WO2011137745A1

WO2011137745A1 - Method and device for transmitting iq data

Info

Publication number: WO2011137745A1
Application number: PCT/CN2011/073673
Authority: WO
Inventors: 宁科; 武杰; 钟政良; 李强
Original assignee: 华为技术有限公司
Priority date: 2010-05-06
Filing date: 2011-05-05
Publication date: 2011-11-10
Also published as: CN102238628A; CN102238628B

Abstract

A method for transmitting IQ data and the corresponding device are provided in the embodiments of the present invention. The method includes: determining the target sample point value for the k sample points in the IQ data to be transmitted, wherein k is an integer greater than or equal to 2; determining, according to the target sample point value, the public part of the k sample points and its index, and the mantissa part of the value of each sample point in the k sample points which relatives to the public part; transmitting the index of the public part and the mantissa part of each sample point through Common Public Radio Interface (CPRI) frames. Determining the common public part of multiple sample points and calculating the mantissa part of each sample point according to the public part can reduce the bits of the sample points, and enhance the transmission efficiency of the IQ data.

Description

IQ data transmission method and device The present application claims priority to Chinese Patent Application entitled "IQ Data Transmission Method and Apparatus", filed on May 6, 2010, the Chinese Patent Office, Application No. 201010172860. This is incorporated herein by reference. Technical field

The present invention relates to the field of wireless communication technologies, and in particular, to an IQ data transmission method and apparatus. Background technique

CPRI (Common Public Radio Interface) is a communication between REC (Radio Equipment Controller) and RE (Radio Equipment) in wireless base station equipment developed by several communication equipment manufacturers. Interface standard.

Taking the GSM system as an example, the existing IQ data transmission format of CPRI is shown in Fig. 1. Each sample point is represented by I and Q, and is directly interleaved and transmitted according to the quantization bit widths I and Q.

The inventors have found that the prior art has the following disadvantages: In the case of an increase in bandwidth or carrier, direct transmission of IQ data may cause a surge in traffic on the CPRI interface, and it is necessary to physically increase the number of CPRI transmission interfaces or increase the rate of CPRI so as to be able to adapt. The need for larger IQ data traffic. This leads to a significant increase in hardware costs. Summary of the invention

In view of this, an aspect of the present invention provides an IQ data transmission method, including: determining a target sample value for k samples in the I Q data to be transmitted, where k is an integer greater than or equal to 2;

Determining an index and a common portion of the common portion of the k samples according to the target sample value, and determining a value of each of the k samples relative to a mantissa portion of the common portion; The index of the common portion and the mantissa portion of the respective samples are transmitted through a common public wireless interface CPRI frame.

In still another aspect of the present invention, an IQ data transmission apparatus is provided, including:

a selection module, configured to determine a target sample value for k samples in the IQ data to be transmitted, where k is an integer multiple of 2;

a compression module, configured to determine an index and a common portion of the common portion of the k samples according to the target sample value, and determine a value of each of the k samples relative to a mantissa portion of the common portion ;

And a sending module, configured to send the index of the common part and the mantissa part of the respective samples through a universal public radio interface CPRI frame.

The above method and device determine a common common part for a plurality of samples, and calculate a mantissa portion of each sample according to the common portion, thereby reducing the number of bits required for the sample and improving the transmission efficiency of the IQ data. DRAWINGS

The drawings described herein are provided to provide a further understanding of the invention, and are in no way of limitation.

1 is an I Q data interleaving diagram of CPR I in the prior art;

2 is a schematic flowchart of a method according to another embodiment of the present invention;

Figure 3 is a comparison diagram of IQ data compression in the above embodiment;

4 is another comparison diagram of the I Q data compression in the above embodiment;

FIG. 5 is a schematic flowchart of a method according to another embodiment of the present invention;

FIG. 6 is a schematic diagram of a device according to still another embodiment of the present invention. detailed description

The objects, technical solutions and advantages of the embodiments of the present invention are more clearly understood. The embodiments of the present invention are further described in detail in the embodiments and the accompanying drawings. The illustrative embodiments of the present invention and the description thereof are intended to explain the present invention, but are not intended to limit the invention.

Referring to FIG. 2, a method according to an embodiment of the present invention includes:

S101: Determine a target sample value for k samples in the I Q data to be transmitted.

The k samples here are continuously transmitted IQ data, which can be regarded as one frame or one block.

k samples, which can refer to the total number of I data and Q data. For example, if k is equal to 8, it means that there are 4 I data and 4 Q data in the sample. Since I and Q data always appear in pairs, Therefore k is an integer multiple of 2.

Or k samples can be just k I data or k Q data. k is greater than or equal to 2, for example, k can take 2, 4, 6, 8, 10, etc., k samples can be consecutive k I data, or consecutive k Q data, or continuous k/2 I data And continuous k/2 Q data.

The target sample value is used for normalization and can be the absolute value of the sample with the largest absolute value among the k samples, or it can be larger than the value of the sample with the largest absolute value.

S102: Determine an index of a common part of the k samples according to the target sample value, and determine a value of each of the k samples relative to a mantissa part of the common part.

Each sample value can be expressed in scientific notation, for example, can be expressed in this form,

I(Q = Q''2 ^P (Formula, where I(Q) is the actual value of the I or Q sample, and ρ is the common part. You can use the target sample value to find an appropriate value as the Ρ, in the use The actual value of each sample is divided by 2 ^ρ to obtain the mantissa portion of each sample value.

Where ρ can be calculated from a function related to the bit width of the sample.

For example, _p= (exp ₊ i W ₊ ML-2-.W (formula _{Π )}

Among them, w can take 1, 05 or 0.25 and so on. W takes different values and can slightly improve the signal-to-noise ratio SNR performance of the signal. And exp is the index of the public part, the log of T is obtained by log2 to get exp. For example, it can be calculated using the following formula: 2( _2Μ )

Exp= ^Ceil( zW (Formula III)

Where T is the target sample value for normalizing each sample; Μ is the actual bit width of a single sample of the k samples, and N is the bit width of the exponent exp of the common portion , ce il means round up.

In order to better understand the above formula, let w take 1 , then the above formula can be evolved into: τ

Exp = ceil(log ₂ (- r)) - 1

I\Q ^, ) = I{Q)l 2 ^exp+ ' ^+M - ^L - ^lN . The I (Q) is the value of a single sample in the k samples, Γ ((, ) is the 1 ^ The fraction of a single sample relative to the mantissa of the common portion, the L being the bit width of the mantissa portion, and N being the bit width of exp.

The formula provided in this embodiment is only used to understand how to calculate the index, common part and mantissa part of the common part.

According to the nature of the logarithmic function, the above formula can be expressed as: exp = ceil(log ₂ T) - t

Where t is an adjustment factor used to adjust the exp size, which can be related to the compressed bit width of each sample value, and can be set according to actual performance requirements.

The public part actually adds or subtracts an adjustment factor to exp, which is also related to the actual bit width of the compressed sample.

When w is equal to other values, it can be obtained by performing the corresponding transformation according to the above formula, and details are not described herein again.

S103: Send the common part and the mantissa part through a universal public radio interface CPRI frame It can be executed by the CPRI sender, such as REC, or RE. The common part and the mantissa part are sent to the opposite end through the CPRI frame. The CPRI peer end can recover the corresponding data through the common part and the mantissa part in combination with the corresponding algorithm.

Exemplarily, the algorithm may be preset to the CPRI sender and the CPRI receiver, or may be negotiated and determined.

The method provided in this embodiment determines a common part for a plurality of samples, and calculates a mantissa part of each sample according to the common part. Since each sample share a common part, IQ data compression efficiency can be improved, thereby improving transmission efficiency and reducing hardware consumption.

The following is another embodiment provided by the present invention, and the basic steps are as follows:

Make 2n (n=l, 2, 3. . . ) sample points into a frame. The frame includes n I data, n Q data, and use the logarithmic function and the exponential function to compress the data to form a new one. The frame is transmitted, thereby reducing the amount of data transmitted by the IQ data stream.

Each sample point of each frame can extract the common part of all the samples by the logarithmic function, extract the mantissa part of each sample point by the exponential function, and transmit separately at the transmitting end, the exponent and the mantissa part of the common part and the common part. The calculation formulas are as follows: = ceil(log ₂ )) - 1

I(Q) = I'(Q') .2 ^{exp+ +M} - ^L - ^lN In the above formula, the meaning of each character is as follows:

I (Q) represents the actual value of the sample; (Q, ) represents the mantissa part of the sample compression, and exp represents the index of the common part of each sample, ^ex P ^{+ 1 + M} - - ^2W represents the common part; Max represents the absolute value of the sample with the largest absolute value in each sample in the formed frame; M represents the actual bit width of the sample, that is, the bit width of the sample before compression; L represents the bit position of the sample after compression width.

In the compressed frame, there are two parts, one is the exponent exp of the common part, and the other Each is a mantissa part; and each sample includes two parts, a common part and a mantissa part, wherein there is only one common part, which can be obtained by adding a coefficient according to the index of the common part, and the common part is shared by each sample, and The mantissa parts are different.

Referring to FIG. 3, FIG. 3 is a comparison diagram of compression using the method provided by this embodiment.

Among them, Fig. 301 is the actual sample point, and there are K samples in each frame (the IQ data appears in pairs, so the number of I samples is K/2 and the number of Q samples is K/2). As for lj (0: M-1 ), it means the jth I data, j = l , 2~k/2_l , (0: M_l ), the bit of the I data of the jth bit is 0 to M_l; (0: M-1 ) means the jth Q data, j = l , 2~k/2- 1 , and its bit is 0 to M-1.

After compression, a frame as shown at 302 is formed. Where exp (0: Nl) represents the compressed common part, (0: N-1) represents the compressed bit width; and Γ j (0: L-1) represents the jth I data after compression The mantissa part, j = l , 2~k/2-l , (0: L-1 ) indicates that the occupied bit width is 0 to L-1 tail; Q, j (0: L_l ) how to represent j The mantissa part of the Q data. Since the frame is compressed, the value of L may not match the value of M.

Referring to FIG. 4, FIG. 4 is another comparison diagram of compression by the method provided by the embodiment.

The meaning of the letter of Fig. 4 is consistent with the meaning of Fig. 3 above, where exp is represented by two lines, and (0: N/2-1) represents the value of the 0 to N/2 - 1 bit of the exponent exp of the common part, And (N/2-1: N-1) represents the value of the bit of the common part index N/2-1 to N-1.

At the receiving end, the data IQ data is decompressed, and the recovered signal may have some distortion, and the signal performance loss is small. In addition, since the transmission according to the new data frame may generate some calculation delay, it may be based on the actual transmission bandwidth. As well as the tolerance of delay and performance loss, select the parameters K, N, L in the formula.

It can be seen from the above description that by determining a common common part for the I data and the Q data to be transmitted, the bit width occupied by the IQ data can be maximized and the transmission efficiency can be improved. Further, by the conversion of the scientific notation, larger data can be expressed with smaller data, so that the bit width actually occupied by the IQ data can be further saved. And, after compression, EVM/SNR The performance changes are small and the data transfer rate can be increased.

Referring to FIG. 5, a method provided for another embodiment includes:

S401. The CPRI transmitting end obtains the absolute value of the sample with the largest absolute value from the k samples.

S403. The CPRI sender acquires an index of the common part of the k samples, a common part, and a mantissa part of each sample.

The exponent exp of the common part can be calculated by using the above formula I I I; then using the formula I I , substituting exp to calculate the common part p, substituting into formula I to calculate the mantissa part of each sample.

It can be understood that, in this embodiment, the common portion exp is a common index portion of each sample.

S405. The CPRI sender uses the CPRI frame to transmit the exponent and mantissa portions of the common part.

Based on the exponent and mantissa of the common part, the CPRI receiver can use the formula I and the formula I I to restore the values of the samples.

In the solution provided by this embodiment, by setting a common common part for the I sample and the Q sample, and using the mantissa part to represent each sample point, the bit width occupied by each sample point can be saved, and the transmission efficiency can be improved.

Another embodiment of the present invention further provides an IQ transmission apparatus, which can be used to execute the foregoing method embodiments, including: a selection module 601, a compression module 603, and a transmission module 605.

Each of the above modules is used to perform the steps corresponding to the foregoing method embodiments.

For example, the selection module 601 is configured to select a target sample value; the compression module 603 is configured to compress, and generate an exponential, common portion, and mantissa portion of the common portion; and the transmitting module 605 transmits the exponent and mantissa portions of the common portion through the CPRI frame. To the opposite end.

As an example, in the apparatus provided by this embodiment,

a selection module 601, configured to determine a target sample value for the k samples in the IQ data to be transmitted, where the k is greater than or equal to 2;

a compression module 603, configured to determine an index and a common part of the common part of the k samples according to the target sample value, and determine a value of each sample of the k samples relative to a mantissa of the common part section; The sending module 605 is configured to send the common part and the mantissa part through a universal public radio interface CPRI frame.

The compression module 603, specifically the above formula I, the formula I I and the formula I I I , calculates the exponent exp, the common part p and the mantissa part except the common part. For the meanings and values of the parameters in the formula, refer to the description of the foregoing method embodiments, and no further details are provided.

The apparatus provided in this embodiment determines a common common part for the I data and the Q data, and calculates a mantissa part of each sample according to the common part. Therefore, the bits occupied by the I data or the Q data can be greatly reduced, and the transmission efficiency of the IQ data can be improved.

Further, the common part and the mantissa part can be calculated by the above formula I or formula II, the common part is calculated by the logarithmic function, and the mantissa part is calculated by the exponential function, and larger data can be expressed by the smaller data, which can further save The bit width actually occupied by the IQ data.

The steps of a method or algorithm described in the above-disclosed embodiments can be implemented directly in hardware, a software module executed by a processor, or a combination of both. Software modules can be placed in random access memory (RAM), memory, read only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or technical field. Any other form of storage medium known.

The above is only the preferred embodiment of the present invention, and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection of the present invention. Within the scope.

Claims

Rights request

1. An IQ data transmission method, comprising:

Determining a target sample value for k samples in the I Q data to be transmitted, k being an integer greater than or equal to 2;

Determining an index and a common portion of the common portion of the k samples according to the target sample value, and determining a value of each of the k samples relative to a mantissa portion of the common portion;

The index of the common portion and the mantissa portion of the respective samples are transmitted through a general public radio interface CPRI frame.

1. The method according to claim 1, wherein the common portion and the mantissa portion satisfy the following conditions:

Wherein I (Q) is a value of an I sample or a Q sample in the k samples, and the Γ (Q, ) represents a tail portion of the I sample or the Q sample, and p is The public part.

3. The method according to claim 2, wherein said p = (exp + l) - W + ML - 2 ^{N -} W ^ 2 ( 2 _Μ )

The exp= ^eeil ( ^zW , the exp is an index of a common part, the T is the target sample value, and the M is an actual bit width of a single sample of the k samples, the N For the bit width of the exp, ceil represents rounding up, 1 is an adjustment coefficient, L is the bit width of the mantissa portion, and w is a weight value.

4. The method of claim 3, wherein the target sample value is greater than or equal to an absolute value of any of the k samples.

5. The method according to any one of claims 1 to 4, wherein the k samples comprise k I points, or k Q points, or k/2 I points and k/2 Qs point.

6. An IQ data transmission device, comprising:

7. The apparatus according to claim 6, wherein the common portion and the mantissa portion satisfy the following conditions:

8. The method according to claim 7, wherein said p = (exp + l) - W + ML - 2 ^{N -} W ^ 2 ( 2 _Μ )

9. The apparatus of claim 8, wherein the target sample value is greater than or equal to an absolute value of any of the k samples.

10. The apparatus according to any one of claims 6-9, wherein the k samples comprise k I points, or k Q points, or k/2 I points and k/2 Qs point.