OFDM partial transmit sequence question

Started by October 4, 2006
Hello, maybe someone can shed a little light on this subject. I've been researching the different PAPR reduction techniques and I feel like I have a pretty good handle on the subject although I've reached a road block. I've successfully implemented the selected mapping technique in Matlab but when I implemented the partial transmit sequence technique I got poor results, only a very small PAPR reduction. Then I went through all of my programs and rechecked the scaling and everything seemed to be working out good. Ok, so I started analyzing the actual ofdm signal and how the partial transmit sequence technique would actually change it and now it doesn't make any sense to me. Supposedly in the technique you separate the data into blocks, do the IFFT first and then multiply by a phase sequence. How would this reduce the PAPR??? All this would do would be to shift the peak from say positive side to the negative side or vice versa. In the selected mapping technique the data is rotated before IFFT which makes complete sense to me.
Then I saw a simple example in a paper where they describe the phase rotations as being after the ifft but when they do the example they do the phase rotations before. What gives?