Dependence of optically preamplified receiver sensitivity on optical and electrical filter bandwidths-measurement and simulation

Pfennigbauer, Martin; Strasser, Martin; Pauer, M.; Winzer, Peter J.

doi:10.1109/lpt.2002.1003108

Cited by 48 publications

(14 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, the optimum low-pass (LP) filter bandwidth is computed, which is about 0.7 times the transmission symbol rate. This result supports a commonly used LP filter bandwidth design rule, for example, the LP filter bandwidth in experiments and simulations of fiberoptic systems is often chosen to be 0.6 to 0.8 times the transmission rate [18][19][20]. Furthermore, according to our theoretical analysis, two simple system modifications are proposed to improve the whole system performance by just adjusting coefficients of the mapping unit at the transmitter or decision boundaries at the receiver.…”

Section: Introductionsupporting

confidence: 72%

Unequal bit error probability in coherent QPSK fiber‐optic systems using phase modulator based transmitters

Zhao

Agrell

Karlsson

2007

Trans Emerging Tel Tech

View full text Add to dashboard Cite

SUMMARYWe report on the occurrence of unequal bit error probability in a coherent quadrature phase shift keying (QPSK) fiber-optic system. The bit error rates (BER) of two QPSK bits are derived individually based on the developed system model, and they turn out to differ by more than an order of magnitude for a phase modulator based transmitter. The phenomenon, previously unreported, arises because such a transmitter introduces a controlled form of inter-symbol-interference (ISI), and the receiver low-pass filters affect this ISI differently for the two bits. The optimum bandwidth of the receiver low-pass filter is obtained from the analytic derivation, which is about 0.7 times the symbol rate. We propose two simple system modifications, one in the transmitter and one in the receiver, to compensate for the phenomenon and equalize the two BER's. Those modifications improve the system performance by about 2 dB without adding any extra hardware.

show abstract

Section: Introductionsupporting

confidence: 72%

Unequal bit error probability in coherent QPSK fiber‐optic systems using phase modulator based transmitters

Zhao

Agrell

Karlsson

2007

Trans Emerging Tel Tech

View full text Add to dashboard Cite

show abstract

“…The carrier wavelength is 1550 nm and the bandwidth of the optical BP filter is 40 GHz. The Butterworth LP filters have an order of N = 3 and a cutoff frequency of f c = 7 GHz in all simulations [29,32].…”

Section: Simulation and Ser Resultsmentioning

confidence: 99%

“…This noise is commonly modelled as a zeromean additive white Gaussian noise (AWGN) with singlesided noise power spectral density N 0 = F/2hνG, where F is the amplifier noise figure and hν is the photon energy. Although the optimum bandwidth of the optical BP filter is around 2-3 times the data rate depending on the modulation format, the transmitter characteristics [23,29], a sufficiently stable narrow-band optical filter is not a realistic solution. The BP filter in this paper is chosen to have a 3 dB bandwidth of 40 GHz and a Gaussian frequency response.…”

Section: Receivermentioning

confidence: 99%

Intersymbol interference in DQPSK fibre‐optic systems

Zhao¹,

Agrell²,

Karlsson³

2009

Trans Emerging Tel Tech

View full text Add to dashboard Cite

SUMMARYThe limited bandwidth of high-speed transmitter and receiver hardware gives rise to intersymbol interference (ISI) in digital communication systems. The influence of such ISI is investigated in a low-complexity fibreoptical setting, using differential quaternary phase-shift keying (DQPSK) modulation. The ISI is approximated by a memoryless, stochastic model, for use in applications where neither equalisation nor sequence detection can be afforded. The channel capacity of this model is calculated and shown to depend strongly on the transmitter calibration. More than 3 dB is gained, at a target bit error rate (BER) of 10 −6 , by allowing the transmitted phase levels to deviate from their nominal multiples of π/2 by up to 7%.

show abstract

“…In the second approach, formulae are derived for the mean and standard deviation of the received current from which Gaussian approximations to the electrical pdfs are obtained [20], [21]. Even though the pdfs are not Gaussian, simulations have shown that for amplitude-shift-keyed modulation formats, the BER values obtained using the Gaussian pdfs are accurate to within a factor of 2 [22], [23].…”

Section: Introductionmentioning

confidence: 99%

“…Building on the work of Lima et al [13] and Pfennigbauer et al [22], we derive formulae for the mean and variance of the received electric current in each bit, from which we estimate the -factor and the BER. As such, the analytical formulae we derive using the second approach to receiver modeling are complementary to the formulae that Zhang et al [16] obtained using the first approach.…”

Section: Introductionmentioning

confidence: 99%

Receiver Model for Depolarized Signal Due to Polarization-Mode Dispersion and Partially Polarized Noise Due to Polarization-Dependent Loss in an Optical Fiber Communication System

Jiao

Zweck

et al. 2009

J. Lightwave Technol.

View full text Add to dashboard Cite

Abstract-We systematically investigate the combined effect on the system performance in an optical fiber communication system of a signal that is depolarized due to polarization-mode dispersion (PMD) and noise that is partially polarized due to polarization-dependent loss. We derive a formula for the variance of the electric current of the signal due to the signal-noise beating between a depolarized signal and partially polarized noise. We validate this theoretical formula by comparing the -factor calculated using the theory to results obtained from Monte Carlo simulations and experiments. We show that the system performance strongly depends on the power-splitting ratio, the degree of polarization of the noise, and the angle between the states of polarization of the signal and the polarized part of the noise. Although the theoretical formula is derived assuming that the optical fiber only has first-order PMD, we show that for arbitrary fiber, this formula still produces a reliable estimate of the -factor provided that the second-order PMD is on the order of 300 ps 2 or less.

show abstract

Dependence of optically preamplified receiver sensitivity on optical and electrical filter bandwidths-measurement and simulation

Cited by 48 publications

References 12 publications

Unequal bit error probability in coherent QPSK fiber‐optic systems using phase modulator based transmitters

Unequal bit error probability in coherent QPSK fiber‐optic systems using phase modulator based transmitters

Intersymbol interference in DQPSK fibre‐optic systems

Receiver Model for Depolarized Signal Due to Polarization-Mode Dispersion and Partially Polarized Noise Due to Polarization-Dependent Loss in an Optical Fiber Communication System

Contact Info

Product

Resources

About