SLIDE 1
Optimizing the Efficiency of the Transfer Mechanism in SCA-based Radio Systems
Shan Wang
National University of Defense Tech chnology Research ch Center of Softw are Defined Radio Engineering
Optimizing the Efficiency of the Transfer Mechanism in SCA-based - - PowerPoint PPT Presentation
Optimizing the Efficiency of the Transfer Mechanism in SCA-based - - PowerPoint PPT Presentation
Optimizing the Efficiency of the Transfer Mechanism in SCA-based Radio Systems Shan Wang National University of Defense Tech chnology Research ch Center of Softw are Defined Radio Engineering Outline Problem of CORBA Linux-kernel-based
SLIDE 2
SLIDE 3
Some Premises
In SCA 4.1 specifications, the transfer mechanism becomes more flexible and diversified. As a classical transfer mechanism, CORBA has been widely used in SCA-based software radio systems. Communication between components can be classified by their locations : inter-chip and intra-chip, we call them remote-call and local-call separately.
SLIDE 4
Application Environment
ARM Cortex-A7 （Zynq70xx） Linux-3.17 1GB RAM CF-SCA4.1 TAO & omniORB Delay (μs)
968.6 961.9 960.6 971 967.7 965.84 173.3 175.2 173.6 173.8 173.3 173.8 200 400 600 800 1000 1200 1 2 3 4 5 Avg TAO
- mniORB
Problem of CORBA
SLIDE 5
Problem of CORBA
Application Environment
Intel i7-6700 Linux Ubuntu14.02 4GB RAM CF-SCA4.1 TAO & omniORB Delay (μs)
255.8 265.3 276 272.2 270.2 267.9 44.2 43.4 41.3 44.3 43.1 43.26 50 100 150 200 250 300 1 2 3 4 5 Avg TAO
- mniORB
SLIDE 6
Results show that the efficiency of the ORB-based middleware, such as TAO and omniORB, has become one of the challenges that constrain SCA to be further widely applied.
Why?
SLIDE 7
The advantage is that it loosens the difference of program languages and can work in distributed networks. However, the disadvantage is also obvious, especially in what we call the intra-chip application environment.
In many Linux based applications, Core Framework starts all components in process mode, including device/service components, and waveform components.
SLIDE 8
Local-call Characteristics Thread Mode
How to integrate process mode with thread mode flexibly.
SLIDE 9
GPP Device, FPGA Device, DSP Device, Ethernet Device, Serial Device, Platform Service, etc. NET Component, LLC Component, Security Component, etc.
Device Application
Domain Manager Device Manager
SLIDE 10
Application Environment
Intel i7-6700 Linux Ubuntu14.02 4GB RAM CF-SCA4.1 Process VS. Thread
255.8 265.3 276 272.2 270.2 267.9 2.1 2.3 2.1 2 2.1 2.12 50 100 150 200 250 300 1 2 3 4 5 Avg. Process Mode Thread Mode 44.2 43.4 41.3 44.3 43.1 43.26 0.3 0.2 0.2 0.2 0.2 0.22 5 10 15 20 25 30 35 40 45 50 1 2 3 4 5 Avg. Process Mode Thread Mode
- mniORB’s Delay (μs)
TAO’s Delay (μs) 126 times 196 times
SLIDE 11
Application Environment
ARM Cortex-A7 Linux-3.17 1GB RAM CF-SCA4.1 Process VS. Thread
- mniORB’s Delay (μs)
TAO’s Delay (μs)
968.6 961.9 960.6 971 967.7 965.84 36.1 36.3 36 37.2 37.3 36.56 200 400 600 800 1000 1200 1 2 3 4 5 Avg. Process Mode Thread Mode 173.3 175.2 173.6 173.8 173.3 173.8 5.9 6.1 6.01 6.03 6.05 6.02 20 40 60 80 100 120 140 160 180 200 1 2 3 4 5 Avg. Process Mode Thread Mode
26 times 29 times
SLIDE 12
Process Mode
This mode is suitable for waveform resource deployment based on multi- channel equipment or control interactions among different nodes.
Thread Mode
Components must run in the same program space. So, it is suitable for small size SoC system or waveform applications located in the same processor.
SLIDE 13