STOCHASTIC MODEL OF THE RADIO SPECTRUM ALLOCATION PROCESS WITH VARIABLE CHANNEL OCCUPANCY

Abstract

The article develops a stochastic model of the radio spectrum allocation process with variable channel occupancy, focused on use in cognitive radio networks with dynamic access to frequency resources. The relevance of the study is due to the increasing density of wireless devices, the development of 5G/6G and IoT services, and the limited radio frequency spectrum, which requires increasing the efficiency of its use. Traditional approaches to static frequency allocation do not take into account the temporal unevenness of traffic and the variable activity of primary users, which leads to fragmented channel loading and overloading of individual spectrum sections.

To formalize the access process, a Markov model with transitions between the "free" and "busy" states, which describe the behavior of the primary user, is proposed. Based on the Kolmogorov equation system, analytical expressions for the steady-state probability of channel availability, the probability of collision during transmission, and the effective fraction of the spectrum that can be used by secondary terminals without creating interference were obtained. The model takes into account the intensity of the primary user appearance λ, the intensity of channel clearance μ, and the transmission duration T, which allows us to study the influence of load modes on the quality of service indicators.

The numerical experiments conducted confirmed the adequacy of the analytical dependencies and the consistency of the theoretical results with the modeling in a wide range of parameters. It was established that an increase in the intensity of channel clearance and a decrease in the transmission duration provide an increase in the efficiency of spectrum use to 0.87 with an average collision coefficient of less than 5%. The obtained results demonstrate the possibility of formally determining the optimal access conditions and minimizing spectrum losses in conditions of stochastic channel behavior.

The proposed model can serve as the basis for building adaptive algorithms for dynamic radio resource allocation in cognitive networks, as well as be used when designing next-generation systems with increased requirements for spectral efficiency and transmission reliability.