| The efficient management of radio resource is essential to the success of wireless networks. One important component of radio resource management is power control; which is a flexible mechanism to provide Quality of Service to individual users, and can be used as a vehicle for implementing on-line network operations. However, many existing distributed power control algorithms may diverge when infeasibility arises. In this work; we first deal with the infeasibility resulting from the admission of new users. Based on a system parameter called the discriminant, we propose two distributed admission control algorithms for a power-controlled system. The discriminant also provides a good measure for system robustness, and finds applications in power allocation problems of wideband cellular systems with link adaptation. We then show that a hard constraint on SIR requirement is the basic reason for the power divergence in infeasible systems. By reformulating the problem using a softened SIR requirement, we present a power control framework called utility-based power control (UBPC). Under this framework; a user will automatically decrease its target SIR (and may even turn off transmission) when it senses that traffic congestion is building up. By adjusting parameters in UBPC, we can satisfy different service requirements of users (such as delay, bit error rate, and fairness). UBPC maintains major merits of existing power control algorithms, but is exempt from the divergence problem. Moreover; it provides a unified way to deal with both voice users and data users, and demonstrates high flexibility for radio resource management. |
