Abstract: For the underlay cognitive radio networks with multiuser under parameter uncertainties, a robust distributed robust power control algorithm was proposed. This algorithm was formulated to minimize total transmit power of SUs under the interference temperature and SINR constraint. Based on euclidean ball-shaped uncertainty, the robust distributed solution of the optimization problem was obtained by using Lagrange dual decomposition theory. The simulation results illustrate that, compared with the non-robust algorithm and traditional SOCP algorithm, the proposed algorithm achieves a better performance and guarantees the QoS requirement for both SUs and PUs under the parameter uncertainty.