ICTACT Journal on Communication Technology

Dhananjay Kumar ¹, Pavithra Hari ¹, Panbhazhagi Selvaraj ¹, Sharavanti Baskaran ¹

Affiliations
1 Department of Information Technology, Anna University, MIT Campus, Chennai, IN

Abstract

Machine learning schemes can be employed in cognitive radio systems to intelligently locate the spectrum holes with some knowledge about the operating environment. In this paper, we formulate a variation of Actor Critic Learning algorithm known as Continuous Actor Critic Learning Automaton (CACLA) and compare this scheme with Actor Critic Learning scheme and existing Q-learning scheme. Simulation results show that our CACLA scheme has lesser execution time and achieves higher throughput compared to other two schemes.

Keywords

Markov Decision Process, Reinforcement Learning, Q–Learning, Actor–Critic Learning, CACLA.

Full Text

     

Dhananjay Kumar ¹, A. Magesh ¹, U. Mohammed Hussain ¹

Affiliations
1 Department of Information Technology, Madras Institute of Technology, Anna University Chennai, IN

Abstract

Most common works for rendezvous in cognitive radio networks deal only with two user scenarios involving two secondary users and variable primary users and aim at reducing the time-to-rendezvous. A common control channel for the establishment of communication is not considered and hence the work comes under the category of 'Blind Rendezvous'. Our work deal with multi-user scenario and provides a methodology for the users to find each other in the very first time slot spent for rendezvous or otherwise called the firstattempt- rendezvous. The secondary users make use of the history of past communications to enable them to predict the frequency channel that the user expects the rendezvous user to be. Our approach prevents greedy decision making between the users involved by the use of a cut-off time period for attempting rendezvous. Simulation results show that the time-to-rendezvous (TTR) is greatly reduced upon comparison with other popular rendezvous algorithms.

Keywords

Cognitive Radio, Rendezvous, Time to Rendezvous, First Attempt Rendezvous, Channel Hopping.

Full Text