Power/Bit Loading in OFDM-Based Cognitive Networks With Comprehensive Interference Considerations: The Single-SU Case

Conflict-free power allocation for secondary users (SUs) in cognitive networks is a challenging problem because the accessible spectrum of the SUs is shared with the primary users (PUs). The problem becomes particularly difficult when the objective is to achieve maximum network capacity considering the mutual interference between the PUs and the SUs. In this paper, we specify the case with a single SU and multiple PUs, where both the SU and the PUs are orthogonal frequency-division multiplexing (OFDM) modulated. The power allocation for the subcarriers of the SU is modeled into a constrained optimization problem, where the mutual interference between the SU and the PUs is comprehensively formulated as restrictions on the SU's transmission power. As a result, the proposed modeling scheme restrains the interference to the PUs, as well as maximizing the capacity of the SU. A novel iterative power-loading algorithm with low computational complexity is proposed to realize the power loading. On this basis, a suboptimal integral-bit-loading algorithm is further presented. In a simplified scenario, simulation results are exhibited to confirm the efficiency of the proposed power and bit-loading algorithms. Finally, the influence of the mutual interference on the SU's power/bit loading and the system capacity is illustrated.