Research On Algorithms Advanced Receiver Of HSPA

WCDMA technology is studied maturely, and has been the most widely used technology, and most of the network has been updated to HSPA(High-Speed Packet Access). The design of HSPA senior receiver is the very important module of HSPA technology. If the channel is frequency-flat, the traditional RAKE receiver is optimal. However, in the high-speed wireless communication link, the multipath delay spread can cause interference between signals increase, which leads the channel to become a frequency-selective fading channel. The existence of the multi-path destroys the orthogonality of the spreading codes, that causes the performance of the RAKE receiver seriously degrades because of the influence of IPI(Interpath Interference), makes RAKE receiver can’t meet the requirements of the system.GRAKE receiver can effectively remove the interference of the multi-path and multi-user, providing the guarantee for high-speed packet business of HSPA. It converts the colored noise to white noise at the receiving end by increasing the noise path and different combining coefficient on the basis not changing the RAKE receiver structure. At the same time, the chip-level equalizer is proved to be an effective method, it can restore the orthogonality of spreading codes, suppress multiple access interference, improve the performance of the system.In the first place, this paper focuses on the HSPA key technologies, MIMO technology and Precoding algorithm. Secondly, we introduce the channel estimation and multi-path detection in the receiver. Then we introduce the basic principle of RAKE receiver, MRC and the basic principle of GRAKE receiver, the selective of noise path and the principle of chip-level equalizer, the block iteration chip-level equalizer algorithm. What’s more, we have a simulation for the performance of the receiver. We can see from the simulation results that the performance of GRAJKE receiver is much better than RAKE receiver, and the performance of chip-level equalizer is obviously superior to RAKE receiver. When the number of users is big, the equalizer can well restrain MAI, while RAKE receiver will counter the "platform" effect because of the influence of the autocorrelation not ideal.The innovation of this paper has three points:a new type of Block Iterator LMMSE equalizer Algorithm is presented, because the channel is sparse channel, so use the gauss-seidel iterative method to calculate matrix inversion, the complexity is significantly lower than the gauss elimination method of LMMSE equalizer. From the simulation result, its performance is superior to RAKE receiver, but similar with LMMSE equalizer algorithm.