Space-time coding digital transmission systems and methods

• US 7,477,696 B2
• Filed: 07/03/2001
• Issued: 01/13/2009
• Est. Priority Date: 07/04/2000
• Status: Expired due to Fees

• Current Assignee: Thales SA
• Original Assignee: Thales SA
• Inventors: Ferreol, Anne, Balp, Hugues
• Primary Examiner(s): Tran; Khanh C
• Application Number:
  US10/312,179
• Publication Number:
  US 20040042560A1
• Time in Patent Office: 2,751 Days
• Field of Search: 375260-261, 375265-267, 375/295, 375298-299, 375/316, 375/350, 455/101, 370/329, 370/345, 370/349
• US Class Current: 375/267
• CPC Class Codes: H04B 7/0669 using different channel coding between antennas space-time coding H04L1/0618
• CPC Class Codes: H04B 7/0848 Joint weighting
• CPC Class Codes: H04L 1/0618 Space-time coding

Abstract

The invention concerns digital signal transmission. In particular, it concerns high speed transmission using layered space-time encoding architecture adapted to all types of propagation channels.

The invention therefore proposes a digital signal transmission system comprising:

• • a space-time encoder (1) receiving a flow of data to be transmitted d[i], formatting this data d[i] as symbol vectors v[k] of dimension P(P>1) and generating said symbol vectors v[k], and
  • modulator-transmitters {2^p}_(1≦p≦P), each receiving one component of the symbol vector m[k] output from the space-time encoder (1), applying the constellation of a predetermined modulation to said symbol m_p[k], and converting the symbol obtained a_p[k] into a signal s_p(t) presenting time diversity transmitted on said antenna (24^p) connected to said transmitter (2^p).

To demodulate in parallel the Q signals of the space-time observation

$\underset{\_}{y}\left[k\right]=\sum _{j=0}^{J-1}{H}_y\left(\mathrm{jTs}\right)·\underset{\_}{a}\left[k-j\right]+{\underset{\_}{b}}_y\left[k\right]$
where ā[k] is the symbol vector transmitted at instant t=kTs+i, H_y(t) the transfer function taking into account at least the transmission-reception, modulation, channel filters and the transmission-reception antenna gains and b_y(t) the noise, the invention proposes a two dimensional suitable estimator-demodulator.

Claims

1. A digital signal transmission system comprising:

a space-time encoder receiving a flow of data to be transmitted d[i], encoding this data d[i] as symbol vectors m[k] of dimension P(P>

1) and generating said symbol vectors m[k], and

P modulator-transmitters {2^P}<sub>(1≦

p≦

P)</sub>, each receiving one component m_p[k] of the symbol vector m[k] output from the space-time encoder, applying the constellation of a predetermined modulation to said symbol m_p[k] to obtain symbol a_p[k], and converting the symbol obtained, a_p[k], into a signal s_p(t) transmitted on said antenna (24^p) connected to said transmitter (2^p)

wherein the transmitters are adapted to transmit signals s(t) with time diversity,

wherein P modulator-transmitters {2_p};

each produce said symbol a_p[k] in parallel at instant k,

each form a filter of function h_p(t) comprising a delay element τ

_ρ

with τ

₁≠

τ

₂≠

. . . τ

_ρ

, such that h_p(t)=h_p(t−

τ

₉₂ ) for all values of p, such that the function h_p(t) of the transmitter (2^p) is different from those of the other transmitters {2^q}<sub>(q≠

p)</sub>;

h₁(t)≠

h₂(t)≠

. . . ≠

h_p[(t),

each generate at their respective transmission antennas the signal s_p[k] corresponding at least to the filtering by the function h_p(t) of the symbols a_p[k]; and

wherein the digital signal transmission system further comprises a wave form h_p, wherein h₁≠

h₂≠

. . . ≠

h_P for all values of p wherein (1≦

p≦

P).

2. The transmission system according to claim 1, wherein the function h_p(t) has one or more of the following features:

any wave form h not identical to the function h_q(t) of the transmitter (2^q, q≠

p) and a delay τ

_p delaying the transmission of the received symbol a_p[k] by said duration τ

_p, such that the function h_p(t)=h(t−

τ

_p) with τ

₁≠

τ

₂≠

. . . ≠

τ

_p for all values of p(1≦

p≦

P),

any wave form h not identical to the function h_q(t) of the transmitter (2^q) (q≠

p) and a frequency shift f_p such that the function of the filter h_p(t)=h·

exp(j2π

f_pt) with f₁≠

f₂≠

. . . ≠

f_p for all values of p(1≦

p≦

P),

a wave form h_p is one of type NRZ, Nyquist with roll-off α

or α

_p.

3. The transmission system according to claim 1:

wherein said space-time encoder comprises at least a demultiplexer with P channels generating a symbol vector m[k] AND or OR at least one or more of the following devices;

an encoder generating a symbol vector c[k],

a device used to add at least an apprenticeship sequence app known by the receiver to the useful symbol vectors m[k] or encoded symbol vectors c[k] in order to create the symbol vectors v[k],

AND or OR wherein each modulator-transmitter comprises one or more of the following devices;

a linear or linearizable modulator,

a modulator with or without time memory,

a BPSK or GMSK modulator,

a device applying the constellation of said predetermined modulation to the received symbols v_p[k] generating the symbols a_p[k],

a device used to add at least the p^th component of an apprenticeship sequence app known by the receiver to the symbols a_p[k],

a filter filtering the constellated symbols a_p[k],

a filter filtering the vector of oversampled symbols

$\begin{array}{cc}{\underset{\_}{u}}_p^K<br><br>\left(t=\mathrm{kTs}+i\right)=\left[\begin{array}{c}{u}_p<br><br>\left(t\right)\\ \mathrm{⋮<br><br>}\\ u_p<br><br>\left(t-K+1\right)\end{array}\right]=\left[\begin{array}{c}{\underset{\_}{0}}_i\\ a_p<br><br>\left[k\right]\\ {\underset{\_}{0}}_{\mathrm{Ts}-1}\\ a_p<br><br>\left[k-1\right]\\ \mathrm{⋮<br><br>}\end{array}\right]& \mathrm{for}<br><br><br><br>0\le <br><br>i<<br><br>\mathrm{Ts}\end{array}$ $\mathrm{with}<br><br><br><br>{\underset{\_}{0}}_i={\left[\begin{array}{ccc}0& \mathrm{\cdots <br><br>}& 0\end{array}\right]}^T<br><br><br><br>\mathrm{and}<br><br><br><br>\dim <br><br>\left({\underset{\_}{0}}_i\right)=i\times <br><br>1$

an element to modulate the signal to be transmitted on the carrier frequency f₀.

4. An estimator-demodulator receiving in parallel N signals y(t) formed from L samples resulting from the transmission of digital signals by a transmission system according to claim 1, wherein these signals y(t) represent a space-time observation since each of the N spatial components comprises L samples:

wherein the estimator-demodulator comprises at least the following devices;

a first recursive two dimensional filter Ĥ

_y^Q receiving an estimation of coefficients and the symbols already detected ā

(k−

Q) . . . ā

(k−

J+1),

an adder used to remove the vector which is the result of the first filter applied to the received observations y(t) and obtain v_q(t),

a second transverse two dimensional filter Ŵ

_Q(t) receiving y_Q(t) and an estimation of coefficients and generating estimated received symbols â

_Q[k],

a detector of the modulation states of the estimated symbols {circumflex over (ā

[k]=Ŵ

_q y(t) generating the detected symbols {tilde over (ā

[k], then

a demodulator generating the symbols ^v[k].

an estimator of the coefficients Ŵ

_Q of the transverse filter and Ĥ

_y^Q of the recursive filter.

5. The estimator-demodulator of the signals transmitted by a transmitter according to claim 4, wherein the estimator-demodulator comprises at least a two dimensional estimator-demodulator of the symbols {a[k] . . . a[k−

J+1]} using a Viterbi algorithm.

6. A digital signal reception system comprising:

a receiver comprising at least a network of N reception antennas and an estimator-demodulator according to claim 4, and

a space-time decoder wherein

said receiver comprises at least;

N reception devices {31ⁿ}<sub>(1≦

n≦

N)</sub>, comprising at least an element used to put the received signal in the baseband, generating an observation vector x(t) of dimension N,

a windower producing from observations x(t) the discrete observations x[kTs+i] with t=kTs+i and 0≦

i≦

Ts given that the observations x[kTs+i] depend on the transmitted signal vectors a[k] to a[k−

J+1] and generating a space-time observation

$\underset{\_}{y}<br><br>\left(t\right)=\left[\begin{array}{c}\underset{\_}{x}<br><br>\left(t\right)\\ \underset{\_}{x}<br><br>\left(t-1\right)\\ \mathrm{⋮<br><br>}\\ \underset{\_}{x}<br><br>\left(t+L-1\right)\end{array}\right]$

from N observations x(t),

AND or OR wherein the space-time decoder comprises at least one or more of the following devices;

an element capable of removing the apprenticeship sequences app,

a decoder with P channels in input/output,

a multiplexer with P channels in input; and

a wave form h_p, where in h₁≠

h₂≠

. . . ≠

h_p for all values of p wherein (1≦

p≦

P).

7. A digital signal reception system according to claim 6, wherein the number of transmission antennas P of said transmission system is greater than or equal to the number of paths M(P≧

M).

8. A digital signal reception system according to claim 6, wherein the number of transmission antennas P of said transmission system is less than or equal to the number of paths M (P≧

M).

9. An estimator-demodulator receiving in parallel N signals y(t) formed from L samples resulting from the transmission of digital signals by a transmission system according to claim 1, wherein these signals y(t) represent a space-time observation since each of the N spatial components comprises L samples, wherein the estimator-demodulator comprises at least the following devices:

a two dimensional Wiener filter estimator,

an estimated two dimensional filter Ŵ

_Q receiving the observations y(t) and the estimation of the coefficients of said filter Ŵ

_Q,

a detector of the modulation states of the estimated symbols {circumflex over (ā

[k]=Ŵ

_Q y(t) generating the symbols detected ā

[k], then

a demodulator generating the symbols ^v[k]; and

a wave form h_p, wherein h₁≠

h₂≠

. . . ≠

h_P for all values of p wherein (1≦

p≦

P).

10. The estimator-demodulator according to claim 9, wherein the estimator-demodulator further comprising the filter estimator is an estimator either in the sense of maximum resemblance, the sense of least squares, or using a Viterbi algorithm,

the demodulator corresponds to a modulation of one or more of the following types:

linear, linearizable, with time memory, without time memory, BPSK, GMSK,

at least some of the received signals v(t) represent an apprenticeship sequences app known by said estimator-demodulator allowing one or more of the following operations;

estimation of said filter,

estimation in the sense of the least squares of a recursive filter such that coefficients Ĥ

_y^Q are the first Q columns of the matrix

${\hat{H}}_y^Q=R_{y·<br><br>{\underset{\_}{\mathrm{app}}}_y}<br><br>R_{{\underset{\_}{\mathrm{app}}}_y·<br><br>{\underset{\_}{\mathrm{app}}}_y}^{-1},$

estimation with the zero-forcing method of a transverse filter such that coefficients are

${\hat{W}}_Q=R_{{\underset{\_}{\mathrm{app}}}_y·<br><br>y}<br><br>R_{y·<br><br>y}^{-1},$

estimation using the Wiener method in the sense of maximum resemblance of a filter such that coefficients are

${\hat{W}}_Q=R_{{\underset{\_}{\mathrm{app}}}_y·<br><br>y}<br><br>R_{y·<br><br>y}^{-1},$

initialization of the estimation of filter(s)

initialization of filtering,

initialization of the Viterbi algorithm of the estimator-demodulator.

11. The digital signal transmission system comprising transmission system according to claim 1, and a reception system comprising an estimator-demodulator receiving in parallel N signals y(t) formed from L samples, wherein these signals y(t) represent space-time observation since each of the N spatial components comprises L samples, comprising, in addition, at least a transmission channel such that a signal s_p(t) transmitted by said transmission system takes M separate paths (M≧

1) in said transmission channel before reaching said reception system.

12. The estimator-demodulator of signals transmitted by a transmitter according to claim 11, wherein the estimator-demodulator comprises at least the following devices:

a two dimensional Wiener filter estimator,

an estimated two dimensional filter Ŵ

_Q receiving the observations y(t) and the estimation of the coefficients of said filter Ŵ

_Q,

a detector of the modulation states of the estimated symbols {circumflex over (ā

[k]=Ŵ

_Q y(t) generating the symbols detected ā

[k], then

a demodulator generating the symbols ^v[k],

13. The estimator-demodulator of the signals transmitted by a transmitter according to claim 11, wherein the estimator-demodulator comprises at least the following devices:

a first recursive two dimensional filter Ĥ

_y^Q receiving said estimation of coefficients and the symbols already detected ā

(k−

Q) . . . (k−

J+1),

an adder used to remove the vector which is the result of the first filter applied to the received observations y(t) and obtain v_q(t),

a second transverse two dimensional filter Ŵ

_Q receiving y_Q(t) and said estimation of coefficients and generating the estimated received symbols â

_Q[k],

a detector of the modulation states of the estimated symbols {circumflex over (ā

[k]=Ŵ

_q y(t) generating the detected symbols {tilde over (ā

[k], then

a demodulator generating the symbols ^v[k],

an estimator of the coefficients Ŵ

_Q of the transverse filter and Ĥ

_y^Q of the recursive filter.

14. The estimator-demodulator of the signals transmitted by a transmitter according to claim 11, wherein the estimator-demodulator comprises at least a two dimensional estimator-demodulator of the symbols {a[k] . . . a[k−

J+1]} using a Viterbi algorithm.

15. The estimator-demodulator according to claim 11, wherein the estimator-demodulator comprises:

the filter estimator is an estimator either in the sense of maximum resemblance, the sense of least squares, or using a Viterbi algorithm,

the demodulator corresponds to a modulation of one or more of the following types;

linear, linearizable, with time memory, without time memory, BPSK, GMSK,

at least some of the received signals v(t) represent an apprenticeship sequences app known by said estimator-demodulator allowing one or more of the following operations;

estimation of said filter,

estimation in the sense of the least squares of a recursive filter such that its coefficients Ĥ

_y^Q are the first Q columns of the matrix

${\hat{H}}_y^Q=R_{y·<br><br>{\underset{\_}{\mathrm{app}}}_y}<br><br>R_{{\underset{\_}{\mathrm{app}}}_y·<br><br>{\underset{\_}{\mathrm{app}}}_y}^{-1},$

estimation with the zero-forcing method of a transverse filter such that coefficients are

${\hat{W}}_Q=R_{{\underset{\_}{\mathrm{app}}}_y·<br><br>y}<br><br>R_{y·<br><br>y}^{-1},$

estimation using the Wiener method in the sense of maximum resemblance of a filter such that coefficients are

${\hat{W}}_Q=R_{{\underset{\_}{\mathrm{app}}}_y·<br><br>y}<br><br>R_{y·<br><br>y}^{-1},$

initialization of the estimation of filter(s),

initialization of filtering,

initialization of the Viterbi algorithm of the estimator-demodulator.

16. A digital signal transmission method comprising the steps of:

a space-time encoding step comprising at least the encoding of the flow of data to be transmitted d[i] as symbol vectors m[k] of dimension P)P>

1), and

a modulation-transmission step comprising;

application in parallel of the constellation of a predetermined modulation to the P symbols m[k] to obtain constellated symbols a_p[k],

transmission in parallel of the P signals s(t) obtained from the constellated symbols a[k] from P spatially separate points,

wherein the modulation-transmission step is adapted to transmit the signals s(t) with time diversity,

wherein P modulator-transmitters {2^p};

each produce a symbol a_p[k] in parallel at instant k,

each form a filter of function h_p(t) comprising a delay element τ

_ρ

with τ

₁≠

τ

₂≠

. . . τ

_ρ

, such that h_p(t)=h_p(t−

τ

_ρ

) for all values of p, such that the function h_p(t) of the transmitter (2^p) is different from those of the other transmitters {2^q}<sub>(q≠

p)</sub>;

;

h₁(t)≠

h₂(t)≠

. . . ≠

h_P[(t),

each generate at their respective transmission antennas the signal s_p[k] corresponding at least to the filtering by the function h_p(t) of the symbols a_p[k]; and

each generates a waveform h_p, wherein h₁≠

h₂≠

. . . ≠

h_P for all values of p wherein (1≦

p≦

P).

17. The transmission method according to claim 16, wherein the modulation-transmission step comprises, in addition, at least, on each channel p(1≦

p≦

P), filtering of symbols a_p[k] generating a signal s_p(t) such that the filtering of channel p is different from carried out on the other P−

1 parallel channels.

18. The transmission method according to claim 17, wherein said filtering carried out on channel p has one or more of the following features:

any wave form h identical or not to the channel q, for all values of p and q(1≦

q≠

p≦

P) and a delay of duration τ

_p different from channel q,

any wave form h identical or not to the channel q, for all values of p and q(1≦

q≠

p≦

P) and a frequency shift f_p different from channel q, a form h_p different from channel q, for all values of p and q(1≦

q≠

p≦

P); and

a wave form h_p, wherein h₁≠

h₂≠

. . . ≠

h_P for all values of p wherein (1≦

p≦

P).