to managing access for multiple users and devices while maintaining high levels of confidentiality and authentication. These issues could be resolved by using a Zero Trust orchestrator like Twingate and its automation with Infrastructure as Code could help to integrate the architecture in a home network for a secure environment. This research explores and evaluates the principle of Zero Trust which is “never trust, always verify” to prevent security risks that involve confidentiality and authentication with automation. It shows how the Zero Trust framework plays its role in home security including methods of Multi-Factor Authentication, Access Control, Biometrics, Encryption mechanisms like AES256 and KMS, and network segmentation to prevent data leakage and ensure confidentiality. Moreover, it allows verified and authenticated users from outside the network to get into the network and use its resources, which could not be possible by using a Virtual Private Network. To justify the Zero Trust Model in this research. A Cloud platform AWS has used in the network with its services like for computing EC2 and S3 for encrypted storage and Access Control List for policy enforcement. These services are tested to provide a secure environment for a user in a home network and its whole configuration is done by using Infrastructure as Code that automates everything and provides a scalable, repeatable Zero Trust model with fewer chances of human error involvement and time-saving The results in the form data encryption, network segmentation and access management with user and its device authentication validates that Zero Trust in resolving the security issues critical in a home network and provide resilient environment that could be integrated with automation instead of using traditional VPN solutions. This research not only goes through the Zero Trust principles but also validates its automation that can be used for both enterprise and personal networks to overcome the challenges of confidentiality and authentication
Integrating Explainable AI (XAI) for Improved Malware Detection and Analysis
No full textMalware has significantly evolved over the decades, transitioning from simple viruses to complex threats such as Advanced Persistent Threats (APTs). This evolution requires robust and advanced detection methods. Traditional methods, including signature-based malware detection, struggle with obfuscated and novel malware. This research integrates machine learning (ML) models: Logistic Regression, Support Vector Machine (SVM), and Random Forest with Explainable (XAI) techniques, specifically LIME (Local Interpretable Model-Agnostic Explanations), to improve malware detection system’s accuracy and interpretability. Using a malware memory dump dataset, the Logistic Regression model achieved the highest accuracy of 99.94%, while the Random Forest model showed signs of overfitting. To utilise the full potential of this XAI-based malware detection system, an email alert system was incorporated to send alerts to the administrator with proper explanations made by the XAI technique whenever the system detects potential malware
In-Depth Analysis of Machine Learning for Securing Internet of Things devices using CIC IoT & Net Flow Dataset
No full textThe rise of Internet of Things (IoT) has made the digital landscape transformed into providing more services but also introduced significant cybersecurity challenges by expanding potential vulnerabilities. Security systems such as Data Loss Prevention (DLP), Intrusion Detection Systems (IDS) and firewalls are struggling to keep up with these modern threats. They often produce a high number of false positives and lack the capability to identify more advanced, evolving attacks. To address these shortcomings, our research presents a machine learning models where we compare the ML models and analyse in detail as to which model produces the best accuracy. This approach has been documented each model and made sure the results are best for the networks for processing sequential data. The hybrid design improves detection accuracy and reduces the rate of false positives.
Furthermore, we developed these machine learning models are two datasets in order to figure out the best results. The novelty of this research lies in the hyper parameter tuning of machine learning models to achieve best results. The main contributions of this research are the extensive machine learning and deep learning algorithms for detecting and classifying malicious traffic. Our research conducted five novel machine learning models on UQ-NIDS dataset and added hyper parameter tuning in order to determine how well the model performs on those IoT attacks and for CICIoT dataset, our research is performed extensively on the classification of attack categories and implemented three novel algorithms to find the best model for each attack categories. And our research was able to find that Random Forest and Logistic Regression performs well on both the datasets and also identifying different categories in IoT attacks