Ranjan Das
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ranjan Das has a total of 70 co-authors. Here are the top co-authors by number of shared publications:
- Qingfeng Zhang
Southern University of Science and Technology, Shenzhen, China
12 shared publications - Thomas Schneider
Technische Universität Braunschweig
8 shared publications Abhishek Kandwal
SUSTech, Department of Electrical and Electronics Engineering, China
6 shared publications- Abhishek Kandwal
School of Chips, Xi’an Jiaotong-Liverpool University, Suzhou, China
6 shared publications Jayanta Mukherjee
Electrical Engineering Department, Indian Institute of Technology Bombay, Mumbai, India
5 shared publications- Louis Wai Yip Liu
Faculty of Engineering, Vietnamese German University, Main Gate, Ring Road 4, Quarter 4, Thoi Hoa Ward, Ben Cat City 75911, Binh Duong Province, Vietnam
5 shared publications
Latest Publications
- Open Access
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Journal
- Vol. 18 (5) , 6601305
Silicon Photonic Paper-Clip Spiral Waveguides With Ultra-Low Propagation and Delay Loss
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Journal
- Vol. 18 (5) , 6601305
Open AccessSilicon Photonic Paper-Clip Spiral Waveguides With Ultra-Low Propagation and Delay Loss
- Open Access
- 13 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Journal
- Vol. 18 (3) , 0600808
Noise Characterization of All-Silicon Avalanche Photodetectors Operating at 1550 nm
In this work, we characterize the noise associated with the operation of all-silicon avalanche photodetectors integrated on a silicon-on-insulator platform, designed to operate at 1550 nm. The silicon waveguide photodetectors are sensitized to sub-bandgap light through inert ion implantation and low-temperature annealing. The devices are subsequently operated in the avalanche mode. The excess noise factor is correlated with the mid-gap defect concentration, with an associated saturated value $k=0.1$, where $k$ is defined as the ratio of hole to electron impact ionization coefficient. In addition, the noise equivalent power (NEP) is measured to be 30 pW/$\sqrt{\text{Hz}}$, achieved in the device with lowest ion implantation dose.
- 13 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Journal
- Vol. 18 (3) , 0600808
Open AccessNoise Characterization of All-Silicon Avalanche Photodetectors Operating at 1550 nm
In this work, we characterize the noise associated with the operation of all-silicon avalanche photodetectors integrated on a silicon-on-insulator platform, designed to operate at 1550 nm. The silicon waveguide photodetectors are sensitized to sub-bandgap light through inert ion implantation and low-temperature annealing. The devices are subsequently operated in the avalanche mode. The excess noise factor is correlated with the mid-gap defect concentration, with an associated saturated value $k=0.1$, where $k$ is defined as the ratio of hole to electron impact ionization coefficient. In addition, the noise equivalent power (NEP) is measured to be 30 pW/$\sqrt{\text{Hz}}$, achieved in the device with lowest ion implantation dose.
- Open Access
- 1 May 2026
- Published by Springer Nature
- p. 241-256
Integrated Photonic-Assisted Signal Processing and Thermal Crosstalk
Integrated photonic-assisted signal processing has multiple applications such as signal amplification, multiplexing, and high-Q filtering in optical communication systems, optical sensing systems, and also microwave photonics. We will review recent works on integrated photonic-assisted signal processing for sinc-shaped Nyquist pulse generation, high-bandwidth Nyquist signal detection with low bandwidth devices, arbitrary waveform generation and measurement, and on-chip photonic frequency decoding. However, in such photonic integrated circuits (PICs), the photonic components are placed very close to each other on the chip, resulting in thermal crosstalk which degrades the system performance. Air-filled oxide and deep trench designs have proven to be very effective in mitigating the thermal crosstalk for various frequently deployed photonic devices like Mach-Zehnder modulators (MZMs), ring resonators, optical switches, and photodetectors designed on a standard silicon-on-insulator (SOI) platform. In this chapter, we will additionally review the basics of optical signal processing and some results for such trench-enhanced thermal crosstalk resilient circuits.
- 1 May 2026
- Published by Springer Nature
- p. 241-256
Open AccessIntegrated Photonic-Assisted Signal Processing and Thermal Crosstalk
Integrated photonic-assisted signal processing has multiple applications such as signal amplification, multiplexing, and high-Q filtering in optical communication systems, optical sensing systems, and also microwave photonics. We will review recent works on integrated photonic-assisted signal processing for sinc-shaped Nyquist pulse generation, high-bandwidth Nyquist signal detection with low bandwidth devices, arbitrary waveform generation and measurement, and on-chip photonic frequency decoding. However, in such photonic integrated circuits (PICs), the photonic components are placed very close to each other on the chip, resulting in thermal crosstalk which degrades the system performance. Air-filled oxide and deep trench designs have proven to be very effective in mitigating the thermal crosstalk for various frequently deployed photonic devices like Mach-Zehnder modulators (MZMs), ring resonators, optical switches, and photodetectors designed on a standard silicon-on-insulator (SOI) platform. In this chapter, we will additionally review the basics of optical signal processing and some results for such trench-enhanced thermal crosstalk resilient circuits.
- 16 April 2026
- Published by Optica Publishing Group in Journal of the Optical Society of America B
- Vol. 43 (5) , 967
Crosstalk-assisted thermo-optic switching on a silicon photonics platform
Key performance metrics for thermo-optic switches often exhibit inverse relationships. This study examines a symmetric thermo-optic Mach–Zehnder interferometer (TO-MZI) switch across a 3.5–41.5 µm interferometric waveguide gap range, and an asymmetric TO-MZI switch across a 7.5–22.5 µm interferometric waveguide gap range, evaluating the correlated impact of footprint, thermal crosstalk, switching time, and power consumption. Simulations and experimental investigations lead us to summarize device performance via a defined figure of merit (FoM). We demonstrate that there exists an optimized physical gap between interferometric waveguide arms which results in improved overall performance. Specifically, gap optimization provides a minimum footprint below which switching power increases significantly. We also show the impact of this optimization on switching time, which is reduced as the gap decreases. These findings highlight the trade-offs and benefits of gap tuning in the TO-MZI design.
- 16 April 2026
- Published by Optica Publishing Group in Journal of the Optical Society of America B
- Vol. 43 (5) , 967
Crosstalk-assisted thermo-optic switching on a silicon photonics platform
Key performance metrics for thermo-optic switches often exhibit inverse relationships. This study examines a symmetric thermo-optic Mach–Zehnder interferometer (TO-MZI) switch across a 3.5–41.5 µm interferometric waveguide gap range, and an asymmetric TO-MZI switch across a 7.5–22.5 µm interferometric waveguide gap range, evaluating the correlated impact of footprint, thermal crosstalk, switching time, and power consumption. Simulations and experimental investigations lead us to summarize device performance via a defined figure of merit (FoM). We demonstrate that there exists an optimized physical gap between interferometric waveguide arms which results in improved overall performance. Specifically, gap optimization provides a minimum footprint below which switching power increases significantly. We also show the impact of this optimization on switching time, which is reduced as the gap decreases. These findings highlight the trade-offs and benefits of gap tuning in the TO-MZI design.
- Open Access
- 30 August 2025
- Published by Springer Nature
- p. 257-271
Error Metrics for Photonics-Assisted Signal Processing
For the next generation of wireless communications (6G and beyond), peak data rates of several Tbit/s are foreseen. To keep up with such ever-rising demands, photonics-assisted signal processing has been considered a viable solution where conventional copper-based electronics suffers from excess power dissipation issue. This is also attributed to its high bandwidth, energy efficiency, CMOS compatibility, and compact design. Photonics-assisted analog-to-digital converters (PADCs) have been regarded as an alternative to bandwidth and signal integrity-limited electronic analog-to-digital converter (EADC). For the comparison of the quality, different error metrics such as SINAD, ENOB, and root mean square error (RMSE) are used. In this chapter, we begin by describing and discussing these various metrics, followed by using them to evaluate the performance of PADC and compare it with EADC. We address different error sources including time jitter and laser relative intensity noise (RIN). As we further validate by simulation and experiment, PADC may provide better signal-to-noise and distortion ratio (SINAD), effective number of bits (ENOB), and RMSE than state-of-the-art EADC.
- 30 August 2025
- Published by Springer Nature
- p. 257-271
Open AccessError Metrics for Photonics-Assisted Signal Processing
For the next generation of wireless communications (6G and beyond), peak data rates of several Tbit/s are foreseen. To keep up with such ever-rising demands, photonics-assisted signal processing has been considered a viable solution where conventional copper-based electronics suffers from excess power dissipation issue. This is also attributed to its high bandwidth, energy efficiency, CMOS compatibility, and compact design. Photonics-assisted analog-to-digital converters (PADCs) have been regarded as an alternative to bandwidth and signal integrity-limited electronic analog-to-digital converter (EADC). For the comparison of the quality, different error metrics such as SINAD, ENOB, and root mean square error (RMSE) are used. In this chapter, we begin by describing and discussing these various metrics, followed by using them to evaluate the performance of PADC and compare it with EADC. We address different error sources including time jitter and laser relative intensity noise (RIN). As we further validate by simulation and experiment, PADC may provide better signal-to-noise and distortion ratio (SINAD), effective number of bits (ENOB), and RMSE than state-of-the-art EADC.
- Open Access
- 30 August 2025
- Published by Springer Nature
- p. 289-299
Metrological Analysis of Non-idealities for Photonics-Assisted Signal Processing
Waveform metrology provides the key to understand and analyze the waveform behavior of different systems and environments. To ensure reliable measurements with high precision, this is particularly crucial for calibrating and characterizing electrical systems, like high-speed sampling oscilloscopes, spectrum analyzers, waveform generators, etc. Additionally, full waveform metrology techniques make it possible to derive parameters from devices operating in the time domain like oscilloscopes, as well as devices operating in the frequency domain such as vector network analyzers. This ensures a comprehensive system analysis with a complete transfer function with the full measurement uncertainty represented by a covariance matrix. For high-bandwidth signals, especially for Terahertz (THz) or sub-THz communications, photonics-assisted signal processing is advantageous as it offers a much wider bandwidth compared to its electrical counterpart with no electromagnetic interference. As such, it is important to determine the transfer function (TF) of a complete optical sampling system adapting techniques from waveform metrology. Therefore, in this chapter, we perform the metrological evaluation of the optical system in terms of non-idealities like amplitude ripple, sideband suppression ratio (SSR), and spectrum roll-off factor $$\beta $$ and study parameters like the root mean square error (RMSE), signal-to-noise and distortion ratio (SINAD), and the effective number of bits (ENOB).- 30 August 2025
- Published by Springer Nature
- p. 289-299
Open AccessMetrological Analysis of Non-idealities for Photonics-Assisted Signal Processing
Waveform metrology provides the key to understand and analyze the waveform behavior of different systems and environments. To ensure reliable measurements with high precision, this is particularly crucial for calibrating and characterizing electrical systems, like high-speed sampling oscilloscopes, spectrum analyzers, waveform generators, etc. Additionally, full waveform metrology techniques make it possible to derive parameters from devices operating in the time domain like oscilloscopes, as well as devices operating in the frequency domain such as vector network analyzers. This ensures a comprehensive system analysis with a complete transfer function with the full measurement uncertainty represented by a covariance matrix. For high-bandwidth signals, especially for Terahertz (THz) or sub-THz communications, photonics-assisted signal processing is advantageous as it offers a much wider bandwidth compared to its electrical counterpart with no electromagnetic interference. As such, it is important to determine the transfer function (TF) of a complete optical sampling system adapting techniques from waveform metrology. Therefore, in this chapter, we perform the metrological evaluation of the optical system in terms of non-idealities like amplitude ripple, sideband suppression ratio (SSR), and spectrum roll-off factor $$\beta $$ and study parameters like the root mean square error (RMSE), signal-to-noise and distortion ratio (SINAD), and the effective number of bits (ENOB). - 15 April 2024
- Published by Optica Publishing Group in Optics Letters
- Vol. 49 (8) , 2193-2196
Integrated segmented IQ-modulator for orthogonal sampling and multi-level high-bandwidth signal generation
Photonic-assisted signal processing of high-bandwidth signals emerges as a solution for challenges encountered in electronic-based processing. Here we present a concept for a compact, photonic-assisted digital-to-analog converter (DAC) and optical IQ-modulator in one single integrated device based on two innovative concepts: a segmented Mach–Zehnder modulator and orthogonal sampling. For electrically driving the modulator, only a single radio frequency oscillator and no pulse source or electrical DAC are required. The presented and simulated proof-of-concept device with six segments can generate a multi-level and high-bandwidth signal from low-bandwidth electronic drivers; e.g., we show the generation of a 120 Gbps data rate, 16-quadrature amplitude modulation (16-QAM, 30 Gbaud) signal solely based on low-bandwidth (5 GHz) non-return-to-zero (NRZ) signals. Integrated on a silicon photonic platform, the device provides fixable speed and bandwidth operations, positioning it as a viable solution for diverse communication systems.
- 15 April 2024
- Published by Optica Publishing Group in Optics Letters
- Vol. 49 (8) , 2193-2196
Integrated segmented IQ-modulator for orthogonal sampling and multi-level high-bandwidth signal generation
Photonic-assisted signal processing of high-bandwidth signals emerges as a solution for challenges encountered in electronic-based processing. Here we present a concept for a compact, photonic-assisted digital-to-analog converter (DAC) and optical IQ-modulator in one single integrated device based on two innovative concepts: a segmented Mach–Zehnder modulator and orthogonal sampling. For electrically driving the modulator, only a single radio frequency oscillator and no pulse source or electrical DAC are required. The presented and simulated proof-of-concept device with six segments can generate a multi-level and high-bandwidth signal from low-bandwidth electronic drivers; e.g., we show the generation of a 120 Gbps data rate, 16-quadrature amplitude modulation (16-QAM, 30 Gbaud) signal solely based on low-bandwidth (5 GHz) non-return-to-zero (NRZ) signals. Integrated on a silicon photonic platform, the device provides fixable speed and bandwidth operations, positioning it as a viable solution for diverse communication systems.
- 11 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 45-48
Segmented Mach-Zehnder Modulator for Orthogonal Sampling Based High Bandwidth THz Transmitters
We present the concept for a compact, integrated, photonics-assisted digital-to-analog converter and optical modulator in one single device. The system takes advantage of two ideas namely: A segmented Mach-Zehnder modulator (MZM), and orthogonal sampling. The presented and simulated proof-of-concept device with three segments has a bandwidth of 15 GHz and a sampling rate of 30 GSs. No pulse source is required as it is generated from a single radio frequency oscillator and a properly biased segmented modulator. The system can enable the generation and modulation of high-bandwidth signals within the Terahertz range.
- 11 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 45-48
Segmented Mach-Zehnder Modulator for Orthogonal Sampling Based High Bandwidth THz Transmitters
We present the concept for a compact, integrated, photonics-assisted digital-to-analog converter and optical modulator in one single device. The system takes advantage of two ideas namely: A segmented Mach-Zehnder modulator (MZM), and orthogonal sampling. The presented and simulated proof-of-concept device with three segments has a bandwidth of 15 GHz and a sampling rate of 30 GSs. No pulse source is required as it is generated from a single radio frequency oscillator and a properly biased segmented modulator. The system can enable the generation and modulation of high-bandwidth signals within the Terahertz range.
- 29 February 2024
- Published by Optica Publishing Group in Optics Letters
- Vol. 49 (5) , 1345-1348
Reconfigurable integrated photonic filters for optical signal processing using a silicon photonic platform
We present a systematic photonic filter design approach by deploying pole-zero optimization. The filter transfer function is derived from its specifications by formulating closed-form optimization objective functions and subsequently translating them into optical design parameters. Two distinct filter examples, namely Chebyshev and elliptic filters, are considered for the design and validation. A compact reconfigurable three-pole photonic filter is fabricated on a silicon photonic platform to illustrate the proposed design technique including transmission tunability. Integrated thermal phase shifters coupled with micro-ring resonators are used to reconfigure filter responses. A well-matched experimental demonstration is presented to validate the proposed tuning method. We achieved a sharp out-of-band edge rejection of at least 20 and 40 dB for the elliptic and Chebyshev filter, respectively.
- 29 February 2024
- Published by Optica Publishing Group in Optics Letters
- Vol. 49 (5) , 1345-1348
Reconfigurable integrated photonic filters for optical signal processing using a silicon photonic platform
We present a systematic photonic filter design approach by deploying pole-zero optimization. The filter transfer function is derived from its specifications by formulating closed-form optimization objective functions and subsequently translating them into optical design parameters. Two distinct filter examples, namely Chebyshev and elliptic filters, are considered for the design and validation. A compact reconfigurable three-pole photonic filter is fabricated on a silicon photonic platform to illustrate the proposed design technique including transmission tunability. Integrated thermal phase shifters coupled with micro-ring resonators are used to reconfigure filter responses. A well-matched experimental demonstration is presented to validate the proposed tuning method. We achieved a sharp out-of-band edge rejection of at least 20 and 40 dB for the elliptic and Chebyshev filter, respectively.
- 5 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 42 (2) , 704-712
Modification of the Electrical Properties of a Silicon Waveguide Avalanche Photodetector Operating at 1550 nm via Defect Engineering
Monolithic silicon sub-bandgap avalanche photodetectors are fabricated using the introduction of lattice defects to silicon waveguides through silicon ion implantation and annealing. The devices are characterized using light at a wavelength of 1550 nm. Through control of implantation dose we create a low-absorption device with 0.4 dB (5.4 dB/cm) excess optical absorption, and a high-absorption device with 3.2 dB (42 dB/cm) excess optical absorption. The low-frequency unity-gain bias point, M = 1, is approximately 12 V for both devices. Increasing reverse bias results in electrical gain such that M = 35 at 32 V for the high-absorption device, with associated responsivity of 13 A/W for coupled optical power ranging from −20 dBm to −5 dBm. The electrical bandwidth of this device peaks at 10 GHz for M = 14 and decreases to 6 GHz for M = 35. The RC limited bandwidth of an unimplanted, control device is measured to be 26 GHz. The gain-bandwidth product of the high-absorption device is 230 GHz. Measurement of the excess noise factor provides an effective k -value of 0.1, consistent with the gain-bandwidth product. Primary carriers in the avalanche process are predominatly electrons, with a carrier ionization threshold of ∼ 0.6eV, indicating for the first time in this device type that both the light detection and avalanche process are mediated by the deep-levels.
- 5 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 42 (2) , 704-712
Modification of the Electrical Properties of a Silicon Waveguide Avalanche Photodetector Operating at 1550 nm via Defect Engineering
Monolithic silicon sub-bandgap avalanche photodetectors are fabricated using the introduction of lattice defects to silicon waveguides through silicon ion implantation and annealing. The devices are characterized using light at a wavelength of 1550 nm. Through control of implantation dose we create a low-absorption device with 0.4 dB (5.4 dB/cm) excess optical absorption, and a high-absorption device with 3.2 dB (42 dB/cm) excess optical absorption. The low-frequency unity-gain bias point, M = 1, is approximately 12 V for both devices. Increasing reverse bias results in electrical gain such that M = 35 at 32 V for the high-absorption device, with associated responsivity of 13 A/W for coupled optical power ranging from −20 dBm to −5 dBm. The electrical bandwidth of this device peaks at 10 GHz for M = 14 and decreases to 6 GHz for M = 35. The RC limited bandwidth of an unimplanted, control device is measured to be 26 GHz. The gain-bandwidth product of the high-absorption device is 230 GHz. Measurement of the excess noise factor provides an effective k -value of 0.1, consistent with the gain-bandwidth product. Primary carriers in the avalanche process are predominatly electrons, with a carrier ionization threshold of ∼ 0.6eV, indicating for the first time in this device type that both the light detection and avalanche process are mediated by the deep-levels.