A subspace algorithm for guard interval based channel identification and source recovery requiring just two received blocks

New Blind Block Synchronization for Transceivers Using Redundant Precoders

“…The idea of repetition index is to repeatedly use each received block and has successfully reduced the number of received blocks needed for blind channel estimation problem in ZP and CP systems, as reported in [3], [4], and [10]. By properly applying this idea, we can develop blind block synchronization algorithms using less data.…”

“…Without loss of generality and for convenience of the presentation, we assume the "correct answer" is always . Furthermore, when the effective channel order is strictly smaller than the guard interval length (i.e., trailing zeros or cyclic prefixes), we observe that (3) can all be considered "correct answers" since we can think of the equivalent channel vector as in this case. No interblock interference will occur due to a timing-mismatch ,…”

“…By properly applying this idea, we can develop blind block synchronization algorithms using less data. We shall present here the application of repetition index in blind block synchronization problems and the readers interested in the blind channel estimation algorithms are referred to [3], [4], and [10].…”

“…One of the most important features in recently reported blind channel estimation algorithms is the reduction of the amount of received data needed for an accurate channel estimate which is favorable in a time-varying environment such as a wireless channel. For ZP systems, Manton et al first pointed out that blind channel estimation can be done with fewer received blocks by repeated use of each block [2], [3]. This concept was later generalized by Su and Vaidyanathan [4], [5] using a parameter called repetition index.…”

Generalized Signal Richness Preservation Problem and Vandermonde-Form Preserving Matrices

“…A theoretical treatment of the richness preservation problem has been presented in [10]. More recently, Manton et alproposed another blind identification algorithm for transmitters using ZP that imposes less stringent conditions on input signals [2], [3], requiring only the coprimality property. We propose in a companion paper [5]a generalized algorithm of which both blind identification methods mentioned above are special cases.…”