- All Products
- Software Integrity
- Semiconductor IP
- Verification
- Design
- Silicon Engineering
- Application Security Services
- Dynamic Analysis (DAST)
- Mobile Application Security
- Penetration Testing
- Managed Static Analysis
- All Services
- Software Security
- Silicon Design
- Community
- About Us
< All Products
< All Products
< All Products
< All Products
< All Products
< All Products
< overview
All Products
+ Software Integrity + Semiconductor IP + Verification + Design + Silicon Engineering + Optical Solutions< Software Integrity
< Software Integrity
Application Security Services
Dynamic Analysis (DAST) Mobile Application Security Penetration Testing Managed Static Analysis< Software Integrity
< Software Integrity
< Software Integrity
Blog
< Software Integrity
Support
< overview
Software Integrity
+ Application Security Products + Application Security Services + Resources + Training & Education Blog Support< Semiconductor IP
< Semiconductor IP
Memories & Libraries
Logic Libraries Memory Compliers Duet Packages HPC Design Kit Embedded Test & Repair Non-Volatile Memory< Semiconductor IP
< Semiconductor IP
Processor Solutions
ARC Processors Embedded Vision Processors Development Tools Operating Systems Ecosystems ASIP Tools< Semiconductor IP
< Semiconductor IP
< Semiconductor IP
< Semiconductor IP
< Semiconductor IP
< overview
< Verification
< Verification
< Verification
< Design
< Design
< Design
< overview
< Silicon Engineering
< overview
< main menu
< Solutions
< Solutions
Industry Solutions
Automotive Cloud Computing Financial Services Industrial Controls Systems Healthcare< Solutions
< overview
< All Services
< All Services
< All Services
< overview
< Software Security
< Software Security
Managed Services
< Software Security
Program Development & Design
Maturity Model (BSIMM) Maturity Action Plan (MAP) Policies & Standards Security Metrics Software Security-in-a-Box< Software Security
< Software Security
Support
< overview
Software Security
+ Software Security Services Managed Services + Program Development & Design + Training Support< Silicon Design
< Silicon Design
< Silicon Design
< Silicon Design
< main menu
< Community
< Community
< Community
< Community
Interoperability
ARC Access HAPS Connect HSPICE Integrator In-Sync System-Level Catalyst TAP-in VC Apps Access< Community
< Community
< Community
< overview
Community
+ Community + SNUG + Partners + Interoperability + Academic Programs + Company Blogs + BSIMM< About Us
< About Us
< About Us
< About Us
< About Us
RSoft Application Gallery Note: Electronic Dispersion Compensation
Tools Used: ModeSYS
This example demonstrates the application of electronic dispersion compensation (EDC) in a multimode link. Upgrade of existing Ethernet links from 1 to 10 Gb/s with keeping the 300-m transmission reach faces the challenge of overcoming performance degradation due to the higher modal dispersion penalties. One of the solutions for this problem is being implemented in IEEE standard for 10 Gigabit Ethernet 10GBASE-LRM [1] and is based on applying electronic dispersion compensation to the signal at the receiver. The link configuration under study is:
The OptSim example depicted above demonstrates a 10 Gigabit multimode link. The multimode fiber is a 300-m graded-index fiber with a 50-micron core diameter. The fiber refractive index profile has a dip defect in the center of the core ñ one of the typical profile distortions due to manufacturing imperfections. The fiber profile used is:
The deviations of index profile from an ideal graded-index may cause increase in differential modal delay, and as a result the performance degradation.
The output signal from the transmitter has a single-mode Gaussian pulse profile with 7 micron FWHM. The connector between transmitter and fiber adds 10 micron offset from the fiber core center. As a result, the received eye diagram at the PIN/TIA receiver is distorted and the computed link BER for this eye diagram is as high as 3.1x10-3.
To improve the eye diagram and BER of the link, we apply EDC to electrical signal after the first receiver in the form of equalization filter implemented as an electrical finite-impulse response (FIR) filter. This equalization scheme is also known as linear or forward-feedback equalization (LE or FFE). The FIR filter model used applies the following transformation to the electrical signal:
where x(t) is incoming signal, y(t) the outgoing signal, (2N+1) the number of taps, Ci- weight coefficients or taps, and DT the taps delay.
For the eye diagram before EDC, we found that a FFE parameters setting with 3 taps, tap delay of ?TB, and taps coefficients {-0.3, 1.0, -0.3} is sufficient to improve the eye diagram quality, and hence, the BER. The eye diagram after applying EDC shows a significant improvement. The corresponding computed BER is equal to 3.7x10-13, i.e. satisfies the link BER requirement of 1x10-12 or better.
The electrical signal waveforms before and after the EDC:
The transfer function for the applied FIR filter:
In this example the parameters of equalization filter are optimized for the given offset of laser-fiber alignment. To demonstrate that we can perform the parameters scan simulation for different values of offsets. Figure 7 shows simulated link BER and Q-factor before and after EDC for different values of offset, ranging from 0 to 20 microns. One can see that the biggest gain from EDC is observed at the 10-micron offset ñ i.e. the one we use in the example. Thus, if one wants to use different offset values, then the FFE parameters have to be modified (optimized) for the best performance.
References
[1] IEEE P802.3aq 10GBASE-LRM Task Force, http://www.ieee802.org/3/aq/public/index.html