| A phased array antenna has advantages of flexible beam shape/size and rapid cycling across the coverage area, by the use of solid state power amplifiers and an antenna patterning matrix. Optimum scheduling of a phased array antenna system can enhance the performance of the communication satellite for bursty high-rate data users over high frequency bands. In this paper, we derive an optimum scheduling policy for a phased array antenna satellite to serve a large number of users with a small number of onboard modulators. Then, from the optimum analytic result, we develop a near-optimum real-time algorithm of performing active user selection, power allocation, and admission control. The algorithm serves users with better channel conditions, less interference, and higher queuing delays. Power is allocated for the selected users to have the same marginal returns of a composite cost function with respect to consumed power. We introduce a total accumulated delay constraint for admission control and system stability. To suppress interference between close-in users, the algorithm makes a choice between interference suppression (in the form of space division multiplexing) and sequential service (in the form of time division multiplexing), depending on users' geographic distribution. The simulation result indicates that the real-time algorithm can achieve a throughput close to the analytic steady-state upper bound. |