Towards an Intelligent Internet: Surveying the Influence of Cognitive Technologies on Network Architecture and Performance.

Internet of Things (IoT) is a heterogeneous, mixed and uncertain ubiquitous network, the application prospect of which is extensive in the field of modern intelligent service. Having done a deep investigation on the discrepancies between service offering and application requirement, we believed that current IoT lacks enough intelligence and cannot achieve the expected increasing applications performance. By integrating intelligent thought into IoT, I presented a new concept of Cognitive Internet of Things (CIoT) in this paper. CIoT can apperceive current network conditions, analyze the perceived knowledge, make intelligent decisions, and perform adaptive actions, which aim to maximize network performance. We model the topology of the CIoT network, develop technologies related to the cognitive process, and analyze the results of cooperative intelligence based on game theory. This shows that these new designs can bring intelligence to IoT and significantly improve system performance. Finally, application cases based on the CIoT concept were presented.