This paper presents an adaptive cross-layer resource allocation problem with the fairness in multi-user OFDM-MISO communication systems, and provides two solutions with Hopfield Neural Network (HNN) and Genetic Algorithm (GA) for the problem. We utilized HNN’s characteristics such as parallel processing, fast convergence speed and easy convergence to the optimum, to solve this problem under the conditions of proportional fairness for satisfying system performances and users’ requirements. The method is simplified in the computation by dividing the bit-loading matrix into three matrixes. The simulation results show that HNN and GA can effectively solve optimization problems of resource allocation in such system, and results of selected HNN and GA methods are more effective than that of the traditional method.