AFGT-Net-LSTM: AI-Driven Mobility and Energy Prediction-Based Routing in Energy-Harvesting Mobile AdHoc Networks

H. Mary Shyni; U. Sakthivelu; H. Shalma; K. Theivanai

doi:10.3991/ijim.v20i19.63177

Authors

H. Mary Shyni SRM Institute of Science and Technology, Vadapalani Campus, Chennai, Tamil Nadu, India https://orcid.org/0000-0002-6386-5947
U. Sakthivelu Vel Tech Rangarajan Dr. Sagunthala R & D Institute of Science & Technology, Tamil Nadu, Chennai, India https://orcid.org/0009-0004-4971-5338
H. Shalma CMR Institute of Technology, AECS layout, Kundanahalli, Bangalore, India https://orcid.org/0009-0003-5737-7634
K. Theivanai Sona College of Technology, Salem, Tamil Nadu, India https://orcid.org/0009-0007-3337-7970

DOI:

https://doi.org/10.3991/ijim.v20i19.63177

Keywords:

Mobile Ad Hoc Networks (MANETs), Energy Harvesting Networks, Graph Attention Networks (GAT), LSTM-based Energy Prediction, Fuzzy Logic Routing, AI-driven Routing Protocols, Link Stability Optimization, 6G Wireless Networks

Abstract

Mobile ad hoc networks (MANETs) that operate in environments where energy harvesting takes place are confronted with three constant routing impediments: temporal blindness when predicting when energy will deplete, overestimation of link stability due to user mobility, and a lack of topological awareness resulting in fragile route selection. To address these issues, we propose adaptive fuzzy graph temporal network (AFGT-Net), an all-inclusive artificial intelligence (AI) based routing structure that comprises an long short-term memory (LSTM) model to predict the availability of future energy based on the past availability of energy; a Mamdani-type fuzzy inference system to support uncertainty-based decision-making by weighting different factors when making routing decisions; and multiple graph attention networks (multi-headed GATs) to support topological sensitivity in the routing decision process. A joint model was created to improve energy dynamics, mobility uncertainly, and structural robustness through joint training. Experiments were performed with physics-based simulations created in Python and network simulations created with NS-3 (v3.40) respectively. The experimental results revealed that AFGT-Net achieved an overall packet delivery ratio (PDR) of 89.6% while reducing the root mean squared error (RMSE) for energy prediction to be 0.052 joules and was able to extend the route lifetime (TR) by 1.62 times compared to baseline routing protocols. When compared to other AI (i.e.: GNN-R, DRL-R, and FedMANET) based routing protocols, the results demonstrated a consistent increase in all three throughput, delay and energy efficiency metrics. The proposed framework has been developed to provide a scalable and robust IoT network, disaster response system and future 6G wireless environment.

References

[1] K. Akkaya and M. Younis, “A survey on routing protocols for wireless sensor networks,” Ad Hoc Networks, vol. 3, no. 3, pp. 325–349, May 2005, doi: 10.1016/j.adhoc.2003.09.010.

[2] M. M. Alam and S. Moh, “Survey on Q-Learning-Based Position-Aware Routing Protocols in Flying Ad Hoc Networks,” Electronics, vol. 11, no. 7, p. 1099, Mar. 2022, doi: 10.3390/electronics11071099.

[3] Sakthivelu, U., and C. N. S. Vinoth Kumar. "Enhanced Cyberthreat Detection and Classification with CMCOADL-TDC Using Deep Learning and Optimization Techniques." Tehnički vjesnik 32.3 (2025): 891-899.

[4] X. M. Gao and J. H. Xia, “Mobility Prediction-Based Link Stability Routing Protocol for MANET,” Applied Mechanics and Materials, vol. 556-562, pp. 5926–5930, May 2014, doi: 10.4028/www.scientific.net/amm.556-562.5926.

[5] G. J. Alqahtani and F. Bouabdallah, “Energy-Efficient Mobility Prediction Routing Protocol for Freely Floating Underwater Acoustic Sensor Networks,” Frontiers in Communications and Networks, vol. 2, Aug. 2021, doi: 10.3389/frcmn.2021.692002.

[6] B. Chowdhuryan, “A Survey of Methods and Architectures for Optimized Causal Dilated Convolutional Neural Networks-Based Energy-Efficient and Delay-Sensitive Routing Paths Using Mobility Prediction in Mobile WSN,” Multidisciplinary Journal of Research in Engineering and Technology, vol. 12, no. 2, pp. 1–8, Sep. 2025, doi: 10.65521/mjret.v12i2.1910.

[7] R. Saravanan, “A Comprehensive Review of an Optimized Causal Dilated Convolutional Neural Network-Based Energy-Efficient and Delay-Sensitive Routing Paths Using Mobility Prediction in Mobile WSN,” International Journal of Electrical, Electronics and Computer Systems, vol. 12, no. 1, pp. 15–20, Jan. 2023, doi: 10.65521/ijeecs.v12i1.2613.

[8] V. S. Murugan and B. Unhelkar, “Optimizing Mobile Ad Hoc Network cluster based routing: Energy prediction via improved deep learning technique,” International Journal of Communication Systems, vol. 37, no. 10, Apr. 2024, doi: 10.1002/dac.5777.

[9] K. Sivakumar and A. Ashikali, “Hierarchical Energy Harvesting Aware Adaptive Fuzzy Routing with Data Compression for Energy Harvesting WSN,” International Journal of Membrane Science and Technology, vol. 10, no. 5, pp. 834–843, Oct. 2023, doi: 10.15379/ijmst.v10i5.3552.

[10] J. Park, S. Kim, and J. Choi, "GATER: Graph attention network for topology-aware efficient routing in dynamic wireless mesh networks," IEEE Trans. Mobile Comput., vol. 22, no. 11, pp. 6578–6593, Nov. 2023.

[11] T. Hieu, L. Dung, and B.-S. Kim, “Stability-Aware Geographic Routing in Energy Harvesting Wireless Sensor Networks,” Sensors, vol. 16, no. 5, p. 696, May 2016, doi: 10.3390/s16050696.[14]

[12] A. Zappone, M. Di Renzo, F. Lacava, T. Q. Duong, and M. Debbah, "User association and link scheduling for energy harvesting heterogeneous networks with proactive routing," IEEE J. Sel. Areas Commun., vol. 40, no. 5, pp. 1515–1529, May 2022.

[13] S. Temel, I. Bekmezci, and F. Alagoz, "GAMANET: Game-theoretic adaptive MANET routing under energy harvesting constraints," IEEE Trans. Commun., vol. 70, no. 9, pp. 6034–6047, Sep. 2022.

[14] R. Huang, V. Bhatt, and J. Li, "Lifetime-maximizing cooperative routing in energy harvesting wireless sensor networks: A graph-theoretic approach," IEEE Trans. Wireless Commun., vol. 21, no. 12, pp. 10487–10501, Dec. 2022.

[15] K. Rajandekar and B. Sikdar, "A survey of MAC layer issues and protocols for machine-to-machine communications in energy-harvesting IoT networks," IEEE Internet Things J., vol. 9, no. 22, pp. 22385–22404, Nov. 2022.

[16] O. M. Bushnaq, A. Celik, H. Elsawy, M. Alouini, and T. Y. Al-Naffouri, "Aeronautical communication transition from L-band to C-band: How feasible is it?" IEEE Trans. Wireless Commun., vol. 21, no. 3, pp. 1636–1651, Mar. 2022.

[17] T. D. Nguyen, M. Y. Chung, and T. M. Ho, "DeepRoute-EH: Deep reinforcement learning-based routing for energy-harvesting ad hoc networks using Markov decision processes," IEEE Access, vol. 10, pp. 47821–47836, 2022.

[18] R. Sharma, V. Vashisht, and U. Singh, "Metaheuristic optimization algorithms for energy-efficient routing in mobile ad hoc networks: A comprehensive survey," IEEE Access, vol. 11, pp. 21374–21398, 2023.

[19] X. Hua, Y. Wang, J. Lv, W. Zhang, and S. Zhang, "Multi-agent reinforcement learning for cooperative routing in energy-harvesting vehicular networks," IEEE Trans. Intell. Transp. Syst., vol. 23, no. 12, pp. 24871–24884, Dec. 2022.

[20] H. Kim, Y. J. Kim, and S. Han, "DRLR: Deep reinforcement learning-based adaptive routing for large-scale mobile ad hoc networks," IEEE Trans. Netw. Service Manag., vol. 20, no. 2, pp. 1487–1502, Jun. 2023.

[21] F. Tang, Y. Kawamoto, N. Kato, and J. Liu, "Future intelligent and secure vehicular network toward 6G: Machine-learning approaches," Proc. IEEE, vol. 110, no. 2, pp. 145–162, Feb. 2022.

[22] Z. Chen, Q. Zhang, K. Chen, and Y. Li, "Cooperative multi-agent deep reinforcement learning for energy-efficient MANET routing," IEEE Trans. Wireless Commun., vol. 22, no. 4, pp. 2531–2546, Apr. 2023.

[23] Sakthivelu, U., & Vinoth Kumar, C. N. S. (2022). An approach on cyber threat intelligence using recurrent neural network. In ICT Infrastructure and Computing: Proceedings of ICT4SD 2022 (pp. 429-439). Singapore: Springer Nature Singapore.

[24] W. Shi, S. Zhou, Z. Niu, M. Jiang, and L. Geng, "FedRoute: Federated learning-based routing optimization for mobile ad hoc networks with heterogeneous energy sources," IEEE Trans. Mobile Comput., vol. 23, no. 1, pp. 423–438, Jan. 2024.

[25] Y. Sun, M. Peng, Y. Zhou, Y. Huang, and S. Mao, "Application of machine learning in wireless networks: Key techniques and open issues," IEEE Commun. Surv. Tut., vol. 26, no. 1, pp. 202–242, 1st Quart. 2024.

[26] M. J. Aguilar Muñoz, C. J. Angel Rueda, A. Escudero-Nahón, and R. Chaparro-Sánchez, “3D Virtual Worlds as Educational Spaces: Analysis of Uses and Affordances,” IETI Transactions on Data Analysis and Forecasting (iTDAF), vol. 4, no. 2, Jun. 2026, doi: 10.3991/itdaf.v4i2.61305.

[27] Khikmatillo Akhmadjanovich Do‘smatov, Davron Aslonqulovich Juraev, and Y. Li, “Feature Based on Detection of Aviation Decoy Targets Using Dynamic Radar Cross Section Signatures,” IETI Transactions on Data Analysis and Forecasting (iTDAF), vol. 4, no. 2, Jun. 2026, doi: 10.3991/itdaf.v4i2.61479.