Parallel Polar Encoding for Faster 5G Channel Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fifth-generation mobile communication systems face processing delays and increased computational complexity due to serial and sequential methods in polar coding, particularly with large input bit sequences, leading to low processing speed and increased computation time.
Innovation Solution
A parallel computation method for polar coding, known as Parallel Computation for Polarization Encoding (PCPE), which divides the input bit sequence into parallel streams and performs simultaneous calculations using Row After Column (RAC) and Sample Repeat and Scalar Multiplication (SRSM) principles, reducing processing time and reusing hardware resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If sequential multiplication method is used for polar coding, then algorithm implementation is straightforward, but processing speed is slow and computation time increases
Solution Approach 1:
The patent divides the input bit sequence into multiple parallel streams and segments the Kronecker matrix multiplication into multiple stages. Each stage processes a subset of the data in parallel, transforming a single long sequential operation into multiple shorter parallel operations. This segmentation enables simultaneous computation across multiple processing units, directly addressing the slow processing speed of sequential methods while managing complexity through structured division of the computation task.
Solution Approach 2:
The patent introduces a parallel dimension to the computation by organizing the multiplication operation across multiple processing channels simultaneously. Instead of processing bits one after another in a single dimension (sequential), the system processes multiple bits in parallel across multiple dimensions (time, space, and processing channels). This dimensional transformation allows the system to achieve higher throughput by utilizing multiple processors or processing units working concurrently on different portions of the input sequence.
2Reliability
If input bit sequence length increases, then coding rate and error correction capability improve, but computation time and processing delay increase
Solution Approach 1:
The patent segments the long input bit sequence into multiple smaller parallel streams that can be processed simultaneously. By dividing the computation into manageable parallel chunks, the system maintains the ability to handle long sequences (which provide better error correction) while reducing the sequential processing time. Each parallel stream undergoes the same polarization encoding operation independently and concurrently, thereby reducing overall computation time without sacrificing the error correction benefits of long block lengths.
Solution Approach 2:
The patent performs preliminary organization of the input bit sequence into parallel streams before the actual multiplication operation. This preliminary arrangement prepares the data in a format that enables simultaneous processing, allowing the system to handle long sequences more efficiently. By pre-organizing the data structure to support parallel computation, the system reduces the computational burden during the actual encoding phase, thereby reducing computation time while maintaining the reliability benefits of longer block lengths.
3Reliability
If Kronecker matrix size increases with larger data models, then coding performance improves, but hardware resource requirements and computational complexity increase
Solution Approach 1:
The patent segments the large Kronecker matrix multiplication operation into multiple smaller sub-operations that can be executed in parallel. Instead of requiring a single large processing unit to handle the entire matrix multiplication, the system divides the matrix operations into smaller blocks that fit within available hardware resources. Each processing unit handles a subset of the multiplication, and results are combined to produce the final encoded output. This segmentation allows the system to achieve the coding performance of large Kronecker matrices while using modest hardware resources through distributed parallel computation.
Solution Approach 2:
The patent designs a universal parallel processing architecture that can handle different sizes of Kronecker matrices and various data model configurations. The same parallel processing framework adapts to different matrix dimensions and input sequence lengths by dynamically configuring the number of parallel streams and processing stages. This multi-functional design allows the system to maintain optimal performance across different coding scenarios without requiring dedicated hardware for each specific matrix size, thereby reducing overall hardware resource requirements while preserving coding performance.
Data Source
AI summary
The invention refers to the parallel calculation method for polarization coding (PCPE) for channel coding technique in 5th next generation mobile communication systems which includes to split N-bits input sequence into X parallel streams, each stream has Y bits; to multiply Y bits at each stream by the columns of the Kronecker matrix GY, the results are displayed in rows according to the principle of bit elimination; and to multiply the matrix obtained with the columns of the Kronecker matrix GX according to the sample repeat and scalar multiplication. In addition, the invention also refers to the polarization coding system according to the Parallel Computation for Polarization Encoding (PCPE) for the channel coding technique in the 5th next generation mobile communication system.