Practical design considerations for next generation high-speed data wireless systems

Abstract

In a high-speed wireless data system, based on a bias treatment (through a scheduler control.) on good RF-condition users, the system could potentially reach high throughput if good RF users have the infinite data to send. But, in reality, most data applications have only finite sizes. With this finite-size constraint, users at poor RE conditions will occupy more timing resources than good RF users to finish the transmission of a data packet call. As a result, we will show analytically that to improve system performance, scheduler algorithms not only should focus on allocating more resources to good RF users but also need to improve low data rate users at the same time. Based on the 1xEV-DO technology, two scheduler design concepts are used to evaluate the impacts on the system throughput for both the infinite data buffer case and the finite data buffer case.