Performance Analysis of SNR-Based Scheduling Policies in Asymmetric Broadcast Ergodic Fading Channel

We analyze the performance of SNR-based scheduling algorithms in broadcast ergodic fading channels where multiuser selection diversity is exploited. At each channel state, the user with the highest weighted signal-to-noise ratio is selected to be transmitted. The use of weights associated to the users allows us to control the degree of fairness among users and to arrange them according to a prescribed quality of service. These weights parametrize the scheduling algorithms so each set of weights corresponds to a specific scheduling algorithm. Assuming Rayleigh fading broadcast channel, we derive a closed-form expression for the achievable user's rates as a function of the scheduling algorithm, the channel fading statistics of each user, and the transmit power. With the help of this expression, we solve some interesting inverse problems. For example, for a given arbitrary channel statistics we obtain the optimum scheduling algorithm to achieve a prescribed set of users' rates with minimum transmit power.