CN107911122A - Lossless compression method for distributed optical fiber vibration sensing data based on decomposition and compression - Google Patents

Abstract

The invention discloses a method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression. The invention performs discrete cosine transformation on the distributed optical fiber vibration sensing data, extracts the energy main part of the discrete cosine coefficient, and reversely After the discrete cosine transform and the difference with the original data, the energy main coefficient and the difference are respectively subjected to linear predictive coding, and finally entropy coding is performed. The invention decomposes data into two parts with different characteristics and then compresses them, integrates a lossy compression strategy into a lossless compression framework, and improves the compression capability without data loss.

Description

Lossless compression method for distributed optical fiber vibration sensing data based on decomposition and compression

technical field

The invention relates to the field of optical fiber technology, in particular to a method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression.

Background technique

Strengthening border protection, improving energy security, and improving social security are the basic requirements for social stability and rapid economic development. The perimeter security monitoring of some important base facilities in various fields such as military defense, large-scale industrial and mining, and civil security is an effective means to avoid major economic losses and maintain social stability and development. With the continuous development of society, people's security awareness is constantly improving, and various security monitoring technologies are also constantly developing. The optical fiber intrusion sensor system based on optical time domain reflectometer OTDR has the advantages of distributed, high sensitivity, wide monitoring range, concealment, and no terrain restrictions. It has great potential in perimeter security intrusion monitoring and has become people's choice. Research hotspots.

Phase-sensitive Optical Time Domain Reflectometry (Phase-sensitive Optical Time Domain Reflectometry, Φ-OTDR) is developed on the basis of the original OTDR distributed sensor. It is a typical fully distributed optical fiber sensing technology with high sensitivity and passive throughout, and can continuously sense the spatial distribution and temporal change information of dynamic parameters such as strain and vibration on the transmission path.

In the actual vibration measurement of Φ-OTDR, due to its high sensitivity and fast response, it often generates a large amount of sensing data. In a typical Φ-OTDR system, assuming that the sampling rate of the data acquisition module is 100MSP/s, and the number of conversion bits is 14bit, then the data flow of the system is 100MSP/s×14bit=175MB/s. Such huge data is not convenient for transmission and storage, so it needs to be compressed. Most of the current technologies are lossy compression using discrete cosine transform or discrete wavelet transform. Lossy compression often achieves a high compression rate, but it will cause a part of the loss of the original data. In practical engineering applications, complete and accurate data collection is necessary.

Contents of the invention

The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression. This method decomposes the original data into two parts with different characteristics and then compresses them. , integrate the lossy compression strategy into the lossless compression framework, and improve the compression capability without data loss.

The present invention adopts the following technical solutions for solving the problems of the technologies described above:

A method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression proposed by the present invention comprises the following steps:

Step 1. Discrete cosine transform is performed on a set of original data m to obtain a set of discrete cosine coefficients M; where m={m _i |i is an integer and 1≤i≤N}={m ₁ ,m ₂ ,.. .,m _N }, m _i is the i-th data in m, N is the number of data contained in m, M={M _j |j is an integer and 1≤j≤N}={M ₁ ,M ₂ , ...,M _N }, M _j is the jth data in M;

Step 2. The total energy of M is Then h is the smallest integer in the interval [1, N] and satisfying the following formula:

Among them, r is the percentage of energy;

Obtain the frequency domain energy principal coefficient X, X=X ₁ , X ₂ ,...,X _N , X _j is the jth data in the frequency domain energy principal coefficient X,

Step 3: Perform inverse discrete cosine transform on X to obtain time-domain energy principal coefficient x, x=x ₁ , x ₂ ,...,x _N , x _i is the i-th data in x;

Step 4, perform differential operation on m and x to obtain difference d, d _i = m _{i -xi} , wherein, d _i is the i-th data in d;

Step 5. Perform linear predictive coding on X to obtain the prediction remainder PX of X; perform linear predictive coding on d to obtain the prediction remainder Pd of d;

Among them, PX _j is the jth data in PX;

Among them, Pd _i is the i-th data in Pd;

Step 6: Perform entropy encoding on PX to obtain entropy encoded compressed data EX of PX; perform entropy encoding on Pd to obtain entropy encoded compressed data Ed of Pd.

As a further optimization scheme of the decompression-based distributed optical fiber vibration sensing data lossless compression method described in the present invention, the compressed data obtained in step 6 is decoded, and the specific steps are as follows:

Step A, performing entropy decoding on EX to obtain PX; performing entropy decoding on Ed to obtain Pd;

Step B, performing linear predictive decoding on PX to obtain X; performing linear predictive decoding on Pd to obtain d;

Step C, performing an inverse discrete cosine transform on X to obtain x;

Step D, add x and d to get m.

As a further optimization scheme of the decompression-based distributed optical fiber vibration sensing data lossless compression method described in the present invention, the formula of the discrete cosine transform in step 1 is:

As a further optimization scheme of the method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression in the present invention, the original data in step 1 is the original distributed optical fiber vibration sensing data.

As a further optimization scheme of a method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression in the present invention, r is 95 in step 2.

As a further optimization scheme of the decompression-based distributed optical fiber vibration sensing data lossless compression method described in the present invention, the inverse discrete cosine transform formula in step 3 is:

As a further optimization scheme of the method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression described in the present invention, the linear predictive coding used in step five is 2nd-order linear predictive.

As a further optimization scheme of the method for lossless compression of distributed optical fiber vibration sensing data based on decomposition and compression in the present invention, the entropy coding adopted in step 6 is arithmetic coding.

Compared with the prior art, the present invention adopts the above technical scheme and has the following technical effects:

(1) The solution of the present invention is lossless compression without any loss of information, and can completely reconstruct the original data of distributed optical fiber vibration sensing, improving the accuracy and adaptability of the distributed optical fiber vibration sensing system;

(2) The present invention adopts a decomposition and compression strategy, which enhances the compression effect under the premise of lossless, and reduces the pressure of data transmission and storage in the distributed optical fiber vibration sensing system.

Description of drawings

Fig. 1 is a principle framework of lossless compression based on decomposition and compression; wherein, (a) is the encoding process, and (b) is the decoding process.

Figure 2 is the vibration signal of Φ-OTDR.

Fig. 3 is the discrete cosine coefficient of the vibration signal.

Detailed ways

Below in conjunction with accompanying drawing, technical scheme of the present invention is described in further detail:

Shown in (a) among Fig. 1, coding process of the present invention comprises the following steps:

Step 1. Discrete cosine transform is performed on a set of original data m to obtain a set of discrete cosine coefficients M; where m={m _i |i is an integer and 1≤i≤N}={m ₁ ,m ₂ ,.. .,m _N }, m _i is the i-th data in m, N is the number of data contained in m, M={M _j |j is an integer and 1≤j≤N}={M ₁ ,M ₂ , ...,M _N }, M _j is the jth data in M;

The formula for discrete cosine transform is:

Figure 2 and Figure 3 show the data curves before and after the discrete cosine transform respectively, and the transformed energy body is concentrated in the front section, indicating the excellent energy compaction characteristics of the discrete cosine transform.

Step 2, the total energy of M is Then h is the smallest integer in the interval [1, N] and satisfying the following formula:

Among them, r is the percentage of energy;

Obtain the frequency domain energy principal coefficient X, X=X ₁ , X ₂ ,...,X _N , X _j is the jth data in the frequency domain energy principal coefficient X,

Step 3, perform inverse discrete cosine transform on X to obtain time-domain energy principal coefficient x, x=x ₁ , x ₂ ,...,x _N , x _i is the i-th data in x;

The formula for the inverse discrete cosine transform is:

Step 4, perform difference operation between m and x to obtain difference d, d _i = m _{i -xi} , where d _i is the i-th data in d;

Step 5: Perform linear predictive coding on X to obtain the prediction remainder PX of X; perform linear predictive coding on d to obtain the prediction remainder Pd of d;

Among them, PX _j is the jth data in PX;

Among them, Pd _i is the i-th data in Pd;

Step 6: Perform entropy encoding on PX to obtain entropy encoded compressed data EX of PX; perform entropy encoding on Pd to obtain entropy encoded compressed data Ed of Pd.

Shown in (b) among Fig. 1, decoding process of the present invention comprises the following steps:

Step 1: Perform entropy decoding on EX to obtain PX; perform entropy decoding on Ed to obtain Pd;

Step 2: Perform linear predictive decoding on PX to obtain X; perform linear predictive decoding on Pd to obtain d;

Step 3, perform inverse discrete cosine transform on X to obtain x;

Step 4, add x and d to get m.

The performance of a lossless compression method is usually evaluated by the compression ratio. The compression ratio (CR) is defined by the following equation:

Among them, S _c is the size of the compressed data, S _o is the size of the original data.

As shown in Table 1, it is the compression rate of Φ-OTDR data of different types and groups, and the compression rate produced by using the scheme of the present invention to compress the Φ-OTDR data of different types and groups can be seen that the scheme of the present invention The Φ-OTDR data is effectively compressed.

Table 1

The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be assumed that the specific implementation of the present invention is limited to these descriptions. For those of ordinary skill in the technical field of the present invention, without departing from the concept of the present invention, some simple deductions or substitutions can be made, which should be regarded as belonging to the protection scope of the present invention.

Claims (8)

1. it is a kind of based on decompose compression distributed optical fiber vibration sensing data lossless compression method, it is characterised in that including with Lower step：

Step 1: one group of initial data m is carried out discrete cosine transform, one group of discrete cosine coefficient M is obtained；Wherein, m={ m_i|i For integer and 1≤i≤N }={ m₁,m₂,...,m_N, m_iIt is i-th of data in m, N is the data amount check included in m, M={ M_j|j For integer and 1≤j≤N }={ M₁,M₂,...,M_N, M_jIt is j-th of data in M；

Step 2: the gross energy of M isThen h is in [1, N] section and meets the minimum integer of following formula：

<mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>h</mi> </munderover> <msubsup> <mi>M</mi> <mi>j</mi> <mn>2</mn> </msubsup> <mo>&amp;GreaterEqual;</mo> <mi>E</mi> <mo>&amp;times;</mo> <mi>r</mi> <mi>%</mi> <mo>;</mo> </mrow>

Wherein, r is percentage of energy；

Obtain frequency domain energy main body coefficient X, X=X₁,X₂,...,X_N,X_jIt is j-th of data in frequency domain energy main body coefficient X,

<mrow> <msub> <mi>X</mi> <mi>j</mi> </msub> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msub> <mi>M</mi> <mi>j</mi> </msub> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mn>1</mn> <mo>&amp;le;</mo> <mi>j</mi> <mo>&amp;le;</mo> <mi>h</mi> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mi>h</mi> <mo>+</mo> <mn>1</mn> <mo>&amp;le;</mo> <mi>j</mi> <mo>&amp;le;</mo> <mi>N</mi> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>;</mo> </mrow>

Step 3: carrying out inverse discrete cosine transform to X, time domain energy main body coefficient x, x=x are obtained₁,x₂,...,x_N, x_iIt is in x I-th of data；

Step 4: m and x are made calculus of differences, difference d, d are obtained_i=m_i-x_i, wherein, d_iIt is i-th of data in d；

Step 5: carrying out linear predictive coding to X, the prediction remainder PX of X is obtained；Linear predictive coding is carried out to d, obtains the prediction of d Remainder Pd；

<mrow> <msub> <mi>PX</mi> <mi>j</mi> </msub> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>X</mi> <mi>j</mi> </msub> <mo>-</mo> <mn>2</mn> <msub> <mi>X</mi> <mrow> <mi>j</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mi>X</mi> <mrow> <mi>j</mi> <mo>-</mo> <mn>2</mn> </mrow> </msub> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mn>3</mn> <mo>&amp;le;</mo> <mi>j</mi> <mo>&amp;le;</mo> <mi>N</mi> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>;</mo> </mrow>

Wherein, PX_jIt is j-th of data in PX；

<mrow> <msub> <mi>Pd</mi> <mi>i</mi> </msub> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>d</mi> <mi>i</mi> </msub> <mo>-</mo> <mn>2</mn> <msub> <mi>d</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mi>d</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>2</mn> </mrow> </msub> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mn>3</mn> <mo>&amp;le;</mo> <mi>i</mi> <mo>&amp;le;</mo> <mi>N</mi> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>;</mo> </mrow>

Wherein, Pd_iIt is i-th of data in Pd；

Step 6: carrying out entropy coding to PX, the entropy coding compressed data EX of PX is obtained；Entropy coding is carried out to Pd, obtains the entropy coding of Pd Compressed data Ed.

It is 2. according to claim 1 a kind of based on the distributed optical fiber vibration sensing lossless date-compress side for decomposing compression Method, it is characterised in that decode, comprise the following steps that to the compressed data that step 6 obtains：

Step A, entropy decoding is carried out to EX, obtains PX；Entropy decoding is carried out to Ed, obtains Pd；

Step B, linear prediction decoding is carried out to PX, obtains X；Linear prediction decoding is carried out to Pd, obtains d；

Step C, inverse discrete cosine transform is carried out to X, obtains x；

Step D, x is added with d, obtains m.

It is 3. according to claim 1 a kind of based on the distributed optical fiber vibration sensing lossless date-compress side for decomposing compression Method, it is characterised in that the formula of discrete cosine transform is in step 1：

<mrow> <msub> <mi>M</mi> <mi>j</mi> </msub> <mo>=</mo> <msqrt> <mfrac> <mn>2</mn> <mi>N</mi> </mfrac> </msqrt> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>N</mi> </munderover> <msub> <mi>m</mi> <mi>i</mi> </msub> <mi>c</mi> <mi>o</mi> <mi>s</mi> <mo>&amp;lsqb;</mo> <mrow> <mo>(</mo> <mi>i</mi> <mo>+</mo> <mfrac> <mn>1</mn> <mn>2</mn> </mfrac> <mo>)</mo> </mrow> <mrow> <mo>(</mo> <mi>j</mi> <mo>+</mo> <mfrac> <mn>1</mn> <mn>2</mn> </mfrac> <mo>)</mo> </mrow> <mfrac> <mi>&amp;pi;</mi> <mi>N</mi> </mfrac> <mo>&amp;rsqb;</mo> <mo>.</mo> </mrow>

It is 4. according to claim 1 a kind of based on the distributed optical fiber vibration sensing lossless date-compress side for decomposing compression Method, it is characterised in that initial data is original distribution formula optical fiber vibration sensing data in step 1.

It is 5. according to claim 1 a kind of based on the distributed optical fiber vibration sensing lossless date-compress side for decomposing compression Method, it is characterised in that r is 95 in step 2.

It is 6. according to claim 1 a kind of based on the distributed optical fiber vibration sensing lossless date-compress side for decomposing compression Method, it is characterised in that inverse discrete cosine transform formula is in step 3：

<mrow> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>=</mo> <msqrt> <mfrac> <mn>2</mn> <mi>N</mi> </mfrac> </msqrt> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>N</mi> </munderover> <msub> <mi>X</mi> <mi>j</mi> </msub> <mi>c</mi> <mi>o</mi> <mi>s</mi> <mo>&amp;lsqb;</mo> <mrow> <mo>(</mo> <mi>i</mi> <mo>+</mo> <mfrac> <mn>1</mn> <mn>2</mn> </mfrac> <mo>)</mo> </mrow> <mrow> <mo>(</mo> <mi>j</mi> <mo>+</mo> <mfrac> <mn>1</mn> <mn>2</mn> </mfrac> <mo>)</mo> </mrow> <mfrac> <mi>&amp;pi;</mi> <mi>N</mi> </mfrac> <mo>&amp;rsqb;</mo> <mo>.</mo> </mrow>

It is 7. according to claim 1 a kind of based on the distributed optical fiber vibration sensing lossless date-compress side for decomposing compression Method, it is characterised in that the linear predictive coding that step 5 uses is 2 rank linear predictions.

It is 8. according to claim 1 a kind of based on the distributed optical fiber vibration sensing lossless date-compress side for decomposing compression Method, it is characterised in that the entropy coding that step 6 uses is the coding that counts.