An intelligent intrusion detection and performance reliability evaluation mechanism in mobile ad-hoc networks

A Mobile Ad-Hoc Network (MANET) is a widely used temporary network. It is simple to install and handle. However, its dynamic nature is more vulnerable to routing attacks than fixed networks. Many intrusion detection methods are available to tackle its vulnerability and attacks. These works used different statistical parameters and measured various performances to validate their methods. Existing works have different claims or validations. However, these works show ambiguities in performance comparison. There are many reasons for performance ambiguities, such as imbalanced distribution of samples, number of training samples, number of labels, etc., whereas these are executed on the same dataset. This paper presents deployment of the network, data generation, sample labeling, feature extraction, an intrusion detection method, and a performance reliability evaluation model. The evaluation model analyzes the performance and hardware dependability of intrusion detection methods, and it computes the performance reliability of related methods using a fuzzy logic system. The outcome shows that when one statistical performance increases, another performance decreases due to an imbalanced sample ratio. Therefore, the proposed evaluation model has analyzed the two best performances. It evaluates intrusion detection mechanisms and provides a comparative analysis with an approximate performance as scheme reliability. The experimental results show that the performance of the proposed detection method is better than existing methods in terms of maintaining high scheme reliability.