In numerous Internet of Things (IoT) scenarios (such as Intelligent Transportation System (ITS) or environment sensing), the ad hoc mode is widely used. Ad-hoc networks have several benefits including infrastructureless and spontaneous deployment, and the opportunity to set up temporary and dynamic networks cost-effectively. However, some of their characteristics such as the unreliability of wireless links, dynamic topology and resource constraints in energy and processing power of objects, make new protocols’ development ignore security requirements of applications in order to satisfy some quality of service. Consequently, there is a lack of standardization works in the field of secure routing among IoT devices. Furthermore, recently proposed secure routing approaches (mostly from the academic community) do not meet the current needs and requirements of secure routing for IoT applications. In this context, we propose a lightweight and secure routing scheme that combines multipath routing and trust management, that can adapt to different scenarios in the IoT environment. To analyze and evaluate our scheme, we propose a probabilistic model that considers two types of events that affect routing performance for a given node: mobility and uncooperative behavior. Based on this mathematical model, we compared the performance of our proposed routing scheme with two other multipath routing protocols namely, SMORT and DMRP. Our performance results show that our scheme is highly scalable, more efficient, and resilient compared with previous related works and is particularly suitable for insecure large and dense IoT networks.

在众多的物联网（IoT）场景中（例如智能运输系统（ITS）或环境感应），自组织模式被广泛使用。临时网络具有多种优势，包括无基础架构和自发部署，以及以经济高效的方式建立临时和动态网络的机会。但是，它们的某些特性，例如无线链路的不可靠性，动态拓扑以及对象的能量和处理能力中的资源限制，使得新协议的开发忽略了应用程序的安全性要求，从而无法满足某些服务质量。因此，在物联网设备之间的安全路由领域缺乏标准化工作。此外，最近提出的安全路由方法（大部分来自学术界）不能满足当前物联网应用安全路由的需求。在这种情况下，我们提出了一种轻量级且安全的路由方案，该方案结合了多路径路由和信任管理，可以适应IoT环境中的不同场景。为了分析和评估我们的方案，我们提出了一种概率模型，该模型考虑了两种影响给定节点的路由性能的事件：移动性和不合作行为。基于此数学模型，我们将建议的路由方案的性能与其他两个多径路由协议进行了比较，可以适应IoT环境中的不同场景。为了分析和评估我们的方案，我们提出了一种概率模型，该模型考虑了两种影响给定节点的路由性能的事件：移动性和不合作行为。基于此数学模型，我们将建议的路由方案的性能与其他两个多径路由协议进行了比较，可以适应IoT环境中的不同场景。为了分析和评估我们的方案，我们提出了一种概率模型，该模型考虑了两种影响给定节点的路由性能的事件：移动性和不合作行为。基于此数学模型，我们将建议的路由方案的性能与其他两个多路径路由协议进行了比较，SMORT和DMRP。我们的性能结果表明，与以前的相关工作相比，我们的方案具有更高的可扩展性，效率和弹性，并且特别适用于不安全的大型密集物联网网络。