Adaptive Fixed Point Mapping for 5G Fronthaul Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 5G wireless networks face challenges in efficiently managing data traffic and latency due to high throughput requirements between baseband processing units and remote radio units, leading to suboptimal quantization performance and increased costs with traditional fixed uniform quantization methods.
Innovation Solution
Implementing adaptive fixed point mapping by using a shift bit to determine the type of data mapping, allowing for dynamic compression of both precoded and non-precoded data streams, reducing the need for multiple compression algorithms and optimizing data transmission between DU and RU.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fixed uniform quantization methods are used, then implementation is simple, but throughput requirements between DU and RU are high and quantization performance is suboptimal
Solution Approach 1:
The patent applies dynamics by transitioning from fixed uniform quantization to adaptive fixed-point mapping that dynamically adjusts quantization parameters based on data characteristics. The system determines whether data is precoded or non-precoded and applies appropriate mapping strategies, enabling the quantization process to adapt to different data types and traffic conditions, thereby improving quantization performance while reducing throughput requirements
Solution Approach 2:
The patent changes quantization parameters based on data type identification. By detecting whether data is precoded or non-precoded, the system adjusts the fixed-point mapping parameters accordingly, applying different quantization strategies for different data characteristics. This parameter adaptation enables optimal quantization performance for each data type while reducing overall throughput requirements
2Measurement precision
If multiple compression algorithms are used for different data types, then compression performance is optimized, but system complexity increases
Solution Approach 1:
The patent implements a universal fixed-point mapping framework that handles both precoded and non-precoded data through a single system. The unified approach uses data type detection to select appropriate mapping strategies within the same framework, eliminating the need for separate compression algorithms while maintaining optimal compression performance for different data types
Solution Approach 2:
The patent creates a composite quantization approach by combining different fixed-point mapping strategies within a unified framework. The system integrates precoded data mapping and non-precoded data mapping into a single adaptive system that selects the appropriate strategy based on data characteristics, achieving the benefits of multiple specialized algorithms without their complexity
Data Source
AI summary
Compression techniques can reduce the fronthaul throughput in split radio access network (RAN) architectures for next generation designs. Adaptive fixed-point mapping can reduce the throughput requirements between a baseband unit (DU) and a remote radio unit (RU). Thus, a bit or plurality of bits can indicate the type of data being passed over the fronthaul. Consequently, adaptive mapping between precoded downlink data and non-precoded downlink data suited to the type of data passed over the fronthaul can achieve high compression ratios.