Exploiting Machine Learning to Detect Malicious Nodes in Intelligent Sensor-Based Systems Using Blockchain

Table 2

Feature comparison with respect to proposed model.

Contributions	Ref. No.	Similarities	Differences	Limitations
Registration and authentication	[8]	Distributed network authentication	Hybrid blockchain different levels of nodes registered at different platforms	Complex to manage two blockchains
	[30]	Authentication	Use of multiblockchains and cluster manager	Difficult to manage
Routing	[7]	Routing, MN detection	Calculate through Qlearning	Consumes more energy while discovering the route
	[14]	Distributed network, routing	Localization-based routing	Select relatively longer path, which causes rapid energy
MN detection	[10]	Distributed network, registration, MN detection	Detection through network parameters, calculation of node’s reputation	Uses computationally complex consensus algorithm
	[29]	Detection through ML, distributed network	SDN-based network, multilevel detection	Considered very few attacks