Volume-7, Issue-3, May-June 2017 International Journal of Engineering and Management Research Page Number: 227-237 Simulative Evaluations of in Band and Out of Band Crosstalk Penalties for Advanced Modulation Formats in Wavelength Division Multiplexed (WDM) Networks Malik Efshana Bashir 1, Karamdeep Singh 2, Shivinder Devra 3 1,2,3 Department of Electronics Technology, Guru Nanak Dev University Amritsar, INDIA ABSTRACT Optical wavelength division multiplexing (WDM) networks have come forward as a unique solution to the rapidly increasing network demands. However, WDM systems suffer from a severe limitation of linear and nonlinear inter-channel crosstalk. In this paper, the effects of particularly linear inter-channel crosstalk are enlightened and analyzed. Bit error rate (BER) analysis in the scenario of in band and out band crosstalk for various advanced modulation formats is performed by calculating power penalties of the received signals compared with back to back analysis. In order to evaluate the crosstalk induced on a test channel by neighbor aggressor channels, a detuning (out band) crosstalk measurement technique is used. In this technique, the sliding of central frequency of an aggressor channel towards a test channel is carried out and the impact of crosstalk at the test channel is evaluated. In the long run we came to the result that channel spacing and power level affects the quality of a signal in terms of BER. Also, it is seen that in band analysis shows maximum distortion and interchannel crosstalk compared to out band analysis. Keywords-- In band crosstalk, out band crosstalk, power penalties, channel spacing I. INTRODUCTION The Wavelength Division Multiplexing (WDM) technique is the hopeful solution for successful utilization of optical spectrum and the efficient construction of the coming generation of Broadband Optical Network [1, 2]. However WDM optical networks experience serious physical impairments during transmission which include linear as well as non linear impairments. The most essential issue in the design of WDM optical networks is the inter-channel crosstalk, which leads to exchange of power among the channels and degrades system performance. This power transfer can happen because of the optical fiber nonlinearities and leads to nonlinear crosstalk (e.g. Cross Phase Modulation, Four Wave Mixing etc). However, certain crosstalk can occur in a perfectly linear channel too, due to the imperfect nature of various WDM components like de-multiplexers and optical filters [3, 4] called linear crosstalk. Linear crosstalk may be either in band (homodyne) or out band (heterodyne) depending upon whether it occurs between the channels at same or different wavelength respectively. Homodyne crosstalk can be further subdivided into coherent crosstalk between phase correlated signals, and incoherent crosstalk, between signals which are not phase correlated [5, 6]. 1. Linear impairments Optical out band crosstalk as the name defines is outside the pass band of an optical filter but in each node optical in-band crosstalk is added and gets increased up along the transmission. In WDM networks, optical out-band crosstalk appears from channels of different wavelengths. This type of optical crosstalk is not so harmful, as it can be greatly suppressed by using a narrow-band optical filter. When a signal and interferers have closely-valued wavelengths, optical in band crosstalk takes place. As a result, the signal and interferers are within the pass band of an optical filter, usually located in front of an optical receiver and cause a serious degradation in the performance of a receiver. Optical in-band crosstalk cannot be removed by an optical filter and will propagate with desired WDM channels along optical networks [7]. 2. Non-linear impairments In case of WDM systems, all channels use the same fiber and refractive index changes result in phase variations in those channels. This produces effects such as self-phase modulation (SPM), cross-phase modulation (XPM or CPM), and four-wave mixing (FWM) or four-photon mixing (FPM) [8]. Stimulated brillouin scattering (SBS), and stimulated Raman 227 Copyright 2017. Vandana Publications. All Rights Reserved.
scattering (SRS) involve non-elastic scattering mechanism. These impairments set an upper limit on the amount of optical power that can be launched into an optical link [7]. However, now-a-days certain nonlinear effects are being used for various practical purposes. Some researchers have demonstrated all-optical half-adder and half- sub tracter based on exploitation of non-linear effects such as cross gain modulation (XGM) & FWM [9, 10, 11, 12]. Also certain nonlinear devices like Ultra-fast Non- Linear Interferometer (UNI) and Highly Nonlinear Fiber (HNLF) have been reported for the implementation of several All-Optical digital logic gates and modules [13, 14]. In literature, several studies have been reported to address various crosstalk contributions in optical communication system. G.R. Hill, et al [15] and D. J. Blumenthal, et al [16] have analyzed out band coherent crosstalk in space photonic switches. The results revealed that when there is an increase in crosstalk power level over a critical value, the error rate due to this mechanism takes over all other effects inducing errors, and lead to a bit error rate (BER) floor. This type of crosstalk is important in WDM photonic switch networks where channels with same wavelength are routed in the same photonic switch fabric. S. D. Dods, et al [17] and [18], the combined effects of coherent and incoherent, in band crosstalk, in WDM ring and bus networks using add drop multiplexers has been investigated by the authors. A Monte Carlo simulation was used which showed that when both the crosstalk effects are considered, BER floors are likely to be higher than previously appreciated C. X. Yu, et al [19], the authors studied system degradation due to multipath coherent crosstalk that too by means of a Monte Carlo simulation and experimental measurements. The analysis was performed by considering a WDM network having add drop multiplexers as nodes, and showed that coherent crosstalk does not induce a deterministic power penalty for the network, but instead gives rise to multipath interference and fading. Eugenio Iannone, et al [20], the authors focus on backbone WDM networks, having optical cross-connect (OXC) as a key element. In this case, the most important crosstalk element is the in-band crosstalk. They analyze three different simulation models which allow the impact of in-band crosstalk on the transmission performances of WDM networks to be evaluated: an accurate simulation (AS), a Gaussian correlated noise (GCN) model, and a Gaussian white noise (GWN) model. In the AS, an interfering channel is generated at each node and added to the signal after an optical filtering. The final result must be averaged with respect to the message transmitted on the interfering channels and to the phases of the optical carriers. Oliveira, et al [21], the authors present an extensive experimental evaluation of in-band and detuning crosstalk in a flexible grid with mixed line rate scenarios. On the basis of aggressor channels with 10G- NRZ OOK, 40G-NRZ-DPSK, 112G-NRZ-DP-QPSK, 112GRZ -DP-QPSK, 224G-NRZ-DP-16QAM, and 224G- I. RZ/RZ-DP- 16QAM modulation, we experimentally quantify in-band and detuning crosstalk penalties for a test channel with 112G-DPQPSK modulation, with both RZ and NRZ pulse shapes. Detuning crosstalk analysis demonstrated that penalties induced by 224G DP-16-QAM as aggressor channel is higher than 112G DP-QPSK even with a similar spectral occupancy. However, the study we performed has not been carried out in the literature so far. We performed bit error rate (BER) analysis using Returnto-Zero ON-OFF Keying (RZ OOK), Carrier Suppressed Return-to-Zero (CS RZ), Modified Duo-binary Non- Return-to-Zero (MDB NRZ) and Modified Duo-binary Return-to-Zero (MDB RZ) modulation formats, at different bit rates, channel spacing for different power levels over varying single mode fiber (SMF) taking into account both in band and out band crosstalk. We used test and aggressor channels for our analysis. In order to evaluate the crosstalk induced on a test channel by neighbor aggressor channels, a detuning (out band) crosstalk measurement technique is used. In this technique, the sliding of central frequency of an aggressor channel towards a test channel is carried out and the impact of crosstalk at the test channel is evaluated. This paper includes various sections as under: Section 1 includes introduction part. It gives idea about nonlinear and linear (in band and out band) crosstalk. A brief literature survey is also included in this part. Section 2 covers the simulation setup and results. Section 3 concentrates on the conclusion from this work and future recommendations. II. SIMULATION SETUP In our study, we consider the following schematic experimental model as shown in Fig. 1. The model is shown for only one aggressor. We can follow the same guideline for other aggressors also. For performing analysis at various power levels an optical amplifier and an attenuator are used at the transmitting side after modulation. The test signal and an aggressor signal are then fed to a multiplexer before sending through a SMF. We used a Pseudo-Random Bit Sequence Generator (PRBS) of the order of 6 and having sequence length of 64 bits. A continuous wave (CW) laser having line width of 10MHz and power of 0dBm is used. The various parameters used for SMF are mentioned in Table 1 below. At the receiving side we use a Gaussian band pass filter of order 1 having 0dB insertion loss, an attenuator and an optical receiver having responsivity of 1A/W and sensitivity of -18dBm followed by a BER analyzer. 228 Copyright 2017. Vandana Publications. All Rights Reserved.
TABLE 1 SMF parameters used in the study Fig. 1 Experimental model used in the study of in band and out band crosstalk analysis We perform BER analysis using RZ OOK, CS RZ, MDB NRZ and MDB RZ modulation formats, at different bit rates, channel spacing for different power levels over varying link lengths. We take into account both in band and out band crosstalk and use test-aggressor channels for our analysis. In order to evaluate the crosstalk induced on a test channel by the neighboring aggressor channels, a detuning (out band) crosstalk measurement technique is used. In this technique, the sliding of central frequency of an aggressor channel towards a test channel is carried out and the impact of both in band and out band crosstalk at the test channel is evaluated. The analysis is carried out by plotting received power in dbm on X axis and log(ber) along Y axis. Also, we keep the BER threshold line at 10-9. III. RESULTS AND DISCUSSIONS The results and discussions part is divided into two sections: the out band analysis followed by in band analysis. (a) Out band analysis The various cases we performed are summarized in Table 2 as under: 229 Copyright 2017. Vandana Publications. All Rights Reserved.
TABLE 2 Illustration of various cases performed for out band analysis using test signal at wavelength, 1546 nm at 10, 20Gbps bit rate over 10, SMF. RZ OOK as RZ OOK as RZ OOK as test test at test test at 5dBm and 0, test at 5dBm and 0, at 15dBm and 15dBm and at 15dBm 2dBm power, left shift having wavelengths as 1545nm 1545.2nm 1545.4nm 1545.8nm 2dBm power, right shift having wavelengths as 1546.2nm 1546.4nm 1546.8nm 1547nm RZ OOK as test at 15dBm and aggressors at 0, 5, 10dBm power having wavelengths as CS RZ at 1545nm MDB NRZ at 1546.4nm MDN RZ at 1546.8nm We have performed the analysis by replacing the aggressor channels also like first CS RZ at 1545nm, then MDB NRZ at 1545nm frequency. The out band analysis is illustrated in Table.3 below. First we perform back to back BER analysis at the test transmitter for different bit rates without using fiber. After this, fiber is connected and BER analysis at the receiving side for the test signal is aggressors at 0, 5, 10dBm power having wavelengths as MDB NRZ at 1545nm CS RZ at 1546.4nm MDN RZ at 1546.8nm aggressors at 0, 5, 10dBm power having wavelengths as RZ OOK at 1545nm MDB NRZ at 1546.4nm MDN RZ at 1546.8nm and aggressors at 0, 5, 10dBm power having wavelengths as MDB NRZ at 1545nm RZ OOK at 1546.4nm MDN RZ at 1546.8nm examined. In order to evaluate the system performance it is useful to determine the power penalty due to crosstalk. The power penalty of the system is defined as the extra power required at the receiver to counteract the effect of the crosstalk [6]. We note different power penalties at various cases. TABLE 3 Illustration of out band test-aggressor analysis at 0ver 10 and SMF Test Aggressors RZOO CS RZ (1545nm) MDB NRZ (1546.4nm) MDB RZ (1546.8nm) K at 0dBm 5dBm 10dBm 0dBm 5dBm 10dBm 0dBm 5dBm 10dBm 1546n m Test CSRZ at 1546n m 10Gb ps 10Gb ps Aggressors RZ OOK (1545nm) MDB NRZ (1546.4nm) MDB RZ (1546.8nm) 0dBm 5dBm 10dBm 0dBm 5dBm 10dBm 0dBm 5dBm 10dB m 10Gb ps 10Gb ps 230 Copyright 2017. Vandana Publications. All Rights Reserved.
From the aforementioned cases 1 and 2, we note that for bit rate, and link length, we get negative power penalties. We thus discard those cases. For 20Gbps left as well as right shift from the test wavelength, bit rate 20Gbps, fiber length we find the results as shown in Fig. 2 and Fig. 3 below. From Fig.2 and Fig.3 it is illustrated that we used RZ OOK as test signal at 5dBm power and wavelength 1546nm and CS RZ signal as 0 and 2dBm power and we calculate the respective power penalties. Aggressor CS RZ is left shifted first at 1545, 1545.2, 1545.4, 1545.8nm and then right shifted at 1546.2, 1546.4, 1546.8 and 1547nm frequency. We note the power penalties of the test signal if aggressor is left shifted, under various cases as in Table 4 below. Similarly, we note the power penalties of test if aggressor is right shifted as in Table 5. TABLE 4 Left shift power penalty of RZ OOK as test at 5dBm and 0dBm 2dBm, 20Gbps SMF Test 1546nm, Power in dbm Power penalty 0dBm Back to Back power 24 0 Power at CS RZ 1545nm 22.3 1.7 Power at CS RZ 1545.2nm 22.2 1.8 Power at CS RZ 1545.4nm 22.1 1.9 Test 1546nm, Power in dbm Power penalty 2dBm Back to Back power 24.2 0 Power at CS RZ 1545nm 22.3 1.9 Power at CS RZ 1545.2nm 22.2 2 Power at CS RZ 1545.4nm 22.1 2.1 From Table 4, it is seen that in case of left shift, power penalty is minimum if CS RZ used as aggressor is at 1545nm wavelength, 0dBm( 1.7dB) and maximum if CS RZ is at 1545.4nm, 2dBm ( 2.1dB). Thus, CS RZ used at channel spacing 1545nm, 0dBm power level is appreciative rather than at 2dBm power level in neutralizing the effect of crosstalk. Fig.2 RZ OOK as test at 5dBm power, 1546nm wavelength and left shifted aggressor (at 1545, 1545.2, 1545.4, 1545.8nm frequency) at (a) 0dBm power, 20Gbps SMF (b) 2dBm power, 20Gbps SMF TABLE 5 Right shift power penalty of RZ OOK as test at 5dBm and 0dBm 2dBm, 20Gbps SMF Test 1546nm, 0dBm Power in dbm Power penalty Back to Back power 24.5 0 231 Copyright 2017. Vandana Publications. All Rights Reserved.
Power at CS RZ 1546.4nm 21.8 2.7 Power at CS RZ 1546.8nm 22.2 2.3 Power at CS RZ 1547nm 22 2.5 Test 1546nm, 2dBm Power in dbm Power penalty Back to Back power 24.5 0 Power at CS RZ 1546.4nm 22 2.5 Power at CS RZ 1546.8nm 22.5 2 Power at CS RZ 1547nm 22.4 2.1 Fig. 3 RZ OOK as test at 5dBm power, 1546nm wavelength and right shifted 1546.2, 1546.4, 1546.8, 1547nm) at (a) 0dBm power, 20Gbps SMF and (b) 2dBm power, 20Gbps SMF. Cases 3, 4, 5 and 6 mentioned before are Similarly, in Table 5, right shift, power penalty is illustrated in Fig. 4 below. maximum for 1546.4nm, 0dBm ( 2.7dB) and minimum for CS RZ at1546.8nm, 2dBm ( 2dB). Thus the latter counteracts crosstalk more. Fig. 4 Illustration of out band crosstalk analysis using RZ OOK and tests individually at 15dBm, aggressors as CS RZ, MDB NRZ, MDB RZ in case of test RZ OOK and RZ OOK, MDB NRZ, MDB RZ in case of test at 10,. In (a) CS RZ is at 1545nm and in (b) MDB NRZ is at 1545nm frequency. Similarly in (c) MDB NRZ is at 1545nm and in (d) RZ OOK is at 1545nm wavelength. 232 Copyright 2017. Vandana Publications. All Rights Reserved.
The power penalties in case of Fig.4 (a), (b) and Fig.4 (c), (d) are illustrated in Table 6 and Table 7 respectively. TABLE 6 Power penalty for RZ OOK as test at 15dBm and aggressors at 0, 5, 10dBm,, and SMF Test RZ OOK at 15dBm 1546nm, Power in Power penalty aggressors at 0dBm, CS RZ at dbm 1545nm, MDB NRZ at 1546.4, MDB RZ at 1546.8nm Back to Back power 25.6 0 0dBm 24.5 1.1 0dBm 25.5 0.1 5dBm 25.3 0.3 5dBm 25.4 0.2 10dBm 24.5 1.1 10dBm 24.5 1.1 Test RZ OOK at 15dBm 1546nm, Power in Power penalty aggressors at 0dBm, MDB NRZ at dbm 1545nm, CS RZ at 1546.4, MDB RZ at 1546.8nm Back to Back power 25.6 0 0dBm 24.6 1 0dBm 25.5 0.1 5dBm 25.3 0.3 5dBm 25.4 0.2 10dBm 24.6 1 10dBm 24.6 1 In Table 6, minimum power penalty is found in RZ OOK test in both (a) and (b) if aggressors are kept at 0dBm power, at distance ( 0.1, 0.1). And maximum, in case aggressors are kept at 0dBm, 10dBm 10 and SMF (for (a) 1.1 and for (b) 1). Thus for RZ OOK as test, aggressors at 10dBm power affect the test signal poorly and lead to crosstalk. Similarly in Table 7 the effect of aggressors to test signal CS RZ is examined. TABLE 7 Power penalty for test at 15dBm and aggressors at 0, 5, 10dBm,, and SMF Test CS RZ at 15dBm 1546nm, Power in dbm Power penalty aggressors at 0dBm, MDB NRZ at 1545nm, RZ OOK at 1546.4, MDB RZ at 1546.8nm Back to Back power 26 0 0dBm 22 4 0dBm 21 5 5dBm 21.5 4.5 5dBm 20 6 10dBm 10dBm Test CS RZ at 15dBm 1546nm, aggressors at 0dBm, RZ OOK at 1545nm, MDB NRZ at 1546.4, MDB RZ at 1546.8nm No results Power in dbm Power penalty Back to Back power 26.5 0 0dBm 22 4.5 0dBm 21 5.5 5dBm 21.5 5 5dBm 20.5 6 233 Copyright 2017. Vandana Publications. All Rights Reserved.
10dBm 10dBm From Table 7, it is illustrated that, minimum power penalty is found in CS RZ test signal if aggressors are kept at 0dBm ( 4, 4) in both (c) and (d) and maximum in case aggressors are at 5dBm power, fiber ( 6, 6). For test, maximum crosstalk is due to aggressors at 10dBm over 10, SMF, which does not touch the BER threshold line at 10-9. However, test is affected most by the aggressor channels compared to RZ OOK as a test. No results Since crosstalk is also analyzed by the eye patterns, thus we also demonstrate various eye patterns analyzing the effect of aggressors on the test channel selected. The eye pattern for out band analysis is illustrated in Fig. 5 below. Fig. 5 Illustration of (1) RZ OOK test at 15dBm with aggressors at (a) 0dBm, (b) 5dBm, (c) 10dBm power level (2) CS RZ test at 15dBm with aggressors at (d) 0dBm, (e) 5dBm, (f) 10dBm power level over a length of SMF. From Fig.5 (1), it is seen that as power level increases eye pattern becomes distorted as from Fig.5 (a) to (c). At the same time it is seen that Fig.5 (1) (d), (e), (f), eye diagrams are quite distorted compared to Fig.5 (2). This shows that CS RZ signal as test is affected more by the aggressor signals compared to RZ OOK signal as test. (b) In band analysis In band analysis is carried out by keeping all the channels test as well as aggressors at the same wavelength of 1546nm. We performed different cases given in Table 8: TABLE 8 Illustration of various cases performed for in band analysis at bit rate and over 10, SMF RZ OOK as test at 15dBm and aggressors at 0, 5, 10dBm power as RZ OOK as test at 15dBm and aggressor as test at 15dBm and aggressors at 0, 5, 10dBm power CS RZ CS RZ at 0dBm RZ OOK MDB NRZ CS RZ at 5dBm MDB NRZ MDB RZ CS RZ at 10dBm MDB RZ Keeping test, no results were found for in band analysis showing erroneous transmission and crosstalk. In band analysis is illustrated in Fig. 6 below. 234 Copyright 2017. Vandana Publications. All Rights Reserved.
Fig. 6 Illustration of in band analysis using RZ OOK as test at bit rate over 10, SMF in both (a) and (b). Fig. 6(a) shows aggressors CS RZ, MDB NRZ, MDB RZ at 0, 5, 10dBm power at wavelength 1546nm. Fig.6 (b) shows aggressor alone at 0, 5, 10dBm at 1546nm wavelength. The calculated power penalties from the Fig.6 are mentioned in Table 9. From Table 9 (a), it is seen that the power penalty is minimum in case of CS RZ 0dBm power, SMF and maximum at power 10dBm, SMF. Thus the former case affects RZ OOK test signal the least. From Table 9 (b), power penalty is negative in case of CS RZ at 0dBm. TABLE 9 Illustrates in band power penalty using RZ OOK as test at bit rate and 10, SMF RZ OOK test, CS RZ, MDB NRZ, MDB RZ as aggressors at 0, 5, Power in dbm Power penalty in db 10dBm1546nm wavelength at bit rate, 10, SMF Back to back 25.5 0 0dBm 24.5 1 5dBm 24.3 1.2 5dBm 21 4.5 10dBm 19.5 6 RZ OOK test, aggressor Power in dbm at 0, 5, 10dBm at 1546nm wavelength Back to back 25.3 0 0dBm 25.4-0.1 5dBm 20 5.3 Power penalty in db The eye patterns of in band analysis using RZ OOK test signal are shown in Fig.7. Fig.7 (a) shows 0dBm power over SMF, (b) shows power 5dBm over SMF, (c) shows power 10dBm over SMF. Fig.7 (d) shows power 0dBm over SMF, (e) shows power 5dBm over SMF, (f) shows power 10dBm over SMF. Eye pattern is clear in Fig.7 (a) and worst in Fig.7 (f). Also, it is seen from the analysis that in band analysis shows maximum distortion and crosstalk compared to out band analysis. 235 Copyright 2017. Vandana Publications. All Rights Reserved.
Fig. 7 Illustration of eye diagram of in band analysis (1) RZ OOK as test over SMF with aggressors at (a) 0dBm, (b) 5dBm, (c) 10dBm power level and (2) RZ OOK as test over SMF with aggressors at (d) 0dBm, (e) 5dBm, (f) 10dBm power, showing clear eye pattern for aggressors at 0dBm (a) and distorted pattern signifying crosstalk if aggressors are used at high power of 10dBm (f). IV. CONCLUSION REFERENCES In this paper we experimentally evaluated and quantified the impact of in-band and out band crosstalk for RZ OOK as test, CS RZ, MDB NRZ, MDB RZ as aggressors and test, RZ OOK, MDB NRZ, MDB RZ as aggressors. We used a technique in which the neighboring aggressor channels are shifted towards right and left of the test channel to see the effect on the test signal. Keeping RZ OOK at 1546nm wavelength as test we see that the left shifted aggressor CS RZ at 1545nm, 1545.2nm, 1545.4nm and 1545.8nm wavelengths induced lower penalties than right shifted 1546.2nm, 1546.4nm, 1546.8nm and 1547nm wavelengths. Out band crosstalk analysis also demonstrated that penalties induced on CS RZ at 1546nm wavelength as test channel is higher than RZ OOK as test channel. The in-band crosstalk analysis exhibited BER penalty at worst case at higher bit rates and longer distances with CS RZ, MDB NRZ and MDB RZ as aggressors. In-band crosstalk analysis also demonstrated negative penalty with CS RZ pulse shapes alone as compared with CS RZ, MDB NRZ and MDB RZ taken in combination as aggressors. It is also indicated that in band crosstalk scenarios are more affected by the aggressor channels than out band crosstalk scenarios. [1] Haydar Cukurtepe, et al. Impairment aware lightpath provisioning in mixed-line-rate optical networks, Advanced Networks and Telecommunications Systems (ANTS) 2012 IEEE International conference on, pp. 18-23, 2012, ISSN 2153-1676 [2] Manisha, Vikas Malik. Effect of Crosstalk in Optical Component, www.ijraset.com Volume 3 IssueV, May 2015 IC Value: 13.98ISSN: 2321-9653 [3] Maninder Singh, Maninder Lal Singh. Impairment Aware Routing and Wavelength Assignment Model Employing Binary Logic Operators, 2016 International Conference on Optical NetworkDesign and Modeling (ONDM), Cartagena, 2016, pp.1-6.doi: 10.1109/ONDM.2016.7494081 [4] Fiber-Optic Communications Systems, Third Edition. Sons, Inc. [5] Eugenio Iannone, et al, Modeling of In-Band Crosstalk in WDM Optical Networks. Journel of Lightwave Technology, Vol. 17, NO. 7, JULY 1999 [6] P. S. Andre, et al Performance Degradations due to Crosstalk in Multiwavelength Optical Networks Using Optical Add Drop Multiplexers Based on Fibre Bragg Gratings 236 Copyright 2017. Vandana Publications. All Rights Reserved.
[7] Siamak Azodolmolky, et al, A survey on physical layer impairments aware routing and wavelength assignment algorithms in optical networks Elseiver [8] Optical communications and essentials by Gred Keiser. [9] Karamdeep Singh, et al, A Single As 2 Se 3 Chalcogenide Highly Non-Linear Fiber (HNLF) based Simultaneous All-Optical Half-Adder and Half- Subtracter Optical Fiber Technology, Elsevier Science, Vol. 24, August 2015, pp. 56-63 (Thomson Reuters ISI 2016 Impact Factor- 1.6). (Available atwww.sciencedirect.com) [10] Karamdeep Singh, et al, Simultaneous all-optical half-adder, half-subtracter comparator, and decoder based on nonlinear effects harnessing in highly non-linear fibers Optical Engineering, SPIE, Vol. 55, No. 7, July 2016, pp. 077104 (Thomson Reuters ISI 2016 Impact Factor- 0.984) [11] Karamdeep Singh, et al, Performance analysis of an all-optical half-subtracter based on cross-gain modulation (XGM) in semiconductor optical amplifier (SOA) at 20 Gbps Optoelectronics and advanced materials Rapid communications, Vol. 11, No. ¾, pp. 189-196 (Thomson Reuters ISI 2016 Impact Factor- 0.412). [12] Karamdeep Singh, et al, Enhanced performance of an all-optical half-subtracter based on cross-gain modulation (XGM) in semiconductor optical amplifier (SOA) by accelerating its gain recovery dynamics Photonic Network Communications, Springer, (Acceptedin press) (Thomson Reuters ISI 2016 Impact Factor- 0.557). [13] Karamdeep Singh and Gurmeet Kaur, Interferometric Architectures based All-Optical Logic Design Methods and their Implementations. Optics and Laser Technology, Elsevier Science, Vol. 65, June 2015, pp. 122-132 (Thomson Reuters ISI 2016 Impact Factor- 1.879) (Available at www.sciencedirect.com) [14] Karamdeep Singh, et al, A cascadable all-optical half-subtracter based on cross modulation effects in a single highly non-linear fiber (HNLF) Optical and Quantum Electronics, Springer, Vol. 48, No. 9, September 2016, pp. 418 (Thomson Reuters ISI 2016 Impact Factor- 1.29). [15] G. R. Hill et al., A transport network layer based on optical network elements, J. Lightwave Technol., vol. 11, pp. 667 679, May/June 1993. [16] D. J. Blumenthal, et al, BER floors due to heterodyne coherent crosstalk in space photonic switches for WDM networks, IEEE Photon. Technol. Lett., vol. 8, pp. 284 286, Feb. 1996. [17] S. D. Dods, et al, Homodyne crosstalk in WDM ring and bus networks, IEEE Photon. Technol. Lett., vol. 9, pp. 1285 1287, Sept. 1997. [18] Corrections, IEEE Photon. Technol. Lett., vol. 10, p. 303, Feb. 1998. [19] C. X. Yu, et al, System degradation due to multipath coherent crosstalk in WDM network nodes, J. Lightwave Technol., vol. 16, pp. 1380 1386, Aug. 1998 [20] Eugenio Iannone, et al, Modeling of In-Band Crosstalk in WDM Optical Networks. Journel of lightwave technology VOL. 17, NO. 7, JULY 1999. [21] Oliveira, J. C. R. F., et al. Crosstalk penalties analysis in mixed line transmission rates (10G-OOK/40G- DQPSK/112G-DP-QPSK/224G-DP-16-QAM) optical flexible grid networks. Microwave and Optical Technology Letters 55.1 (2013): 119-122. 237 Copyright 2017. Vandana Publications. All Rights Reserved.