Monitoring proportional fairness in cdma2000® high data rate networks

As 3G cdma2000^® networks with high data rates are deployed, it is interesting to study the impact of the scheduling algorithm used in these networks, and to improve its performance. We consider the setting where this scheduling algorithm is used in heterogenous channels, meaning channels for which the fading properties are not the same from one user to the next. Some users might encounter relatively steady conditions while others suffer from fast fading, for instance. In heterogenous channels, it is known that the proportional fairness algorithm falls short of its goal to provide users with a fair throughput that is proportional to their mean rate. We offer an easy way to implement improvements to this shortcoming. The idea for these improvements is to add an independent performance monitoring function to the proportional fairness algorithm that corrects unfairness should it appear.