US20060103559A1 - Method and apparatus of two stage scaling and quantization for coded communication systems - Google Patents
Method and apparatus of two stage scaling and quantization for coded communication systems Download PDFInfo
- Publication number
- US20060103559A1 US20060103559A1 US10/987,343 US98734304A US2006103559A1 US 20060103559 A1 US20060103559 A1 US 20060103559A1 US 98734304 A US98734304 A US 98734304A US 2006103559 A1 US2006103559 A1 US 2006103559A1
- Authority
- US
- United States
- Prior art keywords
- signal
- scaling
- circuit
- bit
- quantizer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000013139 quantization Methods 0.000 title claims description 19
- 238000000034 method Methods 0.000 title claims description 12
- 238000004891 communication Methods 0.000 title description 5
- 230000004044 response Effects 0.000 claims abstract description 17
- 238000010586 diagram Methods 0.000 description 6
- 238000005562 fading Methods 0.000 description 5
- 238000013461 design Methods 0.000 description 3
- 239000000654 additive Substances 0.000 description 2
- 230000000996 additive effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000011664 signaling Effects 0.000 description 2
- 101100458289 Drosophila melanogaster msps gene Proteins 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/85—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression
- H04N19/89—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression involving methods or arrangements for detection of transmission errors at the decoder
Abstract
Description
- The present invention relates to communication systems generally and, more particularly, to a method and/or apparatus for implementing two stage scaling and quantization for coded communication systems.
- To achieve better performance in a conventional digital communication system employing channel coding, decoder inputs are implemented as variables that reflect the reliability of the received data. This is sometimes referred to as soft decoding. Another type of decoding, sometimes referred to as hard decoding, uses only the sign of the received coded symbol as an input. To provide such a soft input to a decoder, the received coded symbol is usually quantized to be represented with a certain number of bits. The more bits used, the higher the quantization accuracy. However, implementing more bits uses more resources to store and process the additional bits.
- Conventional coded systems also implement interleaving to improve error rate performance, especially in fading channels. The interleaved data is the soft value after demodulation. The size of the interleaver memory is determined by the coded symbol frame size and the bit-width of each soft symbol. For a higher data rate, the size can be significant in the overall receiver cost consideration.
- A common configuration for a digital receiver is that the demodulator output is quantized into certain number bits (e.g., n). The whole frame of the coded symbols, each represented by the n bits, are stored into the deinterleaver memory. The deinterleaver reads the data out in a different order and feeds the data to a channel code decoder. The parameter n controls the size of the deinterleaver and the input bit-width of the decoder. Decreasing the parameter n reduces the size of the deinterleaver and decoder, but tends to degrade the decoding performance.
- It would be desirable to implement a system that uses multi-stage quantization and/or scaling to decrease overhead and improve performance compared to conventional approaches.
- The present invention concerns an apparatus comprising a first quantizer circuit, a memory and a second quantizer circuit. The first quantizer circuit may be configured to generate a first intermediate signal in response to (i) an input signal and (ii) a first scaling signal. The memory may be configured to (i) store the first intermediate signal and (ii) present a second intermediate signal, in response to an address signal. The second quantizer circuit may be configured to generate an output signal in response to (i) the second intermediate signal and (ii) a second scaling signal. In general, the second quantizer circuit has a bit-width greater than the bit-width of the first quantizer circuit.
- The objects, features and advantages of the present invention include providing a quantizer system that may (i) achieve optimum error performance in a fading channel without increasing interleaving memory size, (ii) use slow changing physical channel information to scale a high rate decoder input, and/or (iii) be implemented as a multi-stage quantization method and/or apparatus.
- These and other objects, features and advantages of the present invention will be apparent from the following detailed description and the appended claims and drawings in which:
-
FIG. 1 is a block diagram of a receiver implementing two stage quantization in accordance with a preferred embodiment of the present invention; -
FIG. 2 is a more detailed block diagram of the present invention; and -
FIG. 3 is a more detailed diagram of the demodulator and combiner circuit. - Referring to
FIG. 1 , a block diagram of acircuit 100 is shown in accordance with a preferred embodiment of the present invention. Thecircuit 100 generally comprises a block (or circuit) 102, a block (or circuit) 104, a block (or circuit) 106, and a block (or circuit) 108, a block (or circuit) 110 and a block (or circuit) 112. Thecircuit 102 may be implemented as an input circuit. Thecircuit 104 may be implemented as a quantization circuit. Thecircuit 106 may be implemented as a memory. Thecircuit 108 may be implemented as a quantization circuit. Thecircuit 110 may be implemented as a decoder circuit. Thecircuit 112 may be implemented as a control circuit. The present invention may implement two stages of quantization. Thequantizer circuit 104 generally has a first bit-width. Thequantizer circuit 108 generally has a second bit-width that is greater than the first bit-width for thequantizer circuit 104. - Referring to
FIG. 2 , a more detailed diagram of thecircuit 100 is shown. Thecircuit 102 generally comprises a block (or circuit) 118 and a block (or circuit) 120. Thecircuit 118 may be implemented as an automatic gain control circuit. Thecircuit 120 may be implemented as a demodulator and combiner circuit. While thecircuit 102 is shown comprising thecircuit 118 and thecircuit 120, other circuits may also be implemented. For example, thecircuit 102 may also include a frequency conversion circuit, an analog to digital conversion circuit, etc. However, the additional circuitry is not generally necessary for the understanding of the present invention and has been omitted for clarity. - The
circuit 104 generally comprises a block (or circuit) 122 and a block (or circuit) 124. Thecircuit 122 may be implemented as a multiplier circuit. Thecircuit 124 may be implemented as a quantizer circuit. In one example, thecircuit 124 may be implemented as a n1-bit quantizer circuit. - The
circuit 106 may be implemented as a memory, such as a deinterleaver memory. Thecircuit 108 generally comprises a block (or circuit) 130 and a block (or circuit) 132. Thecircuit 130 may be implemented as a multiplier circuit. Thecircuit 132 may be implemented as a quantizer circuit. In one example, thecircuit 132 may be implemented as a n2-bit quantizer circuit. Thecircuit 110 may be implemented as a decoder circuit. - The
circuit 112 generally comprises a block (or circuit) 140, a block (or circuit) 142, a block (or circuit) 144 and a block (or circuit) 146. Thecircuit 140 may be implemented as a scaling value calculation circuit. Thecircuit 142 may be implemented as a deinterleaver address generation circuit. Thecircuit 144 may be implemented as a modular operator circuit. Thecircuit 144 may perform an out=in mod N operation. Thecircuit 146 may be implemented as a second stage scaling buffer circuit. - An input signal (e.g., DATA) may be received by the automatic gain control (AGC)
circuit 118 of thecircuit 102. Thecircuit 102 presents an intermediate signal (e.g., A) to thecircuit 104. Thecircuit 104 presents an intermediate signal (e.g., B) to thecircuit 106. Thedemodulator circuit 120, depending on the signaling design, may include fingers and maximum ratio combining, as in a spread spectrum communication system. The intermediate signal A generally comprises a number of demodulated and combined symbols. The demodulated and combined symbols of the signal A are scaled by thecircuit 122. Thecircuit 124 quantizes the symbols into n1 bits. The scaled and quantized symbols are saved into thedeinterleaver memory 106. A certain number of second stage scaling variables (e.g., a number equal to frame length/N where N>1) are saved in thememory 146 of thecontrol circuit 112. When the deinterleaveraddress generation circuit 142 generates the read address to be presented to the deinterleaver memory 126, the same address is presented to thecircuit 144. Thecircuit 144 may implement a modular N operation to read out the corresponding second stage scaling value, which is applied to the deinterleaved data. The data are further quantized to n2 (where n2>n1) bits before being sent tochannel decoder 110. - A scaling value may be calculated differently according to different signaling specifications. One implementation of the scaling levels may be where the scaling of the circuit 104 (e.g., a
stage 1 scaling) keeps the noise level constant at thequantizer 124 to minimize quantization loss. The scaling of the circuit 108 (e.g., astage 2 scaling) is applied such that the signals sent to thedecoder 110 are represented as optimal variables. The quantization loss is less severe since the bit-width n2 is larger than the bit-width n1 by design. - In one example, an L path fading channel with additive Gaussian noise may be used. At the
receiver 102, the signal DATA is first processed by the automaticgain control circuit 118, then by the demodulation andcombiner circuit 120. Thecircuit 120 contains one or more fingers (to be described in more detail in connection withFIG. 3 ) to extract the corresponding signal power of each path. The 1-th finger output may be represented by the following equation EQ1:
y k (l) =A k √{square root over (E s )} c kαk (l) +A k n k (l) EQ1
Here 1=0, 1, 2, . . . L−1 for L signal paths, k is the time index for symbol sequence where: -
- Ak is the AGC gain at time k,
- ck is the coded symbol (+/−1),
- Es is coded symbol energy,
- αk (l) is the channel gain for
path 1, - nk (l) is Gaussian noise with 0 mean and variance N0 (i)/2
- Referring to
FIG. 3 , a more detailed diagram of thecircuit 120 is shown. Thecircuit 120 may be used to obtain a decision variable with L=2. Thecircuit 120 generally comprises a number offinger circuits 130 a-130 n. Each of thefinger circuits 130 a-130 n generally operates according to the equation EQ1. -
FIG. 3 illustrates an estimation of the channel gain Pk, and is usually obtained from the pilot channel (as in CDMA2000) or pilot symbols provided by the transmitter. The channel gain Pk may be expressed with the following equation EQ2:
pk (l) =A k √{square root over (E p )}α k (l) K+ñ k ≈A k √{square root over (E p )}α k (l) EQ2
Here Ep is the pilot power from a transmitter. When the average length for estimation is large enough, the estimation noise ñk is negligible. - The output of the block 120 (e.g., vk) or the intermediate signal A is proportional to an optimum decision variable (e.g., rk or a soft value) which may be derived from the maximum likelihood criteria, which may be expressed with the following equation EQ3:
- The variable vk and the variable rk hold an approximately proportional relationship as shown in the following equation EQ4:
- Thus, the decision variable vk may be used as a decoder input for the same optimal performance. Due to the AGC gain that varies with channel fading, the noise variance of the optimal decision variable also varies even with constant channel noise variance. The first stage scaling is to normalize the noise variance of the decision variable to reduce quantization noise (or equivalently to reduce bit width requirement) at the following n1 bit quantizer. The scaling value is to be proportional as shown in the following equation EQ5:
Analysis and simulation show this normalization reduce the bit-width requirement of the quantization for the same quantization and decoding performance. - The quantized n1-bit values are saved into deinterleaver memory and deinterleaved. The deinterleaver output is scaled by the second stage scaling unit using the deinterleaved version of the reciprocal value of the equation EQ5 as shown in the following equation EQ6:
- The values in the equations EQ5 and EQ6 change according to the channel condition. When the channel condition is estimated for every modulation symbol, the normalization values S1 and S2 are also updated once per soft value. However, for a high rate data transmission over a wireless channel where the modulated symbol may be as high as several Mega (e.g., million) symbols per second, the channel condition changes relatively slowly since the channel condition is mostly caused by motion in the receiver (e.g., Doppler spread).
- For example, a mobile phone operating with a 2 GHz carrier frequency and traveling at 120 km/h will introduce a Doppler spread of about 220 Hz, which is 10000 times slower than 2.2 Msps data. The channel condition may be assumed to be unchanged for at least 1000 such symbols. When QPSK modulation is used, the number of scaling values can be 2000 times less than the number of soft values that are saved in the
deinterleaver memory 106. - The values in equation EQ6 are normally saved in the second stage scaling buffer (block 146), where N in the
block 144 is 2000 in one example. With such an implementation, thememory 106 may be approximately 2000 times smaller than a conventional deinterleaver memory in terms of address space. The value is accessed similarly by the deinterleaver address except that one value corresponds to N different deinterleaver addresses. The second stage scaling is followed by a n2-bit quantizer where n2 is generally larger than n1. The quantized value is fed into thedecoder 110 for further processing. - The advantages of such a configuration may be illustrated as follows. The costs associated with implementing an increase in bit-width are different for the
deinterleaver memory 106 than for thedecoder 110. The size of thedeinterleaver memory 106 is generally proportional to bit-width. For thedecoder 110, the input bit-width only affects part of the decoder size. Also for some implementations of the decoder 110 (e.g., Viterbi with different repetition rate, turbo decoder that has feedback), the architecture of thedecoder 110 already has most of the capability of handling higher number of bits at the input. By using lower number of bits for thedeinterleaver memory 106 and a higher number of bits at the input of thedecoder 110, both the cost and performance of thememory 106 may be optimized. - The quantized soft value carries information from two origins, the coded symbol, and the physical channel information (e.g., additive noise, fading, receiver front end distortion etc.). Although both need to be properly represented to achieve optimal performance, the relative slow rate of change of physical channels means these information may be treated the same for data over a period of time. This reduces the amount of data need to be stored and applied at the second stage of scaling and quantization.
- In one example, the present invention may be used in a CDMA2000 or WCDAM mobile communication system. However, the present invention may be easily implemented in other designs.
- While the invention has been particularly shown and described with reference to the preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made without departing from the spirit and scope of the invention.
Claims (19)
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/987,343 US7049996B1 (en) | 2004-11-12 | 2004-11-12 | Method and apparatus of two stage scaling and quantization for coded communication systems |
CN200510115341.4A CN1790968B (en) | 2004-11-12 | 2005-11-14 | Method and apparatus of two stage scaling and quantization for coded communication systems |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/987,343 US7049996B1 (en) | 2004-11-12 | 2004-11-12 | Method and apparatus of two stage scaling and quantization for coded communication systems |
Publications (2)
Publication Number | Publication Date |
---|---|
US20060103559A1 true US20060103559A1 (en) | 2006-05-18 |
US7049996B1 US7049996B1 (en) | 2006-05-23 |
Family
ID=36385727
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/987,343 Expired - Fee Related US7049996B1 (en) | 2004-11-12 | 2004-11-12 | Method and apparatus of two stage scaling and quantization for coded communication systems |
Country Status (2)
Country | Link |
---|---|
US (1) | US7049996B1 (en) |
CN (1) | CN1790968B (en) |
Families Citing this family (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7200194B2 (en) * | 2002-10-11 | 2007-04-03 | Spreadtrum Communications Corporation | Receiver signal dynamic range compensation based on received signal strength indicator |
US8205637B2 (en) * | 2009-04-30 | 2012-06-26 | Baker Hughes Incorporated | Flow-actuated actuator and method |
US8671974B2 (en) * | 2009-05-20 | 2014-03-18 | Baker Hughes Incorporated | Flow-actuated actuator and method |
US7967076B2 (en) * | 2009-05-20 | 2011-06-28 | Baker Hughes Incorporated | Flow-actuated actuator and method |
US8047293B2 (en) * | 2009-05-20 | 2011-11-01 | Baker Hughes Incorporated | Flow-actuated actuator and method |
US10477249B2 (en) * | 2009-06-05 | 2019-11-12 | Apple Inc. | Video processing for masking coding artifacts using dynamic noise maps |
US8826108B1 (en) * | 2010-11-03 | 2014-09-02 | Marvell International Ltd. | Pre-scaling soft-decoder inputs |
CN104601274B (en) * | 2013-11-04 | 2018-07-13 | 芯光飞株式会社 | Demapping device and method |
US10868561B2 (en) * | 2017-11-02 | 2020-12-15 | B.G. Negev Technologies And Applications Ltd., At Ben-Gurion University | Digital resolution enhancement for high speed digital-to-analog converters |
CN112564859A (en) * | 2020-11-24 | 2021-03-26 | Oppo广东移动通信有限公司 | Soft bit quantization method, device, terminal and storage medium |
CN113206673B (en) * | 2021-05-24 | 2024-04-02 | 上海海事大学 | Differential scaling method and terminal for signal quantization of networked control system |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5543844A (en) * | 1992-11-25 | 1996-08-06 | Matsushita Electric Industrial Co., Ltd. | Method and apparatus for coding image data |
US5959675A (en) * | 1994-12-16 | 1999-09-28 | Matsushita Electric Industrial Co., Ltd. | Image compression coding apparatus having multiple kinds of coefficient weights |
US20030202582A1 (en) * | 2002-04-09 | 2003-10-30 | Canon Kabushiki Kaisha | Image encoder, image encoding method, image encoding computer program, and storage medium |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6047035A (en) * | 1998-06-15 | 2000-04-04 | Dspc Israel Ltd. | Method and device for quantizing the input to soft decoders |
CN1120577C (en) * | 1999-07-06 | 2003-09-03 | 华为技术有限公司 | Method for estimating code error rate of wireless channel in digital mobile communication system |
US6792055B1 (en) * | 2000-07-21 | 2004-09-14 | Rockwell Collins | Data communications receiver with automatic control of gain or threshold for soft decision decoding |
-
2004
- 2004-11-12 US US10/987,343 patent/US7049996B1/en not_active Expired - Fee Related
-
2005
- 2005-11-14 CN CN200510115341.4A patent/CN1790968B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5543844A (en) * | 1992-11-25 | 1996-08-06 | Matsushita Electric Industrial Co., Ltd. | Method and apparatus for coding image data |
US5959675A (en) * | 1994-12-16 | 1999-09-28 | Matsushita Electric Industrial Co., Ltd. | Image compression coding apparatus having multiple kinds of coefficient weights |
US20030202582A1 (en) * | 2002-04-09 | 2003-10-30 | Canon Kabushiki Kaisha | Image encoder, image encoding method, image encoding computer program, and storage medium |
Also Published As
Publication number | Publication date |
---|---|
US7049996B1 (en) | 2006-05-23 |
CN1790968A (en) | 2006-06-21 |
CN1790968B (en) | 2010-05-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP1221772B1 (en) | Pre-decoder for a turbo decoder, for recovering punctured parity symbols, and a method for recovering a turbo code | |
EP1290821B1 (en) | Methods and systems for decoding a received signal having a transmitter or channel induced coupling between bits | |
JP4101653B2 (en) | Scaling demodulated data in interleaver memory | |
JP2596392B2 (en) | Data rate detector | |
JP4194015B2 (en) | Communication system and method | |
US6381728B1 (en) | Partitioned interleaver memory for map decoder | |
US7159168B2 (en) | Iterative decoding with likelihood weighting | |
KR19990080592A (en) | And apparatus for channel coding / decoding using an iterative decoding controller of a communication system | |
CN101529884A (en) | Sliding window method and apparatus for soft input/soft output processing | |
JPH08195683A (en) | Data receiving device | |
US7049996B1 (en) | Method and apparatus of two stage scaling and quantization for coded communication systems | |
US6662331B1 (en) | Space-efficient turbo decoder | |
US6359935B1 (en) | Method for iterative demodulation and decoding for a system with coding and differential demodulation | |
KR20020064337A (en) | Turbo-code decoder and turbo-code decoding method with iterative channel parameter estimation | |
EP1612947B1 (en) | Method to adaptive scale the input to a channel decoder | |
AU758505B2 (en) | Method and system for fast maximum a posteriori decoding | |
US8000403B2 (en) | Coding device, decoding device, transmitter and receiver | |
US7734983B2 (en) | Input control device and input control method | |
US7634703B2 (en) | Linear approximation of the max* operation for log-map decoding | |
US8284874B2 (en) | Soft decision processing | |
Orten et al. | Sequential decoding in future mobile communications | |
US6263474B1 (en) | Data decoding apparatus and data decoding method | |
JP2002217746A (en) | Method and system for calculating branch metric used by soft decision decoding algorithm | |
EP1455457A1 (en) | Simple early stopping criterion for maximum likelihood soft output decoding algorithms | |
US20050086577A1 (en) | Method and arrangement for enhancing search through trellis |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: VIA TELECOM CO., LTD., CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SHEN, QIANG;REEL/FRAME:015998/0097 Effective date: 20041110 |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
AS | Assignment |
Owner name: INTEL CORPORATION, CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:VIA TELECOM CO., LTD.;REEL/FRAME:037096/0075 Effective date: 20151020 |
|
FEPP | Fee payment procedure |
Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FEPP | Fee payment procedure |
Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.) |
|
LAPS | Lapse for failure to pay maintenance fees |
Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.) |
|
STCH | Information on status: patent discontinuation |
Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362 |
|
FP | Lapsed due to failure to pay maintenance fee |
Effective date: 20180523 |