Subscription music platforms, like many web applications, increasingly rely on Progressive Web Apps (PWAs) to enhance user experience. These PWAs function by exchanging data with servers or REST APIs. However, the current reliance on REST APIs poses significant security risks due to vulnerabilities in data transmission. To address this issue, this research integrates the El Gamal cryptographic algorithm into the architecture of a subscription music platform. By incorporating the El Gamal cryptographic algorithm, this research endeavors to fortify the security posture of data exchanged between users and servers through REST APIs. This encryption method was selected for its robust resistance to various cryptographic attacks, providing a strong defense against unauthorized interception and tampering of sensitive information. To evaluate the efficacy of the El Gamal integration, a rigorous white box testing regimen was employed, encompassing metrics such as cyclomatic complexity and basic path testing. These assessments comprehensively examined the code's structure and execution to identify potential vulnerabilities and ensure the correct implementation of the cryptographic algorithm. The findings of the white box testing unequivocally demonstrated the successful integration of El Gamal cryptography on both the client and server components of the subscription music platform, effectively safeguarding the confidentiality and integrity of data transmitted via REST APIs. This research contributes to the advancement of secure communication protocols within web applications, particularly subscription-based platforms. Through the implementation of robust encryption, the study enhances data integrity and confidentiality, ultimately building user trust. 

The last few years have witnessed exponential rise in internet-of-things (IoT) systems for communication; yet, ensuring quality-of-service (QoS) and transmission reliability over mobile topology has remained challenge. Despite the fact that the use of software defined networks (SDN) have enabled IoTs to achieve resource efficiency and reliability; it doesn’t guarantee optimality of the solution over the network with high dynamism and non-linearity. Moreover, the major at hand SDN-IoT protocols have applied standalone node parameters to perform routing and allied transmission decision that confine its robustness over dynamic network topologies. Interestingly, none of the state-of-art SDN-IoT protocols could address the problem of iterative link-outage and corresponding network discovery cost. Furthermore, even multi-path selection strategies too failed in addressing the problem of joined shortest path selection and allied iterative link-outage due to the common node failure. Considering it as motivation, in this paper a novel and robust Heuristic-Driven Disjoint Alternate Path Switching -based Fault-Resilient Multi-Constraints Routing Protocol for SDN-mIOT system (HDAP-SDNIoT) is proposed. HDAP-SDNIoT exploits multiple dynamic parameters like medium access control information, flooding and congestion probability information. HDAP-SDNIoT exploits aforesaid node parameters to perform node profiling that serves multi-constraints best forwarding path selection. The proposed model retrieves multiple best alternating paths which are fed as input to the Adaptive Genetic Algorithm (AGA) that retains three disjoint best forwarding paths. HDAP-SDNIoT protocol at first avoids any malicious node(s) to become forwarding node, while it provides auto-switching capability to the forwarding node to select a disjoint forwarding alternate path in case of any link-outage in current forwarding path. _Simulation results confirm robustness of the proposed model in terms of high packet delivery rate of 96.5%, low packet loss rate 3.5% and low delay of 211 ms that affirms its suitability towards real-time SDN-mIoT applications.

In the contemporary digital landscape, the proliferation of malware presents a significant threat to the security and integrity of computer systems and networks. Traditional signature-based detection methods are increasingly ineffective against the evolving landscape of sophisticated malware variants. Consequently, there is a pressing need for innovative approaches to malware detection that can adapt to emerging threats in real-time. This research aims to develop a malware detection system using machine learning algorithms. Random Forest classifier and Logistic regression were deployed for the classification of malware based on the features extracted from the CIC-MalMem-2022 dataset. The Malware detection system model was implemented using the Python programming language and evaluated using major performance metrics like F1-score, precision, recall, and accuracy to assess the model’s performance. A comparison between the logistic regression model and the random forest model showed that the Random Forest model approach performed better than the logistic model in detecting malware, achieving accuracies of 98% and 94% respectively. In summary, the report concludes that the developed Malware Detection System using Machine Learning, specifically the Random Forest and Logistic regression models, shows promise in effectively detecting malware and highlights the importance of leveraging Artificial Intelligence for combating malware threats in the computing community.

Microstrip Patch Antennas (MPAs) play a critical role in modern wireless communication systems due to their compact size, easy integration, and capability to ensure reliable communication across wide frequency ranges. This paper introduces enhanced designs of rectangular MPAs aimed at overcoming the narrow bandwidth limitation commonly found in traditional designs. Three innovative configurations are proposed: one featuring a simple rectangular slot on the ground plane, another integrating polygonal Defected Ground Structures (DGS), and a third utilizing rectangular DGS. These antennas are optimized at a frequency of 4 GHz using High Frequency Structural Simulator (HFSS) software to significantly improve antenna performance. The MPA without DGS showed a return loss of -21.124 dB at a resonant frequency of 4 GHz, with a Voltage Standing Wave Ratio(VSWR) of 4.8038 and a gain of 3.88 dBi. In contrast, the MPA with Polygonal DGS exhibited significant improvements, achieving a return loss of -26.87 dB at a resonant frequency of 4.1 GHz, along with a VSWR of 1.3721 and a gain of 4.38 dBi. Similarly, the MPA with Rectangular DGS demonstrated superior characteristics, with a return loss of -27.08 dB, resonance at 3.825 GHz, a VSWR of 1.4399, and a gain of 4.00 dBi. These results underscore the effectiveness of DGS in broadening the bandwidth and improving the performance of MPAs for applications below 6 GHz, making them highly suitable for next-generation wireless communication systems.

In today's interconnected world, the threat of intrusion activities continues to rise, making it imperative to deploy effective security measures such as Intrusion Detection Systems (IDS). These systems play a vital role in monitoring network and system activities to identify unauthorised or malicious behaviour. The focus of this research is on evaluating the efficiency of different IDS in detecting anomalies in network traffic, specifically targeting Denial of Service (DDoS) attacks that exploit server vulnerabilities using IP addresses. The study utilises the CIC-DDoS 2019 dataset to analyse the performance of various IDS, particularly Network Intrusion Detection Systems (NIDSs), in predicting DDoS attacks accurately. To combat the diverse range of DDoS threats, a collective classifier is introduced, which combines four top-performing algorithms to enhance detection capabilities. By transforming the problem into a multilabel classification issue, the researchers aim to address the complexity of DDoS attacks effectively. Several machine learning (ML) and artificial intelligence (AI) algorithms are employed in the study, including Random Forest Classifier, Decision Tree Classifier, Support Vector Machine (SVM), Naïve Bayes, Multi-Layer Perceptron, Long Short-Term Memory (LSTM), and XGBoost Classifier. Evaluating the performance and computational efficiency of these algorithms is crucial to determining the most effective approach to detecting DDoS attacks. The results of the research highlight the effectiveness of the Random Forest Classifier and Multi-Layer Perceptron in accurately detecting DDoS attacks, as evidenced by their high accuracy rates on the test dataset. These findings underscore the importance of leveraging advanced ML algorithms to enhance the security of networks and systems against evolving cybersecurity threats. In conclusion, the study emphasises the significance of deploying robust IDS equipped with sophisticated ML algorithms to safeguard against intrusion activities like DDoS attacks. By continuously evaluating and improving the performance of these systems, organisations can enhance their cybersecurity posture and mitigate the risks posed by malicious actors in the digital landscape.