Frequency-Domain Front-Haul Compression in PHY-Split Cloud RAN
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing C-RAN architecture faces challenges with high front-haul data-rate requirements due to the use of CPRI interfaces, leading to prohibitively expensive deployments and inefficient resource utilization, as well as poor compression performance due to lack of statistical averaging and exploitation of signal redundancy.
Innovation Solution
The PHY-split C-RAN architecture divides base-band physical-layer processing between a BBU Pool and an enhanced RRH, employing frequency-domain compression with uniform scalar quantization and variable-rate Huffman coding, which exploits LTE signal redundancy and user scheduling information to reduce front-haul data-rate requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If CPRI interfaces are used in C-RAN architecture, then reliable communication between RRH and BBU pool is ensured, but front-haul data-rate requirements become prohibitively high and deployment costs increase
Solution Approach 1:
The patent extracts only the essential information from the original I/Q samples by performing FFT transformation and keeping only the magnitude spectrum, which contains the critical signal characteristics. This extraction approach reduces the data volume significantly while maintaining the necessary information for base-band processing at the BBU pool.
Solution Approach 2:
The patent changes the representation parameters of the signal from time-domain I/Q samples to frequency-domain magnitude spectrum. This parameter transformation reduces the data rate requirements by exploiting the redundancy in the signal representation, while still preserving the essential signal characteristics needed for processing.
2Loss of information
If I/Q samples are transmitted over front-haul link, then complete signal information is preserved, but data compression performance remains poor due to lack of statistical averaging
Solution Approach 1:
The patent performs preliminary FFT transformation and magnitude calculation at the RRH before transmission. This preliminary processing prepares the data in a form that enables better compression at the BBU pool by creating a representation that exhibits statistical properties more amenable to compression algorithms, while still preserving the essential signal information.
3Productivity
If centralized base-band processing is implemented in BBU pool, then resource utilization efficiency improves, but front-haul data-rate requirements increase due to full I/Q sample transmission
Solution Approach 1:
The patent extracts only the magnitude spectrum information from the I/Q samples and transmits this reduced representation to the BBU pool. This extraction enables centralized processing to proceed with sufficient information for resource allocation and scheduling decisions, while dramatically reducing the front-haul data rate requirements compared to transmitting complete I/Q samples.
Data Source
AI summary
Systems and methods disclosed herein describe a centralized-processing cloud-based RAN (C-RAN or cloud-RAN) architecture that offers reduced front-haul data-rate requirements compared to common-public-radio-interface (CPRI) based C-RAN architectures. Base-band physical-layer processing can be divided between a BBU Pool and an enhanced RRH (eRRH). A frequency-domain compression approach that exploits LTE signal redundancy and user scheduling information can be used at the eRRH to significantly reduce front-haul data-rate requirements. Uniform scalar quantization and variable-rate Huffman coding in the frequency-domain can be applied in a compression approach based on the user scheduling information wherein a lossy compression is followed by a lossless compression.