BNBikash Nakarmi

Bikash Nakarmi

*This is an algorithmically generated author record based on publicly available information.

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


Bikash Nakarmi has a total of 52 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Linewidth Enhancement of Range Profile in Multi-Chirp Radar via Beat Signal Chirp and Phase Normalization

    by Aigerim Ashimbayeva, Sun Zhijie, Mohd Rashidi Che Beson, T. D. Subash, Ikechi Augustine Ukaegbu, Shilong Pan, Carlo Molardi, Bikash Nakarmi

    In this paper, we present a new approach for chirp and phase normalization for enhancing linewidth and beat-signal coherence in a multi-chirp microwave photonic radar system. The proposed Beat Signal Chirp and Phase Normalization (BSCPN) technique introduces adaptive corrections to inter-chirp phase and chirp-rate drift of the received signal, effectively aligning multiple successive chirps and eliminating phase discontinuities in the post-processing. This normalization enables coherent integration across chirps, yielding improved range profiles and detection fidelity. Beat Signal Chirp Normalization (BSCN) and full-BSCPN methods were implemented for detecting multi-target configurations for two-, three-, and four-chirp photonic radar systems with varying object spacings of 3.5 cm, 3.0 cm, and 2.7 cm. Through post-processing compensation of chirp-to-chirp phase discontinuities and chirp normalization, the results show an improvement in linewidth reduction of up to 70-75%, an improvement in range accuracy to sub-centimeter levels, and a peak sidelobe level (PSL) reduction of approximately 6-12 dB, depending on the multi-chirp configuration. Additionally, the four-chirp radar operation (73-74 GHz, 74-75.5 GHz, 75.5-78 GHz, 78-80 GHz; each with a duration of 0.375 μs) achieves up to 0.22 cm range resolution and a 6-8 dB PSL reduction in comparison to the uncompensated case, confirming the effectiveness of BSCPN for higherorder multi-chirp operation. These results demonstrate that receiver-side post-processing normalization through BSCPN restores inter-chirp beat coherence, reflected in aligned beat-frequency components, reduced linewidth, and improved range resolution in multi-chirp photonic radar systems. Furthermore, additional analyses of computational complexity and processing latency, along with low-SNR robustness, have been incorporated to enhance the practical applicability of the proposed framework.


  • Target Detection Using Photonics-Driven Robust Waveforms in Interference-Prone Radar Networks

    by Bikash Nakarmi, S. M. Rezwanul Islam, Sun Zhijie, Ikechi Augustine Ukaegbu, Aigerim Ashimbayeva, Carlo Molardi, T. D. Subash, Mohd Rashidi Che Beson, Shilong Pan

    This article explores the impact of interference and intrusion on the performance of target object detection and range measurement in an interference-prone radio detection and ranging (RADAR) environment. As RADAR applications expand, managing interference becomes crucial to prevent false detections and resolution deterioration, ensuring safe and secure operations in multiradar environments. Conventional linear frequency-modulated (LFM)-based radar under interference experiences significant degradation in radar detection performance, including false detections and increased noise levels. To address these challenges, we used photonics-driven robust waveforms (PDRWs), specifically frequency-hopping LFM (Ph-FHLFM)- and multichirp LFM (Ph-MCLFM)-based, achieved by manipulating optical beam intensity in a semiconductor laser. In this article, we used both types of PDRWs (Ph-FHLFM and Ph-MCLFM) to demonstrate their interference resilience capability in target detection through theoretical analysis, MATLAB simulation, and the laboratory experiment under interference conditions, irrespective of the type of interference. The analysis shows the interference resilience capability of both types of PDRWs regardless of the interference types. The results illustrate that both waveforms effectively mitigate coherent and non-coherent interferences, thus maintaining accurate target detection with a maximum range error of 3 mm and maximum linewidth of 2.5 cm, while achieving minimum peak-to-sidelobe level (PSL) improvement of over 10 dB compared to conventional Ph-LFM radar. Additionally, we show the advantages of photonics-based radar with PDRWs over other technologies and hence underscore the potential of PDRWs in enhancing the resilience of RADAR system against interference in complex operational environments.

  • Open Access

    Efficient Back-Projection Technique for Multiple Objects Detection and 2D Imaging Through Photonics-Based LFM Radar

    by T. D. Subash, Sun Zhijie, Mohd Rashidi Che Beson, S. M. Rezwanul Islam, Ukash Nakarmi, Ikechi Augustine Ukaegbu, Bikash Nakarmi

    Detection and imaging play a vital role in modern radar technology and have seen significant interest across various fields, including autonomous vehicles, defense, environmental monitoring, and weather forecasting. This work presents a method for the detection and imaging of multiple objects positioned at uniform and non-uniform distances within a defined region of interest. Linear Frequency Modulated (LFM) waveform, which features excellent pulse compression capability is employed in this work. We use an injection of external beam into a Distributed Feedback (DFB) laser to generate LFM waveform and further detect and estimate the range of multiple targets along the longitudinal axis. Additionally, by taking multiple range measurements along the lateral axis, we extend this approach to estimate lateral distances, thereby achieving a comprehensive 2D imaging solution. Using an efficient back-projection technique, we reconstruct a two-dimensional image of the detected objects by leveraging range measurements from a single dimension. Specifically, we leverage filtering and intensity projection to improve classical back-projection. For a proof-of-concept validation, an 8 GHz bandwidth LFM signal centered at 12 GHz is employed to detect objects within the designated space which is further implemented for imaging using proposed efficient back-projection technique. Our proposed imaging method solves the problem of signal bleeding and collapsing and enhances image resolution by utilizing received signal data, achieving an error margin within 3 mm along the longitudinal axis and 9 mm along the lateral axis relative to the actual value with single-dimension measurements. Furthermore, the proposed technique obtains the maximum error reduction of 100% in X-direction and 95% in the Y-direction compared to the conventional method.

  • Open Access

    Peduncle Detection of Ripe Strawberry to Localize Picking Point Using DF-Mask R-CNN and Monocular Depth

    by Niraj Tamrakar, Bhola Paudel, Sijan Karki, Nibas Chandra Deb, Elanchezhian Arulmozhi, Jung Hoo Kook, Myeong Yong Kang, Dae Yeong Kang, Oluwasegun Moses Ogundele, Bikash Nakarmi, et al.

    Accurate localization of picking points and depth estimation is critical for implementing a robotic strawberry harvesting system. Due to the delicate nature of strawberries, harvesting must be performed without bruising or damage, typically by grasping and cutting the peduncle of the ripe strawberry. However, accurately detecting and localizing the thin peduncle in a cluttered environment is a significant challenge. This study proposed depth fused Mask R-CNN (DF-Mask R-CNN), which integrates depth information of the scene with the RGB image to enhance the detection, localization, and segmentation of strawberries and their peduncles in a greenhouse environment. To generate a dense depth map, a cutting-edge monocular depth estimator, ZoeDepth was used. The proposed DF-Mask R-CNN with ResNet101-FPN exhibited superior instance segmentation performance, with an overall mAP of 81.9%, with mAPsmall at 33.3%, mAPmedium at 78.79%, mAPlarge at 88.8 and APIOU=0.5 at 98.1%. In tests with 300 ripe strawberry samples, the method demonstrated a robust picking point detection, with a mean absolute error and root mean square error of 1.98 cm and 2.12 cm, respectively. These results highlight the effectiveness of the DF-Mask R-CNN model combined with the ZoeDepth estimator in enhancing the detection, localization, and segmentation of strawberries and their peduncles. This approach enables precise picking point localization and depth estimation for efficient vision systems for robotic strawberry harvesting.


  • Open Access

    Target Detection Using Fused Unidentical Photonics-Based LFM Sub-Band Radar Signals With an Adaptive Feed Forward Network Equalizer

    by Bikash Nakarmi, S. M. Rezwanul Islam, Hum Nath Parajuli, Ikechi Augustine Ukaegbu, Aigerim Ashimbayeva, Carlo Molardi, T. D. Subash, Xiangchuan Wang, Shilong Pan

    This paper proposes and experimentally demonstrate sub-bands fusion of photonically generated linear frequency modulated (Ph-LFM) radar signals with unidentical bandwidth and chirp rates using an adaptive, delay-less feed-forward network equalizer (FFNE). We demonstrate this using optical injection in a semiconductor laser to generate Ph-LFM signals at different IEEE X-KA radar sub-bands: 19.25–23.94 GHz and 24.06–28.31 GHz (bandgap 0.12 GHz), 19.69–23.06 GHz and 23.625–27 GHz (bandgap 0.56 GHz), and 8–11.5 GHz and 12.75–17 GHz (bandgap 1.25 GHz). Time-frequency analysis (TFA) was used to obtain sub-bandgap signals of 0.12 GHz, 0.56 GHz, and 1.25 GHz which are coherently fused using the FFNE with a particle swarm optimization (PSO) algorithm to optimize complex-valued weights. The method is evaluated by measuring the range resolution and peak-to-sidelobe level (PSL) in detecting two objects separated by 2 cm and 3 cm. The FFNE achieves significant improvement over 10 dB in PSL and resolves previously unresolvable distances, with maximum range resolutions of 1.8 cm, 2.2 cm, and 2 cm, closely matching theoretical values for full-band LFM signals with bandwidth of 9.06 GHz, 7.31 GHz, and 9 GHz, respectively. Experimental results demonstrate the FFNE’s superior performance in enhancing range resolution and PSL in multi-sub-band radar systems compared to scenarios without FFNE.


  • Open Access

    Realization of Arithmetic Logic Units Using Electro-Optic Microring Resonators in Photonic Circuits

    by Assylkhan Nurgali, Bikash Nakarmi, Carlo Molardi, Ikechi Augustine Ukaegbu

    This work presents a new approach to designing Arithmetic Logic Units (ALUs) using electro-optic microring resonators. The methodology covers the demonstration of full and half addition and subtraction, parity checking, and dynamic logic gates functionalities at the data speed of 10 Gbps. In this work, integrated electro-optic circuits have been meticulously designed to have minimum footprints using microring resonators (MRRs) with low losses, thus, ensuring efficient and scalable optical computing system. A unique design for electro-optic full adder/subtractor and three input odd-even parity checkers were implemented and validated successfully within the Lumerical simulation environment. Simulation results demonstrate the successful alignment of optical outputs with their respective truth tables, confirming the reliability and accuracy of the designed circuits. These advancements represent significant progress in the field of optical computing, offering notable improvements in computational speed, efficiency, and scalability. This research contributes to the ongoing exploration and development in optical computing, with the potential to revolutionize computational efficiency and performance in various applications.


  • Multi-Radar Interference Mitigation in Photonics-Based Radar With Sliding Window LSTM Recurrent Neural Network

    by Hum Nath Parajuli, Aigerim Ashimbayeva, U. Nakarmi, Ikechi Augustine Ukaegbu, B. Gaudel, Shilong Pan, Carlo Molardi, B. Nakarmi

    In this paper, we propose and experimentally demonstrate a technique of mitigating multi-radar interference in a photonic radar using a sliding window long short-term memory (SW-LSTM)-based recurrent neural network. For the proof-of-concept experimental demonstration, a photonics-based victim radar (VR) with an 8 GHz linear frequency modulated (LFM) signal is generated using optical injection in a semiconductor laser. This VR is used to detect two objects separated by 10 cm in the presence of multiple interference sources. To perform the analysis of mitigating interference in the multi-radar environment, we generated random interference LFM signals (IR-LFM) with diverse bandwidths (4 GHz to 10 GHz) and chirp rates (±1 GHz/μs to ±20 GHz/μs), both experimentally and synthetically. The generated IR-LFM signals consist of both coherent and non-coherent interferences and are added to the echo signal and trained using SW-LSTM. During the training stage, learning rate, stacked layers, nodes, batch size, and window size are optimized. In the test stage, we use signals with both coherent and non-coherent interferences, which are distinct from the interferences in the training stage. The proposed model is evaluated for one, four, and eight interference test scenarios, in which it successively achieves a mean R 2 value of > 93% for all test scenarios, taken over ten trials in each test scenario. Furthermore, a mean error vector magnitude( EVM )of < 0.07 and a mean signal-to-interference plus noise ratio ( SINR ) of > 11 dB are obtained. These results indicate a strong similarity between the range profile calculated in the interference-free and the multi-interference environments with our proposed scheme of mitigating interference using SW-LSTM.


  • Interference Mitigation in Multi-radar Environment Using LSTM-Based Recurrent Neural Network

    by Hum Nath Parajuli, Galymzhan Bakhtiyarov, Bikash Nakarmi, Ikechi Augustine Ukaegbu


  • Open Access

    4 × 4 bit Programmable Optical Memory Array With Digital Addressing Using Micro-Ring Resonators

    by Assylkhan Nurgali, Bikash Nakarmi, Carlo Molardi, Ikechi Augustine Ukaegbu

    The development of photonic integrated circuit (PIC) based computing systems requires efficient optical memory units, which can be challenging to implement due to the need for precise routing of optical signals. To address this challenge, we propose a $4\times 4$ bit photonic non-volatile memory array system using micro ring resonators for fast and flexible memory addressing. The system uses phase change material (PCM) as a memory unit and implements wavelength division multiplexing (WDM) techniques to access memory cells through active and passive filtering by micro-ring resonators. Results show a system access speed of 30 Gbps with 50Mbps for erasing and 1.6Mbps for writing with accurate memory unit routing. This demonstration shows the potential for developing photonic non-volatile memory storage with reconfigurable memory addressing, which could facilitate the development of advanced PIC-based computing systems.


  • Multi-Chirp LFM Waveforms Generation With Reconfigurable Chirp Rates Using Optical Injection in a Semiconductor Laser

    by B. Nakarmi, Bai Yan Song, Ikechi Augustine Ukaegbu, Hum Nath Parajuli, Aigerim Ashimbayeva, U. Nakarmi, Xiangchun Wang, Shilong Pan

    We propose and experimentally demonstrate a photonics-based technique for generating simultaneous up-down multiple chirp-rate linear frequency modulated (SUDMC-LFM) waveforms using a dual beam injection technique in a distributed feedback (DFB) laser. To generate waveforms with multiple chirps, drive waveforms of different amplitude-time slopes for each sub-interval time are designed in an arbitrary waveform generator (AWG) and fed into the intensity modulator, which in turn, controls the intensity of the injected beam. Consequently, the rate of redshift of the emission wavelength is controlled. In the experiment, the intensity of only one beam is controlled with the designed AWG signal. For the proof-of-concept demonstration, we generate continuous multiple chirps (CMC), simultaneous up-down multiple chirps (SUDMC), and simultaneous up-down frequency hopped multiple chirps (SUDFHMC) LFM waveforms. Furthermore, re-configurability of the generated waveforms in terms of center frequency, bandwidth, number of chirps, and chirp-rates are obtained by either changing the wavelength of the injected beams or by changing the parameters of the AWG signal. In the experiment, the SUDFHMC waveform with four frequency-hopping with sub-interval chirps of 4 GHz/μs, 1.2 GHz/μs, 2.4 GHz/μs, and 2.8 GHz/μs for both the up-chirps and down-chirps, respectively, is generated. The auto-ambiguity analysis at −3 dB full-width half maximum (FWHM) shows the time-bandwidth product (TBWP) of >4545, and unambiguous Doppler of <0.75 MHz for CMC, SUDMC, and SUDFHMC signals with three chirps. These results indicate that the proposed waveforms have the potential to provide high performance in multi-radar and multi-target applications.