Navigating the complex terrain of security challenges and crafting robust solutions in Mec-enabled IOT ecosystems for future societies

The accelerated growth of Internet of Things (IoT) devices and their convergence with Mobile Edge Computing (MEC) networks has transformed healthcare, smart cities, and industrial automation among other sectors. Nonetheless, this development presents notable security vulnerabilities, including Distributed Denial of Service (DDoS) and zero-day attacks against resource-limited edge nodes. This work puts forward a new hybrid deep learning model combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to improve real-time anomaly detection in MEC-based IoT deployments. The proposed model can capture both the spatial and temporal features in network traffic patterns and allow accurate malicious pattern classification. Experiments are conducted using the publicly released IoT-23 dataset, which contains real-world traffic samples from compromised and benign IoT devices. The data was preprocessed with information and ratio-based feature selection techniques. It was implemented in Python with TensorFlow libraries to support scalable and efficient training of the model. The suggested model exhibited a remarkable improvement in detection performance in terms of classification accuracy, achieved an accuracy of 99%, outperforming conventional 1D CNN and gradient boosting models in terms of detection performance. Experimental outcomes indicate that the CNN-LSTM model is capable of accurately identifying DDoS attacks with fewer false positives and higher recall, rendering it ideal for real-time deployment in edge computing contexts. The model’s versatility with lightweight processing enables it to be a resilient security solution for IoT systems of the future.