Cognitive Radio Made Practical: Forward-Lookingness and Calculated Competition

• Main Authors: Jie Ren, Kai-Kit Wong
• Format: Article
• Language: English
• Published: IEEE 2019-01-01
• Series: IEEE Access
• Subjects: Cognitive radios; negotiation mechanism; noncooperative game theory; OFDMA
• Online Access: https://ieeexplore.ieee.org/document/8574896/

Cognitive radio is more than just radio environment awareness, and more importantly, has the ability to interact with the environment in the best way possible. Ideally, cognitive radios will form a self-regulating society of mobile radios achieving maximum spectrum utilization. However, challenges arise as mobile radios tend to compete with one another for spectrum, generating harmful interference, and damaging performance individually and for the network as a whole. In this paper, we present a framework that allows competing radios to teach and learn from each other&#x2019;s action so that a desirable equilibrium can be reached. The heart of cognition to establish this is the <italic>forward-looking</italic> ability, which enables competing radios to see beyond the present time, negotiate and optimize their actions toward a more agreeable equilibrium. Technically speaking, we adopt a belief-directed game where each mobile radio, regarded as player, formulates a belief function to project how the radio environment as a whole would respond to any of its action. This model facilitates engineering of the equilibrium by different choices of the players&#x2019; belief functions. Under this model, players will negotiate naturally through a sequence of calculated competition (i.e., cycles of teaching and learning with each other). We apply this methodology to a cognitive orthogonal frequency-division multiple-access radio network where mobile users are free to access any of the subcarriers and thus compete for radio resources to maximize their rates. The results reveal that the proposed negotiation-by-forward-looking competition mechanism guides users to converge to an equilibrium that benefits not only the individual users but the entire network approaching the maximum achievable sum-rate.