Creating training sequences for space-time diversity arrangements
View Patent ↗A training sequence is created for space-time diversity arrangement, having any training sequence length, while limiting the training sequence to a standard constellation. Given a number of channel unknowns that need to be estimated, L, a training sequence can be creates that yields minimum means squared estimation error for lengths N t =kN PRUS +L−1, for any positive integer k≧1, where N PRUS is a selected perfect roots-of-unity sequence (PRUS) of length N. The training sequence is created by concatenating k of the length N perfect roots-of-unity sequences, followed by L−1 initial symbols of that same PRUS. Good training sequences can be created for lengths N t that cannot be obtained through the above method by concatenating a requisite number of symbols found through an exhaustive search.
1. A space-time diversity transmitter that includes (a) n transmitting antennas, where n is greater than one, (b) an encoder of an applied information stream, said encoder developing n symbol streams, and (c) a constellation mapper responsive to said n symbol streams, for mapping symbols of each of said n symbol streams into a standard signal constellation to create a mapped stream, said constellation mapper applying each of the created n mapped streams to a different one of said n transmitting antennas, the improvement comprising:
a generator of a symbols sequence, s;
an encoder for creating n training sequences from said symbols sequence, where each of the created n training sequences has an impulse-like autocorrelation function and zero cross correlation relative to the other of the n created training sequences; and
a training sequence mapper that maps said n training sequences onto a standard constellation to develop n mapped training sequences, and applies the n mapped training sequences to said n antennas wherein said n training sequences are of length N t ≧2L−1, and N t =kN PRUS +L−1, where L is an integer corresponding to a number of channel parameter unknowns that need to be estimated through which channel said transmitter communicates with a receiver, k is an integer greater than or equal to 1, and N PRUS is a length of a perfect roots-of-unity sequence of length N.
2. A method for developing a training sequence of length, N t , for estimating L channel parameter unknowns that need to be estimated comprising the steps of:
selecting a Perfect Roots-of-Unity Sequences (PRUS) of length N PRUS equal to or greater than L, and
constructing said training sequence by forming a sequence of length N t =kN PRUS +L−1 where k is an integer that is equal to or greater than 1.
3. The method of claim 2 where said step of constructing concatenates k instances of said selected PRUS of length N PRUS forming a subsequence, and forming said training sequence by appending to said subsequence L−1 initial members of said PRUS of length N PRUS .
4. The method of claim 2 where said step of constructing concatenates k instances of said selected PRUS of length N PRUS forming a subsequence, and forming said training sequence by appending to said subsequence L−1 terminal members of said PRUS of length N PRUS to said subsequence.
5. A method for developing a training sequence of length N t for estimating L channel unknowns comprising the steps of:
selecting a Perfect Roots-of-Unity Sequence (PRUS) of length N PRUS equal to or greater than L,
selecting an integer k≧1 such that integer M in equation N t =kN PRUS +L−1+M is equal to, or greater than, 0; and
forming said training sequence from a combining a first subsequence composed of k instances of said PRUS of length N PRUS , a second subsequence composed of L−1 initial, or terminal members of said PRUS of length N PRUS , and a third subsequence composed of M symbols.
6. The method of claim 5 where said PRUS is selected to minimize M.
7. The method of claim 5 where said M symbols are chosen based on a search that minimizes a preselected error criterion.
8. The method of claim 5 where said combining of said first subsequence and said second subsequence, when said second subsequences is composed of L−1 initial members of said PRUS of length N PRUS , comprises said first subsequence followed by said second subsequence.
9. The method of claim 5 where said combining of said first subsequence and said second subsequence, when said second subsequences is composed of L−1 terminal members of said PRUS of length N PRUS , comprises said second subsequence followed by said first subsequence.