COHERENT OPTICAL COMMUNICATION SYSTEMS IN DIGITAL SIGNAL PROCESSING

Transcription

1 228 COHERENT OPTICAL COMMUNICATION SYSTEMS IN DIGITAL SIGNAL PROCESSING K. Nordin 1, N. M. Z. Hashim 2, S. Idris 3, A. M. Darsono 4, A. Salleh 5, N. R. Mohamad (Faculty of Electronic and Computer Engineering UniversitiTeknikal Malaysia Melaka (UTeM), Hang Tuah Jaya, Melaka, Malaysia) ABSTRACT For the future of optical communication systems, Digital signal processing (DSP) is an enabling technology. This system is the most promising widely viewed in the next generation for long-haul transport systems. In past, the digital carrier phase estimation demonstration in coherent receivers has stimulated an extensive attention in coherent optical communications again. In this paper, the emergence of DSP tutorial will discussed before surveying in the algorithms required in digital coherent transceivers for optical communication systems. Keywords Coherent, Detection, Digital Signal Processing (DSP), Signal I. INTRODUCTION Digital Signal Processing (DSP) emerged as an enabling technology for high capacity optical transmission systems since optical communication has undergone the evolution over the last decade. As a promising technique, digital signal processing is under consideration for optical signal modulation, fiber transmission, and signal detection and dispersion compensation. Coherent detection and DSP combination is expected to become part of optical communication systems generation and provide new capabilities that were not possible without the phase of the optical signal detection. Lately, digital coherent optical communication is the main technology for optical transport networks [1]. In future of optical networks, capacity increase will need a flexible bandwidth transmitter which is dynamically allocates bandwidth by user demand. The function is to further increase the network s spectral efficiency [2]. DSP has increase become embedded into optical transceivers when the first application specific integrated circuit (ASIC) implementing advanced algorithms appeared. ASIC are designed for Gsymbol/s polarization-multiplexed QPSK signal has been developed, and digital coherent receiver real time operation at 46 Gbit/s bit rate has been demonstration striated by using an ASIC. In modern coherent optical communications history, the achievement is really milestone [3]. Nascent technology culmination is the digital coherent receivers for DSP essential to operate with allow 40Gbits/s, 100Gbit/s and to deployed in core networks over fiber. It is due to polarization-mode dispersion (PMD) not support 10Gbit/s by using conventional technique. The basic of the principle coherent optical detection is coherent measure the complex amplitude of the optical signal with the shot noise limited sensitivity. This is how information on the state of polarization can be extracted by polarization diversity uses. Coherent detection associated of DSP can be very advantageous. To increase optical receiver sensitivity and permitting a greater span loss to be tolerated, coherent detection is a promising technology. Quadrature amplitude modulation (QAM) and quadrature phase shift keying (QPSK) are an example of supported of more spectrally efficient modulations format by coherent detection. Coherent detection also allows digital signal processing for transmission impairments compensation such as chromatic dispersion (CD), PMD, signal carrier offset and spectrum restricting instead of requesting expensive physical impairments compensation links. Next generation of optical transmission systems need adaptive fitting for time varying transmission impairments. For example, channel spectrum narrowing and random phase noise. DSP become more powerful solution for the future optical transmission links. For this tutorial, the aim is to outline the development and issues associate details with realizing a digital coherent transceiver. Fig. 1: Generic coherent system [8]

2 229 In Figure 1, in a real arrangement, this black box embodies a number of complex functions. As well the purposes that are additionally performed in a classical arrangement, such as data aggregation, coding, and framing, supplementary steps demand to be gave in a transmitter for convoluted modulation formats, as it is normally utilized in a consistent system. Early of all, the data have to be mapped to constellation points and, in case of several messenger systems, to frequencies. Often, the data are additionally differentially preceded to cope alongside period slips across receiver-side messenger synchronization. In a subsequent pace, the mapped data could be processed digitally, for instance, it could be pre-distorted to compensate for the nonlinearity of amplifiers and modulators, or it might be precompensated for deterministic fiber results, for instance, chromatic dispersion (CD). The processing of the preceding delineated data aftermath in four digital data streams those afterward demand to be modified into analog data. In case of single-carrier QPSK indicating, every single data stream carries merely a solitary bit each signal, and consequently, does not need a DAC. This reduces the intricacy, and therefore, power consumption of the transmitter significantly. But even for multicarrier arrangements [6] or modulation formats alongside higher order than QPSK, the presentation necessities alongside respect to resolution and conversion speed are normally less restrictive for the send DAC than for the accord AD converter (ADC). For instance, for a 16-QAM transmitter lacking preprocessing, only 2 bits (four levels) at a conversion speed equivalent to the symbol rate are needed, as at the receiver side, typically 6 8 bits at a sampling rate of twice the data rate are needed [8]. II. INTERFACING ANALOG AND DIGITAL DOMAINS To apply DSP, the key element has been the availability of high speed, time interleaved CMOS digital to analog and the important is analog to digital converters (ADC). The real-time sampling oscilloscope is currently test equipment that led the way in the ADC development technology. The first generation 20GSa/s 8 bit ADCs in 2003 were emerged in test equipment by using time-interleaving in 180nm CMOS. Test equipment has move to 160GSa/s in 2012, increase ADC rate sampling at steady rate of 25% per annum over the same period. 60GSa/s is common place for 100GbE transceivers using PDM-QPSK while ADCs sampling rates lag test equipment. Test equipment with rates of 50GSa/s is only emerging even though 40GSa/s and 60GSa/s digital to analog converters (DAC) s already exist in commercial products in contrast for DAC [4]. III. DIRECT DETECTION SYSTEMS IN DSP The detect detection systems is a first application of digital signal processing. The primary application increases the range of uncompensated transmission. Maximum likelihood sequence estimators (MLSE) include at the receiver to pre-compensation of chromatic dispersion by coherently modulating both the phase and amplitude at the transmitter in addition. The exponential increase in complexity with transmission distance resulting in commercial MLSE is one of the limitations of MLSE. It is being unable to achieve uncompensated transmission over more than metropolitan distances. With transmission distance since it acts on the coherent field, the complexity of precompensation at the transmitter. Coherent detection is the most advanced detection method. The receiver computes decision variables based on the recovery of the full electric field. It s contains amplitude and phase information. In modulation formats, coherent detection allows the greatest flexibility that information can be encoded in amplitude and phase or in both in-phase (I) and quadrature (Q) components of a carrier. Knowledge of the carrier phase need in the coherent receiver detection because of received signal is demodulated by a local oscillator (LO). LO function to serves as an absolute phase reference. Usually, carrier synchronization has been performed by a phase-locked loop (PLL). In optical systems, it can used an optical PLL (OPLL) which synchronizes the frequency and phase of the LO laser with the TX laser. The system also can used an electrical PLL where the function to down conversion using a free-running LO laser. It followed by a second-stage demodulation by an analog or digital electrical VCO whose frequency and phase is synchronized. Duplex systems can have an advantage by using an electrical PLL which the transceiver can use one laser as both TX and LO. Delay requirement is difficult to satisfy due to PLLs are sensitive to propagation delay in the feedback path. To fix this problem, the feed forward (FF) carrier synchronization has the solution. In addition,, FF can achieve better performance than a PLL as a FF synchronizer uses both past and future symbols to estimate the carrier phase. FF can only employ past symbols due to it is function as a feedback system. Normally, to perform in software, DSP has enabled polarization alignment and carrier synchronization. A

3 230 coherent transmission system and the model system are shown in Figure 1 and Figure 2. With both quadrature s detected employing full coherent detection that shown in Figure3, to gain 3 db in receiver sensitivity is possible. Polarization multiplexing is available for additional spectral efficiency and compensates ability to an arbitrarily high degree from linear impairments. The most attractive possibility for full coherent detection is about 100 Gb/s and above. Fig. 2 Coherent transmission system implements Fig. 3 System model Fig. 4 Quadrature coherent detection IV. COHERENT TRANSMITTERS IN DSP Coherent transmitters allow a conventional direct detection receiver to be employed at coherent system in DSP. Modulated complex optical fields are used in this situation. It was shown in a Cartesian Mach- Zehnder modulator. An electrical drive signal modulates the real part of the field and another electrical drive signal modulates the imaginary part of the field. By using finite impulse response (FIR) filter, it can achieve chromatic dispersion pre-compensation. This purpose can implement the complex transfer function with the inverse chromatic dispersion. The directly achievement for FIR filter implementation in the time domain or the frequency domain are using overlap methods. The time domain implementation can be more efficient for the length of filter typically required for core if the networks at the transmitter are also more efficient. At the transmitter, in contrast to the general case, it occurs since the inputs are binary removing the need for multiplication. With merely summing a suitable combination of the tap weights, convolution can be achieved. To control the optical spectrum for non-binary modulation formats something like 16- QAM, transmitter DSP has primarily focused on pulse shaping. It implement by using FIR filters. V. COHERENT RECEIVERS IN DSP Coherent detection of optical communication signals has long been recognized to propose several performances advantages over direct detection [5], but to date it has not been used in fiber optic networks. Coherent detection is sensitive to the amplitude and phase of an optical signal and it can be used to notice phase-encoded modulation formats like quadrature phase shift keying (QPSK), binary phase shift keying (BPSK) and quadrature amplitude modulation (QAM). These formats give better receiver sensitivity than easy on-off modulation. QPSK and QAM permit many bits each signal to be transmitted without a substantial degradation in receiver sensitivity, by transmitting the information in both quadrature components. Coherent detection replies only to the optical spectrum in the instant area of the local oscillator, so it is equivalent to encompassing a narrow optical filter in the receiver. In fact coherent detection is the only method that can notice information spectral densities approaching the Shannon limit. One more advantage of coherent over direct detection is the potential of correcting fiber propagation impairments like chromatic dispersion (CD) in the mechanical domain. These advantages are all priceless in today s fiber optic transmission systems, which use high channel count wavelength division multiplexing above multi- 1000km distances. The outstanding disadvantage of coherent detection is the complexity, and consequently the price, of the receiver. Fig 5 Typical DSP subsystems in a digital coherent

4 231 Based on Figure 5, inside a digital coherent receiver there are several key subsystems, that could be loosely tear in to those concerned alongside equalization and those concerned alongside synchronization more details of that could be discovered in. Equalization algorithms are primarily concerned alongside compensating for the imperfections in the optical channel in supplement to those present in the physical transmitter and receiver, fluctuating from timing and amplitude correction for the high speed time interleaved analog to digital converter and also to digital compensation of residual chromatic dispersion. After the signal has been equalized it is probable for the synchronization algorithms to align the oscillators, both mechanical and optical, thereby mitigating the encounter of the difference in phase and frequency amid the transmitter and receiver. Naturally there are numerous probable variations of the structural design of the DSP, for example one could select to digitally compensate for the frequency offset, compare than estimate the frequency offset to provide a control signal for the local oscillator to hold the frequency offset to within the range of the carrier phase estimation and correction algorithm. As the last way, simplifies the DSP the feedback induces coupled dynamics and consequently care have to be seized to ensure that time constants for every single subsystem are optimized to ensure stable operation. Fig. 6 DSP signal flow model Hence for fast acquisition a feed onward structure is needed, albeit at the price of higher DSP complexity [6]. In the end the demodulated signal is optimally decoded to produce the best estimate of the sequence of bits encoded by the transmitter alongside the coding overhead optimized for the working condition of the system [7]. Estimation of power consumption is challenging due to the largely parallel nature of the DSP that frequently pushes the design instruments to their limits. One of the challenges in approximating power consumption is the paucity of data considering the parameters for the power consumption and the algorithms utilized in business digital coherent receivers. The notable exclusion is the early commercially obtainable 40Gbit/s application specific integrated circuit (ASIC) projected by Nortel (now Ciena), for that a significant amount of detail has been published. Instituted on a preceding scrutiny [8] for the ASIC projected in 90nm CMOS, the manipulation consumption was manipulated by multiplications alongside 2.7pJ utilized each real multiplication. VI. OPTICAL COMMUNICATION SYSTEMS IN DSP Applying DSP to optical transceivers has revolutionized optical fiber communication systems enabling transceivers to evolve to become software defined. With this new possibilities such as cognitive transceivers emerge, permitting for both software describes webs and cognitive webs both in the core and the admission networks. As we have concentrated primarily on core webs going onward, DSP is probable to be a key technology for optical admission webs, just as it has completed for wireless admission networks [9-18]. VII. CONCLUSION Digital signal processing (DSP) is an enabling technology for future optical communication system. Over the last decade optical communications has undergone a revolution, as digital signal processing (DSP) emerged as an enabling technology for highcapacity optical transmission systems when the first application specific integrated circuit (ASIC) implementing advanced algorithms appeared, DSP has increasingly become embedded into optical transceivers. The culmination of this nascent technology is the digital coherent receiver for which DSP is essential to its operation. ACKNOWLEGMENT We are grateful to Centre for Telecommunication Research and Innovation (CeTRI) and UniversitiTeknikal Malaysia Melaka (UTeM) through PJP/2013/FKEKK (29C)/S01215 for their kind and help for supporting financially and supplying the electronic components and giving their laboratory facility to complete this study. REFERENCES [1] Chongjin Xie, Chromatic dispersion estimation for single-carrier coherent optical communication, IEEE Photonics Technology Letters, vol. 25, no. 10, pp , May. 15, 2013.

5 232 [2] M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, Y. Tanaka, and A. Hirano, Distanceadaptive spectrum resource allocation in spectrumsliced elastic optical path network, IEEE Commun. Mag. 48(8), (2010). [3] K. Kikuchi, Coherent Optical Communications: Historical Perspectives and Future Directions, pp. 8-9, [4] C. Laperle, Advances in High-Speed ADCs, DACs and DSP for Optical Transceivers, Paper OTh1F.5, OFC/NFOEC [5] Govind P. Agrawal, "Fiber-optic communication systems," Wiley (2nd ed., 1997). [6] Seb J. Savory, Digital Signal Processing for Coherent Optical Communication Systems, International Conference on Photonics in Switching, [7] S.R. Desbruslais and S.J. Savory, "Optimizing the FEC Overhead in a 100 GbE PDM-QPSK Digital Coherent Transmission System," Jnl. Lightwave Tech, vol.29, no.8, pp , (2011) [8] Leven, N. Kaneda and S. Corteselli, Real-Time Implementation of Digital Processing for Coherent Optical Digital Communication Systems, IEEE Journal, Vol 6, [9] Hamzah, R. A., Hamid, M. S., Rosly, H. N., &Hashim, N. M. Z., A distance and pixel intensity relation for disparity mapping in region of interest in IEEE 3rd International Conference on Communication Software and Networks (ICCSN), 2011, pp [10] Hamzah, R. A., Hamid, M. S., Rosly, H. N., &Hashim, N. M. Z., An analysis of multiple size region of interest in disparity mapping for stereo vision application in IEEE Symposium on Computers & Informatics (ISCI), 2011, pp [11] Hamzah, R. A., Hamid, M. S., Rosly, H. N., Hashim, N. M. Z., &Napiah, Z. A. F. M., An Aligned Epipolar Line for Stereo Images with Multiple Sizes ROI in Depth Maps for Computer Vision Application, International Journal of Information and Education Technology, Vol. 1, No. 1, 2011, pp [12] Mohamad, N. R., Soh, A. S. A. M., Salleh, A., Hashim, N. M. Z., Aziz, M. A., Sarimin, N., Othman. A. &Ghani, Z. A., Development of Aquaponic System using Solar Powered Control Pump, IOSR Journal of Electrical and Electronics Engineering (IOSR-JEEE), Vol. 8, Issue 6,2013, pp [13] Husin, S. H., Ngahdiman, A. A., Hashim, N. M. Z., Yusop, Y., &Ja afar, A. S., Home Electrical Appliances Smart System, International Journal of Computer Science and Mobile Computing, Vol.2, Issue.9, 2013, pp [14] Aziz, K.A.A., Ramlee, R.A., Hashim, N.M.Z., Rahman, R. A., Machine Vision Based Height Measuring System, International Journal of Research in Advent Technology, Vol.2, No.8, 2014, pp [15] Aziz, K. A. A., Mohamood, N. &Hashim, N. M. Z., Sliding Window for Radial Basis Function Neural Network Face Detection, International Journal of Science and Engineering Applications, Volume 3 Issue 4, 2014, pp [16] Salleh, A., Mohamad, N. R., Aziz, M. A. A., Hashim, M. N. Z., Misran, M. H., & Othman, M. A., Design the High Gain and Low Power Amplifier for Radio over Fiber Technology at 2.4 GHz, American Journal of Engineering Research (AJER), Volume-02, Issue-09, 2013, pp [17] Salleh, A., Mohamad, N. R., Aziz, M. A., Misran, M. H., Othman, M. A., &Hashim, N. M. Z., Simulation of WiMAX System Based on OFDM Model with Difference Adaptive Modulation Techniques, International Journal of Computer Science and Mobile Computing, Vol.2 Issue. 9, 2013, pp [18] Salleh, A., Aziz, M. A., Mohamad, N. R., Misran, M. H., Othman, M. A., &Hashim, N. M. Z., Simulation of 2.4 GHz Low Power RF Front End Design for Radio over Fiber Technology, International Journal of Emerging Trends in Engineering and Development, Vol.5, Issue 3, 2013, pp