This thesis proposes a modified summing average filter, which named weighted window correlator for P1 detection. P1 detection is operated as finding a maximum value of correlation between main information and redundancy information. Correlation is the...
This thesis proposes a modified summing average filter, which named weighted window correlator for P1 detection. P1 detection is operated as finding a maximum value of correlation between main information and redundancy information. Correlation is the result of multiplying and averaging between P1 symbol redundancies. And P1 symbol is detected at maximum correlation point.
The conventional P1 detection scheme specified in DVB-T2 implementation guideline uses a threshold based peak detection algorithm. Since the plat region of correlation result make the ambiguity of deciding peak position, the offset can be decreased as removing the plat region. This method especially can make large offset in delay channel case.
The proposed weighted window of correlator make the correlation result have peak point. Moreover because the gradient of correlation result is different centering around peak position, the peak is maintained when correlation of delayed signals was added.
In addition a peak decision process is proposed in this thesis for multiple peak case. Multiple peak of correlation result can occur when large delayed signal is added. At this time, the process help find the dominant peak and peak detection algorithm is reliable.
In this thesis, the DVB-T2 system model has been implemented including the P1 symbol and the performance of synchronization using P1 symbol has been analyzed extensively by simulation in different channel conditions. P1 symbol based synchronization scheme can achieve coarse timing. Various moving sum windows to find the correlation peak of P1 are compared in this thesis.