Channel Equalization and Phase Noise Compensation Free DAPSK-OFDM Transmission for Coherent PON System

Similar documents
Novel OBI noise reduction technique by using similar-obi estimation in optical multiple access uplink

Phase Modulator for Higher Order Dispersion Compensation in Optical OFDM System

Performance Analysis Of Hybrid Optical OFDM System With High Order Dispersion Compensation

COHERENT DETECTION OPTICAL OFDM SYSTEM

60 Gbit/s 64 QAM-OFDM coherent optical transmission with a 5.3 GHz bandwidth

Single channel and WDM transmission of 28 Gbaud zero-guard-interval CO-OFDM

Utilizing Self-Seeding RSOA with Faraday Rotator Mirror for Colorless Access Network

Phase Noise Compensation for Coherent Orthogonal Frequency Division Multiplexing in Optical Fiber Communications Systems

Lecture 13. Introduction to OFDM

System Impairments Mitigation for NGPON2 via OFDM

Optical Fiber Technology

Optical signal suppression by a cascaded SOA/RSOA for wavelength reusing reflective PON upstream transmission

Performance analysis of direct detection and coherent detection system for optical OFDM using QAM and DPSK

Performance Improvement of OFDM System using Raised Cosine Windowing with Variable FFT Sizes

Performance Evaluation of Wireless Communication System Employing DWT-OFDM using Simulink Model

Performance Evaluation using M-QAM Modulated Optical OFDM Signals

An OFDM Transmitter and Receiver using NI USRP with LabVIEW

Lecture 3: Wireless Physical Layer: Modulation Techniques. Mythili Vutukuru CS 653 Spring 2014 Jan 13, Monday

Simulative Investigations for Robust Frequency Estimation Technique in OFDM System

Local Oscillators Phase Noise Cancellation Methods

Blind symbol synchronization for direct detection optical OFDM using a reduced number of virtual subcarriers

A WDM passive optical network enabling multicasting with color-free ONUs

An Amplified WDM-PON Using Broadband Light Source Seeded Optical Sources and a Novel Bidirectional Reach Extender

Simple Self-Homodyne Detection Scheme for Optical OFDM With Inserted Pilot Subframes and Application in Optical Access Networks

A Novel Multi-band CO-OFDM based Long Reach Passive Optical Network Architecture

Single Carrier Ofdm Immune to Intercarrier Interference

Clipping and Filtering Technique for reducing PAPR In OFDM

An Equalization Technique for Orthogonal Frequency-Division Multiplexing Systems in Time-Variant Multipath Channels

All-VCSEL based digital coherent detection link for multi Gbit/s WDM passive optical networks

CHAPTER 3 ADAPTIVE MODULATION TECHNIQUE WITH CFO CORRECTION FOR OFDM SYSTEMS

A HIGH SPEED WDM PON FOR DOWNSTREAM DPSK ASK SIGNALS AND UPSTREAM OOK SIGNAL WITH BROADCAST CAPABILTY

Power margin improvement for OFDMA-PON using hierarchical modulation

Comparative Study of FLIP-OFDM and ACO-OFDM for Unipolar Communication System

Light Polarized Coherent OFDM Free Space Optical System

DIGITAL signal processing has been revolutionizing communication

Channel Estimation by 2D-Enhanced DFT Interpolation Supporting High-speed Movement

Gigabit Transmission in 60-GHz-Band Using Optical Frequency Up-Conversion by Semiconductor Optical Amplifier and Photodiode Configuration

Joint nonlinearity and chromatic dispersion pre-compensation for coherent optical orthogonal frequency-division multiplexing systems

Implementation and Comparative analysis of Orthogonal Frequency Division Multiplexing (OFDM) Signaling Rashmi Choudhary

Analytical Estimation in Differential Optical Transmission Systems Influenced by Equalization Enhanced Phase Noise

Performance Analysis of OFDM System with QPSK for Wireless Communication

Next-Generation Optical Fiber Network Communication

Adaptation of AMO-FBMC-OQAM in optical access network for accommodating asynchronous multiple access in OFDM based uplink transmission

Emerging Subsea Networks

Performance Evaluation of OFDM System with Rayleigh, Rician and AWGN Channels

Coherent Optical OFDM System or Long-Haul Transmission

Performance analysis of OFDM with QPSK using AWGN and Rayleigh Fading Channel

Practical issue: Group definition. TSTE17 System Design, CDIO. Quadrature Amplitude Modulation (QAM) Components of a digital communication system

Carrier Frequency Offset Estimation Algorithm in the Presence of I/Q Imbalance in OFDM Systems

2.

Lecture 7 Fiber Optical Communication Lecture 7, Slide 1

Colorless Amplified WDM-PON Employing Broadband Light Source Seeded Optical Sources and Channel-by-Channel Dispersion Compensators for >100 km Reach

Design and Implementation of OFDM System and Reduction of Inter-Carrier Interference at Different Variance

Underwater communication implementation with OFDM

ORTHOGONAL frequency division multiplexing (OFDM)

Comparison of ML and SC for ICI reduction in OFDM system

TSTE17 System Design, CDIO. General project hints. Behavioral Model. General project hints, cont. Lecture 5. Required documents Modulation, cont.

Optical Fiber Technology

Optical Wireless Communication System with PAPR Reduction

Techniques for Mitigating the Effect of Carrier Frequency Offset in OFDM

Outline / Wireless Networks and Applications Lecture 7: Physical Layer OFDM. Frequency-Selective Radio Channel. How Do We Increase Rates?

A Research Concept on Bit Rate Detection using Carrier offset through Analysis of MC-CDMA SYSTEM

OFDMA and MIMO Notes

OFDM AS AN ACCESS TECHNIQUE FOR NEXT GENERATION NETWORK

SC - Single carrier systems One carrier carries data stream

ORTHOGONAL frequency-division multiplexing

Study of Turbo Coded OFDM over Fading Channel

ORTHOGONAL frequency division multiplexing

Decrease Interference Using Adaptive Modulation and Coding

Optical Complex Spectrum Analyzer (OCSA)

Digital Modulation Schemes

ANALYSIS OF BER AND SEP OF QPSK SIGNAL FOR MULTIPLE ANENNAS

Signal Conditioning Parameters for OOFDM System

4x4 Time-Domain MIMO encoder with OFDM Scheme in WIMAX Context

S.D.M COLLEGE OF ENGINEERING AND TECHNOLOGY

Simplified Levenberg-Marquardt Algorithm based PAPR Reduction for OFDM System with Neural Network

A Novel On-Channel Repeater for Terrestrial-Digital Multimedia Broadcasting System of Korea

DESIGN, IMPLEMENTATION AND OPTIMISATION OF 4X4 MIMO-OFDM TRANSMITTER FOR

Multi-Carrier Systems

CHAPTER 1 INTRODUCTION

A Novel of Low Complexity Detection in OFDM System by Combining SLM Technique and Clipping and Scaling Method Jayamol Joseph, Subin Suresh

Multi-carrier Modulation and OFDM

Optimized BPSK and QAM Techniques for OFDM Systems

Orthogonal frequency division multiplexing (OFDM)

Noise Plus Interference Power Estimation in Adaptive OFDM Systems

Citation Wireless Networks, 2006, v. 12 n. 2, p The original publication is available at

Bit error rate simulation using 16 qam technique in matlab

Table 10.2 Sensitivity of asynchronous receivers. Modulation Format Bit-Error Rate N p. 1 2 FSK heterodyne. ASK heterodyne. exp( ηn p /2) 40 40

NG-PON2 Optical Components Update. Hal Roberts System Architect

Demonstration of bi-directional LED visible light communication using TDD traffic with mitigation of reflection interference

MITIGATING CARRIER FREQUENCY OFFSET USING NULL SUBCARRIERS

Fiber Nonlinearity Compensation Methods (used by our group)

A New Adaptive Channel Estimation for Frequency Selective Time Varying Fading OFDM Channels

Anju 1, Amit Ahlawat 2

Performance Analysis of Multi-format WDM-RoF Links Based on Low Cost Laser and SOA

OFDMA PHY for EPoC: a Baseline Proposal. Andrea Garavaglia and Christian Pietsch Qualcomm PAGE 1

INTERFERENCE SELF CANCELLATION IN SC-FDMA SYSTEMS -A CAMPARATIVE STUDY

Peak-to-Average Power Ratio (PAPR)

SPARSE CHANNEL ESTIMATION BY PILOT ALLOCATION IN MIMO-OFDM SYSTEMS

Figure 1: Basic OFDM Model. 2013, IJARCSSE All Rights Reserved Page 1035

Transcription:

Compensation Free DAPSK-OFDM Transmission for Coherent PON System Volume 9, Number 5, October 2017 Open Access Kyoung-Hak Mun Sang-Min Jung Soo-Min Kang Sang-Kook Han, Senior Member, IEEE DOI: 10.1109/JPHOT.2017.2729579 1943-0655 2017 IEEE

Compensation Free DAPSK-OFDM Transmission for Coherent PON System Kyoung-Hak Mun, Sang-Min Jung, Soo-Min Kang, and Sang-Kook Han, Senior Member, IEEE Department of Electrical and Electronic Engineering, Yonsei University, Seoul 07322, South Korea DOI:10.1109/JPHOT.2017.2729579 1943-0655 C 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. Manuscript received June 15, 2017; revised July 14, 2017; accepted July 17, 2017. Date of publication July 25, 2017; date of current version August 3, 2017. This work was supported by the ICT R&D program of MSIP/IITP, South Korea, [2014-0-00538, Next-generation coherent optical access physical network]. Corresponding author: Sang-Kook Han (e-mail: skhan@yonsei.ac.kr.) Abstract: Channel equalization and phase noise compensation are required in coherent optical orthogonal frequency-division multiplexing (CO-OFDM) transmission systems for reliable performance. In general, redundant data such as training symbols and pilot tones are necessary for channel and phase noise estimation in quadrature amplitude modulation (QAM) based CO-OFDM systems, which reduces the spectral efficiency of the system. Especially, in the coherent passive optical network (PON), the redundancy increases as the number of subscribers increases. In this letter, we propose a differential amplitude shift keying (DAPSK)-based CO-OFDM transmission for PON system, which does not require channel equalization and phase noise compensation processes. The proposed DAPSKbased CO-OFDM transmission was experimentally demonstrated, and the performance according to receiver launched power and the number of subcarriers was compared with QAM-based CO-OFDM. Index Terms: Coherent communication, fiber optics systems, microwave photonics signal processing. 1. Introduction Coherent optical orthogonal frequency division multiplexing (CO-OFDM)-based passive optical networks (PONs) have drawn attention as a next-generation optical access network technology because of its high spectral efficiency, receiver sensitivity, and flexibility [1], [2]. However, for reliable system performance of CO-OFDM-PON, channel equalization and phase noise compensation are required. Because, in the link, the frequency dependent channel response caused by the transmission channel characteristic and impulse response of various components distorts the received signal; also, the phase noise from transmit and local oscillator (LO) lasers leads to rotation of the received symbol constellation. In order to equalize the frequency dependent channel response, post processing using digital signal processing (DSP) is usually used. For phase noise compensation, an optical phase-locked loop (OPLL), which tracks and matches the phase between transmitter and LO laser source, can be used [3], but installing additional hardware in every optical network unit (ONU) is cost inefficient in PON systems. Therefore, DSP is commonly preferred over OPLL for phase noise compensation, like above mentioned channel equalization. Generally, in quadrature

Fig. 1. (a) CO-OFDM-PON architecture, (b) QAM-OFDM, and (c) DAPSK-OFDM frame structure for coherent-pon system. amplitude modulation (QAM)-based OFDM in coherent-pon system, channel equalization and phase noise compensation using training symbols and pilot tones which are inserted in the OFDM frame at certain intervals according to the channel condition are used [4], [5]. However, because training symbols and pilot tones are redundant data, this approach has a disadvantage in spectral efficiency. Alternatively, a differential modulation technique can be adopted instead of the QAM. Differential amplitude and phase shift keying (DAPSK) is a differential modulation technique to map data to the phase and amplitude and difference between adjacent symbols, and it has been mostly researched in the radio-frequency (RF) communication [6] [9]. The DAPSK demodulation process divides adjacent symbols, so multiplied noise such as channel response and phase noise are canceled out, assuming that the phase and channel response variations between two adjacent symbols are small enough to be negligible, namely, the quasi-stationary condition. Consequently, DAPSK-based CO-OFDM-PON system has advantages in spectral efficiency and system complexity because it does not require channel equalization and phase noise compensation processes that use redundant data. A disadvantage of DAPSK is that differential modulation has a higher signal-to-noise ratio (SNR) penalty than that of the non-differential technique [10]. Fortunately, the optical channel is more stable and less noisy than the RF channel. Furthermore, coherent detection has much higher receiver sensitivity, so a high SNR can be easily achieved in a coherent optical transmission system compared to optical intensity modulation and direct detection (IM/DD) systems. In this letter, we propose a DAPSK-based CO-OFDM transmission technique for PON system, whose performance is experimentally demonstrated. In CO-OFDM-PON system, there are several deterioration factors that increase the overhead, but we focused on channel response and phase noise aspects. Through differential modulation and demodulation, channel response and phase noise are effectively removed, and the data are restored without channel equalization and phase noise compensation after 20 km of coherent optical transmission. In consideration with the quasi-stationary condition, the performance was measured and analyzed in DAPSK-OFDM, which has 64 and 128 subcarriers, with a 5-GHz signal bandwidth because the symbol duration is changed according to OFDM subcarrier spacing. In addition, the performance of DAPSK-OFDM was compared to that of QAM-OFDM, which included channel equalization and phase noise compensation, and we showed that the spectral efficiency of DAPSK-OFDM was better than that of QAM-OFDM. In spite of the 1 2 db power penalty, DASPK-OFDM could satisfy bit error rate (BER) performance for the forward error correction (FEC) limit of 2 10 3 because of the high receiver sensitivity of coherent detection. 2. Schematics Fig. 1(a) shows the CO-OFDM-PON architecture. Subcarriers or bandwidths are allocated according to the needs of each ONU, and the signals of each ONU are divided by a passive power splitter in the down-link and summed in the up-link transmission. Fig. 1(b) represents frame structure for general QAM-OFDM considering channel equalization and phase noise compensation in

Fig. 2. Schematics diagram of the proposed DAPSK-OFDM tansmission for coherent-pon system. coherent-pon system. Training symbols and pilot tones are inserted at some intervals according to the channel condition, and at least one pilot tone should be inserted for each ONU in order to achieve uniform performance because the signals of each ONU have different phase noise characteristics. Therefore, the redundancy increases as the channel characteristic deteriorates and the number of ONUs increases. Consequently, this approach can significantly reduce spectral efficiency in coherent-pon system. Frame structure for the proposed DAPSK-OFDM is represented in Fig. 1(c). Redundant data for channel equalization and phase noise compensation processes is not required except the first frame for dividing the next frame because the processes are performed automatically during demodulation process. Also, the loss of the first frame is meaningless when the frame transmission is continuous enough. The channel equalization and phase noise compensation processes of QAM-OFDM are performed with training symbol and pilot tone interval, whereas demodulation and impairments cancelation of the proposed DAPSK-OFDM is performed independently of all OFDM subcarriers at interval of two frames without redundancy. Therefore, in the proposed system, redundancy does not increase as the number of ONUs increases and the channel condition degrades in terms of channel equalization and phase noise compensation. The proposed transmission scheme can be used in both uplink and downlink, and because of the features of the proposed scheme mentioned above, it is possible to prevent the reduction of the spectral efficiency which is caused by the increase of the number of ONUs. Also, it is possible to reduce the burden of the DSP for the channel equalization and phase noise compensation in the ONU and optical line terminator (OLT). Fig. 2 shows the schematic diagram of the proposed DAPSK-OFDM transmission for coherent- PON system. The differential modulation and demodulation processes are simply added without channel equalization and phase noise compensation processes. In the transmission part, serial bit data are converted to parallel bit sequences, and then each bit sequence is modulated to a DAPSK symbol. Fig. 3(a) shows the 16-DAPSK modulation process. The four-bit input sequence of the i th frame and kth subcarrier (b 1, b 2, b 3, b 4 ) i,k are mapped to a differential complex value B i,k that is expressed as B i,k = a q exp (jn ϕ) q { 1, 0, 1}, n ={0,, 7}, ϕ= 45 (1) where a is amplitude factor, and a q refers to the amplitude transition value. The amplitude transition value depends on the first bit (b 1 ) i,k and the previous DAPSK-OFDM symbol S i 1,k, which is determined according to Table 1.

Fig. 3. (a) 16-DAPSK modulation process and (b) 16-DAPSK demodulation process. TABLE 1 Two-Level DAPSK Amplitude Mapping Table B i,k =a q (b 1 ) i,k 0 1 S i 1,k 1 1, (q = 0) a, (q = 1) a 1, (q = 0) 1/a, (q = 1) In two-level DAPSK, symbol S i 1,k can have an amplitude state of 1, or a, but transition state B i,k can have three kinds of states (a 1,1,a 1 ). The remaining three bits are mapped to the complex value exp(jn ϕ), which is the general 8-Phase Shift Keying (PSK) symbol. Amplitude mapping and PSK are separately performed. Amplitude level and PSK order can be changed, thereby total DAPSK order can be changed simply. The DAPSK-OFDM symbol S i,k is obtained by multiplying the complex value B i,k by the previous DAPSK-OFDM symbol S i 1,k. Consequently, the symbol of the i th frame and kth subcarrier is presented as S i,k = B i,k S i 1,k. (2) Constellations of 4-differential phase shift keying (DPSK), 8-DAPSK, 8-DPSK, 16-DAPSK, and 16-DPSK symbols are presented in Fig. 4(a). Constellation of 8-DAPSK symbol consists of 4-DPSK and two-level differential amplitude shift keying (DASK). 16-DAPSK consists of 8-DPSK and twolevel DASK. Also, the differential complex values B i,k of two level DAPSK modulated symbols which are 8-DAPSK and 16-DAPSK have three amplitude levels as shown in Fig. 4(b). The inverse fast Fourier transform (IFFT) is followed by cyclic prefix (CP) insertion, in order to mitigate the in-phase and quadrature (IQ) timing offset and inter-symbol interference (ISI) caused by fast Fourier transform FFT window timing offset, and then the signal is transmitted after optical IQ modulation. The received DAPSK-OFDM symbol that is obtained after optical coherent transmission and FFT is given as R i,k = S i,k H i,k e j φ PN i,k + n i,k (3)

Fig. 4. Constellations of (a) differential modulated symbols S i,k, and (b) differential complex value B i,k. where H i,k, e j φ PN i,k and n i,k represent the channel response, phase noise, and additive white Gaussian noise (AWGN), respectively. Fig. 3(b) shows the 16-DAPSK demodulation process. The differential demodulation is a simple process that divides R i,k by R i 1,k. Assuming a negligibly low AWGN, the demodulated differential complex value D i,k can be written as D i,k = R i,k R i 1,k = S i,k S i 1,k H i,k H i 1,k e j φpni,k e j φ PN i 1,k = B i,k H i,k e j φpni,k H i 1,k e j φ. (4) PN i 1,k D i,k B i,k if H i,k H i 1,k, e j φ PN i,k e j φ PN i 1,k. (5) Provided that channel response is sustained at least for two-symbol duration, i.e., assuming quasi-stationary, H i,k and H i 1,k are canceled out in (4). In the same manner, if phase fluctuation is very small for two-symbol duration, e j φ PN i,k and e j φ PN i 1,k also can be canceled out. Therefore, without channel equalization and phase noise compensation, the differential complex value B i,k can be restored, as shown in (5). Because demodulation performance depends on the quasi-stationary condition, the shorter the symbol duration, the better the performance. The demodulation process is independently performed in each OFDM subcarrier, so channel response and phase noise can be effectively removed in frequency-selective channel transmission. The amplitude mapped bit (b 1 ) i,k is de-mapped according to Table 1 with D i,k and S i 1,k. Phase mapped bits (b 2, b 3, b 4 ) i,k are de-mapped by 8-PSK demodulation from arg(d i,k ). In the case where AWGN is not negligible, it deteriorates the system performance. Moreover, in the case of the differential system, it has an SNR penalty of approximately 2 db compared with a non-differential system such as QAM-mapped OFDM. Nevertheless, in the optical coherent system that has stable channel conditions and high receiver sensitivity, these drawbacks can be overcome. In addition, system complexity resulting from DSP for impairment compensation can be relaxed by using the differential modulation and demodulation techniques. 3. Experiments The experimental setup for the proposed scheme is shown in Fig. 5. Because the proposed technique focuses on the channel response and phase noise, the experimental demonstration was performed through self-coherent setup which has the carrier frequency offset (CFO)-free condition

Fig. 5. Experimental setup. for proof-of-concept. An external cavity laser (ECL) with a center wavelength of 1550.223 nm and linewidth of 100 khz was used for both the optical source and LO, by using a 3-dB optical splitter. In two-level DAPSK-OFDM, amplitude factor a was set to 2. The digital samples of the D(A)PSK-OFDM signal were made by offline processing, and the analog signal was generated by an arbitrary waveform generator (AWG). The signal was electro-optic (O/E)-converted by the IQ modulator where the bias was controlled by the auto bias controller (ABS). The IQ modulator consisted of three Mach- Zehnder modulators and the signal was modulated at null bias point. Therefore, the insertion loss of the IQ modulator was high, and the fiber launched power was 15 dbm. Polarization controllers were used in front of the IQ modulator and the coherent receiver, which are polarization-dependent components. The optical signal was transmitted through a 20 km single-mode fiber (SMF), and then received by a coherent receiver. In order to emulate optical distribution network (ODN) loss according to splitting ratio in PON system, the received power was tuned by variable optical attenuator (VOA). The received signal was sampled by a mixed-signal oscilloscope (MSO) with a 25-GHz sampling rate. Then, the remaining processes including FFT, D(A)PSK demodulation, and the symbol de-mapping were performed by offline processing without the channel equalization and phase noise compensation processes. The QAM-OFDM demonstration was performed in the same manner as D(A)PSK-OFDM. However, unlike D(A)PSK-OFDM, channel equalization and phase noise compensation algorithm using training symbols and pilot tones were added. Both bandwidth of D(A)PSK-OFDM and QAM-OFDM were 5 GHz, and CP that was 5% of the signal frame length was included in the signals to mitigate the IQ timing offset and ISI caused by FFT window timing offset. Differential modulation method could be more affected by ISI; however, in OFDM, ISI could be easily removed by CP. Thus, there was no difference in performance between QAM-OFDM and DAPSK-OFDM according to ISI when CP was used. For the cases of 4-DPSK, 8-DAPSK, 8-DPSK, 16-DAPSK, and 16-DPSK, BER was measured according to receiver launched power. In addition, in D(A)PSK-OFDM with 64 and 128 subcarriers, the measurement was carried out repetitively in order to evaluate the system performance according to the symbol duration under the quasi-stationary assumption. Symbol durations of an OFDM symbol with 64 and 128 subcarriers were 13.4 ns and 26.9 ns, respectively, including CP. QAM-OFDM transmission performance was also measured by keeping the same order and number of subcarriers as the previously measured D(A)PSK-OFDM in order to compare the performance. The number of training symbols and pilot tones were optimized by considering the channel and frequency characteristics of the experimental setup. Training symbols were inserted at intervals of 100 frames. Pilot tones were inserted at intervals of 20 subcarriers in QAM-OFDM with 128 subcarriers, and at intervals of 10 subcarriers in QAM-OFDM with 64 subcarriers, considering the frequency interval of the subcarrier. Therefore, six pilot tones were used for QAM-OFDM phase noise compensation in both cases.

Fig. 6. Constellations of received D(A)PSK-OFDM symbols R i,k, and demodulated differential complex values D i,k in the case of (a) 64 and (b) 128 subcarriers. 4. Results and Discussion Fig. 6(a) and (b) show constellations of received and demodulated D(A)PSK-OFDM symbols. Constellations of received symbols were distorted because of channel response and phase noise after 20 km coherent optical transmission. ISI was eliminated by using CP. Thus, the main causes of the distortion were phase noise by laser linewidth, and channel response by RF (Radio Frequency) and optical components. Since the fiber launched power was as small as under 15 dbm, the nonlinear effect such as self-phase modulation (SPM) was not significant. SPM is mainly considered in long-haul system, where the fiber input power is more than 0 dbm and the transmission distance is several hundred kilometers. Also, generally, single wavelength based coherent-pon system which has high receiver sensitivity does not require high fiber launched power; therefore, SPM is not an issue. After differential demodulation, constellations of demodulated differential complex value D i,k were restored for every case. In particular, cases of two-level DAPSK were restored to three-level constellations because they had three kinds of transition states. In addition, D(A)PSK-OFDM with 64 subcarriers was restored better than D(A)PSK-OFDM with 128 subcarriers. This was because the OFDM symbol duration for the 64-subcarrier case was shorter than that of 128 subcarriers, in half; in other words, D(A)PSK-OFDM with 64 subcarriers satisfied the quasi-stationary condition better. Fig. 7(a) shows BER performance according to coherent receiver launched power in the case of D(A)PSK and QAM-OFDM with 64 subcarriers. The solid line presents D(A)PSK-OFDM performance. DAPSK-OFDM shows better performance than DPSK-OFDM, which has the same order

Fig. 7. BER performance as a function of received power in the case of (a) 64 and (b) 128 subcarriers. TABLE 2 Spectral Efficiency and Transmission Rate DASPK/DPSK 64sub_QAM 128sub_QAM Order SE Bit rate (Gb/s) SE Bit rate (Gb/s) SE Bit rate (Gb/s) 4 1.90 9.52 1.71 8.54 1.80 8.99 8 2.86 14.29 2.56 12.82 2.70 13.48 16 3.81 19.04 3.42 17.08 3.59 17.97 SE: spectral efficiency (bit/s/hz) because the adjacent symbol distance of DAPSK is longer than that of DPSK. In particular, the symbol distance of 16-DPSK was so short that the symbols could not be distinguished, as shown in Fig. 5; therefore, the FEC limit of 2 10 3 could not be satisfied in 16-DPSK-OFDM. However, the 16-DAPSK symbol was relatively well distinguished, so the 16-DAPSK-OFDM case could satisfy the FEC limit over launched power of 29 dbm. Considering the fiber launched power of 15 dbm, ODN loss of 14 db was acceptable in the set-up. If an IQ modulator which has low insertion loss or optical pre-amplifier was used, the higher power margin could be obtained. D(A)PSK-OFDM had an approximately 1-dB power penalty at a BER of the FEC limit compared with QAM-OFDM having the same order. Nevertheless, at a relatively low launched power, the FEC limit could be satisfied. In addition, because D(A)PSK-OFDM did not use training symbols and pilot tones for channel and phase estimation, a better spectral efficiency and data rate could be achieved compared to QAM- OFDM. In this experiment, training symbols that had intervals of 100 frames and six pilot tones were used for QAM-OFDM receiving; thus, the spectral efficiency and bit rate were degraded compared to D(A)PSK-OFDM, which is summarized in Table 2. Since the number of optimized redundancy in the experiment depends on the channel environment, spectral efficiency may also vary depending on the channel environment. However, If the number of ONUs which is supported in coherent-pon system increases, more training symbols and pilot tones should be used for channel equalization and phase noise compensation in the QAM-OFDM. Therefore, spectral efficiency will be more degraded, and the system complexity for this case will increase. Whereas, D(A)PSK-OFDM based coherent-pon will maintain spectral efficiency and system complexity because it does not require redundancy for channel equalization and phase noise compensation. BER performance of D(A)PSK and QAM-OFDM with 128 subcarriers is presented in Fig. 7(b). In the case of 16-DAPSK-OFDM, the FEC limit could be satisfied at receiver launched power of

24 dbm. Overall, the required launched power for the FEC limit increased by 5 db in comparison with the 64-subcarrier case. This was because the OFDM symbol duration was doubled from 13.4 to 26.9; in other words, the quasi-stationary condition may not be maintained well. As with the results of the 64-subcarrier case, in the 128-subcarrier case, DAPSK-OFDM shows better performance compared to DPSK-OFDM, and D(A)PSK-OFDM has a 2-dB power penalty compared to QAM- OFDM. When the number of subcarriers increased from 64 to 128, the spectral efficiency and bit rate of QAM-OFDM also increased, because the number of training symbols and pilot tones did not change. Nevertheless, the spectral efficiency and bit rate of QAM-OFDM were still lower than those of D(A)PSK-OFDM as summarized in Table 2. 5. Conclusion We have proposed a DAPSK-OFDM transmission scheme for an efficient coherent-pon system. Through the experimental demonstration, it was verified that DAPSK-OFDM coherent optical transmission was possible without the phase noise compensation and the channel equalization with a higher spectral efficiency compared to QAM-OFDM. Although DAPSK-OFDM has 1 2 db power penalty compared to QAM-OFDM, this disadvantage can be overcome in optical coherent systems by its high receiver sensitivity. Furthermore, unlike QAM-OFDM, DAPSK-OFDM can achieve high spectral efficiency regardless of frequency channel stability and the number of subcarriers. Therefore, the proposed DAPSK-OFDM coherent optical transmission technique will be useful in the coherent-pon system that requires low system complexity and high spectral efficiency. References [1] N. Cvijetic, OFDM for next-generation optical access networks, J. Lightw. Technol., vol. 30, no. 4, pp. 384 398, Feb. 2012. [2] K. Kikuchi, Fundamentals of coherent optical fiber communications, J. Lightw. Technol., vol. 34, no. 1, pp. 157 179, Jan. 2016. [3] C. Camatel, V. Ferrero, R. Gaudino, and P. Poggiolini, Optical phase-locked loop for coherent detection optical receiver, Electron. Lett., vol. 40, no. 6, pp. 384 385, Mar. 2004. [4] W. Shieh and I. Djordjevic, Orthogonal Frequency Division Multiplexing for Optical Communications. Burlington, MA, USA: Academic Press, 2009. [5] X. Yi, W. Shieh, and Y. Tang, Phase estimation for coherent optical OFDM, IEEE Photon. Technol. Lett., vol. 19, no. 12, pp. 919 921, Jun. 2007. [6] H. Rohling and V. Engels, Differential amplitude phase shift keying (DAPSK)-a new modulation method for DTVB, in Proc. Int. Broadcast. Conv., Sep. 1995, pp. 102 108. [7] Y. C. Chow, A. R. Nix, and J. P. McGeehan, Error performance of circular 16-DAPSK in frequency-selective Rayleigh fading channels with diversity reception, in Proc. Int. Conf. IEEE Veh. Technol., Chicago, IL, USA, Jul. 1995, pp. 419 423. [8] V. Engels and H. Rohling, Differential modulation techniques for a 34Mbit/s radio channel using orthogonal frequency division multiplexing, Wireless Pers. Commun., vol. 2, no. 1, pp. 29 44, Mar. 1995. [9] D. Z. Liu and C. H. Wei, Channel estimation and compensation for preamble-assisted DAPSK transmission digital mobile radio system, IEEE Trans. Veh. Technol., vol. 50, no. 2, pp. 546 556, Mar. 2001. [10] S. Moriyama, K. Tsuchida, and M. Sasaki, Digital transmission of high bit rate signals using 16 DAPSK-OFDM modulation scheme, IEEE Trans. Broadcast., vol. 44, no. 1, pp. 115 122, Mar. 1998.