Combined interference cancellation and maximum likelihood decoding of space-time block codes
First Claim
1. A method for decoding a set of M signals received at an input interface from a plurality of terminal units that transmit on a given channel, comprising the steps of:
- processing said M signals, received over L time intervals, where L is an integer, with signals having components related to channel coefficients between transmitting points of said terminal units and an input interface, to detect signals transmitted from each of said terminal units by canceling interference from K of said terminal units, where M and K are integers such that M≧
2 and K≦
M, and to identity probable signals transmitted by said terminal units through maximum likelihood detection; and
applying signals developed through said maximum likelihood detection to a location, from which the signals may be applied to post processing that culminates in signals adapted for delivery to users.
1 Assignment
0 Petitions
Accused Products
Abstract
Block-encoded transmissions of a multi-antenna terminal unit are effectively detected in the presence of co-channel interfering transmissions from other multi-antenna terminal units, when the base station has a plurality of antennas, and interference cancellation is combined with maximum likelihood decoding. The signals received in one base station antenna are employed in processing the signals received in a second base station antenna so as to cancel the signals of one terminal unit, while decoding the signals transmitted by the other terminal unit. Zero-forcing and MMSE approaches are presented. In another embodiment of this invention the basic decoding approach is used to obtain an initial estimate for the symbols from each terminal. Assuming that the signals of the first terminal unit has been decoded correctly, the receiver employs this initial decoded signal of the first terminal unit to cancel their contribution to the signals received at the base station antennas while decoding the signals of the second terminal unit. This process is then repeated assuming that the signals of the second terminal unit has been decoded correctly, the receiver employs this initial decoded signal of the second terminal unit to cancel their contribution to the signals received at the base station antennas while decoding the signals of the first terminal unit. The above disclosed techniques are viable for any number K of terminal units concurrently transmitting over a given channel, where each terminal unit is using a space-time block code with N transmit antennas, and a base station has at least K receive antennas.
-
Citations
39 Claims
-
1. A method for decoding a set of M signals received at an input interface from a plurality of terminal units that transmit on a given channel, comprising the steps of:
-
processing said M signals, received over L time intervals, where L is an integer, with signals having components related to channel coefficients between transmitting points of said terminal units and an input interface, to detect signals transmitted from each of said terminal units by canceling interference from K of said terminal units, where M and K are integers such that M≧
2 and K≦
M, and to identity probable signals transmitted by said terminal units through maximum likelihood detection; and
applying signals developed through said maximum likelihood detection to a location, from which the signals may be applied to post processing that culminates in signals adapted for delivery to users. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22)
, where , I is the diagonal matrix, hij is a channel coefficient estimate between a transmitting antenna i of a first transmitting unit and a receive antenna j of said M receiving antennas, and gij is a channel coefficient estimate between a transmitting antenna i of a second transmitting unit and a receive antenna j of said M receiving antennas.
-
-
8. The method of claim 7 where L=2.
-
9. The method of claim 7 where said maximum likelihood detection minimizes a metric ∥
- {tilde over (r)}1−
{tilde over (H)}·
{circumflex over (c)}∥
2 over all potential code vectors {circumflex over (c)}, where {tilde over (r)}1 is one of said plurality of signals formed by said step of multiplying, and {tilde over (H)}=H1−
G1G2−
1H2.
- {tilde over (r)}1−
-
10. The method of claim 9 where said maximum likelihood detection develops an uncertainty measure.
-
11. The method of claim 10 where the uncertainty measure is given by Δ
-
c=∥
{tilde over (r)}1−
{tilde over (H)}·
{circumflex over (c)}∥
2.
-
c=∥
-
12. The method of claim 6 where said processing executes a subroutine ZF.DECODE(rx,ry,Hx,Hy,Gx,Gy) which returns an output {circumflex over (c)} by executing the following calculations
-
13. The method of claim 12 where said processing comprises executing the ZF.DECODE subroutine a first time with argument (r1,r2,H1,H2,G1,G2), and executing the ZF.DECODE subroutine a second time with argument (r2,r1,H2 ,H1,G2,G1).
-
14. The detector of claim 13 where said conditioning matrix is
-
G 2 - 1 - H 2 H 1 - 1 I ] , where H 1 = [ h 11 h 12 h 12 * - h 11 * ] , G 1 = [ g 11 g 12 g 12 * - g 11 * ] , H 2 = [ h 21 h 22 h 22 * - h 21 * ] , G = [ g 21 g 22 g 22 * - g 21 * ] , , I is the diagonal matrix, hij is a channel coefficient estimate between a transmitting antenna i of a first transmitting unit and a receive antenna j of said M receiving antennas, and gij is a channel coefficient estimate between a transmitting antenna i of a second transmitting unit and a receive antenna j of said M receiving antennas.
-
-
15. The detector of claim 14 where said processor also performs maximum likelihood detection on said plurality of signals.
-
16. The detector of claim 15, where said input interface comprises M receiving antennas.
-
17. The method of claim 12 where said subroutine ZF.DECODE also returns an uncertainty measure output Δ
-
c by carrying out the calculation Δ
c=∥
{tilde over (r)}−
{tilde over (H)}·
{circumflex over (c)}∥
2.
-
c by carrying out the calculation Δ
-
18. The method of claim 17 where said processing comprises the steps of:
-
executing the ZF.DECODE subroutine a first time, with argument (r1,r2,H1,H2,G1,G2), to obtain a first version of a probable vector signal of a first terminal unit, {circumflex over (c)}0, and a measure of uncertainty, Δ
c,0;
obtaining a first version of a probable vector signal of a second terminal unit, {circumflex over (s)}0, and a measure of uncertainty Δ
s,0;
executing the ZF.DECODE subroutine a second time, with argument (r2,r1,G2,G1,H2,H1), to obtain a second version of a probable vector signal of the second terminal unit, {circumflex over (s)}0, and a measure of uncertainty, Δ
s,1;
obtaining a second version of a probable vector signal of a first terminal unit, {circumflex over (c)}0, and a measure of uncertainty Δ
c,1;
selecting said first version of said probable vector signals if (Δ
c,0+Δ
s,0)<
(Δ
c,1+Δ
s,1), and selecting said second version of said probable vector signal based otherwise.
-
-
19. The method of claim 1 where M>
- 2 and K>
2, and where said processing comprises the steps of;combining said M received signals for form a signal that is essentially independent of transmissions of all but one of the terminal units, performing a maximum likelihood detection to identify a probably signal vector transmitted by said one of the terminal units, modifying remaining ones of said M received signals to account for contribution of the probable signals vector, and returning to said step of combining as long as there is a terminal unit whose transmission is to be detected, to select another one of said terminal units whose signal is to be detected with said step of maximum likelihood detection.
- 2 and K>
-
20. The method of claim 1 where M>
- 2 and K>
2, and where said processing comprises executing a subroutinewhere rm(a) is a received signal at antenna m, as modified in step (a);
Hk,m(a) is a matrix of channel coefficients between terminal unit k and receive antenna m, at step a; and
(Hk,m(a))+ is the generalized inverse of Hk,m(a).
- 2 and K>
-
21. The method of claim 1 where said processing form a vector of signals from said M signals and said processing includes pre-multiplying said vector by a preconditioning matrix having components related to said channel coefficients.
-
22. The method of claim 21 where said components of the conditioning matrix are related to estimates of said channel coefficients.
-
23. A detector responsive to M received signals, where M≧
- 2, from a plurality of terminal units and canceling interference from K of said terminal units, where K≦
M, comprising;a processor for transforming said M received signals through algebraic operations involving components related to channel coefficients between transmitting antennas of said terminal units and said M received signal to form a plurality of signals, to each of which is substantially unrelated to signals sent by all but one of the terminal units, where M, N, and K are positive integers, and to perform maximum likelihood detection calculations on said plurality of signals; and
a memory coupled to said processor for storing information for controlling said processor. - View Dependent Claims (24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 37, 38, 39)
forms a first version of a first of said plurality of signals, selects a first version of a probable transmitted signal vector of a first of said terminals, and evaluates a first uncertainty measure, then forms a first version of a second of said plurality of signals and in the course of said forming said first version of said second of said plurality of signals nullifies contribution of transmissions of said first version of a probable transmitted signal vector of said first of said terminals, selects a first version of a probable transmitted signal vector of a second of said terminals, and evaluates a second uncertainty measure, then forms a second version of said second of said plurality of signals, without said nullifying, selects a second version of a probable transmitted signal vector of said second of said terminals, and evaluates a third uncertainty measure, then forms a second version of said first of said plurality of signals and in the course of said forming said second version of said first of said plurality of signals nullifies contribution of transmissions of said second version of a probable transmitted signal vector of said second terminal, selects a second version of a probable transmitted signal vector of said first of said terminals, and evaluates a fourth uncertainty measure, then forms a first combination involving said first and second uncertainty measures, forms a second combination involving said third and fourth uncertainty measures, and selects either said first version or said second version of the probable transmitted signal vector of said first of said terminals and the probable transmitted signal vector of said second of said terminals based on whether said first combination is greater than said second combination.
- 2, from a plurality of terminal units and canceling interference from K of said terminal units, where K≦
-
27. The detector of claim 26 where said subroutine is executed not less than K times.
-
28. The detector of claim 23 where M=N=K=2.
-
29. The detector of claim 23 where M>
- 2, and K>
2 .
- 2, and K>
-
30. The detector of claim 23 where M>
- 2, K>
2, and N>
2.
- 2, K>
-
31. The detector of claim 23 where the processor also estimates parameters of channel through which said M signals traverse.
-
32. The detector of claim 23 further comprising a subroutine stored in said memory and executed by said processor, which subroutine performs said transforming of said M received signals.
-
33. The detector of claim 23 further comprising a subroutine stored in said memory and executed by said processor, which subroutine performs said transforming of said M received signals, followed by a maximum likelihood detection.
-
34. The detector of claim 23 further comprising means for estimating channel parameters.
-
35. The detector of claim 23 where said processor forms said plurality of signals by forming a vector from said M received signals and pre-multiplying said vector by a conditioning matrix of coefficients that are related to channel coefficients between said transmitting antennas and an input interface.
-
37. The detector of claim 23 where the processor also calculates an uncertainty measure associated with each maximum likelihood detection.
-
38. The detector of claim 27 further comprising M antennas and amplification and demodulation circuitry interposed between said M antennas and said processor.
-
39. The detector of claim 23 where said processor forms said plurality of signals one at a time, and with each formed signal of said plurality of signals the processor selects a probable transmitted signal vector of one of said terminals and evaluates an uncertainty measure.
-
36. A detector responsive to M received signals, where M≧
- 2, from a plurality of terminal units and canceling interference from K of said terminal units, where K≦
M, comprising;a first means responsive to said M received signals, for transforming said M received signals through algebraic operations involving components related to channel coefficients between transmitting antennas of said terminal units and said M received signal to form a plurality of signals, each of which is substantially unrelated to signals sent by all but one of the terminal units, where M, N, and K are positive integers;
a second means, responsive to said first means, for performing maximum likelihood detection calculations on said plurality of signals; and
means for applying information developed by said second means to a location accessible for further processing and delivery to a user.
- 2, from a plurality of terminal units and canceling interference from K of said terminal units, where K≦
Specification