CN-115664432-B - Self-adaptive iteration number method and system of decoder

CN115664432BCN 115664432 BCN115664432 BCN 115664432BCN-115664432-B

Abstract

The invention provides a self-adaptive iteration number method and a self-adaptive iteration number system of a decoder, which comprise the steps of starting a self-adaptive iteration function, initializing the decoder, setting the iteration number of the decoder as the maximum iteration number, counting the average MCS value and the average signal-to-noise ratio of a physical downlink shared channel in a sliding window, adjusting the iteration number of the decoder as the initial iteration number of the decoder according to the average MCS value and the average signal-to-noise ratio, decoding a unit block by the decoder, recording the actual decoding number used when the decoding is successful, calculating the actual decoding number used when the decoding is successful for the current unit block in the sliding window and the average decoding number of the actual decoding number used when the decoding is successful for the previous unit block, and adjusting the current iteration number of the decoder according to the average decoding number to decode the next unit block. The maximum iteration number of the LDPC decoder is dynamically adjusted, so that unnecessary decoding iteration is reduced, the power consumption of the UE is reduced, and the balance between performance and power consumption is achieved.

Inventors

ZHANG JINHUI

Assignees

成都新基讯通信技术有限公司

Dates

Publication Date: 20260505
Application Date: 20221104

Claims (9)

1. A method for adaptive iteration number of a decoder, comprising: Step A1, starting a self-adaptive iteration function; step A2, initializing the decoder, and setting the iteration number of the decoder as the maximum iteration number; step A3, the average MCS value and the average signal-to-noise ratio of the physical downlink shared channel in the sliding window are counted; step A4, according to the average MCS value and the average signal-to-noise ratio, adjusting the iteration times of the decoder as the initial iteration times of the decoder; Step A5, the decoder decodes the unit blocks, and records the actual decoding times used when the decoding is successful; Step A6, calculating the actual decoding times used when the decoding of the current unit block in the sliding window is successful and the average decoding times of the actual decoding times used when the decoding of the previous unit block is successful; step A7, adjusting the current iteration times of the decoder according to the average decoding times, and returning to the step A5 until the decoding is finished; The step A4 comprises the following steps: step A41, judging whether the average signal-to-noise ratio is higher than a first signal-to-noise ratio threshold value and whether the average MCS value is lower than the first MCS threshold value: If the average signal-to-noise ratio is higher than the first signal-to-noise ratio threshold value and the average MCS value is lower than the first MCS threshold value, continuing to step A42; if the average signal-to-noise ratio is not higher than the first signal-to-noise ratio threshold value, or the average MCS value is not lower than the first MCS threshold value, continuing to step A43; Step A42, adjusting the iteration times of the decoder to be a first preset multiple of the maximum iteration times, and continuing the step A5; Step A43, judging whether the average signal-to-noise ratio is lower than a second signal-to-noise ratio threshold value, and judging whether the average MCS value is lower than a second MCS threshold value: If the average signal-to-noise ratio is lower than the second signal-to-noise ratio threshold value and the average MCS value is lower than the second MCS threshold value, continuing to step A44; If the average signal-to-noise ratio is not lower than the second signal-to-noise ratio threshold value or the average MCS value is not lower than the second MCS threshold value, continuing the step A5; And step A44, adjusting the iteration times of the decoder to be a second preset multiple of the maximum iteration times, and continuing the step A5.
2. The method of adaptive iteration count of a decoder of claim 1, wherein said first predetermined multiple is less than said second predetermined multiple.
3. The method of adaptive iteration count of a decoder of claim 1, wherein said first predetermined multiple is 0.5 and said second predetermined multiple is 0.75.
4. The method of adaptive iteration count of a decoder of claim 1, wherein in step A7, adjusting the iteration count of the decoder according to the average decoding count comprises: step a71, determining whether the average decoding frequency is higher than an iteration threshold: if yes, continue step a72; if not, continuing to step A73; Step A72, increasing the iteration times of the decoder according to a first preset value; And step A73, reducing the iteration times of the decoder according to a second preset value.
5. The method of claim 4, wherein the iteration threshold is a third predetermined multiple of the initial iteration number.
6. The method for adaptive iteration count of a decoder of claim 5, the method is characterized in that the third preset multiple is 0.8.
7. The method of claim 4, wherein the first predetermined value is 1 and the second predetermined value is 1.
8. The method of adaptive iteration number of a decoder of claim 1, further comprising, in performing said steps A5-A7: Step B1, counting the average error rate of the physical downlink shared channel in a sliding window in the decoding process executed by the decoder; step B2, judging whether the average bit error rate is higher than a third preset value: If yes, continuing the step B3; If not, continuing the step B4; Step B3, if the decoder is currently in the self-adaptive iteration function, switching to a common iteration function and then executing decoding; step B4, if the decoder is currently in the self-adaptive iteration function, continuing the steps A5-A7, if the decoder is currently in the common iteration function, switching to the self-adaptive iteration function, and executing the steps A1-A7; the normal iteration function means that the iteration number of the decoder is kept unchanged at the maximum iteration number.
9. A system of adaptive iteration times for a decoder, characterized in that a method of adaptive iteration times for a decoder according to any of claims 1-8 is used, comprising: the starting module starts the self-adaptive iteration function; The setting module is connected with the starting module and used for initializing the decoder and setting the iteration number of the decoder as the maximum iteration number; the first statistics module is used for counting the average MCS value and the average signal-to-noise ratio of the physical downlink shared channel in the sliding window; The first adjusting module is respectively connected with the setting module and the statistics module and is used for adjusting the iteration times of the decoder to be the initial iteration times of the decoder according to the average MCS value and the average signal-to-noise ratio; The decoder is connected with the first adjusting module and is used for decoding the unit blocks by using the current iteration times; the recording module is connected with the decoder and used for recording the actual decoding times used when the unit blocks are successfully decoded; the second statistical module is connected with the recording module and is used for calculating the actual decoding times used when the current unit block in the sliding window is successfully decoded and the average decoding times of the actual decoding times used when the previous unit block is successfully decoded; And the second adjusting module is connected with the second statistical module and the decoder and is used for adjusting the current iteration times of the decoder according to the average decoding times.

Description

Self-adaptive iteration number method and system of decoder Technical Field The present invention relates to the field of LDPC decoding technologies, and in particular, to a method and system for adaptive iteration number of a decoder. Background In an NR wireless communication system, a PDSCH codec at a UE end adopts an LDPC codec, and in the decoding principle, each CB needs to be subjected to iterative decoding for a plurality of times, and the decoding result is finally determined through repeated computation of a check node and a verification node. If the maximum iteration number is still unable to decode correctly, the decoder stops decoding this time and judges decoding failure. LDPC decoding, i.e., low-density parity-check code decoding, is a very complex computational process, and each iteration consumes a lot of time and power consumption of the UE, especially for mobile terminals, which are very important indicators. In the current technical implementation scheme, the maximum iteration number of the LDPC decoder is fixed and cannot be dynamically changed according to the actual channel environment. In an actual external field environment, particularly in the case of good signal-to-noise ratio and low code rate of scheduling data, when the number of LDPC decoding iterations exceeds the average number of iterations of the scene, the CB block decoded at this time still cannot be decoded correctly, and the CB block may be subjected to data anomaly caused by burst interference or decoding anomaly caused by parameter error, and if the LDPC decoder continues to perform multiple iteration attempts, the LDPC decoder may be purely wasteful in power consumption. Disclosure of Invention The invention provides a self-adaptive iteration number method and a self-adaptive iteration number system of a decoder, and aims to solve the technical problems of high UE power consumption and the like in decoding of LDPC decoders in the prior art. A method of adaptive iteration number for a decoder, comprising: Step A1, starting a self-adaptive iteration function; a2, initializing a decoder, and setting the iteration number of the decoder as the maximum iteration number; step A3, the average MCS value and the average signal-to-noise ratio of the physical downlink shared channel in the sliding window are counted; Step A4, according to the average MCS value and the average signal-to-noise ratio, adjusting the iteration times of the decoder as the initial iteration times of the decoder; Step A5, the decoder decodes the unit blocks, and records the actual decoding times used when the decoding is successful; Step A6, calculating the actual decoding times used when the decoding of the current unit block in the sliding window is successful and the average decoding times of the actual decoding times used when the decoding of the previous unit block is successful; and A7, adjusting the current iteration times of the decoder according to the average decoding times, and returning to the step A5 until the decoding is finished. Further, step A4 includes: step A41, judging whether the average signal-to-noise ratio is higher than a first signal-to-noise ratio threshold value and whether the average MCS value is lower than the first MCS threshold value: If the average signal-to-noise ratio is higher than the first signal-to-noise ratio threshold value and the average MCS value is lower than the first MCS threshold value, continuing to step A42; If the average signal-to-noise ratio is not higher than the first signal-to-noise ratio threshold value, or the average MCS value is not lower than the first MCS threshold value, continuing to step A43; Step A42, the iteration times of the decoder are adjusted to be a first preset multiple of the maximum iteration times, and the step A5 is continued; step A43, judging whether the average signal-to-noise ratio is lower than a second signal-to-noise ratio threshold value, and judging whether the average MCS value is lower than the second MCS threshold value: if the average signal-to-noise ratio is lower than the second signal-to-noise ratio threshold value and the average MCS value is lower than the second MCS threshold value, continuing to step A44; if the average signal-to-noise ratio is not lower than the second signal-to-noise ratio threshold value or the average MCS value is not lower than the second MCS threshold value, continuing to step A5; step A44, the iteration number of the decoder is adjusted to a second preset multiple of the maximum iteration number, and step A5 is continued. Further, the first preset multiple is smaller than the second preset multiple. Further, the first preset multiple is 0.5. Further, the second preset multiple is 0.75. Further, in step A7, adjusting the iteration number of the decoder according to the average decoding number includes: Step A71, judging whether the average decoding frequency is higher than the iteration threshold: if yes, continue step a72; if not,