Ghulam Saber
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Ghulam Saber has a total of 83 co-authors. Here are the top co-authors by number of shared publications:
David V. Plant
Department of Electrical and Computer Engineering, McGill University 1 , Montreal, Quebec H3A 0E9, Canada
19 shared publications- Zhenping Xing
Electrical and Computer Engineering, McGill University, Montreal, Quebec, Canada
14 shared publications - Eslam El-Fiky
Department of Electrical and Computer Engineering, McGill University, Montreal, QC, Canada
13 shared publications - Luhua Xu
Department of Electrical and Computer Engineering, McGill University, 5620 Montreal, Quebec Canada H3A 0E9
11 shared publications David Patel
Department of Electrical and Computer Engineering McGill University Montreal Quebec H3A 0E9 Canada
10 shared publications- Maxime Jacques
Electrical and Computer Engineering, McGill University, Montral, Quebec, Canada
9 shared publications
Latest Publications
- 29 October 2026
- Published by Optica Publishing Group in Journal of Optical Communications and Networking
- Vol. 18 (11) , 1333
Transition-aware routing in hybrid hollow-core/single-mode fiber networks: a cost–throughput investigation
Incremental deployment of hollow-core fiber (HCF) in single-mode-fiber (SMF) networks introduces a routing trade-off: reducing HCF–SMF transitions can improve physical-layer feasibility, but overly transition-averse routing incurs harmful path detours. We study it with a common event-driven simulator comparing six protected routing schemes—fiber-blind, generalized signal-to-noise ratio (GSNR)-aware, and explicitly transition-aware—on hybrid HCF/SMF topologies, using a per-transition GSNR penalty and an exploratory splice-failure availability term. Across six reference topologies, five HCF deployment fractions, and dynamic loads, the strongest minimizers—transition-penalty-aware routing (TPAR) and the GSNR/fiber-transition joint scheme (GFJ)—halve the mean transition count at a 20%–25% carried-traffic penalty. Among the intermediate designs, GSNR-maximal routing with transition-aware reranking (GMR-T) cuts transitions by ∼21% versus distance-adaptive routing and spectrum assignment (DA-RSA) at a 6% throughput cost under heavy load, while below the blocking knee bounded-detour TPAR (BD-TPAR) removes ∼12% of them at no measurable throughput cost. Contiguous HCF rollout lowers transitions by ∼42% on average while improving carried traffic, reducing the benefit of aggressive transition-aware routing, and the ranking survives sweeps of the per-transition penalty, of the differential-delay allowance, and of the node-level HCF–SMF interface loss every HCF span carries. These results support GMR-T as the practical default at heavy load, BD-TPAR at moderate load, and TPAR/GFJ only where external transition cost is high.
- 29 October 2026
- Published by Optica Publishing Group in Journal of Optical Communications and Networking
- Vol. 18 (11) , 1333
Transition-aware routing in hybrid hollow-core/single-mode fiber networks: a cost–throughput investigation
Incremental deployment of hollow-core fiber (HCF) in single-mode-fiber (SMF) networks introduces a routing trade-off: reducing HCF–SMF transitions can improve physical-layer feasibility, but overly transition-averse routing incurs harmful path detours. We study it with a common event-driven simulator comparing six protected routing schemes—fiber-blind, generalized signal-to-noise ratio (GSNR)-aware, and explicitly transition-aware—on hybrid HCF/SMF topologies, using a per-transition GSNR penalty and an exploratory splice-failure availability term. Across six reference topologies, five HCF deployment fractions, and dynamic loads, the strongest minimizers—transition-penalty-aware routing (TPAR) and the GSNR/fiber-transition joint scheme (GFJ)—halve the mean transition count at a 20%–25% carried-traffic penalty. Among the intermediate designs, GSNR-maximal routing with transition-aware reranking (GMR-T) cuts transitions by ∼21% versus distance-adaptive routing and spectrum assignment (DA-RSA) at a 6% throughput cost under heavy load, while below the blocking knee bounded-detour TPAR (BD-TPAR) removes ∼12% of them at no measurable throughput cost. Contiguous HCF rollout lowers transitions by ∼42% on average while improving carried traffic, reducing the benefit of aggressive transition-aware routing, and the ranking survives sweeps of the per-transition penalty, of the differential-delay allowance, and of the node-level HCF–SMF interface loss every HCF span carries. These results support GMR-T as the practical default at heavy load, BD-TPAR at moderate load, and TPAR/GFJ only where external transition cost is high.
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 44 (18) , 8069-8083
System-Level Limits of Higher-Order QAM in Hollow-Core Fiber Systems
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 44 (18) , 8069-8083
System-Level Limits of Higher-Order QAM in Hollow-Core Fiber Systems
- Open Access
Beyond Silica Assumptions: Optical Network Design in the Hollow-Core Era
Hollow-core fiber (HCF) is often presented as an incrementally better transmission medium that can be slotted into networks designed around solid-core silica. We argue instead that recent progress—most visibly reported as attenuations below 0.1 dB/km and now approaching 0.05 dB/km, together with a broad low-loss window, reduced propagation delay and very low optical nonlinearity—makes it worth asking which long-standing design conventions are intrinsic to optical communication and which are artifacts of silica. Reviewing physical-layer, transceiver and network architecture implications, we suggest that the most durable gains may come not from treating HCF as a drop-in replacement, but from cross-layer co-design, and we outline the studies and demonstrations needed to test where that advantage is real.
Open AccessBeyond Silica Assumptions: Optical Network Design in the Hollow-Core Era
Hollow-core fiber (HCF) is often presented as an incrementally better transmission medium that can be slotted into networks designed around solid-core silica. We argue instead that recent progress—most visibly reported as attenuations below 0.1 dB/km and now approaching 0.05 dB/km, together with a broad low-loss window, reduced propagation delay and very low optical nonlinearity—makes it worth asking which long-standing design conventions are intrinsic to optical communication and which are artifacts of silica. Reviewing physical-layer, transceiver and network architecture implications, we suggest that the most durable gains may come not from treating HCF as a drop-in replacement, but from cross-layer co-design, and we outline the studies and demonstrations needed to test where that advantage is real.
- Open Access
CO2-Limited Hollow-Core Fiber Links: A Capacity-Map Guide to Pre-Emphasis and Spectral Avoidance
CO2 gas-line absorption is emerging as a major L-band impairment in low-loss hollow-core fiber (HCF) links. We compare two transponder-side mitigation strategies—spectral pre-emphasis and spectral avoidance—over span lengths of 100–300 km and transmission reach of up to 3000 km. The preferred strategy depends on reach, launch power, span length, and the stability of the live-link absorption comb. Pre-emphasis is favored at short reach and for short spans, whereas spectral avoidance is superior at moderate to long reach, with a peak capacity gain of about 4 Tb/s. Pre-emphasis is also more sensitive to mismatch between the design-time and live-link absorption combs: increasing the live absorption peak from 0.10 to 0.35 dB/km reduces capacity by up to 8.5 Tb/s, while tripling the CO2 absorption linewidth reduces capacity by up to 10.3 Tb/s. We further review implementation options for both methods: DGFF-based pre-emphasis at the WSS sites, and DSP-based avoidance via digital subcarrier multiplexing (DSCM) or entropy-loaded orthogonal frequency-division multiplexing (OFDM). These results provide a concise framework for selecting mitigation strategy under realistic operating conditions.
Open AccessCO2-Limited Hollow-Core Fiber Links: A Capacity-Map Guide to Pre-Emphasis and Spectral Avoidance
CO2 gas-line absorption is emerging as a major L-band impairment in low-loss hollow-core fiber (HCF) links. We compare two transponder-side mitigation strategies—spectral pre-emphasis and spectral avoidance—over span lengths of 100–300 km and transmission reach of up to 3000 km. The preferred strategy depends on reach, launch power, span length, and the stability of the live-link absorption comb. Pre-emphasis is favored at short reach and for short spans, whereas spectral avoidance is superior at moderate to long reach, with a peak capacity gain of about 4 Tb/s. Pre-emphasis is also more sensitive to mismatch between the design-time and live-link absorption combs: increasing the live absorption peak from 0.10 to 0.35 dB/km reduces capacity by up to 8.5 Tb/s, while tripling the CO2 absorption linewidth reduces capacity by up to 10.3 Tb/s. We further review implementation options for both methods: DGFF-based pre-emphasis at the WSS sites, and DSP-based avoidance via digital subcarrier multiplexing (DSCM) or entropy-loaded orthogonal frequency-division multiplexing (OFDM). These results provide a concise framework for selecting mitigation strategy under realistic operating conditions.
- 17 November 2025
- Published by Optica Publishing Group in Optics Letters
- Vol. 50 (23) , 7211-7214
OTDR trace distortion correction in Raman-amplified systems using transformer models
We present a Transformer model-based method for correcting optical time domain reflectometer (OTDR) trace distortions in Raman-amplified optical systems. Unlike prior approaches, the proposed model operates without requiring Raman pump power information or a reference trace, relying only on the distorted OTDR trace, fiber length, and pump wavelengths—parameters readily available in deployed networks. By predicting the accumulated stimulated Raman scattering (SRS) gain, the model effectively removes Raman-induced distortions in a single step, eliminating the need for additional curve fitting. Experimental results on a 180 km Raman-amplified span show that the model reduces mean squared error (MSE) by over an order of magnitude, restoring baseline trace features and connector losses with high fidelity. The method enhances robustness to Raman pump power-related uncertainties, simplifies the correction process, and enables scalable, in-service diagnostics for high-capacity fiber links.
- 17 November 2025
- Published by Optica Publishing Group in Optics Letters
- Vol. 50 (23) , 7211-7214
OTDR trace distortion correction in Raman-amplified systems using transformer models
We present a Transformer model-based method for correcting optical time domain reflectometer (OTDR) trace distortions in Raman-amplified optical systems. Unlike prior approaches, the proposed model operates without requiring Raman pump power information or a reference trace, relying only on the distorted OTDR trace, fiber length, and pump wavelengths—parameters readily available in deployed networks. By predicting the accumulated stimulated Raman scattering (SRS) gain, the model effectively removes Raman-induced distortions in a single step, eliminating the need for additional curve fitting. Experimental results on a 180 km Raman-amplified span show that the model reduces mean squared error (MSE) by over an order of magnitude, restoring baseline trace features and connector losses with high fidelity. The method enhances robustness to Raman pump power-related uncertainties, simplifies the correction process, and enables scalable, in-service diagnostics for high-capacity fiber links.
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Technology Letters
- Vol. 38 (1) , 1-4
AI-Assisted Online Compensation of SRS Effects in OTDR Traces for Super C+L-Band Systems
The emergence of wideband optical networks, particularly those utilizing super C+L-band transmission, introduces new challenges for in-service optical time-domain reflectometer (OTDR) due to the effects of stimulated Raman scattering (SRS). As OTDR probe pulses propagate through a fiber carrying high-bandwidth traffic, they experience wavelength-dependent amplification or depletion from SRS, resulting in distorted traces that no longer accurately reflect the fiber’s attenuation profile. In this work, we present a Transformer model-based online approach to correct OTDR trace distortions caused by super C+L-band traffic to restore the fiber loss profile. Notably, the model operates solely on the distorted OTDR trace without requiring knowledge of the fiber launch power distribution or any auxiliary input. The approach is experimentally validated over a 12 THz super C+L-band transmission system by varying the total fiber input power to induce different amount of SRS gain in the OTDR traces. By applying the correction method, the mean absolute errors (MAE) between the OTDR traces distorted by the traffic and the traces when there is no traffic were reduced from 4.7, 3.24, 2.18, and 1.51 dB to 0.072, 0.062, 0.063, and 0.09 dB, respectively, across four launch power settings.
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Technology Letters
- Vol. 38 (1) , 1-4
AI-Assisted Online Compensation of SRS Effects in OTDR Traces for Super C+L-Band Systems
The emergence of wideband optical networks, particularly those utilizing super C+L-band transmission, introduces new challenges for in-service optical time-domain reflectometer (OTDR) due to the effects of stimulated Raman scattering (SRS). As OTDR probe pulses propagate through a fiber carrying high-bandwidth traffic, they experience wavelength-dependent amplification or depletion from SRS, resulting in distorted traces that no longer accurately reflect the fiber’s attenuation profile. In this work, we present a Transformer model-based online approach to correct OTDR trace distortions caused by super C+L-band traffic to restore the fiber loss profile. Notably, the model operates solely on the distorted OTDR trace without requiring knowledge of the fiber launch power distribution or any auxiliary input. The approach is experimentally validated over a 12 THz super C+L-band transmission system by varying the total fiber input power to induce different amount of SRS gain in the OTDR traces. By applying the correction method, the mean absolute errors (MAE) between the OTDR traces distorted by the traffic and the traces when there is no traffic were reduced from 4.7, 3.24, 2.18, and 1.51 dB to 0.072, 0.062, 0.063, and 0.09 dB, respectively, across four launch power settings.
- 4 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- Vol. 40 (2) , 147-155
Physical Layer Standardization for AI Data Centers: Challenges, Progress, and Perspectives
The rapid advancement of artificial intelligence (AI) is driving unprecedented demand on datacenter infrastructure, especially within the physical layer. With the parameter size of large language models doubling every four months in the recent years, physical layer connectivity faces immense pressure to meet new demands for reliability, bandwidth, power efficiency, and latency. This paper begins by examining the architecture of AI datacenters, highlighting key network segments, detailing specific requirements for critical functions such as model training and inference. Next, we review recent progress in input/output (I/O) interconnect technologies for both electrical and optical links, along with the major challenges that remain. Finally, we explore ongoing standardization efforts for electrical and optical interconnects, emphasizing the potential for achieving 400G per lane and assessing the feasibility of convergence across various standards.
- 4 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- Vol. 40 (2) , 147-155
Physical Layer Standardization for AI Data Centers: Challenges, Progress, and Perspectives
The rapid advancement of artificial intelligence (AI) is driving unprecedented demand on datacenter infrastructure, especially within the physical layer. With the parameter size of large language models doubling every four months in the recent years, physical layer connectivity faces immense pressure to meet new demands for reliability, bandwidth, power efficiency, and latency. This paper begins by examining the architecture of AI datacenters, highlighting key network segments, detailing specific requirements for critical functions such as model training and inference. Next, we review recent progress in input/output (I/O) interconnect technologies for both electrical and optical links, along with the major challenges that remain. Finally, we explore ongoing standardization efforts for electrical and optical interconnects, emphasizing the potential for achieving 400G per lane and assessing the feasibility of convergence across various standards.
- Open Access
- 1 April 2021
- Published by AIP Publishing in AIP Advances
- Vol. 11 (4) , 045219
Broadband all-silicon hybrid plasmonic TM-pass polarizer using bend waveguides
A complementary–metal–oxide semiconductor (CMOS) compatible all-silicon TM-pass polarizer using plasmonic bends is proposed. To simplify the fabrication and be compatible with the CMOS process, we employ only two materials: silicon and silicon dioxide. Highly doped silicon is used to support the plasmons. We obtain an extinction ratio and an insertion loss of 45.4 and 1.7 dB, respectively, at 1550 nm and a maximum extinction ratio of 58 dB. This is the highest reported extinction ratio for a TM-pass polarizer to the best of our knowledge. Furthermore, we achieved >20 dB of extinction ratio and μm2. To achieve this, we exploit the properties of tight bends in plasmonic waveguides. Another advantage of the device is that it is robust against fabrication variations.
- 1 April 2021
- Published by AIP Publishing in AIP Advances
- Vol. 11 (4) , 045219
Open AccessBroadband all-silicon hybrid plasmonic TM-pass polarizer using bend waveguides
A complementary–metal–oxide semiconductor (CMOS) compatible all-silicon TM-pass polarizer using plasmonic bends is proposed. To simplify the fabrication and be compatible with the CMOS process, we employ only two materials: silicon and silicon dioxide. Highly doped silicon is used to support the plasmons. We obtain an extinction ratio and an insertion loss of 45.4 and 1.7 dB, respectively, at 1550 nm and a maximum extinction ratio of 58 dB. This is the highest reported extinction ratio for a TM-pass polarizer to the best of our knowledge. Furthermore, we achieved >20 dB of extinction ratio and μm2. To achieve this, we exploit the properties of tight bends in plasmonic waveguides. Another advantage of the device is that it is robust against fabrication variations.
- Open Access
Systematic Performance Comparison of (Duobinary)-PAM-2,4 Signaling under Light and Strong Opto-Electronic Bandwidth Conditions
We present a systematic comparison of PAM-2 (NRZ), Duobinary-PAM-2, PAM-4, and Duobinary-PAM-4 (duo-quaternary) signaling in the context of short-reach photonic communications systems using a Mach–Zehnder modulator as transmitter. The effect on system performance with a relaxed and constrained system’s opto-electronic bandwidth is analyzed for bit rates ranging from 20 to 116 Gb/s. In contrast to previous analyses, our approach employs the same experimental and simulation conditions for all modulation formats. Consequently, we were able to confidently determine the performance limits of each format for particular values of bit rate, system bandwidth, transmitter chirp, and fiber dispersion. We demonstrate that Duobinary-PAM-4 is a good signaling choice only for bandwidth-limited systems operating at relatively high speed. Otherwise, PAM-4 represents a more sensible choice. Moreover, our analysis put forward the existence of transition points: specific bit rate values where the BER versus bit rate curves for two different formats cross each other. They indicate the bit rate values where, for specific system conditions, switching from one modulation to another guarantees optimum performance. Their existence naturally led to the proposal of a format-selective transceiver, a component that, according to network conditions, operates with the most adequate modulation format. Since all analyzed modulations share similar implementation details, signaling switching is achieved by simply changing the sampling point and threshold count at the receiver, bringing flexibility to IM/DD-based optical networks.
Open AccessSystematic Performance Comparison of (Duobinary)-PAM-2,4 Signaling under Light and Strong Opto-Electronic Bandwidth Conditions
We present a systematic comparison of PAM-2 (NRZ), Duobinary-PAM-2, PAM-4, and Duobinary-PAM-4 (duo-quaternary) signaling in the context of short-reach photonic communications systems using a Mach–Zehnder modulator as transmitter. The effect on system performance with a relaxed and constrained system’s opto-electronic bandwidth is analyzed for bit rates ranging from 20 to 116 Gb/s. In contrast to previous analyses, our approach employs the same experimental and simulation conditions for all modulation formats. Consequently, we were able to confidently determine the performance limits of each format for particular values of bit rate, system bandwidth, transmitter chirp, and fiber dispersion. We demonstrate that Duobinary-PAM-4 is a good signaling choice only for bandwidth-limited systems operating at relatively high speed. Otherwise, PAM-4 represents a more sensible choice. Moreover, our analysis put forward the existence of transition points: specific bit rate values where the BER versus bit rate curves for two different formats cross each other. They indicate the bit rate values where, for specific system conditions, switching from one modulation to another guarantees optimum performance. Their existence naturally led to the proposal of a format-selective transceiver, a component that, according to network conditions, operates with the most adequate modulation format. Since all analyzed modulations share similar implementation details, signaling switching is achieved by simply changing the sampling point and threshold count at the receiver, bringing flexibility to IM/DD-based optical networks.