Partial Sum Tables for Wireless Decoder Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile wireless communications devices face challenges in processing memory-based modulations due to increased computational and memory demands, especially when dealing with multipath signals, which can lead to complex trellis structures and intensive on-the-fly computations, making practical implementation difficult in limited power compact devices.
Innovation Solution
A wireless communications device with a decoder that generates partial sum tables based on channel estimates and possible signal values, allowing for the computation of branch metrics without intensive on-the-fly calculations, using an iterative process that exchanges extrinsic information with an outer FEC code, thereby reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If MAP trellis decoder is used to demodulate CPM signals with multipath, then demodulation performance is improved, but computational complexity and memory requirements increase exponentially
Solution Approach 1:
The patent segments the computationally intensive branch metric computation into two parts: (1) pre-computation of partial sums that are independent of received signal values, and (2) final metric computation that uses pre-computed partial sums. This segmentation reduces on-the-fly computational complexity while maintaining demodulation performance.
Solution Approach 2:
The patent performs preliminary computation of partial sums before the actual demodulation process. These partial sums are stored in memory and reused during branch metric computation, eliminating the need to re-compute them for every state and branch, thereby reducing computational complexity.
2Reliability
If iterative MAP decoding with outer FEC code is implemented, then error correction performance is improved, but processing time and computational resources increase
Solution Approach 1:
The patent pre-computes partial sums that are independent of the received signal values and stores them for reuse across iterative decoding processes. This preliminary action reduces the computational burden in each iteration, thereby reducing processing time while maintaining error correction performance.
Solution Approach 2:
The patent creates and stores copies of partial sum values in memory tables, which can be rapidly accessed and reused during iterative decoding without requiring repeated computation, thus reducing processing time across multiple iterations.
3Measurement precision
If on-the-fly branch metric computation is performed for every state and branch, then demodulation accuracy is improved, but power consumption and device complexity increase
Solution Approach 1:
The patent segments the branch metric computation to separate parts that require frequent updates from parts that remain constant. By identifying and pre-computing the constant partial sum portions, the device reduces the number of power-intensive operations performed on-the-fly, thereby reducing power consumption while maintaining demodulation accuracy.
Solution Approach 2:
The patent performs preliminary computation and storage of partial sums before the actual demodulation process. This eliminates redundant computations during operation, reducing power consumption while maintaining the accuracy required for correct demodulation decisions.
Data Source
AI summary
A wireless communications device includes a receiver, and a decoder coupled downstream from the receiver and using a modulation having memory for a received signal and to decode the received signal. The decoder decodes the received signal by at least determining a channel estimate for the received signal, generating partial sum tables based upon the channel estimate and possible values of a transmitted signal, correlating actual values of the received signal to the possible values from the partial sum tables to generate branch metrics associated with the modulation, and demodulating the received signal based upon the branch metrics using an iterative process based upon exchanging extrinsic information with an outer forward error correction (FEC) code.