Performance and Capacity of HSUPA in Lab Environment

Abstract

This paper analyzes the performance of HSUPA based on test results from a UMTS lab trial network. Different aspects from the technology have been studied in a controlled lab environment, including latency reduction by HSUPA, power control efficiency and capacity. HSUPA (high speed uplink packet access) is the next logical step following the deployment of HSDPA to offer a truly 3G high speed data service with higher throughputs and reduced latency. Despite the similarities with HSDPA (such as HARQ, short TTI, NodeB scheduling), HSUPA is fundamentally different because of the use of a dedicated channel per user, called E-DCH. This results in a very different treatment of radio resources management scheme, with UL power control instead of link adaptation, or a more challenging scheduling system. The traffic scheduling is rather random in HSUPA since all data users are not synchronized with each other, and therefore a common pipe cannot be established by the network for all users as it does for the downlink. The only common resource the system can control is the uplink noise rise. In a typical UMTS network, under interference limited condition, downlink is the bottleneck. However, with the introduction of E-DCH, it is important to understand its impact to the uplink performance and capacity, especially under conditions with mixed voice and data services on one carrier. The results from our study shows a good data performance of HSUPA in data only scenarios, however there is a significant degradation on the voice service when they share the same carrier.