Implementation of the Reinforcement Learning Mechanism in the Random Access Channel Procedure

Abstract

An analytical model for establishing a connection in a random access channel is developed. he probabilistic-temporal characteristics are analyzed, such as the probability of a successful and unsuccessful connection establishment, the average delay of a successful connection establishment. The dependence of the listed characteristics on the probability of collision possible with the transmission of the preamble is investigated. a method for selecting a preamble based on a reinforcement learning mechanism is proposed, the probability of choosing the same preambles is determined, a probabilistic-temporal characteristics of the connection establishment procedure are compared with the equally probable and proposed preamble selection methods.