MULTIMODE FIBER TRANSMISSIONS OVER ANY (LOSS-LIMTIED) DISTANCES USING ADAPTIVE EQUALIZATION TECHNIQUES

Transcription

1 1 Gb/s MULTIMODE FIBER TRANSMISSIONS OVER ANY (LOSS-LIMTIED) DISTANCES USING ADAPTIVE EQUALIZATION TECHNIQUES Fow-Sen Choa Department of Computer Science and Electrical Engineering, The University of Maryland Baltimore County, Baltimore, MD 2125, USA (o), (fax), Equalizer 1.

2 1 Gb/s Ethernets The speed of Ethernets increases nearly 1 times for every 3 years It is 1Gb/s now and in the next 1-2 years, it will be focused on the 1Gb/s (meet with SONET) Among the 5 objectives of 1Gb/s GE in the standards meeting (1. 1 m installed multimode fiber, 2. 3m on new multimode fiber, 3. 2 km on single mode fiber, 4. 1 km on single mode fiber, and 5. 4 km single mode fibers) the only one without solution is about the already installed multimode fibers (MMFs). The main source of bit errors in a MMF transmission system is the inter-symbol interference (ISI) caused by the differential mode dispersion (DMD) generated multi-path effects Equalizer 2.

3 Multipath Delay Bin Outputs Amplitude(mV) Time(ns) Amplitude(mV) Time(ns) Amplitude(mV) Time(ns) A single pulse is transmitted through a 1.5 km long MMF at different launching offsets using a 1.55 um gain switched DFB laser. The fundamental mode is at around the time 87 ns. Other higher order modes are followed. Equalizer 3.

4 Approaches to Resolve The Multi-Path/ISI Problems Multi-fiber parallel transmissions 1x 1Gb/s. Multi-Wavelength WDM approaches 4x2.5 Gb/s or 2x5 Gb/s. Multi-level modulation approaches. Subcarrier modulation approaches. Single channel equalization approaches. Equalizer 4.

5 Examples (WDM Approaches) LD-1 LD-2 LD-3 LD-4 4X 1 MM F MMF---SMF coupling Filter SMF---MMF coupling Attenuato r RX Power(dBm) CH.1 CH.2 CH.3 CH.4 7.5nm 8.4nm 6.8nm W avelength(nm) Ch.3, After 2m MM F Ch.3, Back-to-Back Ch.4, Back-to-Back Ch.4, After 2m MM F 1 2 DFB, 2m M MF, angle= DFB, back-back DFB, 5m M MF, angle=.3 DFB, 5m M MF, angle= DFB, 5m M MF, angle=.6 -log(ber) 5 Log(BER) Power (dbm) Receiving Power(dBm) Equalizer 5.

6 Equalization Principle MMF Delay bins Fundamental Detector PreAmp. + Timing and Decision circuits - - τ2 τ1 W1 τ1 W2 τ2 Equalizer 6.

7 Equalization Architecture w t Data sender : regular transmission : training u(n) Multi-mode fiber channel Integrating circuit cp Clock recovery + d(n) w = r(n) t _ u(n) e(n) y(n) in training, or r ISI (n) in regular transmission Decision level = D u(n) D w t u w 1 w 2 w M D D Equalizer 7.

8 Transmission Stages of the Adaptive Equalizer ntraining Periods A training data stream is transmitted A replica of the training data is stored at the receiver end as the input vector u(n) of the adaptive equalizer Desired value d(n) is the detected signal r(n) before decision Weight w(n) is updated adaptively (LMS) to approach the impulse response h(t) of the MMF transmission channel n Regular transmissions The equalizer works with the fixed weights w*(n T ) r ISI (n) = w'* T (n T )û(n-1) uˆ( n) = 1,, if if [ r( n) r [ r( n) r ISI ISI ( n)] > ( n)] < D D Equalizer 8.

9 Nature of the MMF channel Impulse response of the MMF channel h ( t ) = N k = A k δ ( t Impulse response of the inverse system N 1-1 j2πfτ k h ( t) = F Ak e k= Length of filter approaches h(t) M min = (τ max - τ min ) / T The filter approaching h(t) is much shorter than the one approaching h (t) + τ k ) Equalizer 9.

10 Statistics of the Input Tap Vector n Correlation matrix (R = E{u(n)u T (n)}) R = p p 2 p 2 p p 2 p 2 p 2 p 2 p M x M Assuming i.i.d. input samples u (n) n Condition number of the matrix (c(r)) 1 + ( M 1) p χ (R ) = 1 p Conditioning of R becomes worse as M increases Equalizer 1.

11 Performance of the LMS Algorithm n Maximum step size m max for LMS to converge µ max = 2 Mp n n m max becomes smaller and c(r) bigger as M increases Slowing down the convergence of the LMS algorithm M min is proportional to the transmission distance z Longer distance between communication hosts requires more iterations for LMS algorithm to converge Equalizer 11.

12 Simulation Results: Comparison of Converged Filter weights h(t) w 1 (n) 1 w 2 (n) time (ps) weight index w 2 has a simpler structure than w 1 and can be represented by a few nonzero coefficients.1 weight index 2 Equalizer 12.

13 Simulation Results: Comparison of Learning Curves -5-1 Log 1 (J) Standard Equalizer System ID 3 iteration index (n) System ID based approach converges much faster Equalizer 13.

14 Simulation Results (I): Performance of the Proposed Equalization approach h(t) w(n) time (ps) Impulse response of the MMF channel and the converged filter weights 5 1 weight index 15 2 Equalizer 14.

15 Simulation Results (II): Performance of the Proposed Equalization approach 1.5 Received signal waves in NRZ form ideal 1.5 without equalizing with equalizing Units in bit Equalizer 15.

16 Simulation Results (III): Performance of the Proposed Equalization approach No equalization with equalization Eye-pattern diagrams of the received signal 1 2 x time (second) x 1-1 Equalizer 16.

17 Experimental Setup Transmitter Equalizer is composed of Multilink Decision Ckt., Delay lines, Amplifiers (providing phase reversal). Receiver and Part of the Equalizer Equalizer 17.

18 Experimental Results (a ) (b ) (c ) Since the delay path (cables) are long, we send fixed data pattern and use signals in the current frame to cancel multipath copies in the next frame. a. Transmitted Pattern, b. Before equalization, c. After equalization Equalizer 18.

19 Conclusions (I) We have demonstrated using adaptive equalization techniques to overcome the signal degradation caused by differential modal dispersion in a conventional multimode fiber. With this technique, we can not only obtain MMF-based 1 Gb/s GEs but also upgrade all the already installed MMFs (OC-3 backbones) to higher-speed pipes at nearly any (loss limited) lengths. Equalizer 19.

20 Conclusions (II) The good news: The modal diffusion constant in MMF is small and once the initialization process is done no more adaptive processes are required in later transmission. No matter how we change the fiber temperature and stress if the launching condition is not changed, the excited modes will be very stable in the MMF. Ethernet protocol can be boosted after the initialization without any modification to accommodate the adaptive process, since there is no more "adaptive equalization Equalizer 2.