Joint Optimization of Detection and Power Allocation for OFDM-Based Cognitive Radios

Efficient spectrum sensing ensures cognitive radio users opportunistically use the under-utilized frequency band without causing harmful interference to primary users. However, in practice, owing to the non-zero probability of the missed detection and false alarm of active primary transmission, a certain degree of performance degradation of the primary user is unavoidable. In this paper, we consider OFDM-based communication systems and present efficient algorithms to maximize the total throughput of the cognitive radio by optimizing jointly both the detection operation and the power allocation, taking into account the influence of the probabilities of missed detection and false alarm.The optimization problem can be formulated as a two-variable non-convex problem, which can be solved approximately by using an alternating direction optimization method. A novel criterion is introduced to ensure that the performance degradation of the primary user is bounded properly. First, we analyze the case of only one cognitive radio and then we generalize to the case of a two cognitive radio non-cooperative power allocation game, showing that a Nash equilibrium can be achieved in few iterations by using an iterative alternating direction optimization method. Simulation results demonstrate that the proposed solution can considerably improve system performance.