IoT Security Detection Method Based on Multifeature and Multineural Network Fusion

Table 1

Comparison of this model detection method with other model detection methods.

Papers	Scenarios	Model type	Advances
Awajan [26]	Neural network-based IoT detection	DNN-based	(i) Get dynamic access to IoT security data
			(ii) Improved accuracy
Rami Reddy et al. [27]	Feature selection based on an improved version of the grey wolf algorithm	ECHIGWO-based algorithm	(i) Forward search to derive the optimal feature subset
			(ii) Better optimization in class II and class V models
Bhandari et al. [28]	Artificial intelligence-based malware discovery framework	DNN-based ml model	(i) Capture dynamic IoT security status data to proactively uncover vulnerabilities
Thandapani et al. [29]	AI-based IoMT	CNN-based	(i) Multimodel training to improve feature accuracy
Kalutharage et al. [30]	Interpretable artificial intelligence (XAI) detection methods	Based on the automatic coding model	(i) Save data collection costs by capturing dynamic data with distributed deployment ids
			(ii) Detecting diverse IoT security threats
Alzahrani and Alzahrani [31]	IoT anomaly mitigation system based on a multialgorithm model	Three algorithms based on ML models (KNN, EWMA, and CUSUM)	(i) Always acquire IoT data and proactively discover vulnerabilities
			(ii) Streamlined model deployment costs