A method and system for calculating mean value of signals sent by an IDMA system
By calculating the send signal mean by using hyperbolic tangent function and segmented polynomial fitting in the IDMA system, the problems of multiplication and insufficient accuracy in the fixed-point algorithm are solved, and the calculation of high-precision send signal mean on DSP/FPGA hardware is realized.
Patent Information
- Application Number
- CN202210066075.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-20
- Publication Date
- 2024-12-27
- Estimated Expiration
- 2042-01-20
AI Technical Summary
In IDMA systems, it is difficult for the prior art to accurately calculate the average of the transmitted signal on digital signal processing and field programmable gate array hardware, especially in fixed-point algorithms, where multiplication is cross-border and insufficient accuracy.
A method for calculating the sending signal mean value of IDMA system is proposed. By judging the information value after channel decoding and reinterleaving, the sending signal mean value is calculated using hyperbolic tangent function or segmented polynomial fitting, fixed-point polynomial multiplication and normalization processing is performed. , avoid multiplication and improve bit accuracy.
It realizes high-precision calculation of the sending signal average on DSP/FPGA hardware, avoids the problem of cross-border in fixed-point multiplication, and ensures the bit accuracy and accuracy of the algorithm.
Smart Images
Figure CN114491396B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of mobile communications, and in particular relates to a method and system for calculating a mean value of a signal sent by an IDMA system. Background Art
[0002] As a typical non-orthogonal multiple access (NOMA) method, Interleave Division Multiple Access (IDMA) technology can significantly improve the throughput of mobile communication systems. Compared with the traditional Code Division Multiple Access (CDMA) system, IDMA technology can effectively overcome Multiple Access Interference (MAI).
[0003] IDMA allocates different interleavers to different users to share the same time / frequency / space domain resources. Traditional IDMA transmission and reception systems, such as Figure 1 As shown in Figure 2, the interleaver as a user characteristic is usually generated by a random interleaver. The transmitter consists of several channel coding and spread spectrum modules, as well as different interleavers for each user. The information bit d of user k is k First, a low-rate encoder generates c k , c k The different user information x is interleaved with the interleaver specific to each user. k Sent to the wireless channel via the multiple access channel.
[0004] The receiving end adopts the chip-level iterative detection technology. The detection structure is mainly composed of the elementary signal estimator (ESE) and K single-user posterior probability (APP) decoders (DEC). The output information of ESE and channel decoding is about {x k The basic signal estimator ESE performs multi-user joint detection, that is, it uses the extrinsic Log-Likelihood Ratios (LLR) prior probability likelihood ratio information returned by other user channel decoding to estimate the log-likelihood ratio information of the target user signal.
[0005] Combination Figure 1 , in the case of a single-path wireless channel, the received signal is expressed as:
[0006] j=1,2,...,J;
[0007] Among them, h k is the channel impulse response of user k, which can be obtained through channel estimation; x k (j), k = 1, 2, ..., K is the transmitted signal of user k, J is the length of the user's transmitted signal, and K is the total number of users; n(j) represents the variance σ 2 =N0 / 2 additive Gaussian white noise.
[0008] The mathematical expectation E(x) of the signal sent by user k k (j)), expressed as:
[0009]
[0010] in, The information after channel decoding and re-interleaving is the input of the ESE module.
[0011] Since the squares of the positive and negative polarizations of the transmitted signal are equal, the transmitted signal x k (j) The variance is:
[0012] Var(x k (j)) = 1-(E(x k (j))) 2 ;
[0013] According to the formula of the received signal, the mean and variance of the received signal r(j) can be obtained:
[0014]
[0015]
[0016] In the ESE module, the kth user The posterior LLR of is expressed as:
[0017]
[0018] The chip-level ESE multi-user detection process first uses the information after channel decoding and re-interleaving Calculate the mean value E(x k (j)) and variance Var(x k (j)) Use the received signal to estimate the current wireless channel impulse response h k , update the received signal mean E(r(j)) and variance Var(r(j)) by combining the transmitted signal mean and variance, and finally calculate the kth user using the wireless channel impulse response, the transmitted signal and the received signal mean and variance The a posteriori LLR is used as the estimated value of the transmitted signal e ESE (x k (j)). The above process is repeated until the predetermined number of iterations is completed or the receiving end performance index is met.
[0019] The mathematical expectation E(x k (j)) requires the use of the hyperbolic tangent function tanh(·). How to use fixed-point algorithm to accurately calculate this value for the convenience of digital signal processing (DSP) / field-programmable gate array (FPGA) implementation is of great practical significance. Summary of the invention
[0020] In order to facilitate DSP / FPGA implementation, the present invention proposes a method and system for calculating the mean value of a signal sent by an IDMA system. The method is based on the information after channel decoding and re-interleaving. The values of are used to calculate the mean of the transmitted signal, including:
[0021] when When , the mean value of the transmitted signal is always 1, that is,
[0022]
[0023] Here, |·| means taking the absolute value.
[0024] when When the value is within the range, the signal is segmented and the mean value of each segment of the transmitted signal is calculated, which is expressed as:
[0025]
[0026] Wherein, s=1, 2, ..., S represents a sub-segment, and S represents the number of segments into which the signal is divided; ζ s,n Represents the coefficient of the s-th subsegment and n-th order polynomial, n = 1, 2, ..., N, and N represents the maximum order of the polynomial.
[0027] Furthermore, when the bit precision is M, The mean value of the transmitted signal in the value range is the sum of the multiplication results of each sub-segment rounded down. The multiplication result of each sub-segment is the product of the value obtained by dividing the left and right data of the current sub-segment by the maximum value of the current sub-segment, performing the vertex exponential operation and retaining the M-bit value, performing the exponential operation on the maximum value of the current sub-segment and multiplying the value obtained by the corresponding coefficient of the current sub-segment, that is:
[0028]
[0029] in,(·) s,max represents the maximum value of the function in the s subsegment, Indicates rounding down. When , the mean value of the transmitted signal is expressed as:
[0030]
[0031] The present invention also proposes a signal mean value calculation system for an IDMA system, characterized in that the system comprises a first judgement device, a hyperbolic tangent function calculator, a second judgement device and a segmentation calculator, wherein the first judgement device calculates the mean value of a signal transmitted by the IDMA system according to the information after channel decoding and re-interleaving. The value of determines how to calculate the mean value of the transmitted signal. Then use the hyperbolic tangent function calculator to calculate the mean value of the transmitted signal, expressed as: like The average value of the transmitted signal is always 1, that is, E(x k (j)) = 1; if Then the segment calculator is used to calculate the mean value of the transmitted signal, which is expressed as:
[0032] Wherein, s=1, 2, ..., S represents a sub-segment, and S represents the number of segments into which the signal is divided; ζ s,n Represents the coefficient of the s-th subsegment and n-th order polynomial, n = 1, 2, ..., N, and N represents the maximum order of the polynomial.
[0033] The present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor runs the computer program stored in the memory to send a signal mean value calculation method in any one of the aforementioned IDMA systems.
[0034] The present invention firstly utilizes the odd symmetry property of the hyperbolic tangent function to obtain the equivalent negative domain function value of the negative domain function value according to the positive domain function value of the independent variable, and then calculates the function value of each sub-segment based on the piecewise polynomial fitting idea, and finally calculates the fixed-point polynomial multiplication of each sub-segment after normalization according to the maximum value of different sub-segments to avoid the fixed-point multiplication out of bounds, and finally obtains the mathematical expectation fitting value of the signal sent by all users. The multi-segment fixed-point polynomial fitting algorithm not only has high bit accuracy, but also is easy to implement in DSP / FPGA. In summary, the present invention can not only avoid the problem of out of bounds in the fixed-point multiplication process, but also ensure the bit accuracy of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1 Schematic diagram of traditional IDMA sending and receiving system;
[0036] Figure 2 A flow chart of a method for calculating the mean value of a signal sent by an IDMA system of the present invention;
[0037] Figure 3 Comparison diagram of the approximate mean value and theoretical value of the signal sent by the method of the present invention;
[0038] Figure 4 The method of the present invention sends the absolute error between the approximate value of the mean value of the signal and the theoretical value;
[0039] Figure 5 The method of the present invention sends a relative error between an approximate value of a signal mean and a theoretical value. DETAILED DESCRIPTION
[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0041] The present invention proposes a method for calculating the mean value of a signal sent by an IDMA system, based on the information after channel decoding and re-interleaving. The values of are used to calculate the mean of the transmitted signal, including:
[0042] when When , the mean value of the transmitted signal is always 1, that is,
[0043] when When the value is within the range, the signal is segmented and the mean value of each segment of the transmitted signal is calculated, which is expressed as:
[0044]
[0045] when When , the mean value of the transmitted signal is expressed as: Wherein, |·| represents taking the absolute value; s=1,2,...,S represents the sub-segment, and S represents the number of segments into which the signal is divided; ζ s,n represents the coefficient of the nth order polynomial of the sth subsegment, n=1,2,...,N, N represents the maximum order of the polynomial; Indicates when The mean value of the transmitted signal is obtained; E(x k (j)) indicates The mean value of the transmitted signal is The negative number of the mean value of the transmitted signal obtained. Parameters S and N are usually calculated by the Matlab Cftool polynomial fitting toolbox based on the error between the mean value of the transmitted signal and the theoretical value. The larger the value, the smaller the error, but the higher the algorithm complexity. Under the premise that the parameters S and N are determined by simulation, the Matlab Cftool polynomial fitting toolbox is used to calculate the s-th sub-segment and n-th order polynomial coefficient ζ s,n .
[0046] This embodiment decodes the information according to the channel and interleaves it again The scope includes the following steps:
[0047] Step 1: When When , the mean value of the transmitted signal is always 1, that is,
[0048] Step 2: When When the value is within the range, the signal is segmented and the mean value of each segment of the transmitted signal is calculated, which is expressed as:
[0049] Step 3: When When , the mean value of the transmitted signal is expressed as:
[0050] Wherein, |·| represents taking the absolute value; s=1,2,...,S represents the sub-segment, and S represents the number of segments into which the signal is divided; ζ s,n represents the coefficient of the nth order polynomial of the sth subsegment, n=1,2,...,N, N represents the maximum order of the polynomial; Indicates when The mean value of the transmitted signal is obtained; E(x k (j)) indicates The mean value of the transmitted signal is The negative of the mean value of the transmitted signal.
[0051] Preferably, in order to facilitate DSP / FPGA hardware implementation and avoid the problem of out-of-bounds in the multiplication process, the following formula can be used for calculation:
[0052]
[0053] in,(·) s,max represents the maximum value of the function in the s subsegment, M represents the binary bit precision, Indicates rounding down. The order of polynomial multiplication in formula (4) is: first, all data of the current sub-segment are divided by the maximum value of the current sub-segment, and a fixed-point exponential operation is performed, and M bits are retained; secondly, the maximum value of the current sub-segment is subjected to an exponential operation and multiplied with the corresponding coefficient of the current sub-segment; the above two multiplication results are multiplied again to obtain the current factorial result of the current sub-segment.
[0054] Optionally, the order of multiplication of the polynomial in formula (4) is as follows: first, all data of the current sub-segment are divided by the maximum value of the current sub-segment, and an exponential operation is performed, and M bits are retained; secondly, the maximum value of the current sub-segment is subjected to an exponential operation and multiplied with the previous result; the current multiplication result is multiplied with the corresponding coefficient of the current sub-segment to obtain the current factorial multiplication result of the current sub-segment; all factorial multiplication results of the current sub-segment are accumulated and rounded down to obtain the final fixed-point transmission signal mean value E(x k (j)).
[0055] This embodiment also provides a more specific implementation method, such as Figure 1 As shown in the traditional IDMA sending and receiving end system, the total number of users K = 24, the length of the signal sent by each user J = 1024, the channel coding method uses 1 / 2 convolution coding, the interleaving method uses random interleaving, and the wireless channel uses a variance of σ 2 =N0 / 2 additive Gaussian white noise, the mean value E(x k (j)) is used to explain the calculation process in detail. The values are randomly distributed in each calculation process. To facilitate the description of the process, The value range of is calculated separately.
[0056] Step 1: When When , the mean value of the transmitted signal is always 1, that is,
[0057] Step 2: When When the value is within the range, the signal is segmented and the mean value of each segment of the transmitted signal is calculated, which is expressed as:
[0058]
[0059] exist Within the value range, it is further divided into 3 sub-segments, namely Then s = 3, the maximum value of each sub-segment (·) s,max are 0.1, 2, and 4 respectively. Using the matlab cftool toolbox, the average relative error rate of the current fitting data is within 1%, and the fitting third-order polynomial coefficient ζ s,n , then the mean value of each transmitted signal is simplified as:
[0060]
[0061] Preferably, in order to facilitate DSP / FPGA hardware implementation and avoid the problem of out-of-bounds in the multiplication process, when the binary bit precision M=16, the following formula can be used for calculation:
[0062]
[0063] in,(·) s,max Indicates the maximum value of the function in the s subsegment, M = 16 indicates the binary bit precision, Indicates rounding down. The order of multiplication of the above polynomial is: first, all the data of the current sub-segment are divided by the maximum value of the current sub-segment, and a fixed-point exponential operation is performed, and 16 bits are retained; secondly, the maximum value of the current sub-segment is exponentially operated and multiplied with the corresponding coefficient of the current sub-segment; the above two multiplication results are multiplied again to obtain the current factorial result of the current sub-segment.
[0064] Optionally, the order of multiplication of the above polynomial is: first, all data of the current sub-segment are divided by the maximum value of the current sub-segment, and an exponential operation is performed, and 16 bits are retained; secondly, the maximum value of the current sub-segment is subjected to an exponential operation and multiplied with the previous result; the current multiplication result is multiplied with the corresponding coefficient of the current sub-segment to obtain the current factorial multiplication result of the current sub-segment; all factorial multiplication results of the current sub-segment are accumulated and rounded down to obtain the final fixed-point transmission signal mean value E(x k (j)).
[0065] Step 3: When When , the mean value of the transmitted signal is expressed as:
[0066] in, Indicates when The average value of the transmitted signal obtained; E(x k (j)) The mean value of the transmitted signal is The negative of the mean value of the transmitted signal.
[0067] Figure 3 A comparison chart between the approximate mean value of the signal sent by the method of the present invention and the theoretical value is given. It can be seen that the mean value of the signal sent obtained by the method of the present invention can be well consistent with the theoretical value in the range of 0 to 6. Figure 4 The absolute error between the approximate value of the mean value of the signal sent by the method of the present invention and the theoretical value is given. Since the maximum value is 1, the maximum absolute error between the two does not exceed 0.0071, and the error between the two is small enough. Figure 5 The relative error between the approximate value of the mean value of the signal sent by the method of the present invention and the theoretical value is given. It can be seen that the maximum absolute error between the two does not exceed 0.12.
[0068] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A method for calculating the mean value of a signal sent by an IDMA system, characterized in that: According to the information decoded and interleaved again The values of are used to calculate the mean of the transmitted signal, including: when When , the mean value of the transmitted signal is always 1, that is, when When the value is within the range, the signal is segmented and the mean value of each segment of the transmitted signal is calculated, which is expressed as: when When , the mean value of the transmitted signal is expressed as: Wherein, |·| represents taking the absolute value; s=1,2,...,S represents the sub-segment, and S represents the number of segments into which the signal is divided; ζ s,n represents the coefficient of the nth order polynomial of the sth subsegment, n=1,2,...,N, N represents the maximum order of the polynomial; Indicates when The average value of the transmitted signal obtained; E(x k (j)) The mean value of the transmitted signal is The negative of the mean value of the transmitted signal.
2. The method for calculating the mean value of a signal sent by an IDMA system according to claim 1, characterized in that: When the bit precision is M, The mean value of the transmitted signal in the value range is the sum of the multiplication results of each sub-segment rounded up. The multiplication result of each sub-segment is the product of the value obtained by dividing the left and right data of the current sub-segment by the maximum value of the current sub-segment, performing vertex exponent operation and retaining M bits, and performing exponential operation on the maximum value of the current sub-segment and multiplying the value with the corresponding coefficient of the current sub-segment.
3. The method for calculating the mean value of a signal sent by an IDMA system according to claim 2, characterized in that: When the bit precision is M, The mean value of the transmitted signal of the sth sub-segment within the value range is expressed as: in,(·) s,max Indicates the maximum value of the function in the s subsegment.
4. The method for calculating the mean value of a signal sent by an IDMA system according to claim 1, characterized in that: Will be in The value range is divided into three sub-segments: If the average relative error rate of the current fitting data is within 1%, the mean value of each transmitted signal is expressed as:
5. The method for calculating the mean value of a signal sent by an IDMA system according to claim 4, characterized in that: When the binary bit precision M = 16, The average value of the signal sent by each of the three sub-segments within the value range is expressed as: in, Indicates rounding down.
6. An IDMA system sending signal mean value calculation system, characterized in that: The system comprises a first judge, a hyperbolic tangent function calculator, a second judge and a segment calculator, wherein the first judge is based on the information after channel decoding and re-interleaving. The value of determines how to calculate the mean value of the transmitted signal. Then use the hyperbolic tangent function calculator to calculate the mean value of the transmitted signal, expressed as: like The average value of the transmitted signal is always 1, that is, E(x k (j)) = 1; if Then the segment calculator is used to calculate the mean value of the transmitted signal, which is expressed as: Wherein, s=1, 2, ..., S represents a sub-segment, and S represents the number of segments into which the signal is divided; ζ s,n Represents the coefficient of the s-th subsegment and n-th order polynomial, n = 1, 2, ..., N, and N represents the maximum order of the polynomial.
7. The IDMA system sending signal mean value calculation system according to claim 6, characterized in that: When the bit precision is M, Within the value range, the mean value of the signal sent by the sth sub-segment is expressed as: in,(·) s,max Indicates the maximum value of the function in the s subsegment.
8. The IDMA system sending signal mean value calculation system according to claim 6, characterized in that: Will be in The value range is divided into three sub-segments: If the average relative error rate of the current fitting data is within 1%, the mean value of each transmitted signal is expressed as:
9. The IDMA system sending signal mean value calculation system according to claim 8, characterized in that: When the binary bit precision M = 16, The average value of the signal sent by each of the three sub-segments within the value range is expressed as: in, Indicates rounding down.
10. A computer device comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor runs the computer program stored in the memory to implement the method for calculating the mean value of a signal sent by an IDMA system as claimed in any one of claims 1 to 5.