CN102299717A - Research method of special administrative region (SAR) primary data compression error based on quantized interval transition model - Google Patents
Research method of special administrative region (SAR) primary data compression error based on quantized interval transition model Download PDFInfo
- Publication number
- CN102299717A CN102299717A CN2010102184285A CN201010218428A CN102299717A CN 102299717 A CN102299717 A CN 102299717A CN 2010102184285 A CN2010102184285 A CN 2010102184285A CN 201010218428 A CN201010218428 A CN 201010218428A CN 102299717 A CN102299717 A CN 102299717A
- Authority
- CN
- China
- Prior art keywords
- mrow
- adc
- noise
- msub
- baq
- 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.)
- Pending
Links
- 230000007704 transition Effects 0.000 title claims abstract description 25
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000011160 research Methods 0.000 title claims abstract description 9
- 238000013144 data compression Methods 0.000 title claims abstract description 5
- 238000013139 quantization Methods 0.000 claims abstract description 60
- 238000007906 compression Methods 0.000 claims abstract description 29
- 230000006835 compression Effects 0.000 claims abstract description 29
- 238000010606 normalization Methods 0.000 claims abstract description 26
- 230000008569 process Effects 0.000 claims abstract description 9
- 238000012545 processing Methods 0.000 claims abstract description 7
- 238000013507 mapping Methods 0.000 claims description 13
- 229920006395 saturated elastomer Polymers 0.000 claims description 4
- 230000007246 mechanism Effects 0.000 abstract description 4
- 238000011156 evaluation Methods 0.000 abstract description 3
- 230000003044 adaptive effect Effects 0.000 abstract description 2
- 238000002592 echocardiography Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention discloses a research method of a satellite borne special administrative region (SAR) primary data compression error mechanism based on a quantized interval transition model. The research method is applied to compression evaluation of an original data field and comprises the following steps: regarding the processing process of an analog to digital converter (ADC) and block adaptive quantization (BAQ) as a whole; establishing the quantized interval transition model; calculating a quantized interval transition probability caused by a system thermal noise, and solving a total normalization noise power comprising a system thermal noise power, a quantized noise power and a saturation noise power, wherein the quantized interval of each normalization ADC corresponds to a BAQ reconstructed level, and when the system noise exists, a normalization echo signal can transit in the BAQ quantized interval at the adjacent left sides or right sides under the influence of noise, so as to generate a larger quantized noise; and respectively calculating the left or right transition probability of the signal in each normalization ADC quantized interval, so as to obtain the total normalization noise power; and calculating the original data signal to noise ratio after BAQ compression by taking an ADC input signal as a standard.
Description
Technical Field
The invention relates to a research method for a compression error mechanism of satellite-borne SAR (synthetic aperture radar) original data based on a quantization interval transition model, which is applied to original data domain compression evaluation and mainly aims to: meanwhile, system thermal noise, quantization noise and saturation noise are considered, a quantization interval transition model is established, a mapping relation between an input signal-to-noise ratio (ADC) of the satellite-borne SAR raw data on a saturation corpus and a signal-to-noise ratio (BAQ) output signal-to-noise ratio (BAQ) is solved theoretically, and a theoretical basis is provided for selection of a BAQ compression ratio of the satellite-borne SAR and subsequent application analysis.
Background
The original echo signals received by the satellite-borne SAR are usually subjected to two quantization processes. The first time is ADC quantification, and two paths of signals of an I/Q satellite are quantified into Mbit data; the second time is to compress Mbit data to a smaller number of bits in order to meet the real-time data rate constraints.
The compression algorithms currently in common use in engineering are BAQ and fbaq (flexible block adaptive quantization). The BAQ algorithm is proposed by jpl (jet Propulsion laboratory) laboratories, and has been successfully applied to many satellite-borne SAR systems, but the evaluation of the algorithm is still incomplete, and the influence of different compression ratios on the SAR image quality and subsequent applications cannot be theoretically given.
SAR raw data has high information entropy, and ADC quantization and BAQ compression both cause loss of information quantity. In order to match the dynamic range of the echo with that of the ADC to reduce the quantization loss of the ADC, currently, the SAR performs receiver Gain control by using agc (automatic Gain control) and mgc (manual Gain control). Although the dynamic range of the echo can be improved in most cases, there are some limitations. When the contrast of the target changes sharply, for example, at the sea-land junction, weak echoes caused by the time delay of AGC processing are suppressed or strong echoes exceed the dynamic range of the ADC, a truncation effect is generated, data saturation is caused, and the information amount loss is large. For BAQ compression, the larger the compression ratio, the larger the information amount loss, and the worse the quality of the imaged image.
Therefore, on the premise of meeting the application requirements of the system, how to make a compromise between the raw data compression ratio and the imaging quality is an important problem of system design, and the mapping relation between the input signal-to-noise ratio of the raw data domain ADC and the output signal-to-noise ratio of the BAQ is just a bridge for analyzing the problem.
Disclosure of Invention
The invention provides a research method for a mechanism of compression errors of satellite-borne SAR original data based on a quantization interval transition model, which is used for evaluating a compression algorithm in an original data domain more effectively and further providing a theoretical basis for selection of a compression ratio.
In order to achieve the purpose, the invention provides a research method for a compression error mechanism of satellite-borne SAR original data based on a quantization interval transition model, wherein the research method comprises the following steps:
the ADC and BAQ processing process is considered as a whole;
because the BAQ compression normalization adopts the standard deviation of an ADC input signal, the echo is subjected to ADC quantization and then normalization, which is equivalent to the process of firstly performing normalization on the echo and then performing quantization by adopting a normalized ADC;
establishing a quantization interval transition model, calculating the quantization interval transition probability caused by system thermal noise, and solving normalized total noise power including the system thermal noise, the quantization noise and the saturation noise;
the quantization interval of each normalization ADC corresponds to a BAQ reconstruction level, and when system thermal noise exists, the normalization echo signal is influenced by the noise and can jump into the adjacent BAQ quantization interval on the left side or the right side to generate larger quantization noise;
and respectively calculating the probability of signal transition to the left or the right in each normalization ADC quantization interval to obtain the normalization total noise power, and calculating the signal-to-noise ratio of the original data after BAQ compression by taking the ADC input signal as a standard.
The method comprises the following steps: the threshold level and the reconstruction level of the normalized ADC are respectively equal to the values of the threshold level and the reconstruction level of the ADC normalized by adopting the standard deviation of the input signal of the ADC.
In the method, when different compression ratios are adopted, the mapping relation between the ADC input signal-to-noise ratio and the BAQ output signal-to-noise ratio of the satellite-borne SAR original data on the saturation corpus is as follows:
when the ADC is unsaturated, establishing a mapping relation among an ADC input signal standard deviation, an ADC input signal-to-noise ratio and a BAQ output signal-to-noise ratio;
and when the ADC is saturated, establishing a mapping relation among signal saturation, ADC input noise standard deviation and BAQ output signal-to-noise ratio. Wherein, the signal saturation and the standard deviation of the ADC input noise can represent the ADC input signal-to-noise ratio.
In other words, the technical solution adopted by the present invention to solve the technical problem is: the processing of the ADC and BAQ is considered as a whole, taking into account system thermal noise, quantization noise and saturation noise. And calculating the total noise power and the signal-to-noise ratio after BAQ compression by establishing a quantization interval transition model.
The specific operation comprises the following steps: the normalization of BAQ compression adopts ADC input signal standard deviation, so that when the total noise power is calculated, the echo is firstly subjected to ADC quantization and then normalization, which is equivalent to the process of firstly performing normalization on the echo and then performing quantization by adopting a normalization ADC. The quantization interval of each normalized ADC corresponds to the reconstruction level of a BAQ. When system thermal noise exists, the normalized echo signal is influenced by the noise and jumps to the BAQ quantization interval on the adjacent left side or right side, and large quantization noise is generated. And respectively calculating the probability of signal transition to the left or the right in each normalization ADC quantization interval to obtain the normalization total noise power, and calculating the signal-to-noise ratio of the original data after BAQ compression by taking the ADC input signal as a standard.
The method has the advantages that the mapping relation between the ADC input signal-to-noise ratio and the BAQ output signal-to-noise ratio of the original data domain is obtained theoretically when different compression ratios are adopted in the saturation complete set, and theoretical basis is provided for selection of the compression ratios in system design.
Drawings
The invention is further illustrated with reference to the following figures and examples.
FIG. 1 is a basic system model for raw data SNR solution.
Fig. 2 is a theoretical analytical model of the quantization process (Q1 for ADC quantization, Q2 for BAQ quantization).
FIG. 3 is a mapping relationship between ADC input signal standard deviation, ADC input signal-to-noise ratio and BAQ output signal-to-noise ratio when different BAQ compression ratios are adopted under the condition of ADC non-saturation.
Fig. 4 is a mapping relation among signal saturation, ADC input noise standard deviation and BAQ output signal-to-noise ratio when different BAQ compression ratios are adopted under ADC saturation.
Detailed Description
As shown in fig. 1, a SAR receives a raw echo signal x from a distributed scene, where the raw echo signal x includes a signal s and system thermal noise n. Usually, the area covered by the main lobe of the antenna directional diagram comprises a plurality of distributed targets, each transmitted pulse echo is formed by overlapping a plurality of independent scatterer echoes of the distributed targets, according to the central limit theorem, the real part and the imaginary part of s both accord with the mean value of zero, and the standard deviation is sigmasA gaussian distribution of (a). The real part and the imaginary part of n are consistent with the mean value of zero and the standard deviation of sigmanAnd s and n are statistically independent of each other, thus:
where D (x) is the variance of the echo signal x. As shown in (1), the real part and the imaginary part of the echo signal x both have a mean value of 0 and a standard deviation of 0
A gaussian distribution of (a).
Because the original echo I/Q signals have the same statistical rule, only the I signal is considered in the subsequent analysis, and the analysis of the signal-to-noise ratio is not influenced, so that the signal-to-noise ratio of the original echo signal is as follows:
during quantization processing, the signal and the noise are considered separately, and after normalization, the standard deviation of the signal is:
the standard deviation of the noise is:
assuming Mbit quantization of the original echo by the ADC, the threshold level xi＝i，i＝-2M-1+1，…，2M-1-1, quantization interval of 1, quantization interval [ x ]i，xi+1]Has a reconstruction level of xi+0.5, when x ∈ (- ∞, -2)M-1+1]∪[2M-1-1, ∞) the truncation levels are ± (2) respectivelyM-1-0.5); standard deviation of input signal through ADCNormalized, equivalent threshold level
i＝-2M-1+1，…，2M-1-1, quantization interval of
Reestablish the level as
A cutoff level ofThe quantized bit number of BAQ is N, and the threshold level is rm，m＝1，…，2N-1, quantization interval (- ∞, r)1]Has a reconstruction level of y1Quantization interval
At a reconstruction level ofIntermediate quantization interval [ r ]m，rm+1]Has a reconstruction level of ym+1. A theoretical analytical model of the quantification process is shown in fig. 2.
The normalized signal is in the interval [ nxi，nxi+1]And the normalized ADC output signal is located in the interval [ rm，rm+1]The reconstruction level after ADC and BAQ quantization is ym+1. Similarly, each normalized ADC quantization interval corresponds to a reconstruction level of the BAQ. The normalized overall quantization noise after ADC and BAQ quantization processing can thus be found to be:
when there is system thermal noise, interval [ nxi，nxi+1]The normalized echo signal in the region is influenced by noise and transits to p BAQ quantization regions on the adjacent right side or q BAQ quantization regions on the adjacent left side, so that the normalized echo signal is quantized into ym+1-q，…，ym，ym+2，…，ym+1+pAnd large quantization noise is generated. In order to solve the normalized noise power at this time, first, the probability of transition between echo signal intervals is obtained, and a specific solving method is as follows.
The echo signals and the system noise are in Gaussian distribution with the mean value of zero, and the probability density functions of the echo signals and the system noise are bilaterally symmetrical about the longitudinal axis of the coordinate axis, so that the probability of rightward transition and leftward transition of the echo signals in the interval on the left side of the coordinate axis is equal to the probability of leftward transition and rightward transition of the echo signals in the corresponding interval on the right side of the coordinate axis respectively.
Assuming that the normalized echo signal belongs to the interval [ nx ]i，nxi+1]，i＝-2M-1+1, …, -1, BAQ quantization interval corresponding to ADC output is [ rm，rm+1]At this time, the probability of the echo signal transitioning to the right by p sections is:
the probability of the echo signal jumping to the left by q intervals is:
wherein,
then, the interval [ nxi，nxi+1]The normalized noise power of the inner echo signal is:
for interval
Inner signal, the probability of transition to the right and left is:
the normalized noise power generated in the interval is:
thus, the total normalized noise power of the echo signal is:
from this, the SNR of the raw data after ADC and BAQ quantization can be obtainedBAQComprises the following steps:
specific examples are given below:
ADC adopts 8-bit quantization, BAQ adopts 4bit, 3bit and 2bi respectivelyt and 1bit quantization. According to the difference of the ADC input signal power, there are two situations of ADC non-saturation and ADC saturation. Establishing standard deviation sigma of ADC input signal when ADC is not saturatedsThe mapping relation between the ADC input signal-to-noise ratio and the BAQ output signal-to-noise ratio is shown in figure 3; when ADC is saturated, establishing signal saturation and ADC input noise standard deviation sigmanThe mapping relation between the BAQ output signal-to-noise ratio is shown in figure 4.
Wherein, the expression of the signal saturation is: <math>
<mrow>
<mfrac>
<mrow>
<msubsup>
<mo>∫</mo>
<mn>127.5</mn>
<mo>∞</mo>
</msubsup>
<mfrac>
<mn>1</mn>
<mrow>
<msqrt>
<mn>2</mn>
<mi>π</mi>
</msqrt>
<msub>
<mi>σ</mi>
<mi>s</mi>
</msub>
</mrow>
</mfrac>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<mo>-</mo>
<mfrac>
<msup>
<mi>x</mi>
<mn>2</mn>
</msup>
<mrow>
<mn>2</mn>
<msubsup>
<mi>σ</mi>
<mi>s</mi>
<mn>2</mn>
</msubsup>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mi>dx</mi>
</mrow>
<mn>0.5</mn>
</mfrac>
<mo>×</mo>
<mn>100</mn>
<mo>%</mo>
</mrow>
</math>
as can be seen from fig. 3 and 4, when the power of the echo signal is small and the input signal-to-noise ratio is low, the signal-to-noise ratios after 4-bit, 3-bit, 2-bit and 1-bit baq compression are not greatly different; when the echo is a medium power signal and the signal-to-noise ratio is high, the signal-to-noise ratio is deteriorated by about 5dB when the number of the compression bits is reduced by 1 bit; when the echo power is large and the ADC is saturated, the signal-to-noise ratio of quantized original data is reduced in a step mode, and the higher the saturation is, the more serious the signal-to-noise ratio is.
The above-described embodiments are not intended to limit the invention in any way, and all modifications that can be made according to the teachings of the present invention are within the scope of the invention.
Claims (3)
1. A SAR original data compression error research method based on a quantization interval transition model is disclosed, wherein:
the ADC and BAQ processing process is considered as a whole;
because the BAQ compression normalization adopts the standard deviation of an ADC input signal, the echo is subjected to ADC quantization and then normalization, which is equivalent to the process of firstly performing normalization on the echo and then performing quantization by adopting a normalized ADC;
establishing a quantization interval transition model, calculating the quantization interval transition probability caused by system thermal noise, and solving normalized total noise power including the system thermal noise, the quantization noise and the saturation noise;
the quantization interval of each normalization ADC corresponds to a BAQ reconstruction level, and when system thermal noise exists, the normalization echo signal is influenced by the noise and can jump into the adjacent BAQ quantization interval on the left side or the right side to generate larger quantization noise;
and respectively calculating the probability of signal transition to the left or the right in each normalization ADC quantization interval to obtain the normalization total noise power, and calculating the signal-to-noise ratio of the original data after BAQ compression by taking the ADC input signal as a standard.
2. The method of claim 1, wherein: the threshold level and the reconstruction level of the normalized ADC are respectively equal to the values of the threshold level and the reconstruction level of the ADC normalized by adopting the standard deviation of the input signal of the ADC.
3. The method of claim 1, wherein, when different compression ratios are adopted, the mapping relationship between the ADC input signal-to-noise ratio and the BAQ output signal-to-noise ratio of the SAR raw data on the saturation corpus is as follows:
when the ADC is unsaturated, establishing a mapping relation among an ADC input signal standard deviation, an ADC input signal-to-noise ratio and a BAQ output signal-to-noise ratio;
when the ADC is saturated, establishing a mapping relation among signal saturation, ADC input noise standard deviation and BAQ output signal-to-noise ratio; wherein, the signal saturation and the ADC input noise standard deviation are used for representing the ADC input signal-to-noise ratio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102184285A CN102299717A (en) | 2010-06-25 | 2010-06-25 | Research method of special administrative region (SAR) primary data compression error based on quantized interval transition model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102184285A CN102299717A (en) | 2010-06-25 | 2010-06-25 | Research method of special administrative region (SAR) primary data compression error based on quantized interval transition model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102299717A true CN102299717A (en) | 2011-12-28 |
Family
ID=45359910
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102184285A Pending CN102299717A (en) | 2010-06-25 | 2010-06-25 | Research method of special administrative region (SAR) primary data compression error based on quantized interval transition model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102299717A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103840894A (en) * | 2012-11-22 | 2014-06-04 | 中国科学院电子学研究所 | System SAR gain determination method for realizing optimal output signal-to-noise ratio |
-
2010
- 2010-06-25 CN CN2010102184285A patent/CN102299717A/en active Pending
Non-Patent Citations (2)
| Title |
|---|
| 祁海明，禹卫东: "基于实测数据的原始数据压缩对InSAR系统测高影响研究", 《电子与信息学报》 * |
| 祁海明，禹卫东: "针对饱和度全集上SAR原始数据自适应抗饱和BAQ压缩算法", 《自然科学进展》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103840894A (en) * | 2012-11-22 | 2014-06-04 | 中国科学院电子学研究所 | System SAR gain determination method for realizing optimal output signal-to-noise ratio |
| CN103840894B (en) * | 2012-11-22 | 2016-01-20 | 中国科学院电子学研究所 | A kind of SAR system gain defining method towards optimum output signal-to-noise ratio |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107923974A (en) | Method and apparatus for FMCW radar processing | |
| CN115001558B (en) | P code signal regeneration forwarding method of GPS based on W code real-time estimation | |
| KR101677991B1 (en) | A raw data compression method by performing a block adaptive quantization in synthetic aperture radar thereof | |
| CN102299717A (en) | Research method of special administrative region (SAR) primary data compression error based on quantized interval transition model | |
| US9923752B2 (en) | Method for demodulation and demodulator | |
| KR102081373B1 (en) | Method and apparatus for determining compression ratio of raw data in sar(synthetic aperture radar) | |
| CN117741663A (en) | Satellite-borne SAR azimuth multichannel quantization method and device | |
| CN114076945A (en) | Method for establishing ionized layer dispersion effect analytic model of NLFM signal | |
| CN107607926B (en) | Method for detecting and processing low-traffic quasi-signal fusion target of distributed radar | |
| CN101398486B (en) | Azimuth flow processing method and device of spaceborne SAR raw data compression | |
| CN113783575B (en) | Efficient data compression method based on FPGA | |
| CN113630120B (en) | Zero delay communication method combined with 1-bit analog-to-digital converter and application thereof | |
| CN110191430B (en) | Single-bit distributed sparse signal detection method for generalized Gaussian distribution situation | |
| CN102393517A (en) | Density weighting-based sparse synthetic aperture radar imaging method with utilization of simulated annealing | |
| CN109085575B (en) | Signal processing method of OFDM radar communication integrated airborne platform system | |
| US11223399B2 (en) | Base station and modulation method supporting lattice-partition-based non-orthogonal multiple access | |
| DE102005019176A1 (en) | Radio receiver for radio transmission system, has demodulation unit that sets weighting of least significant softbit based on estimated noise power and sets weighting of most significant softbit based on estimated signal power | |
| CN110380803B (en) | Self-adaptive power distribution method based on LDPC code transmission | |
| CN107809399A (en) | A kind of multiple antennas millimeter wave channel estimation methods for quantifying reception signal | |
| CN118311572B (en) | Fine imaging method and system for double-channel single-bit radar | |
| CN114465853B (en) | Multi-symbol incoherent detection method based on neural network | |
| Qi et al. | Performance evaluation of amplitude-phase algorithm for SAR raw data compression | |
| Agrawal et al. | SAR signal processing algorithms | |
| Zhang et al. | Segmented BAQ algorithm for SAR raw data of strong target in weak background | |
| CN102820955A (en) | Frequency-spectrum sensing information non-uniform quantizing method of secondary user nodes |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20111228 |