Hybrid optimization with ensemble learning to ensure VANET network stability based on performance analysis

Division 1

Section 1 contains the results and discussion of proposed and implemented methods to improve machine learning with a hybrid optimization strategy to predict mobility in VANET. The execution of the project (HFSA-VANET) is evaluated and compared to that of the current method (CRSM-VANET). The measured values ​​of delay, power consumption, drop, throughput and fairness index are calculated and compared to the proposal (HFSA-VANET) and the existing (CRSM-VANET)²⁹ methods. Moreover, the implementation is done through NS2 stimulation and comparing the proposed algorithm with these two platforms, as well as the Windows 10 PRO computer, the total RAM capacity of 10 GB and the processor used is Intel^® Core (7M) i3-6100CPU at 3.70 GHz. Performance metrics are discussed in the next section.

Performance indicators

Delays occur as a packet travels from its source to its destination.

This is the number of packets lost as a result of an unauthorized node (DoS attack).

Throughput refers to the amount of packet data established on a destination, which is the aggregate value of packets created by the sending node within a certain time. The formula is:

Results obtained via node

The performance measures of the existing technique and the proposed method are compared in the table below.

The main objective of the performance measures is to assess the ability of the proposed model to predict mobility in VANET. According to Table 1, when compared and reviewed with the existing methodology, the proposed method improves machine learning with a hybrid optimization strategy to predict mobility in VANET is more efficient.

The delay, power consumption, drop, throughput and fairness index of HFSA-VANET and CRSM-VANET are compared below.

The suggested technique achieves 99 J, 0.093690, 0.897708 for power consumption, delay value, and drop value in node 20. Additionally, the new technique achieves a throughput of 31,341, which is superior to the previous approach. The proposed technique has an equity score of 7.000000, while the current method has a value of 8.000000. For power consumption, delay value and drop value in node 60, the proposed approach achieves 47 J, 9.752925, 0.472094. Additionally, the new method achieves a Throughput of 31,341, which is higher than the previous method. The suggested strategy has an equity score of 3.000000, compared to 4.000000 for the current method. The proposed technique achieves 36 J, 10.902826, 0.376633 for power consumption, lag value, and drop value in node 60. Additionally, the suggested technique achieves throughput of 28,423 vs. 26,749 for the existing method. An equity index value of 2.000000 for the proposed method against 4.000000 for the existing method is reached. For power consumption, delay value and drop value in node 80, the suggested approach achieves 11 J, 15.287826, 0.116375. Moreover, compared to the previous approach, the proposed strategy achieves a throughput of 18,197. The proposed approach has a fairness index of 1.000000, while the present method has a fairness score of 2, 000000. Figs. 3, 4, 5, 6 and 7 are the delay, power consumption, drop, throughput, fairness index are obtained through the node, respectively.

Results achieved through speed

The speed of the proposed technique and existing techniques are compared in terms of delay, power consumption, drop, throughput and fairness index. The measured values ​​are shown in the table below. Table 2 shows the speed values ​​of existing and proposed techniques.

The speed is compared to the delay shown in Fig. 8, at the speed vs. energy shown in FIG. 9, at the speed with respect to the fall illustrated in FIG. 10, at the speed versus flow rate shown in FIG. 11 and speed versus equity index shown in FIG. 12 Speed ​​is compared with delay, energy, drop, throughput and fairness index, and the graphical representation is shown below.

In speed 20, the proposed approach achieves 1980 J, 1.873793, 19.954160 in terms of power consumption, delay value and drop value. Moreover, the new method achieves a throughput of 150, which is higher than the previous method. The suggested approach has a fairness score of 6.000000, while the current method also has a number of 6.000000. The suggested technique achieves 1880 J, 390.117000, 18.883762 for power consumption, delay value and 40 speed drop value. Also, the new approach achieves a throughput of 35, which is higher than the existing method . The recommended technique has an equity value of 3.000000, but the current method has a score of 4.000000. At speed 60, the suggested approach achieves 2220 J, 654.169557, 22.597974 in power consumption, delay, and drop value. Moreover, the proposed strategy gives a throughput of 22 against 16 for the current method. The suggested technique has an equity index of 2.000000, while the present method has an equity index of 3.000000. The recommended method achieves 880 J, 1223.026093, 9.309993 for power consumption, delay value and 80 speed drop value. Also, the new technique achieves a rate of 8 and the existing technique achieves a rate of 6 The suggested technique has an equity score of 0.000000, while the current method has one of 2.000000. Figs. 8, 9, 10, 11 and 12 are delay, power consumption, drop, throughput, fairness index are obtained by speed respectively. Section 2 covers the results obtained using MATLAB software.

Division 2

This section covers experimental results obtained via MATLAB (VERSION 2020a) to evaluate performance with the NS2 tool. Moreover, we also include an additional parameter to ensure the network lifetime of the proposed model. Therefore, the performance can be proved as highly effective as the existing technique. Here, the performance of the proposed model is evaluated using various machine learning approaches such as ANN-HFSA-VANET, SVM-HFSA-VANET, NB-HFSA-VANET and DT-HFSA-VANET. Thus, the results of the proposed model can be compared and proven to be more effective than all other existing techniques.

Initially, the proposed model is evaluated separately with ANN-HFSA-VANET, SVM-HFSA-VANET, NB-HFSA-VANET and DT-HFSA-VANET. The following figures. 13, 14, 15 and 16 show graphical results of ANN, SVM, NB and DT, respectively. On the other hand, to show a comparison based on the aggregation of various machine learning techniques that the proposed method compares with the single graphical results.

Parameter analysis of ANN-HFSA-VANET

This section discusses different types of parameters of ANN-HFSA-VANET and is analyzed in the graph shown in Fig. 13.

The above mentioned Figure 13 illustrates the different performance analyzes based on ANN-HFSA-VANET where (a) shows that the proposed technique achieved minimum dropout, (b) shows that the maximum F1 score was achieved using the proposed technique, (c) illustrates that the maximum packet delivery rate was achieved for the ANN-HFSA-VANET, (d) and (e) show that the proposed ANN-HFSA-VANET generated high throughput and a minimum delay, respectively.

Decision Tree Parameter Analysis (DT)-HFSA-VANET

This section deals with the different types of parameters of the decision tree and analyzed in the graphs presented in Fig. 14.

Figure 14 shows the parameter analysis of DT-HFSA-VANET. (a) shows the minimum drop rate of the DT-HFSA-VANET, (b) discusses the maximum score for the F1 score of the DT-HFSA-VANET and its analysis, (c) shows the packet delivery rate of the DT-HFSA-VANET and its plotted values, (d) deals with throughput rate of DT-HFSA-VANET, (e) deals with end-to-end delay of DT-HFSA-VANET. The standard parameters are analyzed and plotted in a graph and the values ​​increased at the end of each graph of the parameter.

Parameter analysis of Navie Baves (NB)-HFSA-VANET

This section discusses the different types of Navie Bave parameters and is analyzed in the graph shown in Fig. 15.

Figure 15a shows that minimum loss, (b) shows maximum F1 score, (c) provides maximum packet delivery rate, (e) shows this minimum delay, respectively for the proposed NB-HFSA-VANET.

Analysis of SVM parameters

This section discusses the different types of an SVM parameter and is analyzed in the graph shown in Fig. 16.

In Figure 16a, the drop rate was obtained with a minimum rate, (b) the F1 score was obtained with a maximum score, (c) the packet delivery rate was obtained with a maximum, and (e ) shows a minimum delay.

Parameters for analyzing different types of data

This section deals with the parameters of the different data and their analysis. The values ​​are plotted in a graph.

From Fig. 17, the parameter analysis value of data types is examined and plotted in a graph where (a) shows the network lifetime obtained for each second, (b) discusses the power consumption of data packets used per second, (c) deals with the throughput ratio of data types and their performance, (d) deals with the packet delivery ratio of different types of data performance.