Bi-directional channel analysis

Description

FIELD OF THE INVENTION

The present invention relates to methods for signal detection in general and to methods for signal detection using known preambles in particular.

BACKGROUND OF THE INVENTION

Methods for detecting a signal and evaluating a channel using a known preamble, are known in the art. Such methods are used in timed framed communication methods such as TDMA communication standard, CDMA communication standard, and the like.

Reference is made to

FIG. 1

which is a schematic illustration of a timed framed TDMA sequence, referenced

1

, known in the art. Frame sequence

1

includes a plurality of slots

2

,

4

,

6

,

8

and

10

. Each of these slots includes bits of data, lasting a predetermined period of time.

In TDMA, several users may use a predetermined communication channel, each at predetermined intervals in time. In the present example, slot

2

is assigned to a first user, slot

4

is assigned to a second user, and so on.

According to common TDMA standards, such as IS-136, IS-54, RCR-27, the preamble (i.e. the first few bits of information) of each slot represents a known synchronization sequence, also known as a sync-word. Each of slots

2

,

4

,

6

,

8

and

10

has a sync-word

12

,

14

,

16

,

18

and

20

, respectively.

Sync-words are used to determine what the channel looks like with regard to several aspects such as amplitude, phase, timing, frequency offset, reflections which produce echoes, represented by channel taps, distortions, interference and the like.

Multiple channel taps cause inter-symbol interference. For example, the value of a sample si sampled at the output of the multi-path channel, is given by

s_{i} = \sum_{n = - L_{1}}^{L_{2}} h_{n} a_{i - n}

wherein the a

i

are the transmitted samples (−∞<i<∞); and

h

n

are the gains of the channel taps (−L

1

≦n≦L

2

). L

1

and L

2

define the length of channel memory which, in turn, defines the number of neighboring symbols which affect each sample.

Conventional methods use the sync-word of each slot to determine the appropriate synchronization to this slot. Thus, the receiver of the first user, receiving slot

2

, will analyze sync-word

12

(S

1

), determine the channel characteristics and proceed analyzing the rest of slot

2

, accordingly.

It will be appreciated that a communication channel may develop in time, which affects the form of a received slot. Thus, an analysis, based on a sync-word at the beginning of a slot, might not be accurate for the last part of the slot.

Accordingly, conventional receivers execute dynamic tracking procedures, which track the channel development in time, while decoding the data contained therein. Temporary fading of a channel may cause considerable degradation in such channel tracking procedures.

Another method known in the art analyzes the channel of a received slot according to the sync-word of the slot and the sync-word of the next adjacent slot. For example, the receiver of the first user, receives slot

2

as well as the sync-word

14

of slot

4

and analyzes the channel from sync-word

12

(S

1

) and

14

(S

2

).

Analyzing the channel for slot

2

using sync-word

12

(S

1

), combined with forward in time channel tracking, is also called forward analysis. Analyzing the channel for slot

2

using sync-word

14

(S

2

), combined with backward in time channel tracking, is also called backwards analysis. According to prior art methods, this can be performed only when S

1

and S

2

are known, fixed sequences.

It will be appreciated that in some communication standards, such as TDMA IS-54, IS-136, RCR-27 and the like, the sync-word of the next slot is not a fixed predetermined sequence, which make the above prior art methods inefficient.

These TDMA standards define a frame which includes six slots. A TDMA standard also defines two modes of channel usage. A first mode is called half-rate, in which a user is assigned one slot (i.e. every sixth frame is reserved for his use). A second mode is called full-rate, in which a user is assigned two slots in a frame (i.e., every third slot).

Reference is now made to

FIG. 2

which is a schematic illustration of a combined full-rate half-rate frame TDMA sequence, generally referenced

48

. Frame

48

includes a plurality of data slots,

50

,

52

,

54

,

56

,

58

,

60

and

62

, each having a preamble

70

,

72

,

74

,

76

,

78

,

80

and

82

, respectively.

The first six slots define a frame, operative for a number of users, which will be repeated further sync-words. This frame is assigned to two half-rate users and two full-rate users.

Slots

54

and

60

are assigned to the first and second half-rate users, respectively. Slots

50

and

56

are assigned to the first full-rate user and slots

52

and

58

are assigned to the second full-rate user.

In the present example, preambles

70

and

76

of slots

50

and

56

which are both assigned to the first full-rate user, have the same sync-word S

1

. Preambles

72

and

78

of slots

52

and

58

which are both assigned to the second full-rate user, have the same sync-word S

2

. Preambles

74

and

80

of slots

54

and

60

, each assigned to a different half-rate user, have different sync-words S

3

and S

6

, respectively.

Accordingly, S

1

is followed by S

2

, while S

2

is followed by either S

3

or S

6

.

Yow-Jong Liu, “Bi-Directional Equalization Technique for TDMA Communication systems over Land Mobile Radio Channels”,

GLOBECOM

1991 IEEE, p 1458-1462, describes a method for using the sync-word of the next slot which decodes the data of a slot according to its sync-word (forward analysis) and according to the sync-word of the next slot (backward analysis). When the sync-word of the next slot may have more than one option, then the data of the current slot is backward analyzed according to each of these options, thereby providing a plurality of data hypotheses. Finally, the most likely hypothesis is selected.

It will be appreciated that an analysis according to this method requires a considerable amount of computing resources and power, which can be critical for mobile communication devices having limited power resources.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be understood and appreciated more fully from the following detailed description taken in conjunction with the drawings in which:

FIG. 1

is a schematic illustration of a TDMA frame, known in the art;

FIG. 2

is a schematic illustration of a combined full-rate half-rate TDMA frame, known in the art;

FIG. 3

is a schematic illustration of a system for analyzing a TDMA signal slot according to multiple sync-words, constructed and operative in accordance with a preferred embodiment of the present invention;

FIG. 4

is a schematic illustration of a method for operating the system of

FIG. 3

, so as to analyze and select between the two postulates available for the next slot sync-word, operative in accordance with another embodiment of the present invention; and

FIG. 5

is a schematic illustration in detail of a portion of the method of FIG.

4

.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

The present invention provides a method which overcomes the disadvantages of the prior art by determining each next slot sync-word as a postulate, analyzing each of these postulates, selecting the most likely one and then analyzing the data of the current slot forward according to its sync-word, and backward according to the selected next slot sync-word.

References are made to

FIG. 3

, which is a schematic illustration of a system for analyzing a TDMA signal slot according to multiple sync-words, generally referenced

100

, constructed and operative in accordance with an embodiment of the present invention.

System

100

includes a front end receiver unit

102

, a current slot sync-word analysis unit

104

, a current slot data buffer

106

, a next slot first sync-word analysis unit

108

and a next slot second sync-word analysis unit

110

, all of which are connected to the front end receiver unit

102

.

System

100

optionally includes additional next slot N sync-word analysis units represented by a next slot last sync-word analysis unit

109

and dotted lines connecting it to the next slot second sync-word analysis unit

110

, the decision controller

112

and the front end receiver

102

.

System

100

further includes a decision controller

112

connected to units

108

and

110

, a forward analysis unit

114

connected to units

104

and

106

, a backward analysis unit connected

116

to buffer

106

and the decision controller

112

and a decoder

118

connected to the forward analysis unit

114

and to the backward analysis unit

116

.

The front end receiver unit

102

receives and samples a TDMA slot as well as the sync-word of the next slot. The front end receiver unit

102

then provides the sync-word of the current slot to the current slot sync-word analysis unit

104

, the data of the current slot to the current slot data buffer

106

and the sync-word of the next slot to the next slot sync-word analysis units

108

and

110

.

The current slot sync-word analysis unit

104

analyzes the sync-word so as to determine the appropriate channel according to which the data stored in the current slot data buffer

106

is to be forward analyzed.

The next slot sync-word analysis unit

108

analyzes the sync-word of the next slot according to a first sync-word postulate, which is optional for the next slot sync-word, thereby producing a respective likelihood value. The next slot second sync-word unit analysis

110

analyzes the sync-word of the next slot according to a second postulate, which is also optional for the next slot sync-word, thereby producing a respective likelihood value. Each of units

108

and

110

also produces a channel analysis according to each of the postulates, respectively.

Units

108

and

110

then provide the results of their analysis to the decision controller

112

, which in turn, determines and selects the postulates which has a better likelihood value.

Then, the decision controller

112

provides the selected sync-word postulate to the backward analysis unit

116

, as the next slot sync-word. At the same time, the current slot data buffer

106

provides the data to the backward analysis unit

116

. Then, the backward analysis unit

116

analyzes the data according to the selected sync-word of the next slot.

At the same time the current slots sync-word analysis unit

104

provides an analysis of the channel according to the current sync-word and provides it to the forward analysis unit

114

. The current slot data buffer

106

also provides the current slot data to the forward analysis unit

114

.

The forward analysis unit

114

analyzes the current slot data according to the channel analysis provided by the current slot sync-word analysis unit

104

.

The forward analysis unit

114

and the backward analysis unit

116

provide the analysis of the data to the decoder

118

which, in turn, decodes the data according to forward and backward data analysis.

Reference is now made to

FIG. 4

which is a schematic illustration of a method for operating system

100

, so as to analyze and select between the two postulates available for the next slot sync-word, operative in accordance with another embodiment of the present invention.

In step

200

, units

108

and

110

calculate an estimated taps value for each of the sync-word postulates.

In step

202

, units

108

and

110

calculate a log likelihood metric for each of the next slot sync-word postulates.

In step

204

, the decision controller

112

selects the sync-word postulate with the best metric and proceeds to step

206

. An example for such a log likelihood calculation is given by the expression:

C (\underline{y}, \underline{h}) = \sum_{m = K}^{N_{s}} {&LeftBracketingBar; y_{m} - \sum_{k = 1}^{K} h_{k} \cdot s_{m - k} &RightBracketingBar;}^{2} = \sum_{m = K}^{N_{s}} {&LeftBracketingBar; y_{m} - {\underline{h}}_{m} \cdot {\underline{s}}_{m} &RightBracketingBar;}^{2}

wherein y represents the signal samples, h represents the channel taps, which relate to channel reflections, s represents the sync-word postulates and m is a time index. This expression represents the Euclidean distance between the received samples and the samples which were estimated, assuming a determined next frame sync-word.

The expression

\sum_{k = 1}^{K} h_{k} \cdot s_{m - k}

represents, generally, a convolution of the channel and the sync-word.

h can be calculated according to a least squares estimation:

{\underline{h}}_{opt} = {(\sum_{l = K}^{N_{s}} {\underline{s}}_{l}^{*} {\underline{s}}_{l}^{t})}^{- 1} \sum_{l = K}^{N_{s}} {\underline{s}}_{l}^{*} y_{l}

or a simple correlation:

{\underline{h}}_{opt} = \frac{1}{N_{s} - K} \sum_{l = K}^{N_{s}} {\underline{s}}_{l}^{*} y_{l}

wherein N

s

is the number of symbols in the sync-word and K is the assumed number of channel taps.

In step

206

, the decision controller

112

further calculates a reliability value for the selected next slot sync-word. This step will be further discussed in detail hereinbelow.

In step

208

, the decision controller

112

determines if the decision of selecting the selected next slot sync-word is to be regarded as reliable. If so, the decision controller

112

proceeds to step

212

where it provides the backward analysis unit

116

with a selected next slot sync-word. Otherwise, in step

210

the decision controller

112

provides a command to the backward analysis unit

116

to disregard the second sync-word and not perform any backward analysis for this slot.

Disregarding of the next slot sync-word may occur in a situation where this next slot sync-word is subject to deep fading of the channel. Thus the data of this next slot sync-word may be corrupt and therefore invalid.

Reference is now made to

FIG. 5

which is a schematic illustration in detail of step

206

of the method of FIG.

4

.

In step

220

, the decision controller

112

detects if the likelihood value of the selected sync-word is substantially greater than the likelihood value of the non-selected sync-word. If the likelihood values are too close, than one is not likely to be preferable over the other. In this case, the likelihood values are often a result of noise rather than distinctive values. In such a case, the decision controller

112

determines the selected sync-word non-reliable (step

226

). Otherwise, the decision controller

112

proceeds to step

222

.

In step

222

, the decision controller

112

detects if the selected next slot sync-word is logically correct, according to historical information stored therein and side information which originates from the standard. For example, referring back to

FIG. 2

, if the slot

70

(the current slot) is assigned to a first user (s

1

), then the next slot

72

can be assigned to the second user (s

2

) or the fifth user (s

5

). If the next slot is assigned to the fifth user (s

5

), then slot

78

which follows an s

1

slot

76

cannot be assigned to the fifth user. It has to be assigned to the second user (s

2

).

If the decision controller

112

determines that the selected next slot sync-word passes the logic test of step

222

, then it determines the selected next slot sync-word reliable (step

224

). Otherwise, decision controller

112

determines the selected next slot sync-word unreliable (step

226

).

It will be appreciated by persons skilled in the art that the present invention is not limited to what has been particularly shown and described hereinabove. Rather the scope of the present invention is defined only by the claims which follow.

Claims

1. A method for analyzing a dynamic channel, the method comprising:calculating from a synchronization sequence in a subsequent slot following an assigned slot of a signal frame received over said dynamic channel an estimated taps value for each of a group of predetermined synchronization sequence postulates; calculating from said synchronization sequence a log likelihood metric value C(y,h) for each of said synchronization sequence postulates wherein y represents the signal samples and h represents the channel taps; selecting one of said synchronization sequence postulates according to its calculated log likelihood metric value.
2. The method according to claim 1 wherein said likelihood metric value C(y,h) is calculated from an expression ∑m=KNs⁢&LeftBracketingBar;ym-∑k=1K⁢hk·sm-k&RightBracketingBar;2,wherein s represents the synchronization sequence postulates, Ns is the number of symbols in the synchronization sequence, K is the assumed number of channel taps axed m is a time index.
3. The method according to claim 2 wherein h is calculated from an expression selected from the group consisting of: (∑l=KNs⁢s_l*⁢s_lt)-1⁢∑l=KNs⁢s_l*⁢yl; ⁢and⁢ ⁢1Ns-K⁢∑l=KNs⁢s_l*⁢yl⁢
4. The method according to claim 1 wherein said likelihood metric value C(y,h) is calculated from an expression ∑m=KNs⁢&LeftBracketingBar;ym-h_m·s_m&RightBracketingBar;2,wherein s represents the synchronization sequence postulates, Ns is the number of symbols in the synchronization sequence, K is the assumed number of channel taps and m is a time index.
5. The method according to claim 4 wherein h is calculated from an expression selected from the group consisting of: (∑l=KNs⁢s_l*⁢s_lt)-1⁢∑l=KNs⁢s_l*⁢yl; ⁢and⁢ ⁢1Ns-K⁢∑l=KNs⁢s_l*⁢yl⁢
6. The method according to claim 1 further comprising the steps of:calculating a reliability value; and determining that said selected synchronization sequence postulate reliable.
7. The method according to claim 6 wherein said reliability value is calculated from the difference between the log likelihood value of the selected synchronization sequence postulate and the log likelihood values of the rest of said synchronization sequence postulates of said group of predetermined synchronization sequence postulates.
8. The method according to claim 6 wherein said selected synchronization sequence postulate is determined reliable when said reliability value exceeds a predetermined threshold value.
9. The method according to claim 6 wherein at least one synchronization sequence postulate of said group of predetermined synchronization postulates is assigned for said subsequent slot and excluded from at least another subsequent slot, wherein said reliability value is calculated by comparing said selected synchronization sequence postulate with said at least one synchronization sequence postulate.
10. The method according to claim 1 further comprising analyzing said channel according to said selected synchronization sequence postulate.
11. The method according to claim 1 further comprising analyzing said channel according to said selected synchronization sequence postulate and a synchronization sequence in said assigned slot.
12. The method according to claim 10 further comprising decoding data in said assigned slot according to channel analysis produced by said analyzing.
13. The method according to claim 11 further comprising decoding data in said assigned slot according to channel analysis produced by said analyzing.

Priority Claims (1)

Number	Date	Country	Kind
120898	May 1997	IL

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is a continuation of allowed U.S. patent application Ser. No. 08/997,603 filed Dec. 23, 1997, now U.S. Pat. No. 6,084,891 which is incorporated herein by reference.

US Referenced Citations (18)

Number	Name	Date	Kind
4763339	Sutphin et al.	Aug 1988	A
5007088	Ooi et al.	Apr 1991	A
5101401	Suzuki et al.	Mar 1992	A
5155742	Ariyavisltakul et al.	Oct 1992	A
5241545	Kazecki et al.	Aug 1993	A
5263026	Parr et al.	Nov 1993	A
5276709	Kazawa et al.	Jan 1994	A
5333148	Tsubaki et al.	Jul 1994	A
5335250	Dent et al.	Aug 1994	A
5363412	Love et al.	Nov 1994	A
5408504	Ostman	Apr 1995	A
5619507	Tsuda	Apr 1997	A
5627864	Yoo	May 1997	A
5677934	Ostman	Oct 1997	A
5705314	Davis et al.	Jan 1998	A
5740204	Nagashima	Apr 1998	A
5757845	Fukawa et al.	May 1998	A
5818876	Love	Oct 1998	A

Non-Patent Literature Citations (4)

Entry
“Bi-Directional Equalization Technique for TDMA Communication Systems over Land Mobile Radio Channels”, Yow-Jong Liu, Globecom, 1991, IEEE, pp. 1458-1462.
“Equalization of a Hard—Limited Slowly—Fading Multipath Signal Using a Phase Equalizer with Time—Reversal Structure”, Sirikiat Ariyavisltakul, IEEE Journal on Selected Areas in Communications, vol. 10, No. 3, Apr. 1992, pp. 589-613.
“A Soft-Output Bidirectional Decision Feedback Equalization Technique for TDMA Cellular Radio”, Yow-Jong Liu, IEEE Journal on Selected Areas in Communications, vol. 11, No. 7, Sep. 1993, pp. 1034-1045.
“A Decision Feedback Equalizer with Time-Reversal Structure”, Sirikiat Ariyavisltakul, IEEE Journal on Selected Areas in Communications, vol. 10, No. 3, Apr. 1992, pp. 599-613.

Continuations (1)

	Number	Date	Country
Parent	08/997603	Dec 1997	US
Child	09/459345		US

Bi-directional channel analysis

Information

Patent Number

Date Filed

Date Issued

Inventors

Original Assignees

Examiners

Agents

CPC

US Classifications

Field of Search

US

International Classifications

Disclaimer