APD SIMD Processing for RF Network Messages

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing mobile computing systems face inefficiencies in processing network messages due to limited computational capabilities of physical layer processors, leading to bottlenecks and reduced quality of service, especially in adverse RF conditions and multipath propagation scenarios.

Innovation Solution

The use of an Accelerated Processing Device (APD) with Single Instruction Multiple Data (SIMD) modules to process various types of network messages, including physical layer messages, by receiving RF signals, converting them to digital data, and assigning them to SIMD modules for accelerated processing, thereby leveraging parallel processing capabilities to improve signal quality and resource allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multiple physical layer processors are used to process network messages, then processing capability is improved, but device complexity and cost increase

Engineering Contradiction:
Improvenetwork message processing capabilityVSAvoidnumber of processors
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent combines multiple physical layer processor functions into a single GPU by mapping physical layer processing tasks onto SIMD (Single Instruction Multiple Data) modules. The GPU's parallel processing architecture allows multiple data streams to be processed simultaneously through unified instruction execution, eliminating the need for separate dedicated processors while maintaining high processing capability.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The GPU is designed to perform multiple functions including both graphics processing and physical layer message processing. By configuring the GPU to handle diverse network message types through its flexible SIMD architecture, the system achieves multi-functionality without requiring specialized dedicated hardware for each processing task.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Speed

If dedicated physical layer processors are used, then processing speed is improved, but the route to CPU becomes long and slow creating a bottleneck

Engineering Contradiction:
Improvemessage processing speedVSAvoiddata transmission delay
Core Design Contradiction:
SpeedVSLoss of time

Solution Approach 1:

The patent merges the physical layer processing function with the GPU, which is already closely integrated with the CPU through the graphics memory interface. This integration creates a shorter data path where processed data can be directly transferred to system memory and CPU without traversing multiple intermediate interfaces, thereby reducing transmission delay.

Inventive Principle:
Principle #5Merging (Combining)

3Productivity

If computational capabilities of physical layer processors are increased, then processing efficiency is improved, but adaptability to frequent changes in communication standards deteriorates

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidadaptability to communication standards
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The GPU's SIMD architecture provides inherent flexibility to adapt to different communication standards through software configuration. The same hardware can be reprogrammed to handle various message types and protocols by loading different instruction sets, providing both high processing efficiency and standard adaptability without requiring hardware changes.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP2880900B1Methods and systems for processing network messages in an accelerated processing device
Publication Date: 2023.05.31 ATI TECHNOLOGIES ULC
  • EP2880900B1 patent drawingFigure 1
  • EP2880900B1 patent drawingFigure 2
  • EP2880900B1 patent drawingFigure 3

AI summary

The present method and system enables receiving a radio frequency (RF) signal. The received RF signal is assigned to a single instruction multiple data (SIMD) module in an accelerated processing device (APD) for processing to extract network messages. The extracted network layer messages are further processed by the SIMD module to obtain data transmitted via the RF signal.