Performance of Coherent Optical OFDM in WDM System Based on QPSK and 16-QAM Modulation through Super channels

Transcription

1 International Journal of Engineering and Technology Volume 5 No. 3,March, 2015 Performance of Coherent Optical OFDM in WDM System Based on QPSK and 16-QAM Modulation through Super channels Laith Ali Abdul-Rahaim 1, Ibraheem Abdullah Murdas 1 and Mayasah Razzaq 1 Electrical Engineering Department, College of Engineering, Babylon University, Babylon, Iraq ABSTRACT In this paper, Polarization Division Multiplexing Coherent Optical Orthogonal Frequency-division Multiplexing (PDM CO-OFDM) system is designed with two modulation formats (QPSK and 16-QAM) at 40 and 100 Gb/s bit rates to show their linear and nonlinear behaver, and the maximum achievement reach at difference data rate. The system is first simulated with a single channel to analyze the performance. Then the system of 8-WDM channels is simulated at 50 GHz (0.4 nm) channel spacing. In order to evaluate the systems performance, spectrum result, electrical constellation diagrams, maximum reach versus the input power curves and BER verses OSNR curves are presented. A reference BER to evaluate the performance of the systems is taken as PDM CO-OFDM QPSK is found to be the best performance for high capacity networks. In the case of single channel the optimal input power is found to be -2 dbm and 1dBm, which corresponding to the maximum reach of 4080 km and 2480 km at 40 Gb/s and 100 Gb/s respectively. For 8-WDM system the optimum input power was found to be -5 dbm and -2dBm, which corresponding to the maximum reach of 2320 km and 1440km at 40 Gb/s and 100 Gb/s respectively. The advantages of PDM CO-OFDM 16-QAM (improve spectral efficiency, decrease electrical bandwidth) come at the expense of reduced maximum reach and optimum input power. In the case of single channel the optimum input power was found to be -4 dbm and -1dBm, which corresponding to the maximum reach of 1760 km and 960km at 40 Gb/s and 100 Gb/s respectively. For 8-WDM system the optimum input power was determine to be -8 dbm and -4dBm, which give the maximum reach of 1040 km and 480km at 40 Gb/s and 100 Gb/s respectively. Simulation results are obtained using Optisystem (version 11.0) software package. KeyWords: Coherent Optical Ofdm, Qpsk And 16-Qam, Pdm, Wdm System. 1. INTRODUCTION Nowadays, with the advent of emerged multimedia applications, such as YouTube and Internet protocol TV (IPTV), those have again continued to drive the bandwidth demand. It does not that the growth of the internet traffic will slow in the foreseeable future [1]. The Cisco s projection of Internet traffic to 2015[2], which shows bandwidth growth of a factor of 2 every 2 years. This phenomenal growth places tremendous pressure on the underlying information infrastructure at every level, from core to metro and access networks [2]. In the following section, several technologies are identified in optical communication networks arising from the rapid increase in IP traffic and merging applications. To cope with the tremendous growth of internet traffic and new services, the optical system that was dominated by a lowspeed, point-to-point transmission a decade ago needs to be upgraded to support massive networking capabilities with a transmission speed approaching 100 Gb/s[1]. The most relevant technologies to increase the speed of optical transmission system are identified as using Wavelength Division Multiplexing (WDM), the transmission capacity can be easily increased by multiplexing the output of several transmitters for the existing fiber links without installation and alternation of the fiber link as shown in fig. (1).The multiplexed signal is fed into the optical fiber for transmission to the second end, where a demultiplexer sends each channel to it's belong receiver. When N channels at bit rates B 1, B 2 B N are transmitted simultaneously over a fiber of length L, the total bit rate of the WDM link becomes as shown in the following eq. [3, 4]: B T = B 1 + B 2 + B N (1) If bit rates in channel were equal, the system capacity is improved by a factor of N. The most important parameters in the design of a WDM system are the number of channels N, the bit rate B at each channel operates, and the frequency spacing Δvch between two Successive channels. The product NB represents the WDM capacity and the product NΔvch denotes the total bandwidth for a WDM system. Such an approach can be considered as one of the most cost efficient ways to increase the optical link throughput [4, 5]. Fig.(1) Configuration for WDM transmission A different approach to realize a high speed optical transmission system is to use polarization division ISSN: IJET Publications UK. All rights reserved. 141

2 multiplexing (PDM). PDM is a way to increasing system capacity or Spectral Efficiency (SE) by two. This done by taking two independently channels with the same wavelength, but orthogonal polarization states are transmitted simultaneously in a same fiber. At the receiver end, the two polarization channels are separated and detected independently [5, 7]. The binary On/Off Keying (OOK) direct intensity modulation format is the simplest modulation format and has been the only modulation format used in the commercial Dense WDM (DWDM) system until 10Gb/s per channel data rate. For the 40Gb/s DWDM system, more advanced modulation formats such as duobinary, Differential Binary Phase Shift Keying (DBPSK), Quadrature Phase Shift Keying (QPSK) as well as Differential Quadrature Phase Shift Keying (DQPSK) have been introduced to enhance the Spectral Efficiency SE [8]. An recent technology on the way to 100Gb/s line rates and to highest SE WDM networks are highorder optical modulation formats, which encodes m = log2 M data bits on M symbols. Transmission symbol rate, which is reduced by M compared with the data rate. This method given higher channel data rates by using existing equipment have lower-speed and so we can increase the speed of present highspeed electronics and Digital Signal Processing (DSP) [9]. By another words, when assuming a given data rate of a channel, the transmission with lower symbol rates will reduce the spectral width of a WDM channel and thus to improve the SE. It can result in a reduced system reach through a higher optical signal-to-noise ratio (OSNR) requirements coupled with a reduced tolerance toward nonlinear impairments [10]. Orthogonal Frequency Division Multiplexing (OFDM) has been preferably for physical-layer interface in wireless communications recent decade. It has been widely studied in mobile wireless communications to solve frequency-selective fading problem and has been development to wireless network standards and Digital Audio and Video Broadcastings (DAB and DVB-T) in Europe, Asia, Australia, and other parts of the world [11, 12]. The using of OFDM in optical communications occurred recently, therefore a huge number of papers on the theoretical and practical performance of OFDM in many optical systems have published. Coherent Optical Orthogonal Frequency Division Multiplexing (CO-OFDM) combines the Characteristics of coherent detection and OFDM modulation for future high-data rate fiber transmission systems [1, 13, and 14]. a. The Chromatic Dispersion (CD) and Polarization Mode Dispersion (PMD) of the transmission system can be effectively estimated and mitigated. b. The spectra of OFDM subcarriers are partially overlapped, resulting in high optical SE which is very suitable for the future high speed optical transmission system [1]. c. OFDM takes advantage of digital signal processing of Fast Fourier Transforms (FFTs) which suggests that the OFDM has superior scalability over the channel dispersion and data rate [1]. d. By using homodyne architecture, the electrical bandwidth requirement can be greatly reduced for the CO-OFDM transceiver, which is extremely attractive for the high-speed circuit design, where electrical signal bandwidth dictates the cost [1]. e. Migration towards higher order modulation in CO-OFDM systems is enabled simply. In contrast, this requires more complicated optical modulator configuration in single-carrier systems [15]. Due to the superior advantages offered by CO-OFDM, it has been considered to be a promising technology for the high speed optical transport system. All the mentioned technologies to increase the speed of optical transmission system are used in this paper and concentrates on the most promising version of these technologies for future optical networks, the CO-OFDM system. The Literature Survey that interest in the field of this paper began with Shieh et al, [8] in 2007, which reported the first experimental demonstration of CO-OFDM systems. 128 OFDM subcarriers with a nominal data-rate of 8 Gb/s were successfully processed and recovered after 1000 km transmission through standard-single-mode-fiber (SSMF) without optical dispersion compensation. Bao et al, [9] in 2007, demonstrated the transmission performance through simulation for WDM systems with CO-OFDM including the fiber nonlinearity effect. The results showed that the system Q of 8 channels at 10 Gb/s is over 13.0 db for a transmission up to 4800 km of SSMF. Also, they presented a novel technique of Partial Carrier Filling (PCF) for improving the nonlinearity performance of the transmission. The system Q of these channels with a filling factor of 50 % at is improved from 15.1 db to 16.8 db for a transmission up to 3200 km of SSMF. Although PCF technique broadens the bandwidth of an OFDM signal, the bandwidth of OFDM signal with this filling is still comparable with conventional intensity modulation and optical duobinary signal. Yao-Jun, [11] in 2011, employed a new method of fiber nonlinearity post-compensation (FNPC) in a 40-Gb/s CO- OFDM system. The FNPC located before the CO-OFDM receiver includes an Optical Phase Conjugation (OPC) unit and a subsequent 80 km High Nonlinear Fiber (HNLF) as a fiber nonlinearity compensator. The OPC unit is based on a four wave mixing effect in a semiconductor optical amplifier. The fiber nonlinearity impairments in the transmission link are post-compensated for after OPC by transmission through the HNLF with a large nonlinearity coefficient. Simulation results showed that the Nonlinear Threshold (NLT) (for Q> 10 db) can be increased by about 2.5 db and the maximum Q factor is increased by about 1.2 db for the single-channel with periodic dispersion maps. In the 7 WDM channels system with 50-GHz channel spacing, the NLT increases by 1.1 db, equating to a 0.7 db improvement for the maximum Q factor. Wang, [12] in 2012, presented a comparison between 16-Amplitude and Phase Shift Keying (APSK) modulated optical OFDM signal and conventional 16-QAM based on 30.2-Gb/s single-channel and Gb/s WDM single-polarization CO-OFDM systems. Simulation results showed that the 16-APSK modulated optical OFDM signal has a higher tolerance toward fiber nonlinearity, such as self-phase modulation (SPM) and cross phase modulation (XPM). Silva, [12] in 2013, demonstrated a transmission of three 450 Gb/s (9x50-Gb/s) superchannels with SE of 3.73 b/s/hz, using PDM-CO-OFDM with QPSK modulation format, over Pure Silica Core Fiber (PSCF) and amplification by Erbium Doped Fiber Amplifiers (EDFAs). A maximum reach of 3842 km was obtained. The results show that this approach can be used to perform 400 Gb/s long haul transmission with SE near to 4 b/s/hz. The references mentioned in this survey do not take into account the performance of PDM CO-OFDM system with QPSK and ISSN: IJET Publications UK. All rights reserved. 142

3 16-QAM modulation formats for high speed application at different bit rate. Also, the linear and nonlinear limits, and the maximum reach for these systems are addressed in this paper carefully. The aims of this paper are to investigate the performance PDM CO-OFDM system with QPSK and 16- QAM modulation formats for high speed applications. The studied data rate at 40 and 100 Gb/s. The system is first simulated with a single channel and then 8-WDM channels are connected at 50 GHz channel spacing. And to investigate the linear and nonlinear limits and the maximum reach for these systems. The performance of the following systems is tested using Optisystem (version 11.0) software package. This paper is organized as follows: In Section 1, an introduction to the subjects of the paper, literature survey, and the objective to the paper are given. In Section 2, a description of channel impairments, an OFDM principles and Optical-OFDM are considered. In Section 3, a design of PDM CO-OFDM systems with QPSK and 16 QAM modulation formats are given. In Sections 4, the results and discussions of numerical results derived from analysis of the simulated system are given. In Section 8, conclusions for the work are summarized. 2. OPTICAL-OFDM The radio frequency (RF) domain OFDM has been studied in last forty years. OFDM has only recently been applied to optical communications. In 2006, three groups independently proposed optical OFDM for long-haul application. Two major research directions appeared Direct Detection Optical-OFDM (DDO-OFDM) and Coherent Optical-OFDM (CO-OFDM) [26]. Both techniques have advantages. CO-OFDM has the highest performance in power and spectral efficiency and robustness against polarization dispersion. DDO-OFDM has simpler optical receiver architecture, but a frequency guard band is needed to prevent the interference from mixing products which reduces the electrical and optical spectral efficiency. Besides, as some power is allocated to the transmitted carrier, DDO-OFDM also requires more transmitted optical power. Currently, there is extensive research into the performance of both systems and on techniques to mitigate the disadvantages [27]. A generic O- OFDM system can be divided into five functional as shown in fig. (2) (i) RF OFDM transmitter, (ii) RF-To-optical (RTO) up-converter, (iii) optical fiber link, (iv) Optical-To-RF (OTR) down-converter, and (v) RF OFDM receiver[1]. of DDO-OFDM system versions proposed and designed for different applications [3]. DDO-OFDM can be classified into two categories according to how the optical OFDM signal is generated: linearly mapped DDO-OFDM (LM-DDO-OFDM), where the optical OFDM spectrum is a replica of baseband OFDM and nonlinearly mapped DDO-OFDM (NLM-DDO- OFDM), where the optical OFDM spectrum does not display a replica of baseband OFDM [3]. NLM-DDO-OFDM aims to obtain the linear mapping between the baseband OFDM and optical intensity even though there is no linear mapping between the baseband OFDM and the optical field. Since the OFDM signal is encoded in the amplitude not the field, and subsequently nonlinearly mapped DDO-OFDM do not have the same capability of the dispersion resilience as linearly mapped DDO-OFDM therefore it is not fit for long-haul transmission. However, because of its simple implementation, it has become a very attractive option for the short-reach SMF application, multimode fiber and optical wireless systems. In this paper we focus our work on long-haul SSMF transmission, thus only linearly mapped DDO-OFDM will be discussed [1]. The baseband OFDM can be made positive either by adding a DC bias, as in DC-biased optical OFDM (DCO-OFDM), or by clipping the bipolar OFDM signal at zero level and removing all the negative going signals, as in asymmetrically clipped optical OFDM (ACO-OFDM). If only the odd subcarriers are loaded with data symbols at the input of IFFT, all of the clipping noise falls on the even subcarriers, and the data carrying odd subcarriers are not impaired [8]. As shown in fig.(3), the optical spectrum of an LM-DDO- OFDM signal at the output of the O-OFDM transmitter is a linear copy of the baseband OFDM spectrum plus an optical carrier. The position of the main carrier can be one OFDM spectrum bandwidth away, or at the end of OFDM spectrum. Detailed signal representations and modeling for DDO-OFDM systems can be found in [3]. In the receiver, a single photodetector is used where the optical carrier mixes with the optical subcarriers to regenerate the electrical OFDM signal. The output of the optical receiver consists of three terms: the first term is a Direct Current (DC) component that can be easily filtered out. The second term is the fundamental term consisting linear OFDM subcarriers that are to be retrieved. The third term is the second-order nonlinearity product of the optical carrier and the OFDM subcarriers as shown in fig.(3) that needs to be removed. Fig.(2) Schematic of a generic Optical-OFDM communication system The common feature for Direct Detection Optical- OFDM (DDO-OFDM) is, of course, the use of direct detection at the receiver. Here there is no need for an LO laser at the receiver like in the case of Coherent-Optical OFDM (CO- OFDM) systems. This is the key advantage of DDO-OFDM, and this is one important reason why there have been varieties Fig.(3) Spectrum of DDO-OFDM signal The following approaches can be used to minimize the pena ISSN: IJET Publications UK. All rights reserved. 143

4 due to the second-order nonlinearity term; the effect of the unwanted nonlinearity term can be avoided by allocating a sufficient guard band. This guard band is set by RF I/Q modulator at the transmitter, as shown in fig. (4). Another way to reduce the distortion of the nonlinearity term to an acceptable level is by reducing the scaling coefficient that describes the OFDM band strength related to the main carrier as much as possible [1,3]. 2.2 Polarization Division Multiplexed of CO-OFDM It is well known that SMF supports two modes in polarization domain. As shown in fig.(6), a Two-Input-Two-Output (TITO) scheme of CO-OFDM is usually applied to support polarization division multiplexed (PDM) transmission. It consists of two set of CO-OFDM transmitters and receivers, each transmitter and receiver pair for a single polarization. In such a scheme, because the transmitted OFDM information symbol can be considered as polarization modulation or polarization multiplexing, the capacity will be doubled compared with Single-Input-Single-Output (SISO) scheme [27]. Fig.(6) PDM CO-OFDM conceptual diagram. Fig.(4) DDO-OFDM with RF I/Q modulator. 2.1 Coherent Optical-OFDM In CO-OFDM the electric field spectrum of the transmitted optical signal is a replica of the baseband RF OFDM signal, with no need for any optical carrier component to be transmitted. Instead, the carrier component needed for OTR conversion is locally generated at the receiver. Fig.(5) shows the conceptual diagrams of CO-OFDM system with homodyne architecture. The RTO converter is simply an optical I/Q modulator comprises two Mach-Zehnder modulators (MZMs) to transform the I and Q components of the baseband OFDM signal to the optical domain directly, as in fig(5). 2.3 Bit Error Rate & Optical Signal to Noise Ratio Bit Error Rate (BER) estimation can be made by treating the noise sources in terms of Gaussian noise statistics. BER for optimum setting of decision threshold for choosing whether bit is a 1 or 0 is given by the Gaussian error function [27] BER = 1 2 erfc [ I 1 I 0 2( 1 0 ) ] (2) Where I 1 and I 0 are the average photocurrent generated by a 1 bit and 0 bit, respectively. 1 and 0 σ are the noise variances for 1 bit and 0 bit, respectively. Optical Signal to Noise Ratio (OSNR) the ASE added by the optical amplifiers in a longhaul transmission link ultimately limits the feasible transmission distance. When we consider a long-haul tot transmission link with N spans, the total ASE power P ASE in tot [W] added by all optical amplifiers along the link equals P ASE = P ASE N spans. This is valid under the assumption that a single amplifier adds a noise power P ASE and that all spans have the same insertion loss. After N spans, the ratio between signal power and ASE power is known as the OSNR and is defined as [13], P out OSNR = (3) P ASE N spans where P out defines the signal power at the output of the last amplifier of the transmission link and has unit watt. 3. SIMULATION OF PDM CO-OFDM In order to increase the transmission reach of the optical signal and improve the system performance, the system must be designed properly by the accurate selection of the various components in the system. This Section introduces the description of the single channel PDM CO-OFDM system Fig.(5) Architecture of CO-OFDM system with homodyne ISSN: IJET Publications UK. All rights reserved. 144 architecture

5 with QPSK and 16-QAM modulation formats from the transmitter to the receiver. Then, the 8-WDM channels system is discussed considering all the system components. a. Single Channel Description In this section, a description of the single channel system is presented. The simulation parameters of this system are shown in table (1). A Pseudo Random Binary Sequence (PRBS) generator is usually required to generate pseudo random binary sequences. PRBS drives 20 and 50 Gb/s bit rate. Using polarization division multiplexing, entire system bit rate will be increased to 40 and 100 Gb/s. The simulated single channel system is shown in fig. (7) with three main parts: PDM CO-OFDM Transmitter, Optical Fiber, and PDM CO- OFDM Receiver. The description of each part of the system is presented in the following subsections. Table (1) Simulation Parameters Parameter Value Units Bit rate 20 and 50 Gb/s Time window e -06 for 20 Gb/s and e -06 for 50 Gb/s s Sequence length bit In this case, polarization multiplexing is used, the laser output is split into two orthogonal polarization components by Polarization Beam Splitter (PBS), which are modulated separately and then combined using a Polarization Beam Combiner (PBC). The parameters of the simulated CW laser are shown in table (2). Table (2) Parameters of the CW laser Parameter Value Units Frequency THz power variable dbm Linewidth 0.1 MHz Initial phase 0 degree The transmitter section of the simulated system for each polarization is built with two parts: RF OFDM transmitter and RTO Convertor. RF OFDM transmitter is built up using Sequence Generator, OFDM modulator and two low pass Roll off electrical filters as shown in fig.(9). It starts with the Sequence Generator to generate two parallel M-ary symbol sequences from binary signals using two forms of sequence generator that is QPSK and 16-QAM. After that, it passes through OFDM modulator component. In it, the complex symbol streams converted from serial to parallel, and then the digital time domain signal is obtained using IFFT, which is subsequently converted into a real-time waveform through a DAC. After that, each OFDM signal component (I and Q) passes through low pass Roll off filter to reduce the influence of the aliasing components on the system performance. The parameters of the simulated RF OFDM transmitter are shown in table (3). Fig.(7) Single Channel PDM CO-OFDM System. Transmitter Structure The transmitter section is shown in fig.(8): Fig.(9) Block diagram of the RF OFDM transmitter Fig.(8) The transmitter section of single channel system. ISSN: IJET Publications UK. All rights reserved. 145

6 Table (3) Parameters of the RF OFDM Transmitter and Receiver. OFDM Modulator Parameter Value Units Sequence generator type QPSK and 16-QAM Bit/sysmbol 2 for QPSK and 4 for 16-QAM Number of Subcarriers 128 Position Array 64 Number of IFFT/FFT points 256 FFT type Radix-2 Baud Rate for 25 for QPSK and 12.5 for Gbaud/s 50Gb/s Baud Rate for 20Gb/s Sampling Frequency for 50Gb/s Sampling Frequency for 20Gb/s 16-QAM 10 for QPSK and 5 for 16- QAM 50 for QPSK and 25 for 16- QAM 20 for QPSK and 10 for 16- QAM Gbaud/s GHz GHz Spectral Efficiency 2 for QPSK and 4 for 16- QAM b/s/hz Low Pass Cosine Role Off Filter Cut off frequency 0.6 * Bit rate for QPSK and 0.3* Bit rate for 16-QAM GHz Insertion Loss 0 db Depth 100 db Roll off factor 0.2 The RTO up-converter (the I/Q optical modulator) is built up using an X-coupler, two Mach-Zehnder modulators, and an optical combiner. The optical signal from the laser source is applied to the first input port of the coupler to yield the I and Q carrier components, at the output ports, which are fed to the MZMs, as shown in fig.(10). Lithium Niubate MZM (LiNb- MZM) with dual-drive type is used. Each MZM is driven by the positive and negative signals of one of the components of the baseband OFDM signal (I or Q) at the two inputs of modulating signal of the MZM. The output signals from the two MZMs are combined by the optical combiner to form the CO-OFDM signal. The parameters of the simulated RTO Converter are shown in table (4). Table (4) Parameters of the RTO Converter. Match Zehnder Modulator Parameter Value Units Electrical gain (1&2) -1 MZMs extinction ratio 60 db MZMs insertion loss 1 db MZMs switching RF voltage 4 V MZMs switching bias voltage 4 V Booster Amplifier EDFA amplifier gain 10 db Noise Figure 4 db b. Optical Fiber Channel The CO-OFDM signal is then lunched in the optical fiber channel. The optical fiber channel as shown in fig.(11) is composed of spans of 80 km of SMF. The fiber has a loss of 0.2 db/km, a dispersion of 17 ps/nm/km, a dispersion slope coefficient of ps/(km.nm 2 ), a nonlinearity coefficient of 2.6 e (-20) m 2 /W and an effective cross-section of 70 μm 2. The value of the PMD parameter is ps/ km and it is expected that CO-OFDM has high PMD tolerance. Fiber dispersion is compensated by the Dispersion Compensation Fiber (DCF) of 16 km. Its attenuation constant is 0.5 db/km, the dispersion coefficient value is -85 ps/(km.nm), the dispersion slope coefficient is -0.3 ps/(km.nm 2 ) and effective area is 22 μm 2. The attenuation of SMF and DCF are compensated by EDFA in each span. The simulation parameters of the optical fibers channel are listed in table (5). Fig.(11) Block diagram of the of the optical fiber channel After that, an EDFA amplifier is used as a booster amplifier to compensate the losses incurred in the transmitter and then boosts the signal in the optical fiber channel. The EDFA is operated in the power control mode with a gain of 10 db and a noise figure of 4 db. Fig.(10) Block diagram of the RTO up-converter ISSN: IJET Publications UK. All rights reserved. 146

7 Table (5) Parameters of the optical fiber channel Parameter Value Units SMF length 80 km Attenuation Constant α 0.2 db/km Dispersion parameter D 17 ps/(nm.km) Dispersion slope S ps/(km.nm 2 ) Effective area 70 μm 2 PMD parameter ps/ km Nonlinear refractive index n e (-20) m 2 /W to be continued SBS threshold 2.33 dbm SRS threshold 31.7 dbm EDFA_1 Gain 16 db Noise Figure 4 db DCF length 16 km Attenuation Constant α 0.5 db/km Dispersion parameter D -85 ps/(nm.km) Dispersion slope S ps/(km.nm 2 ) Effective area 22 μm 2 PMD parameter ps/ km Nonlinear refractive index n e (-20) m 2 /W EDFA_2 Gain 8 db Noise Figure 4 db be used for extracting I and Q components of the OFDM separately. Similarly, the second coupler is used to split the LO signal into two parts to be mixed with the CO-OFDM signal in I and Q branches. The LO signal in one of the branches is phase-shifted with 90 0 by the optical phase shifter to be coupled to the CO-OFDM signal at the third coupler. The outputs from the third coupler are used to be fed to the PDs of the balanced detector to generate I component of the baseband OFDM signal after subtracting the photocurrents output from the two PDs. The Q component of the baseband OFDM signal is generated by the same way at the balanced detector containing the fourth coupler. The simulation parameters of the OTR convertor are listed in table (7). Fig.(12) The Receiver section of the single channel system. c. Receiver Structure The signal is applied to the receiver to detect it and convert it into an electrical signal after the propagation in the optical fiber channel. An optical band pass filter is used in front of the detection process to decrease the noise and the crosstalk in the received signal. The LO is then split into two branches using PBS and the received signal is separately demodulated by each LO component using two single polarization receivers. The parameters of the simulated LO are shown in table (6). Table (6) Parameters of the LO. Parameter Value Units Frequency THz power 0 dbm Linewidth 0.1 MHz Initial phase 0 degree As shown in fig(12), the PDM CO-OFDM receiver section for each polarization consists from two parts as the transmitter section : OTR Convertor and RF OFDM receiver. The OTR down-converter (homodyne receiver) is built up using four X-couplers, a 90 phase shifter, four PIN photodetectors, and two electrical subtractors, as shown in fig. (13). This OTR conversion network employs balanced detectors for noise cancellation. The first coupler is used to split the incoming complex CO-OFDM signal into two parts to Fig.(13) Block diagram of the OTR down converter. Table (7) Parameters of the OTR diagram. Photo detector Parameter Value Units Photodetector type PIN Responsivity 1 A/W Dark current 10 na X-Coupler Coupling Coefficient 0.5 Additional Losses 1 db ISSN: IJET Publications UK. All rights reserved. 147

8 The received electrical signals (I and Q) are then amplified with two electrical amplifiers having a gain of 20dB as shown in fig.(14). After amplification, the signals are passed through Low Pass Roll off filters to eliminate the frequencies above required band, usually, the same characteristics as the low-pass filters used in the RF-OFDM transmitter. Fig.(14) Block diagram of the RF OFDM receiver Then these signals are fed to an OFDM demodulator component. In it, the received baseband OFDM signal is sampled by ADC, and then the FFT of each OFDM subcarrier is taken to find the original transmitted symbol. After that Sequence Decoder decodes two parallel M-ary symbol sequences to binary signals. The decoding process is done by QPSK and 16-QAM Sequence decoder. Then, the binary signal passes through NRZ pulse generator to generate a Non Return to Zero (NRZ) electrical signal. The 3R generator is used to generate the original bit sequence, and the modulated electrical signal to be used for BER analysis. The simulation parameters of the RF OFDM Receiver are the same parameter in table (3). 3.5 WDM System Description The description of the PDM CO-OFDM system with 8-WDM channel can be divided into two parts: Transmitter Structure The block diagram of the transmitter side of the 8-WDM PDM CO-OFDM is shown in fig. (15). The PDM CO-OFDM channels are fed into a multiplexer. The simulation parameters of the multiplexers and the demultiplexers used in the WDM system are shown in table (8). Fig.(15) The transmitter side of the simulated WDM system. Table (8) Parameters of the optical multiplexers and demultiplexers for WDM System. Mux/Demux Parameter Value Units Frequency spacing 50 GHz GHz Bandwidth 40 for QPSK and 20 for 16-QAM GHz Insertion loss 2 db Filter type Gaussian Filter order 4 As in the case of the single channel transmission, an optical booster amplifier with the same parameters is used in the WDM system to accommodate the losses incurred in the signal at the transmitter Receiver Structure Once the propagation of the signal passes through the optical fiber channel, it is then received and detected in the receiver. The receiver structure of the 8WDM channels PDM CO- OFDM system is shown in fig.(16).the channels are demultiplexed with same parameters used in table (8). ISSN: IJET Publications UK. All rights reserved. 148

9 Fig.(16) The receiver side of the simulated WDM system. Fig.(17) RF Spectrum of PDM CO-OFDM QPSK signal (a) before (b) after 1200 km transmission distance. 4. Simulation Results: The simulation results in this paper showed the behaver of PDM CO-OFDM QPSK and 16-QAM system at 40and 100Gbit/s to understand the linear and nonlinear limits, and the maximum achievable reach at each data rate. The data rate of 100Gbit/s was chosen to match the speed of the next generation optical networks. 4.1 Single Channel PDM CO-OFDM System with QPSK modulation:- The main results in this system are: Spectrum Result Fig.(17) shows the RF spectrum analyzer PDM CO-OFDM QPSK signal before and after 1200 km transmission distance at 100 Gb/s. Fig.(18) shows the optical spectrum of PDM CO-OFDM QPSK signal before and after 1200 km transmission distance at 100 Gb/s. The 20 db bandwidth of the signal is around 25 GHz. Fig.(18) Optical spectrum of PDM CO-OFDM QPSK signal (a) before and (b) after 1200 km transmission distance Received Electrical Constellation Diagram Fig. (19) shows the received constellation diagram after removing the chromatic dispersion by DCF for 1200 km with different input power. The constellation of each OFDM symbols corresponding QPSK data points is recovered with four distinct clusters of. The residual noises spreading constellation point are mainly from ASE in transmission system, nonlinearity of optical fibers and PMD. According to the SPM formula, there is no doubt that a higher input power would lead to a larger nonlinearity distortion. As shown in fig(20), when the fiber transmission distance increases, the distortion increases. This simulation result is consistent with the discussions in section 2.2 that the phase shift caused by ISSN: IJET Publications UK. All rights reserved. 149

10 fiber nonlinearity is positively related with fiber transmission length. (C) Fig.(19) Received constellation diagram after 1200 km with different input power. a) -2 dbm, b) 0 dbm, c) 2 dbm (C) Fig(20) Received constellation single channel system with 0dBm input power after different transmission distances (a) 1520 km, (b) 1840 km (c) 2160 km Measurements of Maximum reach The maximum reach was measured as a function of input power at 40 and 100 Gbit/s (Fig.(21)), to characterize the linear and nonlinear limits for QPSK PDM CO-OFDM transmission. This done by assuming the BER = The results are summarized in table (9). For 100 Gbit/s QPSK PDM CO-OFDM transmission the input power had to be increased by ~4dBm to get the same distances in that been get in transmission at 40 Gbit/s. This is because the in-band noise power scales linearly with the symbol rate due to the associated broadening spectral. ISSN: IJET Publications UK. All rights reserved. 150

11 Fig.(21) Measured maximum reach of single channel PDM CO-OFDM QPSK transmission at 40Gbit/s and 100Gbit/s. Table (9) Maximum reach and Optimum Input Power of single channel PDM CO-OFDM QPSK system at 40 and 100Gb/s. Bit rate (Gb/s) Max. Reach (km) Optimum Input Power (dbm) Single channel CO-OFDM PDM QPSK system OSNR Measurement To characterize the PDM CO-OFDM QPSK signal at 40 Gbit/s and 100 Gbit/s, the BER was determined as a function of OSNR, and the results are shown in fig.(22). The required OSNR for the 100 Gb/s is 8.32 db at BER=10-12, which is 2.19 db higher than that for 40 Gb/s. This can be attributed to increase in symbol rate by a factor of more than two resulted in additional 2.19 db OSNR retribution due to the bandwidth increasing. the single amplitude (constant modulus), that improve the tolerance to nonlinearity compared to modulation of multilevel formats. The advantage of generating 100Gbit/s PDM CO-OFDM signals using 16-QAM, however, is has lower symbol rate of 12.5Gbaud requires less by a factor of two compared to QPSK. A lower symbol rate dangles the requirements on the bandwidth of the electronics at the transmitter and the receiver, reducing the cost of electronic and electro-optical components used. 16-QAM can also increase the spectral efficiency in WDM transmission, so can be reduced the spacing between the WDM channels. Another advantage of 16-QAM over QPSK signals is that at higher bitrates (e.g. 200Gbit/s), it becomes impractical to use since a QPSK signal would require a symbol rate of 100Gbaud for which electronics at the transmitter and receiver is not easily available. Therefore, a 25Gbaud PDM CO-OFDM 16-QAM could be better to obtain a 200Gbit/s. Similarly to the QPSK this section shows the comprehensive study of PDM CO- OFDM 16-QAM at 40 and 100 Gb/s in terms of maximum reach. Similar to the QPSK experiments simulate in section (4.1), the main results measured in the 16-QAM PDM CO- OFDM single channel are: Spectrum Result Fig.(23) shows the RF spectrum analyzer PDM CO-OFDM 16-QAM signal before and after 560 km transmission distance at 100 Gb/s. Fig.(24) shows the optical spectrum of PDM CO-OFDM 16- QAM signal at 100Gb/s before and after 560 transmission distance. The 20 db bandwidth of the signal is around 12.5 GHz. Fig.(22) Measured OSNR for 40 Gbit/s and 100 Gbit/s single channel PDM CO-OFDM QPSK. 4.2 Single Channel PDM CO-OFDM System with 16-QAM modulation format Fig.(23) RF Spectrum of 100 Gb/s PDM CO-OFDM 16- QAM signal (a) before (b) after 560 km transmission distance The bit-rate of 100Gbit/s can be achieved using a QPSK PDM CO-OFDM system operating at 25Gbaud. Indeed, such a system can give ultra-long distances, potentially sufficient to use in trans-oceanic routes. The QPSK CO-OFDM PDM signal needed lower OSNR than higher-order modulation formats for the equivalent data rates. It also advantage from ISSN: IJET Publications UK. All rights reserved. 151

12 Fig.(24) RF Spectrum of 100 Gb/s PDM CO-OFDM 16- QAM signal (a) before (b) after 560 transmission distance Received Electrical Constellation Diagrams Fig.(25) shows the received constellation diagram after 560 km at different input power. Fig.(26) shows the received constellation diagram with 0 dbm input power at different distances. (C) Fig(26) Received constellation single channel system with 0dBm input power after different transmission distances (a) 560 km, (b) 720 km(c) 900 km Maximum reach measurements The performance of transmission PDM CO-OFDM 16-QAM signals are simulated in this section. The curves of maximum reach were determined as a function of input power into the fiber (at BER = ) and are shown in fig.(27). In the case of 40Gbit/s the optimum input power was found to be -4dBm, corresponding to the maximum reach of 1460km. At 111Gbit/s PDM CO-OFDM 16-QAM transmission the optimum input power was been -1dBm to given a maximum reach of 960km. (C) Fig.(25) Received constellation diagram after 560 km with different input power. a) -2 dbm, b) 0 dbm, c) 2 dbm. Fig.(27) Measured maximum reach of single channel PDM CO-OFDM 16-QAM transmission at 40Gbit/s and 100Gbit/s It is clear that 16-QAM modulation given (more spectral efficiency and lower electrical bandwidth). This is at the ISSN: IJET Publications UK. All rights reserved. 152

13 expense of decreased maximum reach and optimum input power. The results are summarized in table (10). Table (10) Maximum reach and Optimum Input Power of single channel PDM CO-OFDM 16-QAM system at 40 Gb/s and 100Gb/s. Bit rate (Gb/s) Max. Reach (km) Optimum Input Power (dbm) Single Channel CO-OFDM PDM 16-QAM System OSNR Measurement To characterize the PDM CO-OFDM 16-QAM signal at 40 Gbit/s and 100 Gbit/s, the BER was determined as a function of OSNR, and the results are shown in fig.(28). The required OSNR for the 100 Gb/s is db at BER=10-12, which is 2.4 db higher than that for 40 Gb/s. Fig.(29) RF Spectrum of PDM CO-OFDM QPSK signal for the 4th channel (a) before (b) after 1200 km transmission distance. Fig.(30) shows the optical spectrum of 8-WDM channels PDM CO-OFDM QPSK signal before and after 1200 km transmission distance at 100 Gb/s. Fig.(28) Measured OSNR for 40 Gbit/s and 100 Gbit/s single channel PDM CO-OFDM 16-QAM WDM Channels PDM CO-OFDM system with QPSK modulation format The single-channel systems described in section (4.1) were used to establish the transmission performance of an upperbound for all practical optical fiber transport systems based on WDM systems. WDM allows the optical infrastructure to be shared amongst many channels, thus decreasing the cost of transmitted data information in a fully loaded system. In these operating systems, nonlinear effects become important sources of weaknesses. Because of these additional weaknesses, it is clear that single-channel transmission alone can underestimate the performance of WDM systems with several tens of wavelength channels [3]. The aim of this section is to investigate the performance of PDM CO-OFDM QPSK and 16-QAM signals in the presence of adjacent WDM channels. Similar to the single channel system carried out in section 4.1, the main results measured in the 8-WDM channels PDM CO- OFDM QPSK are: Fig.(30) Optical Spectrum of PDM CO-OFDM QPSK signal for the 4th channel (a) before (b) after transmission Received Electrical Constellation Diagrams Fig. (31) shows the received constellation diagrams for central WDM channel after 1200km at different input power. The constellation is worse with WDM system, which shows the increasing of the nonlinear effects. Fig. (32) shows the received constellation diagrams with -4 dbm input power after different transmission distances Spectrum Result Fig.(29) shows the RF spectrum analyzer for central channel (channel 4) PDM CO-OFDM QPSK signal before and after 1200 km transmission distance at 100 Gb/s. ISSN: IJET Publications UK. All rights reserved. 153

14 (C) Fig.(31) Received constellation diagram after 1200 km with different input power. a) -4 dbm, b) -2 dbm, c) 0 dbm. (C) Fig(32) Received constellation single channel system with -4 dbm input power after different transmission distances (a) 1200 km, (b) 1360 km (c)1520 km Maximum reach measurements To characterize the performance of transmission 8-WDM PDM CO-OFDM QPSK signals at 40 Gb/s and 100Gbit/s, the maximum transmission distance was measured (at BER = ) as a function of the input power per channel (Fig.(33)). The transmission measurements were performed for a central WDM channel (channel 4), because of the central channel experiences the maximum amount of nonlinearity. This causes the worst-case scenario and, hence, the lower bound in the transmission performance. The optimal input power and maximum reach is decreased in 8-WDM system compared with the single channel system. The results are summarized in table (11). Fig. (33) Measured maximum reach of 8-WDM channels PDM CO-OFDM QPSK system, measured at 40 and 100 Gbit/s. ISSN: IJET Publications UK. All rights reserved. 154

15 Table (11) Maximum reach and Optimum Input Power of 8-WDM channels PDM CO-OFDM QPSK system at 40 Gb/s and 100Gb/s. Bit rate (Gb/s) Max. Reach (km) Optimum Input Power (dbm) 8 WDM channels PDM CO-OFDM QPSK System OSNR Measurement Fig.(34) shows the OSNR measurements, measured for the central WDM channel, for both 40 and 100 Gb/s. Fig.(35) RF Spectrum of PDM CO-OFDM 16-QAM signal for 4th channel (a) before (b) after 560 km transmission distance. Fig.(34)Measured OSNR for 8-WDM PDM CO-OFDM QPSK at 40 and 100 Gbit/s. The required OSNR for the 100 Gb/s is db at BER=10-12, which is 2.8 db higher than that for 40 Gb/s, this compares to 2.19 db in a single-channel PDM CO-OFDM QPSK system WDM Channels PDM CO-OFDM system with 16-QAM modulation format Similar to the single channel system carried out in section 4.1, the main results measured in the 8-WDM channels PDM CO- OFDM 16-QAM are: Spectrum Result Fig.(35) shows the RF spectrum analyzer for central channel PDM CO-OFDM 16-QAM signal before and after 560 km transmission distance at 100 Gb/s. Fig.(36) shows the optical spectrum of 8- WDM channels PDM CO-OFDM 16-QAM signal before and after 560 km transmission distance at 100Gb/s. Fig.(36) Optical Spectrum of PDM CO-OFDM 16-QAM signal for 4th channel (a) before (b) after 560 km transmission distance Received Electrical Constellation Diagrams Fig. (37) shows the received constellation diagrams for central WDM channel after 560 km at different input power. Fig. (38) shows the received constellation diagrams with -4dBm input power after different transmission distances. ISSN: IJET Publications UK. All rights reserved. 155

16 4.4.3 Maximum reach measurements To characterize the linear and nonlinear transmission performance of 8- WDM channels PDM CO-OFDM 16-QAM system, the maximum reach was measured as a function of the input power per channel (Fig. (39)). The results are summarized in table (12). Fig. (39) Maximum reach of 8-WDM channels PDM CO- OFDM 16-QAM systems at 40 Gb/s and 100Gb/s. Table (12) Maximum reach and Optimum Input Power of 8-WDM channels CO-OFDM PDM 16- QAM system at 40 Gb/s and 100Gb/s. (C) Fig.(37) Received constellation diagram after 480 km with different input power. a) -4 dbm, b) -2 dbm, c) 0 dbm. Bit rate (Gb/s) Max. Reach (km) Optimum Input Power (dbm) 8-WDM channels CO-OFDM PDM 16-QAM System OSNR Measurement Fig.(40) shows the OSNR measurements, measured for the central WDM channel, for both 40 and 100 Gb/s. Fig. (40) Measured OSNR for 40 Gbit/s and 100 Gbit/s 8- WDM channels PDM CO-OFDM 16-QAM. (C) Fig(38) Received constellation single channel system with -4 dbm input power after different transmission distances (a) 560 km, (b) 640 km(c) 720 km. The required OSNR for the 100 Gb/s is 17.5 db at BER=10-12, which is 3.4 db higher than that for 40 Gb/s, this compares to 2.4 db in a single-channel PDM CO-OFDM 16-QAM system. The results in this Section represent the first investigation to the performance of single channel and 8-WDM channels PDM CO-OFDM QPSK and 16-QAM. Overall, it was clear that maximum transmission distance in the case of WDM system was lower than in single-channel system because of nonlinear effects from the neighboring channels. The results for singlechannel and 8-WDM channels PDM CO-OFDM QPSK system at 40 and 100Gbit/s are summarized in Table (13). ISSN: IJET Publications UK. All rights reserved. 156

17 Table (13) Comparison of maximum reach and optimum input power between single-channel and 8- WDM channels PDM CO-OFDM QPSK system at 40 and 100Gbit/s. Bit rate (Gb/s) Max. Reach (km) Optimum Input Power (dbm) Single channel CO-OFDM PDM QPSK System WDM channels PDM CO-OFDM QPSK System At 111Gbit/s the existence of the additional 7-WDM channels in the case of QPSK reduced the maximum achievable reach from 1440 km to 2480 km, compared to the single-channel experiment. Nevertheless, the maximum achievable transmission distance of 1440 km in the case of WDM means that 100Gbit/s QPSK solution can be potentially used for longhaul transmission. Finally, the 8-WDM channel PDM CO- OFDM 16-QAM transmission at 100Gbit/s was studied. The presence of additional channels reduced the maximum reach from 061km to 480 km compared to the single-channel experiment. Still, 480km is long transmission distance for 100Gbit/s WDM PDM CO-OFDM 16-QAM. The results for single-channel and 8-WDM channels PDM CO-OFDM 16- QAM system at 40 and 100Gbit/s are summarized in Table (14). Table (14) Comparison of maximum reach and Optimum input power between single-channel and 8- WDM channels PDM CO-OFDM 16-QAM system at 40 and 100Gbit/s. Bit rate (Gb/s) Max. Reach (km) Optimum Input Power (dbm) Single channel CO-OFDM PDM 16-QAM System WDM channels PDM CO-OFDM 16-QAM System Conclusions The design of PDM CO-OFDM system operating with QPSK and 16-QAM modulation formats has been investigated for high speed optical transmission system. From the simulation results, the following points can be concluded. For the PDM CO-OFDM system operating with QPSK at 100 Gb/s, the bandwidth of the optical signal is 25GHz. Whereas in the system operating with 16-QAM the bandwidth is 12.5 GHz for the same bit rate. This means that the optical signal of the QPSK system occupies double the bandwidth of the 16-QAM system for the same bit rate. The PDM CO-OFDM QPSK signal also has a higher tolerance to nonlinearity compared to CO-OFDM PDM 16-QAM. The advantages of 16-QAM modulation (increased spectral efficiency, lower electrical bandwidth) come at the expense of reduced maximum reach and optimum input power. For example, in the case of 100Gbit/s PDM CO-OFDM 16-QAM the optimum input power was found to be -1dBm, corresponding to the maximum reach of 960km while for PDM CO-OFDM QPSK transmission the optimum input power was found to be 1dBm, corresponding to a maximum reach of 2480km. Increasing system bit rate from 40 Gb/s to 100 Gb/s leads to increase OSNR penalty. For example, the required OSNR for the 100 Gb/s PDM CO-OFDM QPSK is 8.32 db at BER=10-12, which is 2.19 db higher than that for 40 Gb/s. Also, the required OSNR for the 100 Gb/s PDM CO-OFDM 16-QAM is db at BER=10-12, which is 2.4 db higher than that for 40 Gb/s. This can be attributed to increase in symbol rate by a factor of more than two resulted in an additional OSNR penalty due to the doubling of the bandwidth. The PDM CO-OFDM QPSK signal has a lower required OSNR than 16-QAM for the equivalent bit rates. For example, the required OSNR for the 100 Gb/s PDM CO-OFDM QPSK is 8.32 db at BER=10-12, while for PDM CO-OFDM 16-QAM the required OSNR is db. Increasing the number of channels (increasing the system capacity) leads to worsen the system performance due to additional nonlinear effects from the neighboring channels. For example, at 100Gbit/s the presence of the additional 7- WDM channels in the case of PDM CO-OFDM QPSK decreased the maximum achievable reach from 2480 km to 1440 km, compared to the single-channel experiment. Also, the 8-WDM channel PDM CO-OFDM 16-QAM transmission at 100Gbit/s was studied. The presence of additional channels decreased the maximum reach from 960km to 480 km compared to the single-channel experiment. REFERENCES [1] Y. Tang, High-speed Optical Transmission System Using Coherent Optical Orthogonal Frequency- Division Multiplexing, PhD. thesis, Department of Electrical and Electronic Engineering, University of Melbourne, Australia, June, [2] W. Shieh and I. Djordjevic, OFDM for optical communications, Elsevier, Amsterdam, [3] A. Raheem, Spectral Efficiency Improvement of the Optical Communication Systems, MSC. Thesis, University of Technology, May, [4] Z. Xiang and Y. Jianjun Recent Progress in Highspeed and High Spectral Efficiency Optical Transmission Technology Optical Fiber Communications, July, [5] S. Jansen, D. van den Borne, M. Kuschnerov, Advances in Modulation Formats for Fiber-Optic Transmission Systems, CLEO, Laser Science to Photonic Applications, [6] A. Lowery, L. Du and J. Armstrong, "Performance of optical OFDM in ultralong-haul WDM lightwave systems ", Journal of Lightwave Technology Vol. 25, No. 1, January, [7] W. Shieh, X. Yi, Y. Ma, and Q. Yang, Coherent optical OFDM: has its time come [Invited], Journal of Optical Networking, Vol. 7, No. 3, March, ISSN: IJET Publications UK. All rights reserved. 157