Centralized Decoding HARQ Prediction Antenna Site
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mobile network architectures face challenges in centralizing processing due to non-ideal interfaces, specifically high latency and throughput requirements, which impact HARQ performance and increase operational costs, while also limiting centralization gains and flexibility in resource pooling.
Innovation Solution
Decoding of uplink signals is centralized, with HARQ prediction performed at the antenna site, allowing for autonomous retransmission decisions and reducing latency, and the interface is adapted to carry soft-bits and control information for joint decoding, enabling statistical multiplexing and improved throughput.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If decoding is centralized at the cloud, then processing requirements at antenna sites are reduced and centralization gains are achieved, but interface latency increases and HARQ performance deteriorates
Solution Approach 1:
The system performs HARQ prediction at the antenna site before actual decoding occurs at the cloud. This preliminary action allows the antenna site to predict whether decoding will succeed or fail and make autonomous retransmission decisions without waiting for the decoding result, thereby compensating for the latency introduced by centralized decoding.
Solution Approach 2:
The antenna site is equipped with autonomous HARQ prediction and retransmission decision capabilities, allowing it to self-manage HARQ processes without continuous intervention from the cloud. This self-service mechanism reduces the impact of interface latency on HARQ performance while maintaining centralized decoding benefits.
2Device complexity
If the interface carries I/Q samples for centralized decoding, then maximum centralization gain is achieved, but throughput requirements become excessively high and do not scale with traffic
Solution Approach 1:
The system applies different processing qualities at different locations: the antenna site performs local HARQ prediction and autonomous retransmission decisions, while the cloud performs centralized decoding. This differentiated quality approach allows the interface to carry only necessary data (coded bits and prediction information) rather than all I/Q samples, reducing throughput requirements while maintaining centralization gains.
3Device complexity
If HARQ is implemented at the cloud with centralized decoding, then resource pooling is maximized, but HARQ cycle time increases and scheduling performance suffers
Solution Approach 1:
The antenna site performs HARQ prediction as a preliminary action before the actual HARQ process at the cloud. This allows the system to prepare retransmission decisions in advance based on prediction results, reducing the effective HARQ cycle time and improving scheduling performance while maintaining resource pooling benefits.
Solution Approach 2:
The HARQ function is segmented between the antenna site (prediction and autonomous retransmission decisions) and the cloud (actual decoding). This segmentation allows the time-critical prediction function to be performed locally while the resource-intensive decoding remains centralized, optimizing both scheduling performance and resource pooling.
4Adaptability or versatility
If the interface is standardized for multi-vendor compatibility, then vendor flexibility is improved, but interface performance and customization are limited
Solution Approach 1:
The system uses standardized interfaces (such as CPRI or OBSAI) that support multiple functions: carrying coded bits from the antenna site, transmitting HARQ prediction information, and enabling centralized decoding at the cloud. This multi-functional standardized interface achieves both vendor flexibility and adequate performance for the centralized decoding architecture.
Data Source
Figure 1~4
Figure 5~6
Figure 7~10
AI summary
It is provided a method, comprising predicting if a received coded data block can be successfully decoded by a predetermined decoder adapted to apply a predetermined decoding on the coded data block, wherein the predetermined decoder is comprised in a central device; requesting a retransmission of the coded data block to an apparatus performing the method if it is predicted that the coded data block cannot be decoded by the decoder; forwarding the coded data block and a result of the prediction to the central device without applying the predetermined decoding to the coded data block.