Particle Swarm Optimization Peak-to-Average Power Ratio Suppression Method Based on OFDM Signal Clipping Noise

Abstract

The particle swarm optimization peak-to-average power ratio (PAPR) suppression method based on the clipping noise of OFDM signals of the present invention has the following specific process: First, perform digital clipping operation on the digital baseband signal after upsampling and filtering on the transmitter using the clipping method; Second, subtract the clipped digital baseband signal from the original digital baseband signal to obtain the clipping noise; Third, initialize the particle swarm parameters, use the particle swarm optimization algorithm to iteratively optimize the initial data, obtain the global optimal solution after reaching the maximum number of iterations, and use the global optimal solution as the optimal particle swarm optimization noise; Finally, add the optimal particle swarm optimization noise and the clipping noise together in the reverse direction to the original digital baseband signal for PAPR suppression. The method of the present invention solves the problems that the existing methods for suppressing PAPR will cause excessive error vector magnitude (EVM) and insufficient PAPR suppression ability.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and particularly relates to a particle swarm optimization peak-to-average power ratio (PAPR) suppression method based on OFDM signal clipping noise. Background Art

[0002] The high PAPR characteristic of an OFDM system requires that AD / DA conversion devices, power amplification devices, etc. must have a linear range greater than the interval occupied by the OFDM peak and average power to ensure that the OFDM time-domain signal is not distorted. Especially for the power amplifier in the radio frequency front-end, to ensure a relatively high linear range, one method is to use pre-distortion, and the other is only to use the power back-off technology. Using the pre-distortion method has a high technical difficulty and high implementation cost, so generally the power back-off technology is directly adopted. If the peak-to-average power ratio of the OFDM system is relatively high, using the power back-off technology will result in very low power amplifier efficiency. Suppose the average transmission power of the OFDM system is 20W (43dBm) and the peak-to-average power ratio is 16dB. Then the power amplifier in the radio frequency front-end must use a power amplifier of more than 800W to ensure distortion-free signal transmission. It can be seen that when using a power amplifier tube of more than 800W to transmit 20W of power, its efficiency is only 2.5%, and most of the power consumption is only used for heat dissipation, greatly increasing the thermal design difficulty of the system. For the same system, suppose the peak-to-average power ratio is 10dB. Then the system only needs to use a 200W power amplifier tube, and the transmission efficiency at this time is about 10%, which can greatly reduce the thermal design difficulty of the system. Therefore, for an OFDM system, reducing the peak-to-average power ratio (PAPR) is a very meaningful thing.

[0003] To suppress the high PAPR in an OFDM system, one of the commonly used methods is clipping limit. A threshold is set, and the signal with an amplitude exceeding the threshold is hard-limited. This method is often used in actual projects because of its simple implementation method and low complexity. However, because this method is hard-limiting, for an OFDM signal with a relatively large number of subcarriers, its PAPR is relatively high. Therefore, in some systems with relatively high requirements for PAPR suppression, the clipping operation often makes the average power of the clipped signal worse than that before clipping by more than one time, resulting in an excessive error vector magnitude (EVM). At the same time, the rapid decrease in the average power will also, to a certain extent, make the PAPR suppression effect worse. Obviously, the clipping noise of the clipping method is not perfect.

[0004] Therefore, there is an urgent need to design a PAPR suppression method applicable to an OFDM system. Summary of the Invention

[0005] The object of the present invention is to provide a particle swarm optimization peak-to-average power ratio (PAPR) suppression method based on OFDM signal clipping noise, so as to solve the problems that the existing PAPR suppression methods will cause excessive error vector magnitude (EVM) and insufficient PAPR suppression ability.

[0006] The problems of poor PAPR suppression effect and large EVM loss caused by too low average power due to the clipping method.

[0007] The technical solution adopted by the present invention is a particle swarm optimization PAPR suppression method based on OFDM signal clipping noise, which is specifically implemented according to the following steps:

[0008] Step 1, perform digital clipping operation of the clipping method on the digital baseband signal after upsampling and filtering on the transmitter;

[0009] Step 2, subtract the digital baseband signal after clipping in Step 1 from the original digital baseband signal to obtain the clipping noise;

[0010] Step 3, initialize the particle swarm parameters, use the particle swarm optimization algorithm to iteratively optimize the initial data, obtain the global optimal solution after reaching the maximum number of iterations, and use the global optimal solution as the optimal particle swarm optimization noise;

[0011] Step 4, add the optimal particle swarm optimization noise obtained in Step 3 and the clipping noise obtained in Step 2 in the reverse direction to the original digital baseband signal for PAPR suppression.

[0012] The present invention is also characterized in that

[0013] In Step 1, the digital baseband signal after clipping by the clipping method is:

[0014] (1)

[0015] In formula (1), is the signal after digital clipping by the clipping method, is the original digital baseband signal, and A is a preset clipping threshold value.

[0016] In Step 2, the expression of the clipping noise is:

[0017] (2)

[0018] In formula (2), is the clipping noise, is the signal after digital clipping by the clipping method, is the original digital baseband signal.

[0019] The specific process of Step 3 is:

[0020] Step 3.1, initialize the particle swarm parameters;

[0021] Step 3.2: Use the particle swarm optimization algorithm to iteratively optimize the initial data to obtain the optimized noise, and the expression is:

[0022]

[0023] (3)

[0024] In formula (3), is the direction and distance of the u -th particle's t+ -th iterative movement. u is the -th particle, t is the -th iteration. is the optimal solution searched by the -th particle, is the solution of the -th particle at the -th iteration, are learning factors;

[0025] Step 3.3: Add the noise optimized in Step 3.2 to the clipped noise obtained in Step 2 to obtain the average noise power of the new noise in the data subcarrier part, and calculate the EVM and PAPR values after superimposing the noise optimized in Step 3.2, the clipped noise obtained in Step 2, and the original digital baseband signal;

[0026] The expression of EVM is as follows:

[0027] (5)

[0028] In formula (5), is the original digital baseband signal not affected by distortion, is the signal after superimposing the optimized noise, clipped noise, and the original signal. N is the total number of signal points to be measured;

[0029] The expression of PAPR is as follows, and take the PAPR value at the probability of 1e-4:

[0030] (6)

[0031] The average power of the time-domain signal in the formula is:

[0032] (7)

[0033] In formulas (6) and (7),x ( t ) is the time-domain signal after the superposition of the optimized noise, the clipped noise, and the original signal. t is time;

[0034] Step 3.4: Repeat Step 3.2 and Step 3.3 until the maximum number of iterations is reached, and then select the global optimal solution as the optimal particle swarm optimization noise.

[0035] In Step 3.1, the particle swarm parameters include the particle dimension, learning factor, particle swarm size, number of iterations, maximum inertia weight, minimum inertia weight, upper limit of the particle swarm velocity and upper limit of the particle swarm modulus, and the initial value of the particle swarm.

[0036] In Step 3.2, the inertia weight adopts a linear variation strategy, and the iteration formula is as follows:

[0037] (4)

[0038] In formula (4), and are the current number of iterations and the maximum number of iterations respectively, and are the maximum weight and the minimum weight respectively, W is the inertia weight.

[0039] In Step 3.3, the constraint on the EVM value is: The particle positions with EVM values greater than the EVM value brought by the clipped noise are invalidated, and the particle positions of the previous iteration are directly used to replace the particle positions of the current iteration;

[0040] The constraint on the PAPR value is: The PAPR suppression ability of the optimized particle swarm is always greater than the PAPR suppression ability of the original clipped noise;

[0041] The constraint on the average noise power of the new noise in the data subcarrier part is:

[0042] (8)

[0043] In formula (8), is the average noise power of the new noise in the data subcarrier part, is the clipped noise power.

[0044] The beneficial effects of the present invention are as follows: Based on the digital clipping method for baseband clipping, the present invention optimizes the clipping noise through the particle swarm optimization algorithm, compensates to a certain extent for the problems of deteriorated PAPR suppression effect and deteriorated EVM caused by the too low average power resulting from the original clipping method, and has no impact on the peak value of the original time-domain signal. To a certain extent, it improves the PAPR suppression ability of this method, takes into account the EVM performance and BER performance, and is applicable to the OFDM communication system. While effectively suppressing the PAPR of the system, it can also ensure a small loss of system performance. Description of the Drawings

[0045] Figure 1 It is a flowchart of the method of the present invention;

[0046] Figure 2 It is the signal processing flow of the particle swarm optimization algorithm in the inventive method;

[0047] Figure 3 It is a comparison graph of the EVM and PAPR suppression gain performances between the method of the present invention and the clipping and wave cancellation method;

[0048] Figure 4 It is a comparison graph of the bit error rate performance curves between the method of the present invention and the clipping and wave cancellation method;

[0049] Figure 5 It is the signal processing flowcharts of the transceiver for the bit error rate performance test in the embodiment of the present invention. Detailed Embodiments

[0050] The present invention will be described in detail below in conjunction with the drawings and specific embodiments.

[0051] The particle swarm optimization peak-to-average ratio suppression method for OFDM signal clipping noise of the present invention, as Figure 1 shown, is specifically implemented according to the following steps:

[0052] Step 1, perform digital clipping operation on the digital baseband signal after upsampling and filtering on the transmitter by the clipping method;

[0053] Multiply the preset clipping rate by the average power of the digital baseband signal to obtain the preset clipping threshold, limit the signal modulus value exceeding the clipping threshold to the threshold value, and keep the phase unchanged. The signal with a signal modulus value lower than the threshold value remains unchanged. Then the digital baseband signal after clipping by the clipping method is:

[0054] (1)

[0055] In formula (1), is the signal after digital clipping by the clipping method, is the original digital baseband signal, and A is the preset clipping threshold value;

[0056] Step 2: Subtract the digital baseband signal after clipping in Step 1 from the original digital baseband signal to obtain the clipping noise, and the expression is:

[0057] (2)

[0058] In formula (2), is the clipping noise, is the signal after digital clipping by the clipping method, is the original digital baseband signal;

[0059] As Figure 2 shown, in Step 3: Initialize the particle swarm parameters, and use the particle swarm optimization algorithm to iteratively optimize the initial data. After reaching the maximum number of iterations, obtain the global optimal solution, and use the global optimal solution as the optimal particle swarm optimization noise;

[0060] The specific process is as follows:

[0061] Step 3.1: Initialize the particle swarm parameters, including: particle dimension, learning factor, particle swarm size, number of iterations, maximum inertia weight, minimum inertia weight, upper limit of particle swarm velocity and upper limit of particle swarm modulus, and initial value of the particle swarm;

[0062] Step 3.2: Use the particle swarm optimization algorithm to iteratively optimize the initial data to obtain the optimized noise, and the expression is:

[0063]

[0064] (3)

[0065] In formula (3), is the direction and distance of the u th particle's t+ 1st iteration movement, u is the th particle, t is the th iteration, is the optimal solution searched by the th particle, is the optimal solution searched by the particle swarm at the current number of iterations, is the solution of the th particle at the th iteration,

[0066] Among them, the inertia weight adopts a linear change strategy, aiming to continuously decrease the inertia weight as the number of iterations increases. The iteration formula is as follows:

[0067] (4)

[0068] In Equation (4), and are the current iteration number and the maximum iteration number respectively, and are the maximum weight value and the minimum weight value respectively, W is the inertia weight value;

[0069] Step 3.3: Add the optimized noise in Step 3.2 to the clipped noise obtained in Step 2 to get the average noise power of the new noise in the data subcarrier part, and calculate the EVM (Error Vector Magnitude) value and the PAPR (Peak-to-Average Power Ratio) value after superimposing the optimized noise in Step 3.2, the clipped noise obtained in Step 2, and the original digital baseband signal;

[0070] The expression of EVM is as follows:

[0071] (5)

[0072] In Equation (5), is the original digital baseband signal not affected by distortion, is the signal after superimposing the optimized noise, the clipped noise, and the original signal. N is the total number of signal points to be measured;

[0073] The expression of PAPR is as follows, and the PAPR value at a probability of 1e-4 is taken:

[0074] (6)

[0075] The average power of the time-domain signal in the formula is:

[0076] (7)

[0077] In Equations (6) and (7), x ([[]] t ) is the time-domain signal after superimposing the optimized noise, the clipped noise, and the original signal, t is the time;

[0078] Among them, the constraint on the EVM value is: Discard the particle positions where the EVM value is greater than the EVM value brought by the clipped noise, and directly use the particle positions of the previous iteration to replace the particle positions of the current iteration;

[0079] The constraint on the PAPR value is: Ensure that the PAPR suppression ability of the optimized particle swarm is always greater than the PAPR suppression ability of the original clipped noise;

[0080] The constraint on the average noise power of the new noise in the data subcarrier part is:

[0081] (8)

[0082] In Equation (8), is the average noise power of the new noise in the data subcarrier part, is the clipping noise power;

[0083] By constraint, the signal-to-noise ratio of the signal after adding both the optimized noise and the clipping noise is made larger than that of the signal with only the clipping noise added, thereby improving the BER (Bit Error Rate) performance of the signal receiver. Then, through the above constraints, it can be ensured that the direction of iteratively updating the particle swarm is correct;

[0084] Step 3.4, Repeat Step 3.2 and Step 3.3 until the maximum number of iterations is reached, and then select the global optimal solution as the optimal particle swarm optimized noise;

[0085] The particle in this solution is the optimized noise. During the iterative update process, the optimization index is calculated and the parameter values in the iterative formula are updated accordingly. By continuously updating the positions of the particle swarm, the iteration can continue in the correct direction. Finally, when the maximum number of iterations is reached, the global optimal solution can be selected as the optimized noise;

[0086] Step 4, Add the optimal particle swarm optimized noise obtained in Step 3 and the clipping noise obtained in Step 2 in reverse to the original digital baseband signal for peak-to-average power ratio suppression.

[0087] The comparison diagram of the PAPR gain (GAIN)-EVM performance between the method of the present invention and the clipping and filtering method is as Figure 3 shown;

[0088] (1) The OFDM signal in the simulation is a 2048-point FFT, using 64QAM modulation, with a modulation method of Turbo coding at a code rate of 1 / 3. The number of simulation times is 10K times, and the number of OFDM symbols included in each frame of the signal is 8. The parameters in the particle swarm optimization algorithm are as follows in the table:

[0089]

[0090] The comparison diagram of the bit error rate performance curves between the method of the present invention and the clipping and filtering method is as Figure 4 shown;

[0091] The signal processing flow chart at the transceiver end in the simulation is as Figure 5 shown. Further, the receiver mainly includes twice downsampling, equiripple low-pass filter filtering, OFDM demodulation, 64QAM soft demodulation, Turbo decoding, and descrambling; the number of simulation times is 1M times.

Claims

1. A particle swarm optimization peak-to-average power ratio suppression method based on OFDM signal clipping noise, characterized in that, The implementation is specifically carried out according to the following steps: Step 1: Perform digital clipping operation on the digital baseband signal after upsampling and filtering on the transmitter by the clipping method; Step 2: Subtract the digital baseband signal after clipping in Step 1 from the original digital baseband signal to obtain the clipping noise; Step 3: Initialize the particle swarm parameters, use the particle swarm optimization algorithm to iteratively optimize the initial data, obtain the global optimal solution after reaching the maximum number of iterations, and use the global optimal solution as the optimal particle swarm optimization noise; The specific process of Step 3 is as follows: Step 3.1: Initialize the particle swarm parameters; Step 3.2: Use the particle swarm optimization algorithm to iteratively optimize the initial data to obtain the optimized noise, and the expression is: （3） In Equation (3), is the u direction and distance of the t+ first iteration movement of the u-th particle, where u is the u-th particle and t is the t-th iteration, is the direction and distance of the t-th iteration movement of the u-th particle, is the optimal solution found by the u-th particle, is the optimal solution found by the particle swarm at the current iteration number, is the solution of the u-th particle at the (t + 1)-th iteration, is the solution of the u-th particle at the t-th iteration, , are learning factors; Step 3.3: Add the noise optimized in Step 3.2 to the clipping noise obtained in Step 2 to obtain the noise average power of the new noise in the data subcarrier part, and calculate the EVM and PAPR values after superimposing the noise optimized in Step 3.2, the clipping noise obtained in Step 2, and the original digital baseband signal; The expression of EVM is as follows: （5） In formula (5), is the original digital baseband signal not affected by distortion, is the sum of the optimized noise, clipping noise, and the original signal. Here, N is the total number of signal points to be measured; The expression of PAPR is as follows, and take the PAPR value at the probability of 1e-4: （6） The average power of the time-domain signal in the formula is as follows: （7） In formulas (6) and (7), x ( t ) is the time-domain signal after the superposition of the optimized noise, the clipped noise, and the original signal, t is time; Step 3.4: Repeat Step 3.2 and Step 3.3 until the global optimal solution is selected as the optimal particle swarm optimization noise after reaching the maximum number of iterations; Step 4: Inverse superimpose the optimal particle swarm optimization noise obtained in Step 3 and the clipping noise obtained in Step 2 into the original digital baseband signal for peak-to-average power ratio suppression.

2. The particle swarm optimization peak-to-average ratio suppression method based on OFDM signal clipping noise according to claim 1, characterized in that In Step 1, the digital baseband signal after clipping by the clipping method is: （1） In formula (1), is the signal after digital clipping cancellation by the shear method, is the original digital baseband signal, and A is a preset clipping threshold value.

3. The particle swarm optimization peak-to-average ratio suppression method based on OFDM signal clipping noise according to claim 1, characterized in that In Step 2, the expression of the clipping noise is as follows: (2) In formula (2), is the limiting noise, is the signal after digital clipping cancellation by the clipping method, is the original digital baseband signal.

4. The method for suppressing the peak-to-average ratio by particle swarm optimization based on the clipping noise of OFDM signals according to claim 1, characterized in that, In Step 3.1, the particle swarm parameters include the particle dimension, learning factor, particle swarm size, number of iterations, maximum inertia weight, minimum inertia weight, upper limit of particle swarm velocity and upper limit of particle swarm modulus, and the initial value of the particle swarm.

5. The particle swarm optimization peak-to-average ratio suppression method based on OFDM signal clipping noise according to claim 1, characterized in that In Step 3.2, the inertia weight adopts a linear change strategy, and the iterative formula is as follows: （4） In formula (4), and are the current iteration number and the maximum iteration number respectively, and are the maximum weight and the minimum weight respectively, W is the inertia weight.

6. The particle swarm optimization peak-to-average power ratio suppression method based on OFDM signal clipping noise according to claim 1, characterized in that In Step 3.3, the constraint on the EVM value is: Discard the particle positions with EVM values greater than the EVM value brought by the clipping noise, and directly use the particle positions of the previous iteration to replace the particle positions of the current iteration; The constraint on the PAPR value is: Make the PAPR suppression ability of the optimized particle swarm always greater than the PAPR suppression ability of the original clipping noise; The constraint on the noise average power of the new noise in the data subcarrier part is: （8） In Equation (8), is the average noise power of the new noise in the data subcarrier part, is the clipping noise power.

Images