Frequency domain approach to channel estimation and signal detection in HSUPA

In high speed uplink packet access (HSUPA), higher order modulation methods are one possible solution to support increased peak data rate requirements at high signal-to-interference-plus-noise ratios (SINR). Multipath propagation affects peak data rate performance. At high SINR, relatively large number of multipath components can be exploited for signal detection. It is computationally more cost-efficient to exploit such a number of diversity components in the frequency than in the time domain. Therefore, we have chosen an advanced receiver structure, in which frequency domain processing is utilized. We introduce simple finger allocation and channel estimation algorithms based on the frequency domain despreading. They are proven to offer good performance throughout the receiver stages. To combat intersymbol interference (ISI), single-user chip-level frequency domain equalizer (FDE) is used to detect the high data rate users. Parallel interference cancellation (PIC) is applied in the frequency domain to suppress the interference from data users to speech users. The performance of the whole receiver chain is presented as a function of the number of the fingers, which illustrates the demand for low complexity frequency domain processing. In particular, the simple finger allocation, where a sub-window of consecutive fingers is allocated, is an interesting option for the high SINR region.