Polar Code Triangular Interleaver Using 1D Index Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing method for interleaving and deinterleaving coded bits in polar codes for 5G new radio (NR) has high computation complexity and large memory consumption, which affects latency and efficiency.
Innovation Solution
A one-dimensional interleaving index array is calculated during the preparation phase, based on row and column counter arrays, to efficiently interleave and deinterleave polar coded bits, reducing the need for a two-dimensional matrix and minimizing computational complexity and memory usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a two-dimensional matrix is used for interleaving and deinterleaving coded bits, then the interleaving operation can be performed, but the computation complexity and memory consumption increase significantly
Solution Approach 1:
The patent transforms the traditional two-dimensional matrix-based interleaving approach into a one-dimensional array-based approach. By changing the dimensional representation from 2D matrix to 1D array with calculated indices, the patent eliminates the need for complex matrix operations and large memory allocations, thereby reducing computation complexity and memory consumption while maintaining the interleaving functionality
Solution Approach 2:
The patent segments the interleaving operation into two distinct phases: a preparation phase where the one-dimensional interleaving index array is calculated once, and an execution phase where the actual interleaving is performed using simple array indexing. This segmentation allows the complex index calculation to be done beforehand, making the real-time interleaving operation much faster and less computationally intensive
2Reliability
If the existing two-dimensional matrix method is used for deinterleaving, then coded bits can be deinterleaved, but the operation is 57 times slower compared to the one-dimensional array method
Solution Approach 1:
The patent applies the same dimensional transformation to deinterleaving, converting the complex two-dimensional matrix operations into simple one-dimensional array indexing operations. This reduces the computational burden and memory access patterns, resulting in a 57-fold speed improvement for deinterleaving operations
Solution Approach 2:
The patent performs the complex index calculation work in advance during the preparation phase, storing the results in a one-dimensional interleaving index array. During deinterleaving, only simple array lookups are needed, which are much faster than performing the full matrix operations in real-time, thereby achieving significant speedup
3Reliability
If the existing two-dimensional matrix method is used for interleaving, then coded bits can be interleaved, but the operation is 35 times slower compared to the one-dimensional array method
Solution Approach 1:
The patent transforms the interleaving operation from two-dimensional matrix manipulation to one-dimensional array indexing, eliminating the computational overhead of matrix operations and reducing memory consumption, which results in a 35-fold speed improvement for interleaving operations
Solution Approach 2:
The patent pre-calculates and stores the interleaving indices in a one-dimensional array during the preparation phase. This allows the actual interleaving operation to proceed by simply following the pre-computed index sequence, which is much faster than performing matrix operations during the time-critical transmit chain
4Device complexity
If a one-dimensional interleaving index array is calculated during preparation phase, then processor cycles and memory requirements are reduced, but additional preparation time is required
Solution Approach 1:
The patent performs the computationally intensive index calculation work in advance during the preparation phase, which occurs outside the time-critical transmit and receive chains. This shifts the time cost to a non-critical phase, allowing the actual interleaving and deinterleaving operations to execute quickly with minimal processor cycles and memory requirements
Solution Approach 2:
The patent makes the system dynamic by having the preparation phase execute only when needed (when the interleaving index array needs to be generated or updated), rather than continuously. This allows the system to adapt between performing the heavy preparation work when resources are available and executing the lightweight interleaving operations when time is critical
Data Source
AI summary
A method for use in a wireless transmitter of interleaving coded bits includes: determining an integer number (M) of polar coded bits {e(0), . . . , e(M−1)} for wireless transmission; determining a smallest integer number (T) such that T(T+1)/2>=M; and determining a one-dimensional interleaving index array, interleavingId( ), that represents the column by column non-zero elements of a two-dimensional matrix. The two-dimensional matrix includes a hypothetical T by T matrix containing the polar coded bits {e(0), . . . , e(M−1)} input row by row without using the lower right corner elements of the T×T matrix. The method further includes: interleaving the polar coded bits {e(0), . . . , e(M−1)} using the one-dimensional interleaving index array resulting in the output sequence {f(0), . . . , f(M−1)} wherein f(i)=e(interleavingId(i)); and transmitting the interleaved polar coded bits to a wireless receiver.