Albert E. Manfredi wrote: > The Chinese, in two of three training sequence options, seem to use a partial repetition of the PN sequence. In one case, a 255 symbol sequence, and in the other, a 511 symbol sequence, like the one in ATSC. >"PN420 and PN945 are made up of a complete M-sequence of length 255 and length 511, respectively and with their respective cyclical extensions as pre-ambles and post-ambles, while PN595 is made up of the first 595 symbols of an M-sequence of length 1023 and it has no cyclical characteristic." >I'm not sure how this might be done, but if these extensions help, couldn't we repurpose the three unused (as far as I know) PN63 sequences now in the ATSC training segment? Would it help to create an extension to the PN511 sequence? This has been proposed in the past, but has been stymied by the legacy receiver concern. This is also why McDonald's repetitive PN1023 sequence was not considered further. After reflection I may have been too critical of the Chinese system design. But their sequence 3 suggests that they are simply using the random sequences as training signal for the Wiener method of adaptive equalization, rather than as a basis for computing channnel impulse response by a DFT method or by match filter. Al ---------------------------------------------------------------------- You can UNSUBSCRIBE from the OpenDTV list in two ways: - Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word unsubscribe in the subject line.