Performance analysis and power allocation strategy in overlay cognitive networks

In this paper, we propose a power allocation strategy for overlay cognitive networks, where each node is equipped with single antenna. Owing to the overly strategy, the secondary user (SU) can help transmitting the primary user's (PU) data and meanwhile can convey itself own data with the superposition coding (SC). We first analyze the bit-error-rate (BER) of the PU and the SU Then, the power allocation strategy is devised by minimizing total power, providing that the BER of the PU and that of SU are guaranteed. Since the BER formulations are not convex, the optimization is difficult to conduct. We then propose two new tractable BER approximations for the PU and SU, which can sophisticatedly transfer the design into a convex problem. The solution can thus be obtained. Simulations verify that our power allocation design is validated for different channel environments.