A novel cognitive architecture for QoS/QoE management in NextG Networks based on Q-learning and R-MLP approaches

Next Generation Networks (NGN) principle or 5G technology, is based on the integration of heterogeneous wireless systems and cognitive networks. NGN is expected to have Quality of Service (QoS) as the major challenging issue. More specifically, NGN aims at providing guaranteed QoS levels and Quality of Experience (QoE), especially for multimedia services. However, with the increase in number of connections, services and data amounts, it becomes crucial to ensure an intelligent network management to ensure target goals. This paper proposes a novel cognitive radio network architecture for the management of cognitive networks and 5G technology. The goal of the proposed architecture consists on using artificial intelligence with learning capabilities to optimize network management decisions. The proposed architecture consists of 4 main states: sensing, decision management, action, and learning. It provides the functionality to manage cognitive components and facilitate the interfaces between the network and its surrounding environment. In order to demonstrate the functionality of the proposed architecture, we present simulation results and comparison with existing load balancing method in terms of throughput, latency and QoE.