25 October 2021 in IEEE Access
In recent years, Wireless Sensor Networks (WSNs) have benefitted from their integration with Internet of Things (IoT) applications. WSN usage for monitoring and tracing applications shows massive acceleration, whether indoors or outdoors. WSN is constructed from interconnected sensors, limited resource (battery), which requires considerable importance on deployment and routing strategies, to improve the performance of Quality of Service (QoS) in WSNs. Many of the existing strategies are based on metaheuristics algorithms such as Genetic Algorithms to resolve the problem. This research proposes a new algorithm, Enhanced Non-Dominated Sorting Genetic Routing Algorithm (ENSGRA), to improve the QoS in WSNs. The proposed algorithm relies on Non-Dominated Sorting Genetic Algorithm 3 (NSGA-III), but adjusts reference points through the use of a dynamic weighted clustered scheduled vector to obtain new solutions. Moreover, ENSGRA can be used to find an integration between two parents crossover with multi-parent crossover (MPX), to produce multiple children and improve new offspring to obtain the optimal Pareto Fronts (PF). This algorithm excels when compared with the lagged multi-objective jumping particle swarm optimization, Non-dominated Sorting Genetic Algorithm–II and NSGA-III in terms of the QoS model (31% optimization percentage). Results show that the proposed ENSGRA is superior over other algorithms in evaluation measures for multi-objective algorithms.
1 May 2022 in Indonesian Journal of Electrical Engineering and Computer Science
Multi-objective algorithms are used to achieve high performance for quality of service (QoS) in wireless sensor networks (WSNs) is an important field for researchers because these algorithms improve two or more conflicting objectives and present the best trade-off between the conflicting objectives to solve multi-objective problems (MOPs). Previous research proposed an algorithm that relies on non-dominated sorting genetic algorithm 3 (NSGA-III), namely enhanced non-dominated sorting genetic routing algorithm (ENSGRA). This algorithm is used to optimise three objectives, which include number of worked sensors, energy consuming and node covering area. The fourth objective, which is node load balancing, is added in the current research. Such an addition aims to improve node distribution around cluster heads and decrease network congestion, thus decreasing energy consumption and increasing network lifetime. The ENSGRA algorithm is compared with multi-objective multi-step particle swarm optimisation (MOMSPSO), non-dominated sorting genetic algorithm 2 (NSGA-II), and NSGA-III. The proposed algorithm ENSGRA exceed to MOMSPSO, NSGA-II, and NSGA-III in the proposed QoS model in the final outcomes, as the proposed approach achieved (38%) average combination (optimisation) percentage. Which is the highest percentage over other methods.
23 September 2025 in IEEE Access
With the rapid expansion of the Internet and the increasing reliance on digital services, cyberattacks have become more prevalent and sophisticated. This rise in malicious activities necessitates the development of stronger security measures, particularly across the physical, network, and application layers of the 5G architecture. Given the complexity and scale of modern networks, ensuring robust security mechanisms to mitigate cyber threats is crucial to maintaining service reliability and user trust. This paper proposes a DDoS detection framework capable of self-automating the detection and mitigation of denial-of-service (DoS) attacks as well as distributed denial-of-service (DDoS) attacks in next wireless generations. The framework utilizes machine learning (ML) and deep learning (DL) techniques to increase precision and optimize resource use regarding the detection of attacks on the 5G network’s application and transport layers. To test the efficacy of the suggested framework, a series of tests was conducted using diverse ML and DL models. The findings illustrate advanced cyber threat detection capabilities among some deep learning architectures. In binary classification of an attack, Bidirectional Long Short-Term Memory (BiLSTM) and Gated Recurrent Unit (GRU) indeed achieved the undisputable highest recall rate of 99.98%, confirming the systems’ effectiveness in discerning between normal and attack traffic. Also, while being involved in multi-attack classification, the Multi-Layer Perceptron (MLP) model surpassed other models in recall, reporting an impressive 99.04% rate. It can be observed that the suggested deep learning architectures have high effectiveness for enhancing 5G networks security related to ongoing DoS and DDoS threats. The findings revealed by the proposed framework do unlock a new perspective on enhancing network resilience against sophisticated cyber threats.