Survey on Botnet Detection Techniques: Classification, Methods, and Evaluation

Table 4

Summary of typical botnet detection techniques based on swarm intelligence.

Papers	Mechanism	Algorithm/model	Dataset	Advantage	Drawback
[95]	A hybrid particle swarm optimization (PSO) and voting system botnet detection method (BD-PSO-V) was proposed	PSO	ISOT	(i) Adaptive flow feature selection method	(i) High time complexity
	PSO algorithm was used for feature selection of network stream data. The voting system was used to identify botnets and classify samples	DDN SVM C4.5	Bot-IoT [99]	(ii) Detect during the attack phase
[97]	For Android botnet, a smart adaptive particle swarm optimization support vector machine (SAPSO-SVM) algorithm was used for detection	SAPSO	28 Standard Android Botnet Dataset (28-SABD) [100]	(i) Automatically extract Android botnet features	(i) High time complexity
		SVM		(ii) High detection accuracy
[98]	GWO swarm intelligence algorithm was used to optimize the hyperparameters of OCSVM to detect botnet attacks from damaged IoT devices	GWO	N-BaIoT [101]	(i) Deal with heterogeneous IoT devices	(i) IoT devices are increasing rapidly
		OCSVM		(ii) A new unsupervised evolutionary Internet of Things botnet detection method	(ii) Cannot detect unknown botnets