High efficiency long training field symbol transmission for multiple transmission streams
摘要
說明書
Using smaller P-matrices to process subsets of HE-LTF symbols rather than using a full-sized P-matrix to process the HE-LTF symbols after receiving them all has advantages. For example, after processing group 211 and before receiving groups 212-214, the channel estimates for 4 of 16 streams are available, phase roll computation can begin, channel smoothing and/or interpolation for 4 of 16 streams can be performed, other suitable channel enhancements (e.g., noise whitening and L-LTF-HE/VHTLTF1 combining) can be performed, or any suitable combination thereof. These benefits similarly apply to each group of groups 212-214. Preamble processing 106, in some implementations, finishes processing the preamble with manageable computational complexity in time for additional processing components of receiver 105 to process the data payload of the wireless packet using the determined channel estimates.
The construction method for P16×16 described above is, in some implementations, extendable to any general P-matrix size. For example, P18×18 can be generated using the P16×16 matrix and appending a P2×2, as defined in the standard, at a diagonal position such that it occupies the space defined by the intersection of the 17th and 18th rows and columns. For spatial mapping matrix Q used by the transmitter, in some embodiments, the Q matrix does not have any of its elements as 0. This prevents power from being spread on all transmitted spatial streams during each transmitted LTF symbol. Non-zero elements in the Q-matrix map power allocation by the transmitter to all possible spatial streams usable by the transmitter, which may be a larger number of spatial streams than is used for transmission. Hence, the mapped power allocation by the transmitter is not isolated to transmitted spatial streams.
