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

Table 1

Mapping of identified limitations with proposed solutions and validations.

Limitations identified	Solutions proposed	Validations done
L1: PoW utilizes high computational power [10].	S.1: PoA is used that utilizes low computational power.	V.1: transaction cost, as shown in Figure 6
L2: presence of MN in the network [29]	S.2, S.3: GA-SVM and GA-DT are used for the detection of MNs.	V.2, V.3: accuracy, precision, PDR, PMR, PdR, and PMiR, as shown in Figures 3(a) and 3(b), 4(a) and 4(b), and 5(a) and 5(b)
L3: grayhole attack is possible on routing nodes [7].
L4: long paths deplete nodes’ energy [14].	S.4: the Dijkstra algorithm is used to find the shortest path.	V.4: distance from source to destination is calculated, as shown in Figure 5.
L5: registration consumes more gas due to hybrid blockchain [14].	S.5: lightweight registration and authentication mechanisms	V.5: transaction cost, as shown Figure 6