This paper addresses the problem of Access Point (AP) selection in large Wi-Fi networks. Unlike current solutions that rely on Received Signal Strength (RSS) to determine the best AP that could serve a wireless user's request, we propose a novel framework that considers the Quality of Service (QoS) requirements of the user's data flow. The proposed framework relies on a function reflecting the suitability of a Wi-Fi AP to satisfy the QoS requirements of the data flow. The framework takes advantage of the flexibility and centralised nature of Software Defined Networking (SDN). A performance comparison of this algorithm developed through an SDN-based simulator shows significant achievements against other state of the art solutions in terms of provided QoS and improved wireless network capacity.

Abstract-

Abstract-This paper focuses on addressing the Access Point (AP) selection problem by relying on a centralized controller that provides a global view of the network. This approach follows the Software-Defined Networking (SDN) concept, which has long been considered in the literature as an innovative method to control management functionalities for wired networks and that is also now becoming a hot topic in the context of Wi-Fi networks. The proposed AP selection approach is based on a novel algorithm which relies on the Fittingness Factor (FF) concept, to maximize a function that reflects the suitability of the available spectrum resources to the application requirements. Specifically, this paper describes the development of a framework that implements the FF-based algorithm for smart AP selection in a centralized controller. The simulated performance comparison of this algorithm against a strategy that maximizes the achievable data rate considered in many papers in the literature, illustrates the important achievements that have been obtained in terms of saved bandwidth and users' satisfaction.

Abstract-In spite of the enormous popularity of Wi-Fienabled devices, the utilization of Wi-Fi radio resources (e.g. RF spectrum and transmission power levels) at Access Points (APs) is degraded in current decentralized Radio Resource Management (RRM) schemes. Most state of the art central control solutions apply configurations in which the network-wide impacts of the involved parameters and their mutual relationships are ignored. In this paper, we propose an algorithm for jointly adjusting the transmission power levels and optimizing the RF channel assignment of APs by taking into account the flows' required qualities while minimizing their interference impacts throughout the network. The proposed solution is tailored for an operatoragnostic and Software Defined Wireless Networking (SDWN)-based centralised RRM in dense Wi-Fi networks. Our extensive simulation results validate the performance improvement of the proposed algorithm compared to the main state of the art alternative by showing more than 25% higher spectrum efficiency, satisfying the users' demands and further mitigating the networkwide interference through a flow-based and quality-oriented power level adjustment.

With the unprecedented technological advances witnessed in the last two decades, more devices are connected to the Internet, forming what is called the Internet of Things (IoT). The IoT devices with heterogeneous characteristics and the quality of experience (QoE) requirements may engage in the dynamic spectrum market due to the scarcity of radio resources. We propose a framework to efficiently quantify and supply radio resources to the IoT devices by developing intelligent systems. The primary goal of this paper is to study the characteristics of the next generation of cellular networks with non-orthogonal multiple access (NOMA) to enable connectivity to clustered IoT devices. First, we demonstrate how the distribution and QoE requirements of IoT devices impact the required number of radio resources in real time. Second, we prove that using an extended auction algorithm by implementing a series of complementary functions enhance the radio resource utilization efficiency. The results show a substantial reduction in the number of sub-carriers required when compared with conventional OMA and the intelligent clustering is scalable and adaptable to the cellular environment. Ability to move spectrum usages from one cluster to other clusters after borrowing when a cluster has fewer users or move out of the boundary is another soft feature that contributes to the reported radio resource utilization efficiency. Moreover, the proposed framework provides IoT service providers cost estimation to control their spectrum acquisition to achieve the required quality of service with a guaranteed bit rate (GBR) and non-GBR.