Pub Date : 2025-10-17DOI: 10.1109/JPHOT.2025.3622901
Saadman Yasar;Murat Yuksel
Machine learning (ML) has emerged as a transformative tool in laser-based optics, driving advancements in both fundamental research and industrial applications. This review paper provides a comprehensive analysis of state-of-the-art ML-driven techniques across diverse laser technologies, including beam steering, optical tweezing, adaptive optics, mode-locking and pointing-acquisition-tracking to name among others. By leveraging deep learning, reinforcement learning, and other data-driven approaches, researchers have achieved real-time control, automated optimization, and enhanced system performance beyond the capabilities of traditional methods. The paper categorizes current ML applications in laser optics, addressing challenges such as dynamic aberration correction, precision beam shaping, and high-throughput data interpretation, with illustrative examples including intelligent surgical laser systems, defect-free optical trapping, and self-correcting adaptive optics. Emerging trends such as hybrid ML models, edge computing-based real-time feedback loops, and the development of open-source datasets and benchmarking tools are discussed, underscoring the potential of ML to propel laser-based optical systems to unprecedented levels of precision, efficiency, and adaptability.
{"title":"A Comprehensive Review of Machine Learning Applications in Laser Optics","authors":"Saadman Yasar;Murat Yuksel","doi":"10.1109/JPHOT.2025.3622901","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3622901","url":null,"abstract":"Machine learning (ML) has emerged as a transformative tool in laser-based optics, driving advancements in both fundamental research and industrial applications. This review paper provides a comprehensive analysis of state-of-the-art ML-driven techniques across diverse laser technologies, including beam steering, optical tweezing, adaptive optics, mode-locking and pointing-acquisition-tracking to name among others. By leveraging deep learning, reinforcement learning, and other data-driven approaches, researchers have achieved real-time control, automated optimization, and enhanced system performance beyond the capabilities of traditional methods. The paper categorizes current ML applications in laser optics, addressing challenges such as dynamic aberration correction, precision beam shaping, and high-throughput data interpretation, with illustrative examples including intelligent surgical laser systems, defect-free optical trapping, and self-correcting adaptive optics. Emerging trends such as hybrid ML models, edge computing-based real-time feedback loops, and the development of open-source datasets and benchmarking tools are discussed, underscoring the potential of ML to propel laser-based optical systems to unprecedented levels of precision, efficiency, and adaptability.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-23"},"PeriodicalIF":2.4,"publicationDate":"2025-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11206477","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405383","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
For dynamic optical path provisioning in optical networks, real-time path design based on fast quality-of-transmission (QoT) estimation is highly desirable. We propose an electric field feature propagation based QoT estimation technique using a new feature extraction method. The proposed feature extraction method can reduce the calculation load by a factor of 10000 compared with the conventional approach, enabling comprehensive datasets for various networks to be created in realistic time. This study confirms that the proposed QoT estimation method can accurately estimate the bit error rate (BER) for different combinations of input power, span length, and received optical power.
{"title":"Fast QoT Estimation Based on Probability Density Calculation of Electric Field Waveform","authors":"Ryo Igarashi;Ryo Koma;Kazutaka Hara;Jun-ichi Kani;Tomoaki Yoshida;Tatsuya Shimada","doi":"10.1109/JPHOT.2025.3622629","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3622629","url":null,"abstract":"For dynamic optical path provisioning in optical networks, real-time path design based on fast quality-of-transmission (QoT) estimation is highly desirable. We propose an electric field feature propagation based QoT estimation technique using a new feature extraction method. The proposed feature extraction method can reduce the calculation load by a factor of 10000 compared with the conventional approach, enabling comprehensive datasets for various networks to be created in realistic time. This study confirms that the proposed QoT estimation method can accurately estimate the bit error rate (BER) for different combinations of input power, span length, and received optical power.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-10"},"PeriodicalIF":2.4,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11205962","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-15DOI: 10.1109/JPHOT.2025.3621698
Jun Liu;Qianke Wang;Dawei Lyu;Jian Wang
High-dimensional quantum gates based on spatial mode of photons hold promise for advancing quantum computation. We demonstrate high-fidelity spatial mode quantum gates with fidelities up to 99.6%, using diffractive neural networks (DNNs). These gates are experimentally implemented with programmable phase layers in a compact and scalable device, enabling demonstrations like the Deutsch algorithm. Our approach showcases high reliability, adaptability, and potential for integration in quantum computation.
{"title":"Photonics Breakthroughs 2024: Scalable Quantum Computing on a Single Photon Using Spatial Modes","authors":"Jun Liu;Qianke Wang;Dawei Lyu;Jian Wang","doi":"10.1109/JPHOT.2025.3621698","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3621698","url":null,"abstract":"High-dimensional quantum gates based on spatial mode of photons hold promise for advancing quantum computation. We demonstrate high-fidelity spatial mode quantum gates with fidelities up to 99.6%, using diffractive neural networks (DNNs). These gates are experimentally implemented with programmable phase layers in a compact and scalable device, enabling demonstrations like the Deutsch algorithm. Our approach showcases high reliability, adaptability, and potential for integration in quantum computation.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-5"},"PeriodicalIF":2.4,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11204475","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145612134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-14DOI: 10.1109/JPHOT.2025.3621459
Zhiyong Yang;Yihui Zhang;Zhili Zhang;Zhiwei Zhang;Shun Li;Xiaowei Wang
The key to the researches on the polarization characteristics of targets lies in the precise establishment of the polarization model—the polarized bidirectional reflectance distribution function (pBRDF). Currently, the researches on the pBRDF model for coating targets have the problem of inaccurately describing hemispherical distribution of the degree of linear polarization (DoLP), especially the insufficiency in describing the influence of the azimuth angle on DoLP. In this thesis, firstly, with the aim of minimizing the error in linear polarization degree, through simulation comparative experiments on common microfacet distribution function (NDF), geometric attenuation factor (GAF) and multiple reflection function, it is determined that the Gaussian NDF, the modified integral GAF and the Minnaert model considering roughness are more suitable for coating targets. Secondly, the combined model for coating targets was established by combining the three functions, and it is found that the combined model was insensitive to the azimuth angle. Finally, a pBRDF model incorporating a high-order polynomial of relative azimuth angle is proposed, which improved the problem of excessive error caused by the insensitivity to the azimuth angle. The results of experiments show that the relative errors when adopting the modified model have decreased by 45.8%, 66.7%, 10.5%, and 32.1% respectively. The determination coefficient has reached 0.948, 0.953, 0.917 and 0.930, and the performance indicators are superior to those of the existing models. The research results provide a reference for describing the hemispherical spatial distribution of DoLP for coating targets.
{"title":"A Modified pBRDF Model Considering the Influence of Relative Azimuth Angle for Coating Targets","authors":"Zhiyong Yang;Yihui Zhang;Zhili Zhang;Zhiwei Zhang;Shun Li;Xiaowei Wang","doi":"10.1109/JPHOT.2025.3621459","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3621459","url":null,"abstract":"The key to the researches on the polarization characteristics of targets lies in the precise establishment of the polarization model—the polarized bidirectional reflectance distribution function (pBRDF). Currently, the researches on the pBRDF model for coating targets have the problem of inaccurately describing hemispherical distribution of the degree of linear polarization (DoLP), especially the insufficiency in describing the influence of the azimuth angle on DoLP. In this thesis, firstly, with the aim of minimizing the error in linear polarization degree, through simulation comparative experiments on common microfacet distribution function (NDF), geometric attenuation factor (GAF) and multiple reflection function, it is determined that the Gaussian NDF, the modified integral GAF and the Minnaert model considering roughness are more suitable for coating targets. Secondly, the combined model for coating targets was established by combining the three functions, and it is found that the combined model was insensitive to the azimuth angle. Finally, a pBRDF model incorporating a high-order polynomial of relative azimuth angle is proposed, which improved the problem of excessive error caused by the insensitivity to the azimuth angle. The results of experiments show that the relative errors when adopting the modified model have decreased by 45.8%, 66.7%, 10.5%, and 32.1% respectively. The determination coefficient has reached 0.948, 0.953, 0.917 and 0.930, and the performance indicators are superior to those of the existing models. The research results provide a reference for describing the hemispherical spatial distribution of DoLP for coating targets.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-14"},"PeriodicalIF":2.4,"publicationDate":"2025-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11202864","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405339","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Electro-optic (EO) modulators with ultra-high speed and low driving voltage are essential for high-performance optical computing, optical communication, and broadband microwave photonic links. In this work, we have proposed an electrode optimization method for periodic structures, based on the RGLC transmission line model and exemplified by sawtooth traveling-wave electrodes (STWEs). We have established analytical expressions for characteristic parameters including the resistance, inductance, capacitance, as well as the slot depth and width of STWEs for the first time, achieving synchronous optimization of microwave loss minimization and velocity matching. Validated through simulations on thin-film lithium niobate modulators, the approach achieves an ultra-high EO bandwidth of up to 290 GHz and a low voltage-length product of 2.5 V·cm. The optimized STWEs feature a minimum linewidth of 5 μm and a metal thickness as thin as 500 nm, which are compatible with deep ultraviolet lithography and suitable for large-scale, low-cost fabrication. This work would pave the way for developing scalable ultra-high-speed EO modulator arrays.
{"title":"Optimizing Sawtooth Electrodes for Ultra-High-Speed Electro-Optic Modulators","authors":"Shurui Huang;Deyang Kong;Huijie Sun;Yongzhuo Li;Xue Feng;Wei Zhang;Yidong Huang","doi":"10.1109/JPHOT.2025.3620360","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3620360","url":null,"abstract":"Electro-optic (EO) modulators with ultra-high speed and low driving voltage are essential for high-performance optical computing, optical communication, and broadband microwave photonic links. In this work, we have proposed an electrode optimization method for periodic structures, based on the RGLC transmission line model and exemplified by sawtooth traveling-wave electrodes (STWEs). We have established analytical expressions for characteristic parameters including the resistance, inductance, capacitance, as well as the slot depth and width of STWEs for the first time, achieving synchronous optimization of microwave loss minimization and velocity matching. Validated through simulations on thin-film lithium niobate modulators, the approach achieves an ultra-high EO bandwidth of up to 290 GHz and a low voltage-length product of 2.5 V·cm. The optimized STWEs feature a minimum linewidth of 5 μm and a metal thickness as thin as 500 nm, which are compatible with deep ultraviolet lithography and suitable for large-scale, low-cost fabrication. This work would pave the way for developing scalable ultra-high-speed EO modulator arrays.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-12"},"PeriodicalIF":2.4,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11200483","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145351975","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-08DOI: 10.1109/JPHOT.2025.3619399
Benedictus Yohanes Bagus Widhianto;Hao-Chun Tsui;Jyehong Chen
This paper presents a transition-based decision feedback equalizer (TB-DFE) designed to mitigate bandwidth limitations and transition-dependent distortions in four-level pulse amplitude modulation (PAM-4) optical interconnects using directly modulated vertical-cavity surface-emitting lasers (VCSELs). In contrast to conventional decision feedback equalizers (DFEs), which apply feedback taps regardless of the transition context, the proposed TB-DFE adapts its feedback coefficient based on detected symbol transitions, thereby capturing the non-stationary impulse response variations observed in skew-prone links. Experimental validation was performed using a 26.5625 GBd PAM-4 VCSEL-based transmission system operating over an OM4 multimode fiber at distances up to 500 meters. Results show that, compared to conventional feedforward equalizer (FFE) and FFE-DFE schemes, FFE with TB-DFE improves bit error rate (BER) by nearly an order of magnitude at longer distances. Although its benefit is modest at 100 m and 300 m, a significant gain is observed under severe bandwidth constraints at 500 m. This work demonstrates a 53 Gb/s transmission per lane over 500 m of OM4 multimode fiber using a commercial multimode VCSEL, achieving a pre-forward error correction (FEC) BER below the Reed-Solomon FEC (RS-FEC) threshold (2.4 × 10−4), highlighting the potential of the proposed method to extend transmission reach under skew and bandwidth-limited systems. These findings establish TB-DFE as a practical and low-complexity equalization technique to boost the direct-modulation-based transmitter transmission distance for next-generation short-reach and direct-reach data center applications.
{"title":"Transition-Based Decision Feedback Equalization to Enhance Transmission in VCSEL-Based Short-Reach IM/DD System","authors":"Benedictus Yohanes Bagus Widhianto;Hao-Chun Tsui;Jyehong Chen","doi":"10.1109/JPHOT.2025.3619399","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3619399","url":null,"abstract":"This paper presents a transition-based decision feedback equalizer (TB-DFE) designed to mitigate bandwidth limitations and transition-dependent distortions in four-level pulse amplitude modulation (PAM-4) optical interconnects using directly modulated vertical-cavity surface-emitting lasers (VCSELs). In contrast to conventional decision feedback equalizers (DFEs), which apply feedback taps regardless of the transition context, the proposed TB-DFE adapts its feedback coefficient based on detected symbol transitions, thereby capturing the non-stationary impulse response variations observed in skew-prone links. Experimental validation was performed using a 26.5625 GBd PAM-4 VCSEL-based transmission system operating over an OM4 multimode fiber at distances up to 500 meters. Results show that, compared to conventional feedforward equalizer (FFE) and FFE-DFE schemes, FFE with TB-DFE improves bit error rate (BER) by nearly an order of magnitude at longer distances. Although its benefit is modest at 100 m and 300 m, a significant gain is observed under severe bandwidth constraints at 500 m. This work demonstrates a 53 Gb/s transmission per lane over 500 m of OM4 multimode fiber using a commercial multimode VCSEL, achieving a pre-forward error correction (FEC) BER below the Reed-Solomon FEC (RS-FEC) threshold (2.4 × 10<sup>−4</sup>), highlighting the potential of the proposed method to extend transmission reach under skew and bandwidth-limited systems. These findings establish TB-DFE as a practical and low-complexity equalization technique to boost the direct-modulation-based transmitter transmission distance for next-generation short-reach and direct-reach data center applications.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-10"},"PeriodicalIF":2.4,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11196957","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145351994","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-07DOI: 10.1109/JPHOT.2025.3618943
Ning Chen;Junlong Ma;Qi Zhang;Yaowen Yao;Yi Huang;Yuncai Lu;Jian Shao;Qun Li;Xiaobei Zhang;Tingyun Wang
In this paper, a novel and cost-effective all-glass microphone with enhanced sensitivity and robustness is introduced, enabled by dual diaphragm structure including pressure amplification diaphragm and acoustic sensing diaphragm. Pressure amplification diaphragm is applied to enhance external acoustic wave, with its focal plane located exactly on the plane of sensing diaphragm, thereby introducing enhanced deflection. Then acoustic introduced deflection is further interrogated interferometrically through the light transmitted by optical fiber. Multi-laser processing method is developed for all-glass microphone fabrication, where picosecond laser is applied for precise diaphragm micromachining and welding, and CO2 laser is employed for fixing of optical fiber and fiber housing structure. Experimental results demonstrate that the proposed microphone with dual diaphragm achieves a 2.24-fold enhancement in sensitivity at resonant frequency, that is, 206.314 mV/Pa, with signal-to-noise ratio increased from 66.31 dB to 75.68 dB. With advantages of compact size, high sensitivity and enhanced robustness, the proposed all-glass microphone shows strong potential for harsh environment applications.
{"title":"Dual Diaphragm Structure-Based All-Glass Microphone With Enhanced Robustness and Sensitivity Using Laser Processing","authors":"Ning Chen;Junlong Ma;Qi Zhang;Yaowen Yao;Yi Huang;Yuncai Lu;Jian Shao;Qun Li;Xiaobei Zhang;Tingyun Wang","doi":"10.1109/JPHOT.2025.3618943","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3618943","url":null,"abstract":"In this paper, a novel and cost-effective all-glass microphone with enhanced sensitivity and robustness is introduced, enabled by dual diaphragm structure including pressure amplification diaphragm and acoustic sensing diaphragm. Pressure amplification diaphragm is applied to enhance external acoustic wave, with its focal plane located exactly on the plane of sensing diaphragm, thereby introducing enhanced deflection. Then acoustic introduced deflection is further interrogated interferometrically through the light transmitted by optical fiber. Multi-laser processing method is developed for all-glass microphone fabrication, where picosecond laser is applied for precise diaphragm micromachining and welding, and CO<sub>2</sub> laser is employed for fixing of optical fiber and fiber housing structure. Experimental results demonstrate that the proposed microphone with dual diaphragm achieves a 2.24-fold enhancement in sensitivity at resonant frequency, that is, 206.314 mV/Pa, with signal-to-noise ratio increased from 66.31 dB to 75.68 dB. With advantages of compact size, high sensitivity and enhanced robustness, the proposed all-glass microphone shows strong potential for harsh environment applications.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-7"},"PeriodicalIF":2.4,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11195131","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405386","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-07DOI: 10.1109/JPHOT.2025.3618300
Abhijeet Upadhya;Anu Goel;Vivek K. Dwivedi;Ghanshyam Singh
The research work invokes the long short-term memory (LSTM) deep learning model for combating the issue of outdated channel state information (CSI) during channel estimation on the wireless medium. For demonstration of the concept, the downlink free space optical (FSO)/radio frequency (RF) relaying strategy with outdated CSI has been contemplated. In the considered amplify-and-forward (AF) cooperative relaying, the channel gain has been extracted based on the available outdated CSI at the relay node. Of course, the performance of the FSO/RF downlink system is inferior due to low correlation between the previous (original) channel state and the measured CSI during the next time interval. Since the LSTM can use larger input data sequentially to predict the next target probabilities based on correlation among input variables, they become most suited for CSI estimation from the available outdated CSI. The trained LSTM model becomes accomplished to estimate the previous state, thus serving the relay node to adjust the gain more accurately. The trained LSTM model in the present research work is highly accurate with mean square error (MSE) and root mean square error (RMSE) of $MSE=-43.49$ dB and $RMSE=-13.42$ dB, respectively. The performance of the downlink FSO/RF relay has been presented in terms of outage probability, ergodic capacity and bit error rate (BER). It has been shown in the paper that using the trained deep learning LSTM model, the performance of the relaying system can be made equivalent to that when timing delay exists between the original and the estimated sample values.
{"title":"Outdated Channel Overhead Minimization Using LSTM Based Deep Learning Approach in FSO/RF Relaying Systems","authors":"Abhijeet Upadhya;Anu Goel;Vivek K. Dwivedi;Ghanshyam Singh","doi":"10.1109/JPHOT.2025.3618300","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3618300","url":null,"abstract":"The research work invokes the long short-term memory (LSTM) deep learning model for combating the issue of outdated channel state information (CSI) during channel estimation on the wireless medium. For demonstration of the concept, the downlink free space optical (FSO)/radio frequency (RF) relaying strategy with outdated CSI has been contemplated. In the considered amplify-and-forward (AF) cooperative relaying, the channel gain has been extracted based on the available outdated CSI at the relay node. Of course, the performance of the FSO/RF downlink system is inferior due to low correlation between the previous (original) channel state and the measured CSI during the next time interval. Since the LSTM can use larger input data sequentially to predict the next target probabilities based on correlation among input variables, they become most suited for CSI estimation from the available outdated CSI. The trained LSTM model becomes accomplished to estimate the previous state, thus serving the relay node to adjust the gain more accurately. The trained LSTM model in the present research work is highly accurate with mean square error (MSE) and root mean square error (RMSE) of <inline-formula><tex-math>$MSE=-43.49$</tex-math></inline-formula> dB and <inline-formula><tex-math>$RMSE=-13.42$</tex-math></inline-formula> dB, respectively. The performance of the downlink FSO/RF relay has been presented in terms of outage probability, ergodic capacity and bit error rate (BER). It has been shown in the paper that using the trained deep learning LSTM model, the performance of the relaying system can be made equivalent to that when timing delay exists between the original and the estimated sample values.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-11"},"PeriodicalIF":2.4,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11195091","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145674795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-02DOI: 10.1109/JPHOT.2025.3617297
Jacob N. Bouchard;Marcel W. Pruessner;Nathan F. Tyndall;Steven T. Lipkowitz;Scott A. Holmstrom;Kyle J. Walsh;Michael L. Fanto;Gerald L. Leake;Todd H. Stievater
The thermo-optic effect allows for the modulation of phase with very low induced optical losses in otherwise passive materials used for integrated optical phased arrays. A key drawback of this effect is the required electrical power, typically requiring tens or even hundreds of milliwatts to induce a sufficient phase shift in traditional waveguide materials such as silicon. In this work we describe a foundry-implemented method to increase the efficiency of silicon waveguide-integrated thermo-optic phase shifters by more than an order of magnitude compared to traditional devices. By reducing paths of conductive heat loss through the use of a thermal isolation trench, we experimentally demonstrate efficiency increases of 19x in Si waveguide thermo-optic phase shifting efficiency. We extend this application to OPAs and observe a 15x increase in beam steering efficiency in an 8 channel integrated OPA as well as an effective field of view extension in a wide-pitch 32 channel OPA. We also measure the modulation bandwidth and find a correlation between steering efficiency gains and device speed. Detailed finite-element simulations show excellent agreement with measurements. These simulations provide unique insights into the thermal behavior of these devices and inform tradeoffs between phase shift efficiency and temporal response in future device design.
{"title":"Enhanced Steering Efficiency Through Substrate Removal in 1D Optical Phased Arrays","authors":"Jacob N. Bouchard;Marcel W. Pruessner;Nathan F. Tyndall;Steven T. Lipkowitz;Scott A. Holmstrom;Kyle J. Walsh;Michael L. Fanto;Gerald L. Leake;Todd H. Stievater","doi":"10.1109/JPHOT.2025.3617297","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3617297","url":null,"abstract":"The thermo-optic effect allows for the modulation of phase with very low induced optical losses in otherwise passive materials used for integrated optical phased arrays. A key drawback of this effect is the required electrical power, typically requiring tens or even hundreds of milliwatts to induce a sufficient phase shift in traditional waveguide materials such as silicon. In this work we describe a foundry-implemented method to increase the efficiency of silicon waveguide-integrated thermo-optic phase shifters by more than an order of magnitude compared to traditional devices. By reducing paths of conductive heat loss through the use of a thermal isolation trench, we experimentally demonstrate efficiency increases of 19x in Si waveguide thermo-optic phase shifting efficiency. We extend this application to OPAs and observe a 15x increase in beam steering efficiency in an 8 channel integrated OPA as well as an effective field of view extension in a wide-pitch 32 channel OPA. We also measure the modulation bandwidth and find a correlation between steering efficiency gains and device speed. Detailed finite-element simulations show excellent agreement with measurements. These simulations provide unique insights into the thermal behavior of these devices and inform tradeoffs between phase shift efficiency and temporal response in future device design.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-9"},"PeriodicalIF":2.4,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11189985","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405381","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-02DOI: 10.1109/JPHOT.2025.3617334
Kawsar Ahmed;Md. Mamun Ali;Ruhul Amin;Francis M. Bui;Li Chen;Santosh Kumar
Early, accurate, and cost-effective detection of multiple diseases remains a critical challenge in modern biomedical diagnostics. This study offers a novel computational framework for the simultaneous detection of cancerous and tuberculosis cells using a gold-coated photonic crystal fiber (PCF)-based surface plasmon resonance (SPR) biosensor integrated with deep learning. The proposed dual-channel biosensor structure, optimized through the Finite Element Method (FEM), is designed to detect multi-analyte samples by analyzing confinement loss across different refractive indices (RI). To enhance prediction accuracy and support rapid parameter optimization, a generative adversarial network (GAN) model was designed to estimate confinement loss based on key sensor design features. The GAN model achieved superior performance with a mean squared error (MSE) of 0.0175, a mean absolute error (MAE) of 0.1250, and an R$^{2}$ of 0.9087 compared to traditional machine learning models such as decision trees and random forests. Explainability was incorporated through SHAP (SHapley Additive exPlanations) analysis, which identified critical design parameters that influence model output, thus enhancing transparency and trustworthiness. Extensive ablation studies with various cancer and tuberculosis cells validated the reliability of the proposed model. The predicted confinement loss curves closely aligned with simulated results, confirming the robustness of the proposed model in real-time dual-analyte detection. This study offers a promising data-driven strategy for designing multi-analyte biosensors, paving the way for next-generation noninvasive diagnostic tools.
{"title":"Deep Learning-Based Simultaneous Cancerous and Tuberculosis Cells Detection Biosensor: A Computational Approach","authors":"Kawsar Ahmed;Md. Mamun Ali;Ruhul Amin;Francis M. Bui;Li Chen;Santosh Kumar","doi":"10.1109/JPHOT.2025.3617334","DOIUrl":"https://doi.org/10.1109/JPHOT.2025.3617334","url":null,"abstract":"Early, accurate, and cost-effective detection of multiple diseases remains a critical challenge in modern biomedical diagnostics. This study offers a novel computational framework for the simultaneous detection of cancerous and tuberculosis cells using a gold-coated photonic crystal fiber (PCF)-based surface plasmon resonance (SPR) biosensor integrated with deep learning. The proposed dual-channel biosensor structure, optimized through the Finite Element Method (FEM), is designed to detect multi-analyte samples by analyzing confinement loss across different refractive indices (RI). To enhance prediction accuracy and support rapid parameter optimization, a generative adversarial network (GAN) model was designed to estimate confinement loss based on key sensor design features. The GAN model achieved superior performance with a mean squared error (MSE) of 0.0175, a mean absolute error (MAE) of 0.1250, and an R<inline-formula><tex-math>$^{2}$</tex-math></inline-formula> of 0.9087 compared to traditional machine learning models such as decision trees and random forests. Explainability was incorporated through SHAP (SHapley Additive exPlanations) analysis, which identified critical design parameters that influence model output, thus enhancing transparency and trustworthiness. Extensive ablation studies with various cancer and tuberculosis cells validated the reliability of the proposed model. The predicted confinement loss curves closely aligned with simulated results, confirming the robustness of the proposed model in real-time dual-analyte detection. This study offers a promising data-driven strategy for designing multi-analyte biosensors, paving the way for next-generation noninvasive diagnostic tools.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"17 6","pages":"1-9"},"PeriodicalIF":2.4,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11189987","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145351998","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：