Bionic algorithm is the bionics applied in the field of algorithm. It is an optimization algorithm which is realized by simulating the existing behavior mechanism of nature. From the viewpoint of system complexity, the earth ecosystem involving millions of organisms has an amazing complexity, and thus it is almost impossible to obtain a satisfactory solution by using mathematical analytical methods to achieve a stable balance among these organisms. However, nature realizes the steady dynamic balance of biological system by the simple and smart ecological balance mechanism. Through simulating ecological balance mechanism, this paper proposes a new bionic algorithm: Interdependent Balance Algorithm (IBA), in which the design variables are regarded as the population species and population size is to correspond to the value of design variable. The convergence of IBA is proved by mathematical analytic method. Accordingly, the self-growth function, interdependent function, resource constraint, and algorithm rules have been constructed for IBA. The algorithmic comparison tests demonstrated the reliability and superiority of IBA, which is easy to realize and especially effective to solve large size problems of complex system optimization.