Evaluation of optical ZP-OFDM transmission performance in multimode fiber links

Abstract

Abstract:In this paper, the performance of Zero Padding Orthogonal Frequency Division Multiplexing (ZP-OFDM) on intensity modulationdirect detection (IM-DD) multimode fiber (MMF) links is assessed by means of numerical simulations. The performance of ZP-OFDM is compared to classical Cyclic Prefixed form of OFDM (CP-OFDM) which is known to offer a limited performance in terms of symbol recovery in subcarriers suffering severe fading. Simulations results show that ZP-OFDM is able to reach 29 Gbps in 99.5% of all installed MMF links up to 600 meters compared to 14 Gbps for CP-OFDM when a 64 points fast Fourier transform (FFT) size is used. The use of ZP-OFDM makes it possible to increase the link length up to 1200 and 2400 m with a 25 Gbps data rate if the FFT sizes are increased to 128 and 256 points, respectively; whereas the CP-OFDM scheme will offer a maximum data rate of 10 Gbps in both cases. ZP-OFDM can be an alternative to adaptive loading OFDM schemes without the need of a negotiation between transmitter and receiver, reducing the system deployment complexity and increasing the flexibility in scenarios with multiple receivers.

Combined data detection scheme for zero padded OFDM signals in MMF links

“…2, where NVB is 8 and 12, respectively. It can be seen in both figures how the performance of the proposed data detection receivers for ZP-OFDM is improved respect to the use of the ZF receiver proposed in [4]. It is also shown that the combined OLA+V-BLAST receiver can achieve the performance of V-BLAST with low values of NVB.…”

“…The model employed in the numerical simulations defines a single 1550 nm wavelength unamplified intensity modulated and direct detected (IM-DD) optical MMF link and has been previously used to study optical OFDM [4] [9]. The transmitter models a directly modulated laser (DML) driven by a DMT signal limited in power to 5 dBm to prevent fiber nonlinearities.…”

“…The delay of each mode is modeled as an independent random variable with uniform distribution around the average delay, with a maximum deviation equal to half the maximum differential mode delay (DMD). This one is given by DMD = 2 ns/km and represents the worst 5% of installed MMF links operating at 1550 nm wavelength [4].…”

“…1 exhibit that the proposed combined receiver detecting only 5 carriers with V-BLAST (NVB = 5) has a performance similar to detecting all carriers with V-BLAST. If we set a CDF threshold of 90% to choose the modulation order [4][9], the proposed receivers can employ a 128-QAM constellation whereas the ZF receiver should use 64-QAM, this means that it is possible to increment the data rate from 29.06 Gb/s up to 33.9 Gb/s in more than 99.5% of all installed MMF links up to 600 m. As expected, conventional CP-OFDM offers a quite lower performance, being limited to use 16-QAM (19.37 Gb/s) in the same scenario. The performance of the OLA receiver in ZP-OFDM is the same as the one of CP-OFDM.…”

“…However, there is some optical scenarios characterized by strong frequency selectivity where performance of CP-OFDM is reduced. Zero padded OFDM (ZP-OFDM), where the guard interval is composed of zero samples, has been proposed to overcome this limitation: in single mode fiber (SMF) links it provides higher tolerance to polarization-mode dispersion (PMD) [3] and in MMF links it increases the link capacity or coverage [4].…”