Joint Power Control and Receiver Optimization of CDMA Transceivers using Successive Interference Cancellation

In a broadband uplink code-division multiple-access transmission, the interference among users, depending on the correlation of specific channels and codes, may be significant. Hence, when successive interference cancellation (SIC) is performed at the base station, the minimization of power consumption at the mobile terminals is achieved if a joint optimization of power control and ordering (JOPCO) of user detection is carried out. In this paper, we first investigate a closed-form solution to JOPCO, based on the assumption of perfect interference cancellation and complete knowledge of channel correlation among users. By describing the JOPCO problem as an optimization on a suitable graph, it is seen that the optimum solution is NP-hard, and we approximate it by a greedy algorithm. For a practical implementation, we then model JOPCO as a discrete stochastic optimization problem, and we derive a discrete stochastic approximation (DSA) algorithm that iteratively adapts the transmit powers and the SIC detection order. The DSA solution removes the hypothesis on perfect cancellation and explicit knowledge of the users' correlation. Moreover, by a suitable description of the problem, we obtain a simplified DSA algorithm having reduced requirements in both computational complexity and memory. Performance of the various algorithms is assessed by simulations in a universal mobile telecommunications system scenario.