Modelo adaptativo multivariable de handoff espectral para incrementar el desempeño en redes móviles de radio cognitiva

Abstract

Spectral handoff, in cognitive radio networks, occurs when the secondary user must leave the frequency channel they are using and continue their communication at another spectral opportunity. This process is a key aspect to guarantee adequate quality of service and improve performance in secondary user communications. This research book aims to present a proposal for a multivariate adaptive spectral handoff model for mobile cognitive radio networks. For this purpose, three algorithms were developed for decision-making during a spectral handoff, with different approaches: fuzzy, feedback and predictive; These make up the proposed adaptive multivariate spectral handoff model. To evaluate the level of performance of the developed algorithms, a comparative analysis was carried out between them and the most relevant spectral handoff algorithms in current literature. Unlike related works, the benchmark was validated through a trace of real spectral occupancy data captured in the GSM and Wi-Fi frequency band, which models the real behavior of primary users. In the validation phase, eight evaluation scenarios were proposed, considering two types of networks: GSM and Wi-Fi, two classes of applications: real-time and best-effort, two traffic levels: high and low, and ten evaluation metrics.