(19)
(11)EP 1 101 369 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
29.04.2020 Bulletin 2020/18

(21)Application number: 00935678.3

(22)Date of filing:  29.05.2000
(51)International Patent Classification (IPC): 
H03M 13/29(2006.01)
(86)International application number:
PCT/KR2000/000554
(87)International publication number:
WO 2000/074399 (07.12.2000 Gazette  2000/49)

(54)

APPARATUS AND METHOD FOR ADAPTIVE MAP CHANNEL DECODING IN RADIO TELECOMMUNICATION SYSTEM

VORRICHTUNG UND VERFAHREN ZUR ADAPTIVEN MAP-KANALDEKODIERUNG IN EINEM FUNKTELEKOMMUNIKATIONSSYSTEM

APPAREIL ET PROCEDE DE DECODAGE DE VOIE ADAPTATIF "MAP" DANS UN SYSTEME DE TELECOMMUNICATION RADIO


(84)Designated Contracting States:
DE FI FR GB IT SE

(30)Priority: 28.05.1999 KR 9919476

(43)Date of publication of application:
23.05.2001 Bulletin 2001/21

(73)Proprietor: Samsung Electronics Co., Ltd.
Suwon-si, Gyeonggi-do, 443-742 (KR)

(72)Inventors:
  • KIM, Beong-Jo
    Sungnam-Shi, Kyunggi-Do (KR)
  • KIM, Min-Goo
    Sungnam-Shi, Kyunggi-Do (KR)

(74)Representative: Grünecker Patent- und Rechtsanwälte PartG mbB 
Leopoldstraße 4
80802 München
80802 München (DE)


(56)References cited: : 
EP-A- 0 755 122
US-A- 5 619 537
DE-A- 1 971 546
US-A- 5 933 462
  
  • PIETROBON S S: "IMPLEMENTATION AND PERFORMANCE OF A TURBO/MAP DECODER" INTERNATIONAL JOURNAL OF SATELLITE COMMUNICATIONS, JOHN WILEY AND SONS, US, vol. 16, no. 1, 1998, pages 23-46, XP000856961 ISSN: 0737-2884
  • HEEGARD C; WICKER S B : "Turbo Coding" 1 January 1999 (1999-01-01) , KLUWER INTERNATIONAL SERIES IN ENGINEERING AND COMPUTER , BOSTON, MA , US XP002272827 * paragraph [06.6] * * page 157 - page 158 *
  • PETER HÖHER: "New iterative ("Turbo") Decoding Algorithms" PROCEEDINGS INTERNATIONAL SYMPOSIUM ON TURBO CODES & RELATED TOPICS, 1997, pages 1-8, XP002272828 Brest, France
  • "ITERATIVE DECODING OF PARALLEL CONCATENATED CONVOLUTIONAL CODES" ITERATIVE DECODING OF PARALLEL CONCATENATED CONVOLUTIONAL CODES, XX, XX, 13 January 1999 (1999-01-13), pages 1-12, XP002922269
  • SUMMERS T A ET AL: "SNR mismatch and online estimation in turbo decoding" IEEE TRANSACTIONS ON COMMUNICATIONS, IEEE INC. NEW YORK, US, vol. 46, no. 4, April 1998 (1998-04), pages 421-423, XP002154486 ISSN: 0090-6778
  
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description

BACKGROUND OF THE INVENTION


1. Field of the Invention



[0001] The present invention relates generally to an adaptive MAP (Maximum Aposteriori Probability) channel decoding apparatus and a respective method.

2. Description of the Related Art



[0002] Various channel codes are widely used to conduct communications reliably in a radio telecommunication system (e.g., satellite system, WCDMA, UMTS, and CDMA 2000) The present invention pertains to a turbo code decoding apparatus and method using recursive systematic convolutional codes for data transmission in a bad channel environment

[0003] Turbo code decoding is implemented in a MAP or a SOVA (Soft Output Viterbi Algorithm) method. The former exhibits optimal performance in terms of bit error rate (BER). The performance of a MAP channel decoder is about 0.6-0.7dB higher than that of a SOVA channel decoder in an ideal receiver. To achieve ideal operation of a receiver for MAP or SOVA turbo decoding, the following conditions must be satisfied: (1) the receiver estimates the current channel state accurately; and (2) the receiver uses a sufficient number of bits for inner operations.

[0004] Meanwhile, the receiver can also subject an input signal to RF (Radio Frequency) processing, down-conversion, ADC (analog-to-digital conversion), and symbol demodulation.

[0005] Generally, noise power estimation in a channel environment precedes a MAP operation. Therefore, a MAP channel decoder receives noise power estimated by a channel estimator and the noise power determines decoding performance.

[0006] A plurality of metrics are used as main variables in a MAP algorithm. Hereinbelow, equations required to implement the MAP algorithm will be described.

[0007] A branch metric (BM) is calculated by

where La(dk) is extrinsic information obtained from a feed-back loop in a MAP channel decoder, Lc is channel reliability which can be expressed as 2/σ22 is noise power estimated by a channel estimator), xk is a systematic code, yk is a parity code, i is an expected systematic code, and Yli(m) is an expected parity code.

[0008] A forward state metric (FSM), Alpha is

where Sbj(m) indicates a backward changed state when an information bit is j to set a current state to m. The E function is defined as

In MAP algorithm, x and y can be metric values such as Beta and Alpha as can be seen in (4) and (2)

[0009] A reverse state metric (RSM), Beta is

where

is a state changed forward when an information bit is input in a current encode state m.

[0010] If an information bit is dk at time k of MAP, a log likelihood ratio (LLR) is

where L(dk) is an LLR when an information bit is dk at time k of MAP and

indicates that an operation is performed on NS E functions.

[0011] The MAP decoder calculates a BM from a received code and then an RSM in the above equations, as stated above. The MAP decoder calculates an LLR being a decoded soft output simultaneously with calculation of an FSM.

[0012] As noted from Eq. 1, the MAP channel decoder needs Lc to calculate the BM. That is, the MAP channel decoder can calculate the BM only after it finds out the noise power of a channel environment. According to a general MAP algorithm, therefore, the decoding performance of the MAP channel decoder is sensitively influenced by the operation of a channel estimator.

[0013] If a channel decoder is implemented for turbo decoding, it is preferred that the channel decoder operates stably in a real mobile communication channel environment regardless of the current channel state.

[0014] Particularly, an input signal is multiplied by a scaling factor according to changes in the channel status, like fading, and it is not easy for a receiver to obtain an accurate scaling factor with respect to the current channel state. However, an accurate scaling factor must be reflected to obtain accurate estimation of the channel state. That is, when the channel reliability Lc is degraded, the decoding performance of a MAP algorithm is also degraded, unless the scaling factor is accurately calculated.

[0015] The receiver quantizes an input analog signal with limited bits. As a result, an input signal with strength greater than a predetermined level is clipped. The dynamic range of this input signal seriously influences the performance of the channel decoder, especially in power control.

[0016] S.S. Pietrobon, "Implementation and Performance of a Turbo/Map Decoder", Int. Journal of Satellite Comm., 16, pages 23 to 46, (1998), discloses a log-MAP algorithm in Chapter 2.2 and the sub-MAP algorithm in Chapter 2.3. It is further disclosed that the sub-MAP algorithm is independent of a signal-to-noise ratio (SNR).

Summary of the Invention



[0017] It is, therefore, an object of the present invention to provide a stable MAP channel decoding apparatus and method in the preset mobile communication channel environment.

[0018] It is another object of the present invention to provide an apparatus and method for decoding data stably in a real channel environment of a mobile communication system.

[0019] It is a further object of the present invention to provide a MAP channel decoding apparatus and method in which decoding is stably performed regardless of channel noise power in a channel environment of a mobile communication system.

[0020] It is still another object of the present invention to provide a stable MAP channel decoding apparatus and method in which a MAP channel decoding is performed with a first E function including a log function if the current channel state is a static channel state and with a second E function free of a log function if the current channel state is a time-varying channel state to ensure optimal decoding performance in a real mobile communication channel environment

[0021] These objects are solved by the subject matter of the independent claims.

[0022] Preferred embodiments are the subject matter of the dependent claims.

BRIEF DESCRIPTION OF THE DRAWINGS



[0023] The above and other objects, features and advantages of the present invention will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings in which:

FIG. 1 is a graph illustrating the significance of a new E function according to an embodiment of the present invention;

FIG. 2 is a graph illustrating performance comparisons between a real implemented MAP turbo decoder, sub-MAP turbo decoder, a simulated MAP turbo decoder and sub-MAP turbo decoder;

FIG. 3 is a graph illustrating performance comparisons between a real sub-MAP turbo decoder and a real MAP turbo decoder at various scaling factors at a corresponding Eb/No according to the embodiment of the present invention;

FIG. 4 is a block diagram illustrating an adaptive MAP recursive decoding apparatus depending on a channel state according to the embodiment of the present invention;

FIG. 5A is a graph illustrating a procedure of implementing a log function by referring to a look-up table according to the embodiment of the present invention;

FIG. 5B illustrates the look-up table referred to for implementing a log function according to the embodiment of the present invention; and

FIG. 6 is a flowchart illustrating a MAP decoding method adaptive to a current channel state according to the embodiment of the present invention.


DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT



[0024] A preferred embodiment of the present invention will be described hereinbelow with reference to the accompanying drawings. In the following description, well-known functions or constructions are not described in detail to avoid obscuring the invention in unnecessary detail.

[0025] The present invention provides a MAP channel decoding apparatus and method in which MAP channel decoding is stably implemented regardless of the noise power of a channel in a real mobile communication channel environment when a channel decoder is implemented using a MAP algorithm, which is one of the main algorithms for turbo decoding.

[0026] In the present invention, a general MAP algorithm is implemented in a static channel state and a sub-MAP algorithm is implemented using a two E functions, presented according to the present invention in a time-varying channel state, for stable MAP channel decoding.

[0027] To calculate a BM without Lc in Eq. 1, an E function should be amended in Eq. 3. The E function is defined as the difference between a min function and a log function. The log function shows nonlinear characteristics. Therefore, the present invention defines a new E function (referred to as 2E function hereinafter) obtained by removing the log function from the E function as follows.



[0028] The reason for omitting log(1+e-|x-y|) in Eq. 6 can be found by evaluating Eq. 7.



[0029] Eq. 7 is symmetrical and as |z| increases, F(z) rapidly converges to 0 as illustrated in FIG. 1. FIG. 1 is a graph illustrating the significance of the log function from the new 2E function according to the embodiment of the present invention. The horizontal axis in the graph indicates |z| and the vertical axis, F(z) of Eq. 7. As shown in the graph, F(z) is rapidly converged to 0 as |z| increases.

[0030] It can be noted from FIG. 1 that as |x-y| of Eq. 3 increases, log(1+e-|x-y|) is converged to 0, having no influence on the E function. Accordingly, it can be expected that log(1+e-|x-y|) does not influence the E function unless the difference between x and y is small, that is, a SNR is very small. That is, a change in the performance of a real MAP algorithm without log(1+e-|x-y|) in the conventional E function is negligible only if a certain SNR is maintained for a channel environment.

[0031] FIG. 2 is a graph illustrating performance comparisons between a MAP channel decoder and a sub-MAP channel decoder when they are implemented and simulated. The vertical axis represent a BER (bit error rate) and FER (frame error rate) and the horizontal axis represent Eb/No. Each component decoder has a constraint length k = 4, a code rate r = 1/2, and a frame length of 375 bits. A test was performed on an AWGN (Additive White Gaussian Noise) channel in an ideal channel estimation state.

[0032] As shown in FIG. 2, the gain performance of the sub-MAP algorithm is less than that of the MAP algorithm by about 0.1dB. However, if an SNR is 2.5dB, there is little difference in performance between the MAP channel decoder and the sub-MAP channel decoder. According to the test result, if a channel environment ensures an SNR of a predetermined level, log(1+e-|x-y|) in Eq. 3 converges to approximately near 0.

[0033] Meanwhile, according to the definition of the new E function as Eq. 6, Eqs. 1, 2, 4, and 5 can be changed respectively to

where Lc is offset. Since extrinsic information La obtained from a recursive decoder feed-back loop starts from 0 and thus initial FSM and RSM do not involve Lc multiplication. Therefore, Eq. 8 can be changed to

where Lc is not used in the sub-MAP algorithm according to the embodiment of the present invention as the new E function is defined as Eq. 6. That is, the sub-MAP algorithm is independent of channel noise power.

[0034] Now, the sensitivity of the decoding performances of the MAP algorithm and the sub-MAP algorithm when real channel noise power is mismatched will be discussed. Fixed point simulation is assumed herein for the MAP channel decoder and the sub-MAP channel decoder. That is, a predetermined number of quantization bits are used for an internal metric in a component decoder.

[0035] As a simulation condition, an input signal on an AWGN is scaled as

where S is an input signal, c ∈ {+1,-1} is a code symbol, n is a sample value of normal distribution N(0, 1), g is a scaling factor, and σ is the standard deviation of channel noise. Thus, σ2 is the channel noise power.

[0036] In Eq. 10, c·g is the signal component and σ·n·g is the noise component. When, the scaling factor g is removed, the SNR of the input signal S will be a constant value SNR=(c·g)/(σ·n·g) = c/(σ·n), regardless of g. That is, the scaling in Eq. 10 is based on the assumption that the variation of a channel state is equal to the resulting change in a transmission signal. Therefore, the simulation condition of Eq. 10 maintains the SNR constant via the scaling factor g. To satisfy the above condition, the level of the input signal is artificially increased or decreased. As a result, the input signal deviates from the dynamic range of a quantizer having limited quantization levels and a limited quantization range.

[0037] The sensitivity of the decoding performance of the MAP algorithm and the sub-MAP algorithm will be described in the context of the above simulation condition. Unless Lc as related with scaling is accurately reflected in the MAP algorithm, the decoding performance of the MAP channel decoder will rapidly drop relative to the sub-MAP decoder. If the MAP channel decoder operates within the simulation condition, a component decoder must multiply an input signal by Lc=2/(g·σ)2 where an accurate scaling factor is reflected but receives Lc=2/σ2 instead. Consequently, the decoding performance of the MAP channel decoder is decreased.

[0038] On the other hand, the sub-MAP channel decoder operates regardless of Lc, the decrease in performance when an input signal level deviates from a dynamic range is attributed to the relative change of a quantization range and resolution due to a scaling factor.

[0039] FIG. 3 is a graph illustrating performance comparisons between a real MAP channel decoder and a real sub-MAP channel decoder when a scaling factor in Eb/No is changed according to the embodiment of the present invention. The vertical axis in the graph denotes the BER and FER and the horizontal axis, the value of factor g, is expressed in dB. If the horizontal axis is set to 20xlog(g) and g is 3, 20xlog(3) will be shown as g in the graph. For g=1, log1=0. Therefore, g is 0 on the graph.

[0040] It can be noted from FIG. 3 that the MAP channel decoder is superior to the sub-MAP channel decoder in terms of decoding performance as long as an accurate g is obtained. However, as the scaling factor g is gradually increased or decreased, the decoding performance of the MAP channel decoder rapidly decreases. This implies that the decoding performance of the MAP channel decoder is very sensitive to a current channel state.

[0041] On the other hand, the sub-MAP channel decoder exhibits a slow change in decoding performance with respect to the change of the scaling factor g. That is, the sub-MAP decoding has a wider operation range with respect to the change of the scaling factor g than the MAP decoding. Accordingly, the sub-MAP decoding scheme according to the embodiment of the present invention is more stable with less influence from the current channel state in a real mobile radio environment, as compared to the MAP decoding scheme.

[0042] As shown in FIG. 3, the sub-MAP channel decoder has a stable operation range from about -10dB to +10dB, whereas the MAP channel decoder has a stable operation range from about -3dB to +6dB.

[0043] Depending on the SNRs, MAP channel decoders differ in performance from the sub-MAP channel decoder, as noted from FIGs. 2 and 3. But they are equal at an SNR over a specific level. In this case, the decrease of performance caused by channel noise power mismatch cannot be found in a range from +10 to -10dB.

[0044] FIG. 4 is a block diagram illustrating a MAP channel decoding apparatus adaptive to a channel state according to the embodiment of the present invention. Referring to FIG. 4, a receiver 410 subjects an analog radio signal received through an antenna (not shown) to RF processing, down conversion of the RF signal to IF signal, Analog to Digital Converting the IF signal, and symbol demodulation and outputs an demodulated signal. A channel estimator 420 calculates a channel noise power and a scaling factor from the output of the receiver 410. A controller 430 determines the operation mode of a MAP channel decoder between a static channel mode and a time-varying mode according to the output of the channel estimator 420. The determination is made based on changes in channel noise power and the scaling factor. The controller 430 can determine the operation mode based on BER/FER received from the MAP channel decoder 440. For example, if a predetermined number of successive occurrences of a scaling factor approximate to a predetermined dB are observed, the controller 430 determines the operation mode to be a static channel mode.

[0045] Meanwhile, if the controller 430 determines that the current channel state is a static channel mode, an E function operator 450 of the MAP channel decoder 440 processes inputs x and y in Eq. 3 for optimal decoding performance. The MAP channel decoder 440 decodes the input signal using Eqs. 1, 2, 4, and 5. In the time-varying channel mode, the E function operator 450 processes the inputs x and y in Eq. 6 and then the MAP channel decoder 440 decodes the input signal using Eq. 9.

[0046] A long function in the E function operator 450 can be implemented using a look-up table. The look-up table may be stored in a relatively small amount of memory because the log(1+e-|x-y|) function rapidly drops as |x-y| increases, as stated in FIG. 1.

[0047] FIG. 5A is a graph illustrating the procedure for implementing a log function by referring to a look-up table according to the embodiment of the present invention and FIG. 5B illustrates the look-up table for implementing the log function according to the embodiment of the present invention.

[0048] The horizontal axis in the graph shown in FIG. 5A represents the level of a function input value corresponding to the difference between input values x and y of the E function operator 450 and the vertical axis represents 10g(1+e-|x-y|) versus the function input levels stored in the table.

[0049] The input values x and y have been quantized and there are a limited number of |x-y| values. As shown in FIG. 5B, the look-up table uses the difference between x and y as a memory address and stores an already calculated function value corresponding to the memory address in a data area.

[0050] Due to the nature of the log(1+e-|x-y|), an output value is inversely proportional to an input value in the look-up table according to the embodiment of the present invention. Hence, the look-up table may be stored in a relatively small amount of memory.

[0051] FIG. 6 is a flowchart illustrating a MAP decoding method adaptive to a current channel state according to the embodiment of the present invention. The adaptive MAP decoding method will be described referring to FIGs. 1 to 6.

[0052] The controller 430 estimates a current channel state through the channel estimator 420 in step 610. In step 620, the controller receives accumulated channel noise power or a scaling factor from the channel estimator 420 and determines the current channel state according to the variations in the channel noise power or the scaling factor. In addition, the controller 430 can determine the current channel state according to a change in BER/FER received from the MAP channel decoder 440.

[0053] The controller 430 outputs a mode select signal to the MAP channel decoder 440 to operate the MAP channel decoder 440 in a static channel mode in step 630. That is, the E function operator 450 of the MAP channel decoder 440 processes input x and y using Eq. 3, namely, a general E function in step 630. Thus, the MAP channel decoder 440 decodes the input signal using Eqs. 1, 2, 4, and 5, as illustrated by step 640. Here, the E function operator 450 can implement a log function referring to a look-up table.

[0054] If the controller 430 determines the current channel state to be time varying in step 620, the controller 430 outputs the mode select signal to the MAP channel decoder 440 to operate it in a time-varying channel mode in step 650. That is, the E function operator 450 performs MAP channel decoding using Eq. 9, namely, the 2E function according to the present invention, as illustrated by step 660.

[0055] As described referring to FIGs. 1 to 6, the sub-MAP channel decoding algorithm according to the present invention is independent of a current channel state, which can reduce the sensitivity of a MAP algorithm to a channel state change. In addition, the sub-MAP algorithm has increased stability in a real mobile communication channel environment as compared to the MAP algorithm.

[0056] In accordance with the adaptive MAP channel decoding apparatus and method of the present invention, a current channel state is first checked based on the change of accumulated channel noise power or a scaling factor of a channel estimator, or a change in BER/FER of a MAP channel decoder. If the current channel state is determined to be static, decoding is performed using an E function including a log function to ensure optimal decoding performance. If the current channel state is to be time-varying, the log function is not used in the E function operation regardless of the current channel state to perform stable channel decoding in a real mobile communication channel environment Therefore, MAP channel decoding is implemented adaptively to a channel state,


Claims

1. An adaptive MAP channel decoding apparatus, comprising:

a channel estimator (420) adapted to calculate a channel noise power and a scaling factor;

a MAP channel decoder (440);

a controller (430) adapted to determine an operation mode of the MAP channel decoder (440); and

the MAP channel decoder (440) being adapted to selectively operate in a static channel mode or in a time-varying mode, wherein the MAP channel decoder is adapted to use in the static channel mode the following equation for decoding:

and wherein the MAP channel decoder is adapted to use in the time-varying mode the following equation for decoding:

where x and y are metric values,

wherein a current channel state is determined by the controller to be time-varying or static based on changes in the channel noise power or the scaling factor, and

wherein the time-varying channel mode is determined if the current channel state is determined to be time-varying, and otherwise the static channel mode is determined.


 
2. The adaptive MAP channel decoding apparatus of Claim 1, wherein the MAP channel decoder performs a first MAP channel decoding operation using the following equations while in the static channel mode:

where Dki(m) = a branch metric, BM;

La(dk) = extrinsic information obtained from a feed-back loop in the MAP channel decoder;

Lc = channel reliability which can be expressed as 2/σ2, wherein σ2 is noise power;

xk = a systematic code;

yk = a parity code;

i = an expected systematic code;

Yli(m) = an expected parity code;

Aik(m) = a forward state metric, FSM;

Sbj(m) = a backward changed state when an information bit is j to set a current state to m;

Bki(m) = a reverse state metric, RSM;

Sfjm) = a state changed forward when an information bit is input in a current encode state m; and

L(dk) = a log likelihood ratio, LLR, when an information bit is dk at time k of MAP; and

performs a second MAP channel decoding operation using the following equations while in a time-varying mode:

where Dki(m) = a BM

La(dk) = extrinsic information obtained from a feed-back loop in the MAP channel decoder;

xk= a systematic code;

yk = a parity code;

i = an expected systematic code;

Yli(m) = an expected parity code;

Aik(m) = a forward state metric, FSM;

Sbj(m) = a backward changed state when an information bit is j to set a current state to m;

Bki(m) = an RSM;

Sfj(m) = a state changed forward when an information bit is input in a current encode state m; and

L(dk) = a log likelihood ratio, LLR, when an information bit is dk at time k of MAP.


 
3. The adaptive MAP channel decoding apparatus of Claim 2, wherein the E function operator uses the absolute value of the difference between function input values as a memory address to look-up stored corresponding calculated log function values in a data area of a memory and output the corresponding log function values.
 
4. An adaptive MAP channel decoding method in a communication system, comprising the steps of:

estimating, by a channel estimator, a channel noise power and a scaling factor;

determining a current channel state to be time-varying or static based on changes in the channel noise power or the scaling factor;

controlling an E function operation according to the determined current channel state by:

performing (630) MAP channel decoding using the E function operation according to the following first MAP decoding equation when the current channel state is determined to be static:

where x and y are metric values, and

performing MAP channel decoding using the E function operation according to the following second MAP decoding equation when the current channel state is determined to be time-varying:

where x and y are metric values.


 


Ansprüche

1. Adaptive MAP-Kanaldecodiervorrichtung, umfassend:

einen Kanalschätzer (420), der dazu eingerichtet ist, eine Kanalrauschleistung und einen Skalierfaktor zu berechnen;

einen MAP-Kanaldecoder (440);

eine Steuereinheit (430), die dazu eingerichtet ist, einen Betriebsmodus des MAP-Kanaldecoders (440) zu bestimmen; wobei

der MAP-Kanaldecoder (440) dazu eingerichtet ist, selektiv in einem statischen Kanalmodus oder in einem zeitveränderlichen Modus zu arbeiten, wobei der MAP-Kanaldecoder dazu eingerichtet ist im statischen Kanalmodus die folgende Gleichung zum Decodieren zu verwenden:

und der MAP-Kanaldecoder dazu eingerichtet ist, im zeitvariablen Modus die folgende Gleichung zum Decodieren zu verwenden:

wobei x und y metrische Werte sind,

wobei ein aktueller Kanalzustand von der Steuereinheit basierend auf Änderungen der Kanalrauschleistung oder des Skalierfaktors als zeitveränderlich oder statisch bestimmt wird und

der zeitvariable Kanalmodus bestimmt wird, wenn der aktuelle Kanalzustand als zeitvariabel bestimmt wird, und andernfalls der statische Kanalmodus bestimmt wird.


 
2. Adaptive MAP-Kanaldecodiervorrichtung nach Anspruch 1, bei der der MAP-Kanaldecoder im statischen Kanalmodus eine erste MAP-Kanaldecodiertätigkeit unter Verwendung der folgenden Gleichungen ausführt:

wobei Dki (m) = eine Verzweigungsmetrik, BM, ist;

La(dk) = extrinsische Informationen sind, die von einer Rückkopplungsschleife in dem MAP-Kanaldecoder bezogen werden;

Lc = eine Kanalzuverlässigkeit ist, die als 2/σ2 ausgedrückt werden kann, wobei σ2 Rauschleistung ist;

xk = ein systematischer Code ist;

yk = ein Paritätscode ist;

i = ein erwarteter systematischer Code ist;

Yli(m) = ein erwarteter Paritätscode ist;

Aik(m) = eine Vorwärtszustandsmetrik, FSM, ist;

Sbj (m) = ein rückwärts geänderter Zustand ist, wenn ein Informationsbit j ist, um einen aktuellen Zustand auf m zu setzen;

Bki(m) = eine Umkehrzustandsmetrik, RSM, ist;

Sfj(m) = ein Zustand ist, der vorwärts geändert wird, wenn ein Informationsbit in einem aktuellen Codierzustand m eingegeben wird; und

L(dk) = ein Log-Likelihood-Ratio, LLR, ist, wenn ein Informationsbit zum Zeitpunkt k von MAP dk ist; und

eine zweite MAP-Kanaldecodiertätigkeit unter Verwendung der folgenden Gleichungen in einem zeitvariablen Modus ausführt:



wobei Dki (m) = eine BM ist;

La(dk) = extrinsische Informationen sind, die von einer Rückkopplungsschleife in dem MAP-Kanaldecoder bezogen werden;

xk = ein systematischer Code ist;

yk = ein Paritätscode ist;

i = ein erwarteter systematischer Code ist;

Yli(m) = ein erwarteter Paritätscode ist;

Aik(m) = eine Vorwärtszustandsmetrik, FSM, ist;

Sbj (m) = ein rückwärts geänderter Zustand ist, wenn ein Informationsbit j ist, um einen aktuellen Zustand auf m zu setzen;

Bki(m) = eine RSM ist;

Sfj(m) = ein Zustand ist, der vorwärts geändert wird, wenn ein Informationsbit in einem aktuellen Codierzustand m eingegeben wird; und

L(dk) = ein Log-Likelihood-Ratio, LLR ist, wenn ein Informationsbit zum Zeitpunkt k von MAP dk ist.


 
3. Adaptive MAP-Kanaldecodiervorrichtung nach Anspruch 2, bei der der E-Funktionsoperator den Absolutwert der Differenz zwischen Funktionseingabewerten als Speicheradresse verwendet, um gespeicherte entsprechende berechnete Protokollfunktionswerte in einem Datenbereich eines Speichers nachzuschlagen und die entsprechenden Log-Funktionswerte auszugeben.
 
4. Adaptives MAP-Kanaldecodierverfahren in einem Kommunikationssystem, umfassend folgende Schritte:

Schätzen einer Kanalrauschleistung und eines Skalierfaktors durch einen Kanalschätzer;

Bestimmen eines aktuellen Kanalzustands als zeitveränderlich oder statisch basierend auf Änderungen der Kanalrauschleistung oder des Skalierfaktors;

Steuern einer E-Funktionsoperation gemäß dem bestimmten aktuellen Kanalzustand durch:
Ausführen (630) einer MAP-Kanaldecodierung unter Verwendung der E-Funktionsoperation gemäß der folgenden ersten MAP-Decodiergleichung, wenn der aktuelle Kanalzustand als statisch bestimmt wird:

wobei x und y metrische Werte sind und

Ausführen einer MAP-Kanaldecodierung unter Verwendung der E-Funktionsoperation gemäß der folgenden zweiten MAP-Decodiergleichung, wenn der aktuelle Kanalzustand als zeitvariabel bestimmt wird:



wobei x und y metrische Werte sind.


 


Revendications

1. Appareil de décodage de canal MAP adaptatif, comprenant :

un estimateur de canal (420) adapté pour calculer une puissance de bruit de canal et un facteur d'échelle ;

un décodeur de canal MAP (440) ;

un dispositif de commande (430) adapté pour déterminer un mode de fonctionnement du décodeur de canal MAP (440) ; et

le décodeur de canal MAP (440) étant adapté pour fonctionner sélectivement dans un mode de canal statique ou dans un mode variant dans le temps, dans lequel le décodeur de canal MAP est adapté pour utiliser dans le mode de canal statique l'équation suivante pour le décodage :



et dans lequel le décodeur de canal MAP est adapté pour utiliser dans le mode variant dans le temps l'équation suivante pour le décodage :

où x et y sont des valeurs métriques,

dans lequel un état de canal courant est déterminé par le dispositif de commande comme variant dans le temps ou comme étant statique sur la base de changements de la puissance de bruit de canal ou du facteur d'échelle, et

dans lequel le mode de canal variant dans le temps est déterminé si l'état de canal courant est déterminé comme variant dans le temps, et sinon le mode de canal statique est déterminé.


 
2. Appareil de décodage de canal MAP adaptatif selon la revendication 1, dans lequel le décodeur de canal MAP effectue une première opération de décodage de canal MAP en utilisant les équations suivantes dans le mode de canal statique :







où Dki(m) = une métrique de branche, BM ;

La(dk) = informations extrinsèques obtenues à partir d'une boucle de rétroaction dans le décodeur de canal MAP ;

Lc = fiabilité de canal qui peut être exprimée par 2/σ2, où σ2 est une puissance de bruit ;

Xk = un code systématique ;

Yk = un code de parité ;

i = un code systématique attendu ;

Yli(m) = un code de parité attendu ;

Aik(m) = une métrique d'état direct, FSM ;

Sbj(m) = un état modifié vers l'arrière lorsqu'un bit d'informations est j pour établir un état courant sur m ;

Bki(m) = une métrique d'état inverse, RSM ;

Sfj(m) = un état changé ver l'avant lorsqu'un bit d'informations est entré dans un état de codage courant m ; et

L(dk) = un rapport de vraisemblance logarithmique, LLR, lorsqu'un bit d'informations est dk à un instant k de MAP ; et

effectue une seconde opération de décodage de canal MAP en utilisant les équations suivantes dans un mode variant dans le temps :









où Dki(m) = une BM

La(dk) = informations extrinsèques obtenues à partir d'une boucle de rétroaction dans le décodeur de canal MAP ;

Xk = un code systématique ;

Yk = un code de parité ;

i = un code systématique attendu ;

Yli(m) = un code de parité attendu ;

Aik(m) = une métrique d'état direct, FSM ;

Sbj(m) = un état modifié vers l'arrière lorsqu'un bit d'informations est j pour établir un état courant sur m ;

Bki(m) = une RSM ;

Sfj(m) = un état changé ver l'avant lorsqu'un bit d'informations est entré dans un état de codage courant m ; et

L(dk) = un rapport de vraisemblance logarithmique, LLR, lorsqu'un bit d'informations est dk à un instant k de MAP.


 
3. Appareil de décodage de canal MAP adaptatif selon la revendication 2, dans lequel l'opérateur de fonction E utilise la valeur absolue de la différence entre des valeurs d'entrée de fonction comme adresse de mémoire pour rechercher des valeurs de fonction logarithmique calculées correspondantes stockées dans une zone de données d'une mémoire et pour délivrer en sortie les valeurs de fonction logarithmique correspondantes.
 
4. Procédé de décodage de canal MAP adaptatif dans un système de communication, comprenant les étapes consistant à :

estimer, par un estimateur de canal, une puissance de bruit de canal et un facteur d'échelle ;

déterminer un état de canal courant comme variant dans le temps ou étant statique sur la base de changements de la puissance de bruit de canal ou du facteur d'échelle ;

commander une opération de fonction E en fonction de l'état de canal courant déterminé en :
effectuant (630) un décodage de canal MAP en utilisant l'opération de fonction E conformément à la première équation de décodage MAP suivante lorsque l'état de canal courant est déterminé comme étant statique (

où x et y sont des valeurs métriques, et

effectuant un décodage de canal MAP en utilisant l'opération de fonction E conformément à la seconde équation de décodage MAP suivante lorsque l'état de canal courant est déterminé comme variant dans le temps :

où x et y sont des valeurs métriques.


 




Drawing























Cited references

REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Non-patent literature cited in the description