Table 1 Summary of the related works.
References | Methods used | Advantages | Limitations |
|---|---|---|---|
Federated Generative Adversarial Networks (A3GANS) and two GAN-based algorithms | Effective privacy protection, improved accuracy by 8%, robust data privacy for Android users | Complexity in GAN-based algorithms and federated learning | |
Delay Conscious routing protocol | Effectively used to Multi point relay nodes | Several Problems including limited density, high speed nodes and abrupt dynamic topology. | |
Intrusion detection algorithm based on neighbour information | Better identification of sinkhole nodes, improved network performance, energy-efficient | Dependence on neighbour information, might not scale well in larger networks | |
Human-interactive, visual-based anomaly detection devices | Real-time wormhole attack detection, visual representation of security risks | Limited to IP-enabled WSNs, visual tool scalability concerns | |
Trust-based routing technique, FLION clustering | Enhanced security, energy-efficient routing, extended network lifespan | Potential overhead from multiple trust evaluations, complexity in implementation | |
Lion Swarm Optimization (LSO) with multi-agent frameworks | Improved global exploration capabilities, enhanced accuracy and efficiency | Increased computational overhead, complexity in multi-agent system integration | |
Swarm Intelligence-based Clustering Technique | Improve network throughput | Unbalanced power usage is a common concern in cluster based routing | |
Modified SSA for clustering and routing, polynomial time fitness function | Maximized network lifespan, efficient energy use | Complexity in fitness function formulation, potential overfitting | |
Distributed cluster head selection, multi-hop routing | Balanced energy consumption, avoided hotspot issues, effective for dense WSNs | Complexity in GNN implementation, potential scalability issues | |
Grid-based clustering, Type-3 fuzzy system | Enhanced network longevity, adaptable clustering | Complexity in fuzzy system, potential issues with real-time adaptability | |
Weighted Recurrent Neural Network, Enhanced Golf Optimization Algorithm (EGOA) | Effective vampire attack detection, optimized routing paths | Complexity in RNN and optimization algorithm, computational overhead | |
Boltzmann Selection Probability-centric Gravitational Search Algorithm (BSP-GSA), IECC, MUMLP | Secure data transmission, energy-aware routing, cloud storage integration | Complexity in multiple algorithm integration, potential latency in cloud storage | |
EPK-DNN, LS-BAT, DL-K-Means | High detection accuracy, robust against various attacks | High computational requirements, complexity in multi-algorithm integration | |
Adaptive Neuro-Fuzzy Inference System (ANFIS), MATLAB integration | Effective malicious node detection, adaptable classification | Performance degradation with increasing hostile nodes, complexity in MATLAB integration | |
K-means clustering, hybrid optimized machine learning techniques | High localization accuracy, effective routing threat detection | Dependence on specific datasets, complexity in hybrid model implementation | |
Modified k-means, Philippine Eagle (PE) optimization | Secure and efficient routing, improved detection rate | Potential issues with scalability, computational complexity | |
LEACH, Fuzzy interference rule, ANN | High specificity, accuracy, and sensitivity, effective data transmission | Potential energy consumption from ANN, complexity in the integration of multiple methods | |
Blockchain based identification Strategy | It is used to improve the level of compatibility among blockchain and IoT | End-to-end security is challenging due to the intricacy of IoT communication and different resource capacities. |
