Cognitive Internet of Things: Challenges and Solutions

Abstract

AbstractInternet of Things (IoT) technology will be used in a wide range of applications such as healthcare and transportation. Because of distributed and dynamic nature of IoT-based systems, designing management algorithms for them results in challenging problems. On the other hand, cognitive computing refers to systems that emulate cognitive processes inspired by human thinking processes for solving problems. The Cognitive Internet of Things (CIoT) technology refers to a combination of cognitive systems and IoT. A benefit of this technology is that the management problems of the IoT will be solved by solutions that come from cognitive systems domain. In recent years, CIoT has received much attention. This is because the management algorithms of the IoT are very complex and therefore human thinking processes as computerized cognitive processes are required for fast and accurate decision-making. Existing algorithms reported as management mechanisms for CIoT only focus on utilizing artificial intelligence without considering fundamental issues of cognitive systems such as predictability, controllability, responsibility, safety, security, and energy consumption. Ignoring these issues may result in designing inappropriate cognitive systems that may be faced with a wide range of problems such as hacking and failure. CIoT inherited numerous challenges from artificial intelligence, IoT, and cognitive systems. Therefore, the challenges of these fields should be studied to extract the challenges in designing CIoT. In the literature, there is no study on extracting the challenges considering associated technologies to CIoT. In this paper, the challenges of the associated technologies are summarized. Then, some important challenges in designing CIoT are obtained.KeywordsInternet of ThingsCognitive systemsArtificial intelligenceCognitive Internet of ThingsManagement mechanisms