Have a personal or library account? Click to login
A Novel Scalable Trust-Aware Deep Reinforcement Learning Algorithm for Energy-Efficient and Secure Routing in Software-Defined Wireless Sensor Networks for IoT Cover

A Novel Scalable Trust-Aware Deep Reinforcement Learning Algorithm for Energy-Efficient and Secure Routing in Software-Defined Wireless Sensor Networks for IoT

Open Access
|Dec 2025

Figures & Tables

Fig. 1.

Proposed system model overview.
Proposed system model overview.

Fig. 2.

Proposed system architecture for SDWSN.
Proposed system architecture for SDWSN.

Fig. 3.

COA–HEC secure routing flowchart.
COA–HEC secure routing flowchart.

Fig. 4.

Average throughput analysis.
Average throughput analysis.

Fig. 5.

Analysis of PDR.
Analysis of PDR.

Fig. 6.

Average delay analysis.
Average delay analysis.

Fig. 7.

Energy consumption analysis.
Energy consumption analysis.

PDR comparison [%]_

NodesCOAHBSSEHRIBFAESMRGMPSO
509792908786
1009188858281
1508883807675
2008680787371
2508277747169

Evaluation of average delay comparison [msec]_

NodesCOAHBSSEHRIBFAESMRGMPSO
502.0124.236.347.588.35
1003.0565.977.298.2110.33
1504.376.678.119.3312.96
2005.667.559.1211.5414.77
2506.788.0610.7512.5815.92

Summary of key related works and their limitations_

Method / ProtocolOptimization techniqueSecurity mechanismParameters consideredLimitations
Ex-GWO, I-GWO [5]GWO variantsNoneEnergy, distance, traffic loadPoor adaptability in dynamic networks, high complexity
Fuzzy-GWO [8]Hybrid fuzzy + GWONoneEnergy efficiencyPerformance degradation in heterogeneous networks
SEAMHR [7]Metaheuristic analysisCounter-mode cryptographyDelay, energy, securitySecurity issues due to CTR reuse, high overhead
SEHR [9]Heuristic-based routingLightweight cryptographyEnergy-aware routingIntegrity flaws, limited energy optimization
ESMR [10]Secret sharingKey sharing + multi-hopEnergy, securityNo mobility support, ignores QoS metrics
IBFA [12]Blowfish with CM-MHSymmetric encryptionEnergy, securityNo authentication, high complexity, vulnerable patterns
GMPSO [16]Genetic mutation PSONoneEnergy efficiency, throughputHigh controller overhead, ignores trust, longer flow times
BOA-ACO [15]BOA + ACONoneCluster-head routing, energyNo built-in security, limited scalability
IEE-LEACH [20]Improved LEACH hybrid routingNoneEnergyFocuses only on lifespan, no security
HPSO-ILEACH [18]PSO + Improved LEACHNoneEnergy aggregationLacks robust trust/security
PSOGA [14]PSO + GA hybridNoneEnergy, packet transmissionLimited scalability, no lightweight security

Average throughput comparison [Mbps]_

NodesCOAHBSSEHRIBFAESMRGMPSO
500.930.800.730.660.60
1000.860.710.650.620.55
1500.750.650.560.510.47
2000.640.520.500.430.40
2500.600.500.420.380.35

Details of simulation parameters setup_

ParameterValues
Simulator modelNS-3.26
Sensor nodes count50, 100, 150, 200, 250
Simulation area500×500 m
Optimal path finding protocolCOA
SDN controller count1
Base station3
Size of packet512 bytes
Initial energy50 J
Simulation time300 sec
Transmission range250 m

Energy consumption comparison [J]_

NodesCOAHBSSEHRIBFAESMRGMPSO
500.230.470.550.640.76
1000.350.500.640.730.89
1500.470.720.850.890.97
2000.580.850.991.231.44
2500.630.981.131.451.98
Language: English
Page range: 358 - 365
Submitted on: Sep 21, 2024
|
Accepted on: Oct 13, 2025
|
Published on: Dec 23, 2025
In partnership with: Paradigm Publishing Services
Publication frequency: Volume open

© 2025 Jasmine Alponse, C Yaashuwanth, K Prathibanandhi, published by Slovak Academy of Sciences, Institute of Measurement Science
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.