We model the power allocation interaction between a cognitive radio and a jammer as a two-player zero-sum game. First, we determine the power allocation strategy for the cognitive radio using a modified version of the Q-learning algorithm against fixed jamming strategies. The learned anti-jamming strategy will be compared to the common waterfilling technique. Then, we consider the power allocation game using Q-learning for both the cognitive radio and the jammer. The learned strategies will be compared to the Nash equilibrium found under the assumption of perfect knowledge. Finally, we consider the real scenario of a jammer with imperfect information.
27 March 2021 in International Journal of Machine Learning and Cybernetics
Following 3G and 4G cellular networks, 5G is already deployed in some countries and 6G is under development looking for higher speed and lower latency. This evolution of wireless industry and mobile users requirements enhance the exploration of new spectral access strategies. Even though, millimeter waves offer promising bandwidth, it suffer from limited range due to high path loss. Therefore, new generations of cellular networks are interested in Sub-6GHz spectrum that has adequate propagation properties and has many bands released for unlicensed access. However, these frequencies are used by the majority of commercial wireless technologies such as WIFI, Bluetooth and 3G/4G cellular networks. Their spectral coexistence with new wireless services over unlicensed bands requires mutual interference control based on controlling the power levels of co-located operators. To address this issue, we present a coexistence solution based on power control enabling maximum cell transmission capacity. We formulate the interaction between two co-located operators as a competitive game. Each operator-user pair tries to maximize its sum transmission capacity over the target channels under maximum power constraint. We develop the analytical expressions of the optimal power allocation vectors, then we keep the base stations playing the game iterations until the convergence to a Nash equilibrium. We provide simulation results proving the performance of the proposed model.