University of Peloponnese Adaptive Optical Technologies for Optical

University of Peloponnese Adaptive Optical Technologies for Optical Transmission Systems Maki Nanou, George-Othon Glentis, Kristina Georgoulakis, Chris Matrakidis, Christina (Tanya) Politi, Alexandros Stavdas

Outline v Basic Concepts of Optical Communications v Fiber Impairments & Compensation Techniques v Optical Transmission Simulations v Results v Conclusion University of Peloponnese Dept. of Inform. & Telecommunications

Traffic Growth (*) Cisco Forecast Until 2000: Voice Traffic dominates 60% / year 20% / year After 2004: Data Traffic dominates Traffic growth of 60% per year outstrips the growth in capacity of commercial systems. While the entire traffic in North American core Network could be carried on a single fiber until 2008, in 2011 more than two fibers were required. Every 3 years the required number of fib will double. Increased capacity ↔ Advanced Modulation University of Peloponnese Formats

Non Return To Zero – On – Off Keying Output Intensity quaternary point bias v Most commonly used (widely deployed) v OOK: switching ON and OFF the amplitude of an optical carrier signal (Intensity Modulation Only) Input v External Modulation: biased at the Voltage t LASER MZM optical signal Vπ swing quadrature point of the MZM transfer function, and driven by an electrical binary NRZ-ASK signal with peak-topeak amplitude of Vπ. v Simple Tx/Rx Configurations University of Peloponnese Dept. of Inform. & Telecommunications

Differential Phase Shift Keying - DPSK Output Intensity minimum point bias t LASER MZM o Nearly constant envelope – higher tolerance to non linear effects v Higher receiver sensitivity due to the t 3 d. B lower OSNR requirement to Input Voltage achieve a specific BER. optical signal precoder electrical NRZ data v. Phase Modulation Only 2 Vπ swing v External MZM biased at minimum point and driven with a precoded binary data with twice the switching voltage required for NRZ – OOK (2 Vπ) v More complex Tx /Rx Design University of Peloponnese Dept. of Inform. & Telecommunications

Fiber Impairments in Single Channel Systems Linear Losses Non Linear Dispersion SMF Tx SPM inserts ASE noise DCF G compensates SMF losses G compensates dispersion University of Peloponnese Dept. of Inform. & Telecommunications compensates DCF losses Rx

Chromatic Dispersion Effect Every different f travels with different velocity 1 0 1 input pulse Optical Fibre Dispersion Parameter: DSMF Length of Transmission: LSMF 1 1 1 ISI t Some broadening t Severe broadening v Dispersion tolerance is inversely proportional to the square of the operating bitrate and consequently limitations due to dispersion become more stringent as bit rate increases. v As a linear effect, dispersion can be compensated by means of a DC fibre, providing that the exact amount of dispersion is known in advance. DDCF*LDCF=-DSMF*LSMF University of Peloponnese Dept. of Inform. & Telecommunications

Electronic Equalization v EE attempts to reverse the distortion incurred by a signal transmitted through a channel. v It can be a simple linear filter or a complex algorithm. Electric Filter Clock Recovery y(n) y(t) ADC PIN Receiver (Rx) I(n) Electronic Equalizer v EE are applied after the receiver o no need in intervening in the already installed fibre links v Can cope with variable amounts of dispersion University of Peloponnese Dept. of Inform. & Telecommunications

Electronic Equalization v In our case we investigate the performance of the following equalizers: • Linear Transversal Equalizer – LTE • Decision Feedback Equalizer – DFE • Volterra Decision Feedback Equalizer - VDFE v All equalizers operate at supervised mode, where a training sequence, known by the receiver is transmitted, in order to train the equalizers about the channel characteristics. v Fractional spacing is employed as in this case the performance of the equalizers becomes less sensitive to the sampling phase of the receiver. University of Peloponnese Dept. of Inform. & Telecommunications

Linear Transversal equalizer - LTE is the simplest form of electronic equalizers. The incoming signal is processed by a linear filter. In order to retrieve the transmitted sequence, FS-LTE operates according to: University of Peloponnese Dept. of Inform. & Telecommunications

Decision Feedback equalizer - DFE consists of two parts: a Feed forward part that is driven by the received waveform and a Feedback part that is driven by the estimations of the previous symbols. FS-DFE operates according to: The performance of linear equalizers is constrained when applied to non linear systems.

Non Linear Photodetector v The main reason of non linearity in optical systems is induced by the detector during the conversion of optical to electrical. v Photodiode operates on a square law principle, in which the output of the detector is proportional to the intensity (i. e. , the square of the input signal magnitude). v Although it is a simple circuit, it is nonlinear and as such it is difficult to correct linear distortions such as CD. University of Peloponnese Dept. of Inform. & Telecommunications

Volterra Decision Feedback equalizer - VDFE Simplified VDFE used: University of Peloponnese Dept. of Inform. & Telecommunications

Complexity Calculations University of Peloponnese Dept. of Inform. & Telecommunications

Simulation Setup SMF Tx BER Estimation w/o EDC DCF G G Rx equalizer Transmission Span (x N) BER Estimation with EDC 10 Gb/s bitrate 10 spans x 100 km (1000 km) 40 Gb/s bitrate 3 spans x 100 km (300 km) University of Peloponnese Dept. of Inform. & Telecommunications

Unncompensated Results 380 km 250 km 400 km 200 km 300 km 200 km 150 km University of Peloponnese Dept. of Inform. & Telecommunications

NRZ-OOK Results (1) 94% 70% 98% 80 % 85 % 87. 5 % University of Peloponnese Dept. of Inform. & Telecommunications

NRZ-DPSK Results (1) OCR=70%-90% 10 Gb/s & 40 Gb/s DPSK University of Peloponnese Dept. of Inform. & Telecommunications

Upgrading Scenario Setup Operating at 40 Gb/s Operating at 10 Gb/s SMF Tx BER Estimation w/o EDC DCF G G Rx equalizer Total Length of 1000 km (10 spans x 100 km) 99 % OCR NRZ-OOK Dispersion Tolerance NRZ-OOK NRZ-DPSK Reduces NRZ-DPSK University of Peloponnese Dept. of Inform. & Telecommunications BER Estimation with EDC

Upgrading Scenarios Results Upgrading a system 10 -40 NRZ & DPSK University of Peloponnese Dept. of Inform. & Telecommunications

Conclusion v Low cost, adaptive techniques of optical transmission, consisting of optical and electronic equalization, were studied by simulating configurations with realistic link parameters. v Here, the interplay between optical and electronic techniques for physical impairment mitigation for DD optical transmission with various performance/complexity tradeoffs, is presented. v It has become evident that even in the absence of FEC, low complexity equalizers can perform sufficiently well in conjunction with optical compensation. v Low complexity Volterra equalizers can be used to support the migration of a system from 10 to 40 Gb/s. University of Peloponnese Dept. of Inform. & Telecommunications

Q&A Thank you for your attention! This research was funded by the Operational Program "Education and Lifelong Learning" of the Greek National Strategic Reference Framework (NSRF) Research Funding Program: THALES PROTOMI, grant number MIS 377322. University of Peloponnese Dept. of Inform. & Telecommunications