Academia.eduAcademia.edu

Figure 22 – uploaded by ragab hassen

See full PDFdownloadDownload figure
Fig. 20 Propagation based on tree trunk height that is achieved by the three algorithms

Figure 20 Propagation based on tree trunk height that is achieved by the three algorithms

Related Figures (27)

propagation of the near ground network and its problems as a form of mathematical mod- els. Analytical and empirical methods are used for this purpose. The analytical methods proposed by many researchers such as Li et al. [9], Liao et al. [10], and Li et al. [11], which these methods show high prediction accuracy of communication link quality. However, they are very complicated and their analyses require long computational time and huge computational resource. In addition, these methods may not suitable for ordinary users because of their complexity. On the other hand, the empirical method that is proposed by Goldman et al. [12] gave an easier way of calculating and modeling the path loss in forests, but this model is very site-specific. The researchers in [13] were following the empirical way of path loss measurement [12]. In ad the tree shadowing on the communication dict the path loss of VHF/UHF bands in t link. Joshi et al. [14] propos he forest areas with several Furthermore, a detailed study of empirica path loss in the forest over dition, they focused on studying the impact of ed a method to pre- foliage loss models. VHF/UHF commu- nication link has been conducted in [15] with taking on consideration the lateral wave and ground reflection wave [16] for long range We are going to summarize briefly a fe forested communication and propagation. w studies that use analytica or empirical meth- ods of modeling the signal propagation in woodv/forest areas. We summarize them as
propagation of the near ground network and its problems as a form of mathematical mod- els. Analytical and empirical methods are used for this purpose. The analytical methods proposed by many researchers such as Li et al. [9], Liao et al. [10], and Li et al. [11], which these methods show high prediction accuracy of communication link quality. However, they are very complicated and their analyses require long computational time and huge computational resource. In addition, these methods may not suitable for ordinary users because of their complexity. On the other hand, the empirical method that is proposed by Goldman et al. [12] gave an easier way of calculating and modeling the path loss in forests, but this model is very site-specific. The researchers in [13] were following the empirical way of path loss measurement [12]. In ad the tree shadowing on the communication dict the path loss of VHF/UHF bands in t link. Joshi et al. [14] propos he forest areas with several Furthermore, a detailed study of empirica path loss in the forest over dition, they focused on studying the impact of ed a method to pre- foliage loss models. VHF/UHF commu- nication link has been conducted in [15] with taking on consideration the lateral wave and ground reflection wave [16] for long range We are going to summarize briefly a fe forested communication and propagation. w studies that use analytica or empirical meth- ods of modeling the signal propagation in woodv/forest areas. We summarize them as
Algorithm 1 Non-dominated Sorting Genetic Algorithm (NSGA-ID) Table 2. The pseudocode of SPEA-II
Algorithm 1 Non-dominated Sorting Genetic Algorithm (NSGA-ID) Table 2. The pseudocode of SPEA-II
Table 4 Different countries along with their forests characteristics [46-53]
Table 4 Different countries along with their forests characteristics [46-53]
Table 5 Simulation parameters
Table 5 Simulation parameters
Fig.6 Geometry of radio wave broadcasting in a forest with bare ground region over a wide foliage depth
Fig.6 Geometry of radio wave broadcasting in a forest with bare ground region over a wide foliage depth
Mutation probability 1/d where d is the variable code size Fig. 7 Procedure workflow of evaluation objective functions using NSGA-II algorithm maximize the electromagnetic fields and signal propagation also to minimize both of path loss and signal attenuation. Furthermore, it ranks the population based on non-dominated solutions, after that, it performs selection, crossover, and mutation operations to generate new population (child population). Based on the results of previous operations, the algo- rithm evaluates the same objective functions but using the new generation of the popu- lation (child population). Moreover, the algorithm combines the new generation by the old one, after that, it performs ranking for these populations. At the end, if the number of
Mutation probability 1/d where d is the variable code size Fig. 7 Procedure workflow of evaluation objective functions using NSGA-II algorithm maximize the electromagnetic fields and signal propagation also to minimize both of path loss and signal attenuation. Furthermore, it ranks the population based on non-dominated solutions, after that, it performs selection, crossover, and mutation operations to generate new population (child population). Based on the results of previous operations, the algo- rithm evaluates the same objective functions but using the new generation of the popu- lation (child population). Moreover, the algorithm combines the new generation by the old one, after that, it performs ranking for these populations. At the end, if the number of
Optimal Placement of Near Ground VHF/UHF Radio Communication... Fig.8 Procedure workflow of evaluation objective functions using SPEA-II algorithm
Optimal Placement of Near Ground VHF/UHF Radio Communication... Fig.8 Procedure workflow of evaluation objective functions using SPEA-II algorithm
Fig.9 Procedure workflow of evaluation objective functions using OMOPSO algorithm
Fig.9 Procedure workflow of evaluation objective functions using OMOPSO algorithm
SS a a a I Figure 11 illustrates the results of minimizing the attenuation model using three algo- rithms, including, NSGA-II, SPEA-I, and OMOPSO algorithms. The figure illustrates this objective by using three operators, including, maximum, average, and minimum. The only algorithm with a smaller minimum average value of attenuation was NSGA-II, with ar average of 0.0133556 dB. OMOPSO algorithm had the greatest average value of attenua- tion on 10 times run, with a value of 0.0147 dB. SPEA-II algorithm had the medium aver: age value of attenuation on 10 times run, with a value of 0.01412. Figure 12 illustrates the results of maximizing the electromagnetic fields using three Figure 12 illustrates the results of maximizing the electromagnetic fields using three algorithms, including, NSGA-II, SPEA-I, and OMOPSO algorithms. The figure illus- trates this objective by using three operators, including, maximum, average, and mini- mum. Overall, the NSGA-II algorithm achieved the lowest amount of maximum, average,
SS a a a I Figure 11 illustrates the results of minimizing the attenuation model using three algo- rithms, including, NSGA-II, SPEA-I, and OMOPSO algorithms. The figure illustrates this objective by using three operators, including, maximum, average, and minimum. The only algorithm with a smaller minimum average value of attenuation was NSGA-II, with ar average of 0.0133556 dB. OMOPSO algorithm had the greatest average value of attenua- tion on 10 times run, with a value of 0.0147 dB. SPEA-II algorithm had the medium aver: age value of attenuation on 10 times run, with a value of 0.01412. Figure 12 illustrates the results of maximizing the electromagnetic fields using three Figure 12 illustrates the results of maximizing the electromagnetic fields using three algorithms, including, NSGA-II, SPEA-I, and OMOPSO algorithms. The figure illus- trates this objective by using three operators, including, maximum, average, and mini- mum. Overall, the NSGA-II algorithm achieved the lowest amount of maximum, average,
and minimum for maximizing the electromagnetic fields. For the operator of average, the OMOPSO had the greatest average value of maximizing the electromagnetic field on 10 times run, with a value of 6.53E—04 dB/m?. On the other hand, the SPEA-II algorithm had the medium average value of maximizing the electromagnetic field on ten time runs, which equal to 3.46E—04 dB/m?. SESE RA ate ME ee |e Figure 13 shows the comparison between NSGA-II, SPEA-II, and OMOPSO algorithms in terms of maximizing the signal propagation. The figure presents the maximum, average, and minimum of this objective for the three algorithms. The results show that SPEA-II algorithm achieved a maximum average value of signal propagation compared with other algorithms, which equal to 3.91E+32. We can deduce from the previous figures that NSGA-II outperforms other algorithms
and minimum for maximizing the electromagnetic fields. For the operator of average, the OMOPSO had the greatest average value of maximizing the electromagnetic field on 10 times run, with a value of 6.53E—04 dB/m?. On the other hand, the SPEA-II algorithm had the medium average value of maximizing the electromagnetic field on ten time runs, which equal to 3.46E—04 dB/m?. SESE RA ate ME ee |e Figure 13 shows the comparison between NSGA-II, SPEA-II, and OMOPSO algorithms in terms of maximizing the signal propagation. The figure presents the maximum, average, and minimum of this objective for the three algorithms. The results show that SPEA-II algorithm achieved a maximum average value of signal propagation compared with other algorithms, which equal to 3.91E+32. We can deduce from the previous figures that NSGA-II outperforms other algorithms
Fig. 14 Definition of a knee point based on the reflex angle [59]
Fig. 14 Definition of a knee point based on the reflex angle [59]
Fig. 15 Minimizing path loss based on tree foliage depth that is achieved by the three algorithms
Fig. 15 Minimizing path loss based on tree foliage depth that is achieved by the three algorithms
Fig. 16 Minimizing attenuation based on tree foliage depth that is achieved by the three algorithm:
Fig. 16 Minimizing attenuation based on tree foliage depth that is achieved by the three algorithm:
Fig. 17 Maximizing parameter «, based on tree trunk height that is achieved by the three algorithms
Fig. 17 Maximizing parameter «, based on tree trunk height that is achieved by the three algorithms
Parameter Qt, [dB/m] achieved by NSGA-II Parameter Ot, [dB/m] achieved by SPEA-II
Parameter Qt, [dB/m] achieved by NSGA-II Parameter Ot, [dB/m] achieved by SPEA-II
x= Tree trunk height, Ht=x/2 in [meter]
x= Tree trunk height, Ht=x/2 in [meter]
Fig. 18 Electromagnetic fields based on tree trunk height that is achieved by the three algorithms
Fig. 18 Electromagnetic fields based on tree trunk height that is achieved by the three algorithms
Fig. 19 Maximizing parameter d based on tree trunk height that is achieved by the three algorithms
Fig. 19 Maximizing parameter d based on tree trunk height that is achieved by the three algorithms
Hani Ragab Hassen received his Ph.D. from the university of technol- ogy of Compiegne (UTC), France in 2007. He was a lecturer in Infor- mation Security at the University of Kent, UK from 2007 to 2012, and a lecturer in Network Management and Security at Robert Gordon University, UK from 2013 to 2015. He is currently with Heriot-Watt University. His main research interests include malware analysis, ethi- cal hacking, peer-to-peer systems and key management.
Hani Ragab Hassen received his Ph.D. from the university of technol- ogy of Compiegne (UTC), France in 2007. He was a lecturer in Infor- mation Security at the University of Kent, UK from 2007 to 2012, and a lecturer in Network Management and Security at Robert Gordon University, UK from 2013 to 2015. He is currently with Heriot-Watt University. His main research interests include malware analysis, ethi- cal hacking, peer-to-peer systems and key management.
Related topics:
Computer ScienceDistributed ComputingEvolutionary algorithmsAntennas & Radio Wave PropagationWireless Networks (Computer Science)Network ModelingNetwork PlanningElectrical and Electronic EngineeringSwarm intelligence algorithmsWireless Personal Communications
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved