Adaptive Task Offloading in IoT Networks Using AI and Reinforcement Learning for Energy and Latency Optimization

Abstract

The exponential growth of Internet of Things (IoT) devices has led to increased energy consumption and latency in conventional network architectures, where devices process all tasks locally. This research investigates the use of AI-based edge computing to optimize task allocation in IoT networks, aiming to reduce energy consumption and latency while maintaining adaptability under dynamic conditions. Three strategies are compared: conventional local-only processing, AI-only predictive offloading, and AI combined with Reinforcement Learning (AI + RL) adaptive offloading. The simulation results demonstrate that the AI-only system significantly reduces total energy consumption by approximately 60% and average latency by around 62% compared to conventional processing, while the AI + RL system provides adaptive task allocation that balances energy efficiency with network performance. Additional performance metrics, including maximum latency, minimum latency, latency standard deviation, and task offloading percentage, further validate the effectiveness of AI-based edge computing. These findings indicate that integrating AI and RL into edge computing can substantially improve the efficiency, responsiveness, and robustness of IoT networks, making them suitable for real-time, latency-sensitive applications.

Introduction

The rapid expansion of the Internet of Things (IoT) has increased the need for efficient data processing, but traditional local processing on resource-constrained devices leads to high energy consumption, increased latency, and poor adaptability to dynamic network conditions. Edge computing addresses these issues by enabling task offloading to nearby edge servers, though naïve offloading can introduce new inefficiencies. Integrating Artificial Intelligence (AI) and Reinforcement Learning (RL) enables intelligent, adaptive task allocation that balances energy efficiency and latency based on real-time device and network conditions.

This research evaluates three task-processing strategies in IoT networks: (1) conventional local processing, (2) AI-based predictive offloading using a Multi-Layer Perceptron (MLP), and (3) an adaptive AI + RL method that learns optimal offloading decisions through continuous interaction with the environment. A simulated network of 200 IoT devices is used, each characterized by CPU utilization, data size, battery level, bandwidth, and latency requirements. Energy and latency models are defined for both local processing and edge offloading, and performance metrics include energy consumption, average and maximum latency, latency variance, and offloading percentage.

Results show that the AI-only system is the most energy-efficient, reducing energy usage by ~60% compared to conventional processing. It also achieves the lowest average latency. The AI + RL system provides a balance by adapting to changing conditions, reducing energy and latency compared to conventional processing but not outperforming the AI-only approach in efficiency due to continuous exploration. Overall, the study demonstrates that integrating AI and RL with edge computing significantly improves IoT network performance, offering adaptive, energy-efficient, and low-latency task allocation in heterogeneous environments.

Conclusion

The findings of this study confirm that AI-driven edge computing provides a highly effective solution for improving the performance and sustainability of IoT networks. By intelligently managing computational tasks, the AI-only system achieved remarkable energy savings of nearly 60%, demonstrating its capability to significantly extend the operational life of battery-powered IoT devices. Moreover, both AI-only and AI + RL models achieved substantial latency reductions compared to conventional systems, indicating faster and more efficient task execution. The adaptability of the AI + RL system further underscores its value for dynamic environments, where real-time decision-making and resource optimization are essential. Additionally, the enhanced robustness and consistency of latency performance highlight the reliability of AI-based approaches. Overall, integrating AI and reinforcement learning into edge computing frameworks not only optimizes energy consumption and response time but also strengthens the adaptability and resilience of IoT infrastructures—paving the way for more intelligent, responsive, and energy-efficient next-generation IoT ecosystems.

References

[1] Ahmad, T.; Zhang, D. Using the internet of things in smart energy systems and networks. Sustain. Cities Soc. 2021, 68, 102783. [2] Mahmood, M.; Chowdhury, P.; Yeassin, R.; Hasan, M.; Ahmad, T.; Chowdhury, N.U.R. Impacts of digitalization on smart grids, renewable energy, and demand response: An updated review of current applications. Energy Convers. Manag. X 2024, 24, 100790. [3] Prasad, S.S.; Kumar, C. A Green and Reliable Internet of Things. Commun. Netw. 2013, 5, 44–48. [4] Ravandi, B.; Papapanagiotou, I. A Self-Learning Scheduling in Cloud Software Defined Block Storage. In Proceedings of the 2017 IEEE 10th International Conference on Cloud Computing (CLOUD), Honolulu, HI, USA, 25–30 June 2017; pp. 415–422. [5] Edge to Core and the Internet of Things|Dell Technologies Info Hub. Available online: https://infohub.delltechnologies.com/en-us/t/edge-to-core-and-the-internet-of-things-2/ [6] Cisco Annual Internet Report-Cisco Annual Internet Report (2018–2023) White Paper-Cisco. Available online: https://www.cisco.com/c/en/us/solutions/collateral/executive-perspectives/annual-internet-report/white-paper-c11-741490.html [7] Google and IBM Announce Academic Data Center Collaboration-DCD. Available online: https://www.datacenterdynamics.com/en/news/google-and-ibm-announce-academic-data-center-collaboration/ [8] Nieuwenhuis, L.J.M.; Ehrenhard, M.L.; Prause, L. The shift to Cloud Computing: The impact of disruptive technology on the enterprise software business ecosystem. Technol. Forecast. Soc. Change 2018, 129, 308–313. [9] Kumar Yadav Yanamala, A.; Pointe Blvd, O. Emerging Challenges in Cloud Computing Security: A Comprehensive Review. Int. J. Adv. Eng. Technol. Innov. 2024, 4, 448–479. [10] Ometov, A.; Molua, O.L.; Komarov, M.; Nurmi, J. A Survey of Security in Cloud, Edge, and Fog Computing. Sensors 2022, 22, 927. [11] Singh, S.P.; Nayyar, A.; Kumar, R.; Sharma, A. Fog computing: From architecture to edge computing and big data processing. J. Supercomput. 2018, 75, 2070–2105. [12] Shi, W.; Cao, J.; Zhang, Q.; Li, Y.; Xu, L. Edge Computing: Vision and Challenges. IEEE Internet Things J. 2016, 3, 637–646. [13] Mittal, S.; Negi, N.; Chauhan, R. Integration of edge computing with cloud computing. In Proceedings of the 2017 International Conference on Emerging Trends in Computing and Communication Technologies (ICETCCT), Dehradun, India, 17–18 November 2017; pp. 1–6. [14] Dai, W.; Nishi, H.; Vyatkin, V.; Huang, V.; Shi, Y.; Guan, X. Industrial Edge Computing: Enabling Embedded Intelligence. EEE Ind. Electron. Mag. 2019, 13, 48–56. [15] Zhang, T.; Li, Y.; Chen, C.L.P. Edge computing and its role in Industrial Internet: Methodologies, applications, and future directions. Inf. Sci. 2021, 557, 34–65. [16] Gooi, H.B.; Wang, T.; Tang, Y. Edge Intelligence for Smart Grid: A Survey on Application Potentials. CSEE J. Power Energy Syst. 2023, 9, 1623–1640. Li, J.; Gu, C.; Xiang, Y.; Li, F. Edge-cloud Computing Systems for Smart Grid: State-of-the-art, Architecture, and Applications. J. Mod. Power Syst. Clean Energy 2022, 10, 805–817 [17] Minh, Q.N.; Nguyen, V.H.; Quy, V.K.; Ngoc, L.A.; Chehri, A.; Jeon, G. Edge Computing for IoT-Enabled Smart Grid: The Future of Energy. Energies 2022, 15, 6140. [18] Singh, N.; Buyya, R.; Kim, H. Securing Cloud-Based Internet of Things: Challenges and Mitigations. Sensors 2025, 25, 79. [19] Yousefpour, A.; Fung, C.; Nguyen, T.; Kadiyala, K.; Jalali, F.; Niakanlahiji, A.; Kong, J.; Jue, J.P. All One Needs to Know about Fog Computing and Related Edge Computing Paradigms: A Complete Survey. J. Syst. Archit. 2019, 98, 289–330. [20] Liang, S.; Jin, S.; Chen, Y. A Review of Edge Computing Technology and Its Applications in Power Systems. Energies 2024, 17, 3230. [21] Mansouri, Y.; Babar, M.A. A review of edge computing: Features and resource virtualization. J. Parallel Distrib. Comput. 2021, 150, 155–183. [22] Mell, P.; Grance, T. The NIST Definition of Cloud Computing, Special Publication (NIST SP); National Institute of Standards and Technology: Gaithersburg, MD, USA, 2011. [23] Habeeb, R.A.A.; Nasaruddin, F.; Gani, A.; Hashem, I.A.T.; Ahmed, E.; Imran, M. Real-time big data processing for anomaly detection: A Survey. Int. J. Inf. Manag. 2019, 45, 289–307. [24] Hua, H.; Li, Y.; Wang, T.; Dong, N.; Li, W.; Cao, J. Edge Computing with Artificial Intelligence: A Machine Learning Perspective. ACM Comput. Surv. 2023, 55, 184. [25] AL-Jumaili, A.H.A.; Muniyandi, R.C.; Hasan, M.K.; Paw, J.K.S.; Singh, M.J. Big Data Analytics Using Cloud Computing Based Frameworks for Power Management Systems: Status, Constraints, and Future Recommendations. Sensors 2023, 23, 2952. [26] Aburukba, R.O.; AliKarrar, M.; Landolsi, T.; El-Fakih, K. Scheduling Internet of Things requests to minimize latency in hybrid Fog–Cloud computing. Futur. Gener. Comput. Syst. 2020, 111, 539–551. [27] Caiza, G.; Saeteros, M.; Oñate, W.; Garcia, M.V. Fog Computing at Industrial Level, Architecture, Latency, Energy, and Security: A Review. Heliyon 2020, 6, e03706. [28] Fang, X.; Misra, S.; Xue, G.; Yang, D. Managing smart grid information in the cloud: Opportunities, model, and applications. IEEE Netw. 2012, 26, 32–38. Popeang?, J. Cloud Computing and Smart Grids. Database Syst. J. 2012, III, 57–66. [29] Caiza, G.; Saeteros, M.; Oñate, W.; Garcia, M.V. Fog Computing at Industrial Level, Architecture, Latency, Energy, and Security: A Review. Heliyon 2020, 6, e03706. [30] Fang, X.; Misra, S.; Xue, G.; Yang, D. Managing smart grid information in the cloud: Opportunities, model, and applications. IEEE Netw. 2012, 26, 32–38. Popeang?, J. Cloud Computing and Smart Grids. Database Syst. J. 2012, III, 57–66. [31] Song, Y.; Chen, Y.; Yu, Z.; Huang, S.; Shen, C. CloudPSS: A high-performance power system simulator based on cloud computing. Energy Rep. 2020, 6, 1611–1618. [32] AL-Jumaili, A.H.A.; Mashhadany, Y.I.A.; Sulaiman, R.; Alyasseri, Z.A.A. A Conceptual and Systematics for Intelligent Power Management System-Based Cloud Computing: Prospects, and Challenges. Appl. Sci. 2021, 11, 9820. [33] M. Advancements in intelligent cloud computing for power optimization and battery management in hybrid renewable energy systems: A comprehensive review. Energy Rep. 2023, 10, 2206–2227. [34] Allahvirdizadeh, Y.; Moghaddam, M.P.; Shayanfar, H. A survey on cloud computing in energy management of the smart grids. Int. Trans. Electr. Energy Syst. 2019, 29, e12094. [35] Bagherzadeh, L.; Shahinzadeh, H.; Shayeghi, H.; Dejamkhooy, A.; Bayindir, R.; Iranpour, M. Integration of Cloud Computing and IoT (CloudIoT) in Smart Grids: Benefits, Challenges, and Solutions. In Proceedings of the International Conference on Computational Intelligence for Smart Power System and Sustainable Energy, CISPSSE 2020, Keonjhar, India, 29–31 July 2020. [36] Stergiou, C.; Psannis, K.E.; Kim, B.G.; Gupta, B. Secure integration of IoT and Cloud Computing. Futur. Gener. Comput. Syst. 2018, 78, 964–975 [37] Wang, C.; Bi, Z.; Da Xu, L. IoT and cloud computing in automation of assembly modeling systems. IEEE Trans. Ind. Inform. 2014, 10, 1426–1434. [38] What is Edge Computing?|Cloudflare. Available online: https://www.cloudflare.com/learning/serverless/glossary/what-is-edge-computing/ [39] Satyanarayanan, M. The emergence of edge computing. Computer 2017, 50, 30–39. [40] Hong, X.; Wang, Y. Edge Computing Technology: Development and Countermeasures. Chin. J. Eng. Sci. 2018, 20, 20–26. [41] Khan, W.Z.; Ahmed, E.; Hakak, S.; Yaqoob, I.; Ahmed, A. Edge computing: A survey. Futur. Gener. Comput. Syst. 2019, 97, 219–235. [42] What Is Edge Computing? Introduction to Edge Computing. Available online: https://stlpartners.com/articles/edge-computing/what-is-edge-computing /

Copyright

Copyright © 2025 Rodan Kumar Dharwaiya, Dr. Nidhi Tiwari, Dr. Shivangini Morya. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.