Genetic algorithm with multi-objective optimization strategy was proposed to synthesize the antenna array on LEO satellite according to the requirements of iso-flux beams coverage. Based on the rotation symmetry of beam coverage and elements distribution

antenna elements were partitioned into groups in beam space and element space

and just one third of beam forming matrix coefficients were independent. The simplicity of DBF matrix helped reduce the number of parameters in GA and design the hardware structure of DBF network. Generalized DFT filter bank

multiplex DBF computer and RF calibration in close-loop mode were adopted to design the DBF network

which can reduce the calculation by 95.5% and save the multiplier resource up to 80%. Ultimately a 61-channel phased array transmitter antenna with 16 simultaneous beams was designed. The measurements in anechoic chamber demonstrate that the phased array antenna has good performance of iso-flux beams coverage

which validates the rationality of algorithms and project implementation.