FIS based cognitive radio scheduling
Debjani Mitra, Rupanwita Das Mahapatra · 2013
Cognitive radio is the latest technology paradigm enabling dynamic spectrum access for exploiting unused spectrum resources, cross-layer optimization and achieving optimal system performance using learning and reasoning techniques. This paper derives a generalized decision surface for opportunistic spectrum scheduling by cognitive users in licensed spectrum using Fuzzy Inference System. The proposed FIS has incorporated the PU-SU distance and signal strength inputs in a logical reasoning to balance the tradeoff between spectrum sensing accuracy and interference to PU. The influence of the shape of the MF on the dynamics of the variation in decision making has been evaluated for a spectrum sharing model using different secondary user parameters as input: signal power, spectrum utilization efficiency, degree of mobility and its distance to the primary user. The capabilities of two different rule base frameworks have been compared to develop an optimized FIS structure in terms of the type of membership function for predicting the best scheduled set of secondary users. Simulation results show the feasibility of evaluating CR scheduling with the evaluated possibility factors from built-in fuzzy inference decision surfaces. The superiority of the trapezoidal MF and the distance-factor trade-off inclusive 81 rule base in this regard is also clearly established.