Font Size: a A A

Hsdpa Downlink Channel Equalization Algorithm Research And Simulation

Posted on:2014-02-10Degree:MasterType:Thesis
Country:ChinaCandidate:J ChenFull Text:PDF
GTID:2248330398471561Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Multipath interference which caused by either low spreading gain or undesirable coding autocorrelation of HSDPA is difficult to suppress. To solve this issue, channel equalization is introduced into the system. This paper researches an equalization algorithm applicable to HSDPA system, i.e. Generalized RAKE.Introduce the key techniques, physical channels of HSDPA, and channel coding of HS-DSCH. Then based on RAKE, the basic principle and the interference suppression of G-RAKE are discussed. The finger placement algorithm makes great influence on the performance of G-RAKE. The computing complexity of combining weight is rising due to the large quantity path of traditional fixed finger placement algorithm. To solve this problem, an adaptive finger placement algorithm is proposed, which reduces computing complexity of combining weight by using less fingers, and BLER by the finger selection principle of making maximum improvement on SINR. The combination of MIMO and HSDPA is a hot spot in recent years. According to MIMO HSDPA, This paper proposed a MIMO G-RAKE model. Based on the model, realizes three estimation algorithms for equivalent channel response and interference&noise covariance matrix. The first is ideal and parameters known, named G-1, the second is based on pilot channel, named G-2, and the third is the combination of G-1and G-2, named G-3. As pre-coding can improve the performance of system, this paper realizes a kind of pre-coding based on fixed codebook, and investigates the performance of MIMO G-RAKE under this condition.Through the research and simulation above, first, verify the interference suppression of G-RAKE. Second, according to the simulation results of G-RAKE and RAKE, a conclusion that compared with RAKE, G-RAKE has a9dB gain is reached. Third, for algorithm estimating equivalent channel response and interference&noise covariance matrix, G-3loses2dB gain compared with G-1, while wins2dB gain compared with G-2. At last, the G-RAKE has at least4dB gain compared with RAKE in the presence of pre-coding.
Keywords/Search Tags:HSDPA, G-RAKE, MIMO, Pre-coding
PDF Full Text Request
Related items