Fire detection techniques have received considerable critical attention over the past ten years. Regardless of the progress in the area of fire detection, questions have been raised about the cost, complexity, consumed power from a large number of sensors to analyze sensors’ data. Debate continues about the best strategies for the management of consumed power and how to accelerate the processing of real-time data in fire detection through the internet of things. This paper presents a partial coverage and a power-aware IoT-based fire detection model with different multi-functional sensors in smart cities. In the proposed model, the sleep scheduling approach used for saving sensors energy. This approach will significantly help in saving consumed energy and thus the need for any extra number of nodes required for continuously covering the target area. Moreover, fog computing is applied to process real-time data aggregated from a large number of sensors for running systems more efficiently. Validation of the proposed model was carried out via simulation and experimental testbed implementation with Arduino, sensors, and Raspberry pi. The results obtained indicate how the proposed technique can efficiently determine the sensors to meet the constraints imposed. The most striking finding to emerge from this experimental and simulation study is that the proposed technique helps in ensuring excellent performance in terms of the number of active nodes and the network lifetime, over other state-of-the-art techniques. The most striking finding of this work compared to MWSAC and PCLA, while covering the same area, is the minimization of the number of active nodes by 64.33% and 15%. It raises the network’s lifetime by 91.32% compared to MWSAC and 12% compared to PCLA, respectively.

在过去的十年中，火灾探测技术受到了相当重要的关注。不论火灾探测领域的进展如何，都存在有关大量传感器用于分析传感器数据的成本，复杂性和功耗等问题。辩论将继续探讨有关管理功耗的最佳策略，以及如何通过物联网在火灾探测中加速实时数据的处理。本文介绍了智能城市中具有不同多功能传感器的部分覆盖范围和基于物联网的基于功率感知的火灾探测模型。在提出的模型中，睡眠调度方法用于节省传感器能量。这种方法将极大地帮助节省能耗，因此需要连续覆盖目标区域所需的任何数量的额外节点。此外，雾计算被应用于更有效地处理从大量传感器聚合而来的实时数据，以运行系统。通过使用Arduino，传感器和Raspberry pi进行仿真和实验测试平台实现，对提出的模型进行了验证。获得的结果表明所提出的技术如何能够有效地确定传感器以满足所施加的约束。从这项实验和仿真研究中得出的最惊人的发现是，与其他最新技术相比，所提出的技术有助于确保在活动节点数和网络寿命方面的出色性能。与MWSAC和PCLA相比，这项工作最令人惊讶的发现是，其活动节点的数量最小化了64.33％和15％，尽管它们覆盖相同的区域。