SelectionofPSCHSequences.docVIP

3GPP 3GPP TSG-RAN WG1#48bis R1-071659 St Julian’s, Malta, March 26th – 30th, 2007 Source: Nokia Title: Selection of PSCH Sequences Agenda Item: 7.2 Document for: Discussion and decision Introduction In the RAN1#48 meeting in Sorrento (January, 2007), a working assumption of having 3 different sequence for primary SCH (P-SCH) was agreed [1]. However, the work for the sequence selection for the P-SCH and the secondary SCH (S-SCH) is still ongoing. In this contribution, we study the properties of potential P-SCH sequences which could be used as the primary SCH (P-SCH) sequences. Meanwhile, we also discuss briefly schemes for allocating sequences to P-SCH. Basic requirements for the PSCH sequences The P-SCH is mainly used for symbol timing detection [1]. Although the multiplexing schemes of SCH and the schemes to allocate the sequences are still under investigation, it was agreed that 3 different PSCH sequences will be used. Generally, the first step of the cell search process is done in the time domain, but the PSCH design can still be defined inn either time or frequency domain. Below we have listed some basic requirements that should be considered when defining the P-SCH sequences. (1) Good auto-correlation property in the time domain for SCH timing detection. This is the basic requirement for peak detection to find the symbol timing and the channel impulse response (CIR). (2) Good cross-correlation property in the time domain. First it should have low cross-correlation values to the two other P-SCH sequences. Secondly, it should have low cross-correlation values with the data. Due to the data randomization, the only way to achieve a low cross-correlation value is to increase the correlation length. (3) Simple correlation scheme(s). To reduce the computation load, it may be desired to use separate domain correlations, rather than a full complex domain correlation, or to simplify the complex domain correlation process. Th