A reliable alternate framework using genetic algorithm (GA) hybridized with Sequential quadratic programming (SQP), a powerful nonlinear adaptive technique, is presented for identifying the directions of the source signals incident on the sensor array comprising of smart antenna system. The multiple sources are taken in the far field of sensors array and a cost function is formulated in mean square error sense that requires only a single snapshot. The given scheme is evaluated for several test scenarios using different number of antenna elements and signal sources. The applicability, stability and reliability of the designed approach are validated through statistical analysis based on Monte Carlo simulations. The hybrid approach (GA-SQP) is found to be the most accurate and consistently convergent in comparison with similar stochastic technique applied in this field so far. Beside this, the additional benefits of the proposed scheme is ease in implementation, simplicity of approach, less hardware implementation burden and especially suitable for real-time applications.