Askar, N.A.Habbal, A.2024-09-292024-09-2920242169-3536https://doi.org/10.1109/ACCESS.2024.3456669https://hdl.handle.net/20.500.14619/9484Internet of Things technology is seeing huge success in various fields. Smart objects combined with Internet connection have become an essential part of every aspect of human life. The way people interact with the things around them has inevitably changed. The IoT system presents many challenges. Devices are heterogeneous and limited in energy and memory. The applications used require a continuous and stable connection to transmit information effectively. This results in high energy consumption. Named Data Networking (NDN) is a promising networking concept. Unlike traditional networking, it's a data-driven model. It uses names to identify and retrieve data instead of device addresses. NDN provides a simple and efficient forwarding mechanism which makes it suitable for IOT communication. In this paper, we proposed forwarding strategy based on reinforcement learning for NDN-based IoT communications. The proposal integrates the reinforcement learning algorithm in the path selection strategy to optimize the overall energy consumption and extend the network lifetime. This research consists of two schemes firstly provide the complexities and dynamic nature of real-world IoT environments, finally, enhance the interest forward strategy. Our proposed research is implemented in ndnSIM and compared with state of the-art IOT-NDN forwarding strategies. The obtained results show clearly the effectiveness and robustness of our solution which outperforms the benchmarked methods in terms of energy consumption, network lifetime, retrieval time, and satisfactory rates. © 2013 IEEE.eninfo:eu-repo/semantics/openAccessefficient energy consumptionforwarding strategyInternet of Thingsmachine learningNamed Data networkingQ-learning algorithmArticle10.1109/ACCESS.2024.34566692-s2.0-85204126389Q1