Utilizing Cognitive Radio and Reinforcement Learning for Multiple Preamble Detection in 5G-PRACH

Network congestion in the Fifth Generation (5G) network represents that a major hindrance in the communication process may lead to latency. Such an issue needs to be handled at any cost in order to enable 5G key services like Ultra Reliability and Low Latency (URLLC), massive Machine Type Communication (mMTC), enhanced Mobile Broadband (eMBB) in the network. To specifically handle latency, Cognitive Radio Network (CRN) could be a comprehensive solution to be implemented as in CRN there is a concept of spectrum sharing between licensed (primary user) and unlicensed (secondary users) which solves the problem of congestion and eventually latency gets reduced. 5G already has a complexed architecture with too many complicated use cases, preamble detection in Physical Random Access Channel (PRACH) is one out of those issues. Detecting multiple preamble over PRACH at Uplink communication before hand is a crucial problem, related to spectrum scarcity. In this paper, authors have tried to focus on improving and handling the complication of spectrum congestion in 5G-PRACH using CRN based solution with Reinforcement Learning (RL). Some new obstacles, may occur in successful deployment of 5G system for achieving 5G as a Self Organizing Network (SON). Problem of detecting real primary user and primary user emulator (attacker) at receiver, should be kept in consideration. The main contribution of this paper is to study cognitive radio over PRACH with using RL for taking its benefits to optimize the decision making process.