Tail-Biting Convolutional Coding for Variable UCI Payloads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional channel coding schemes struggle to provide reliable encoding and decoding for Uplink Control Information (UCI) in LTE systems, especially with the increased payload variability due to carrier aggregation, which complicates the use of existing block codes for large payloads.
Innovation Solution
A device and method utilizing a tail biting convolutional code with a constraint length, allowing encoding and decoding of information words with fewer bits, thereby simplifying encoder and decoder structures and avoiding the need for additional channel codes, by using a shift register configuration and appending bits to maintain initial and final state consistency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional block codes (e.g., Reed-Muller codes) are used for small UCI payloads, then encoding is simple, but the codes cannot be generalized to large payloads from 32 carrier aggregation
Solution Approach 1:
The patent applies a single convolutional code structure that can handle both small and large UCI payloads uniformly. The encoder uses a fixed constraint length L=7 and generates codewords of length N=3K for any payload size K, eliminating the need for different code schemes for different payload ranges. This universal approach resolves the contradiction by making the encoder adaptable to all payload sizes while maintaining consistent structural complexity.
2Productivity
If carrier aggregation is increased to 32 component carriers, then throughput is improved, but UCI payload size increases significantly (up to 638 bits), requiring more complex coding schemes
Solution Approach 1:
The convolutional code with constraint length L=7 and codeword length N=3K provides a universal coding solution that scales with payload size without increasing structural complexity. Whether K=2 bits or K=638 bits, the encoder maintains the same operational principles and structural parameters, thus supporting high throughput from 32 carrier aggregation while avoiding increased coding complexity.
Solution Approach 2:
The code parameters (constraint length L=7, codeword length N=3K) are designed to scale with the information word length K. This parameter relationship allows the code to adapt to varying payload sizes from small to large (up to 638 bits) while maintaining consistent encoder structure and decoding complexity, thus supporting high productivity without proportionally increasing device complexity.
3Ease of operation
If tail biting convolutional code is used with constraint length L where K<L-1, then encoding is possible, but conventional approaches require additional channel codes which increase complexity
Solution Approach 1:
The patent implements a unified convolutional encoding approach that works for all K values (including K<L-1) without requiring different code types or additional channel codes. The encoder consistently uses the same shift register structure and generating polynomials regardless of whether K is greater than or less than L-1, thus improving ease of operation while maintaining manageable decoder complexity through a single decoding algorithm.
Data Source
AI summary
An embodiment encoder device for encoding an information word c=[c0, c1, . . . , cK-1] having K information bits, ci, includes an encoder for a tail biting convolutional code having a constraint length, L, where K<L−1; the encoder being configured to receive the K information bits; and encode the K information bits so as to provide an encoded code word. An embodiment decoder device for determining an information word c=[c0, c1, . . . , cK-1], having K information bits, ci, includes a decoder for a tail biting convolutional code having a constraint length, L, where K<L−1; the decoder being configured to: receive an input sequence; compute at least one reliability parameter based on the received input sequence; and determine an information word c based on the at least one reliability parameter.