Machine Learning Enabled Techniques for Protecting Wireless Sensor Networks by Estimating Attack Prevalence and Device Deployment Strategy for 5G Networks

Table 1

Methodology comparison.

References	DDoS attack mechanism	Type of tool and SDN controller used		Experimental limitations
[33]	Intrusion is detected based on statistical analysis	OpenFlow switch table entries attack	OMNeT++	Simultaneous detection cannot be performed when there is an alteration in the headers
[34]	DDoS detection on the basis of entropy	TCP SYN, UDP	Mininet and POX	Only limited to the detection mechanism.
[35]	Prediction based on network bandwidth	DoS attacks	Real testbeds, floodlight	Limited to only DoS attacks.
[36]	PATMOS	DDoS attacks	Mininet and POX	The problem in using clustering.
[37]	Traffic flow migration	DoS attacks	Mininet and Ryu	Consistency issue may occur in case of multiple controllers’ environment
[38]	Network traffic rule-based DDoS mitigation	TCP-SYN traffic flooding attack	Mininet, ODL	May increase the consumption of bandwidth and thus increasing the overhead as well.
[39]	Avant-guard	TCP-SYN traffic flooding	Mininet, POX	For the protocols which are not based on TCP SYN, traffic.