Silicon-based all-optical wavelength converter for on-chip optical interconnection

Shuhong Liu

doi:10.1515/nleng-2024-0059

Artikel Open Access

Silicon-based all-optical wavelength converter for on-chip optical interconnection

Shuhong Liu

Veröffentlicht/Copyright: 10. Februar 2025

Veröffentlicht von

Veröffentlichen auch Sie bei De Gruyter Brill

Manuskript einreichen Informationen für Autor*innen

Aus der Zeitschrift Nonlinear Engineering Band 14 Heft 1

Abstract

The study of silicon-based all-optical wavelength converters (SAOWCs) for on-chip optical interconnect applications holds potential research value and application horizons. The study proposes a SAOWC for on-chip optical interconnect technology. It also designs and optimizes an all-optical wavelength converter based on a quantum dot semiconductor optical amplifier and a Mach-Zönder interferometer in conjunction with photonic integrated circuit technology. The interlayer coupling efficiency is simulated by the finite-difference time-domain method, and the performance of the system is verified in conjunction with experiments. The experimental results showed that the coupling efficiency was as high as 98.44% at 1,550 nm wavelength. The bit error rate (BER) test results showed that the BER rose from 10⁻¹⁷ to 10⁻⁴ when the pulse width increased from 0.1 to 1.0 ps. Meanwhile, the temperature fluctuation and signal strength degradation tests further validated the stability of the device under different operating conditions. These results demonstrate that the converter has significant potential to enhance data transmission capability. The study provides a valuable reference for further improving and optimizing data transmission over all-optical networks, which is expected to promote scientific advancement and technological applications in this field.

Keywords: silicon-based all-optical wavelength converter; optical interconnection technology; data communication; nonlinear optics; integrated circuit design

1 Introduction

The development of the information society has driven the demand for data transmission rate and bandwidth, especially in large data centers and mobile internet. With the increasing demand for data communication, traditional electronic interconnection technology has faced challenges, making optical interconnection technology (OIT) a cutting-edge research field [1,2]. Among them, silicon-based all-optical wavelength converters (SAOWCs) for on-chip optical interconnection (OCOI) play an important role in solving data transmission bottlenecks. An all-optical wavelength converter (AOWC) can achieve wavelength conversion without electronic devices, allowing multiple signals of different wavelengths to be transmitted in the same fiber, greatly improving data transmission capacity [3,4]. SAOWC has become a research focus due to its excellent integration and compatibility. However, the research on SAOWC faces many challenges, involving issues such as nonlinear optics, material science, and integrated circuit design that require in-depth exploration. The innovation is to propose a new design scheme for SAOWC, aimed at improving the conversion efficiency and bandwidth of the device. The device incorporates novel silicon-based materials and process technology to enhance its integration and performance stability. Additionally, it investigates innovative optical wavelength modulation techniques to achieve more precise wavelength conversion. This study promotes the technological progress of SAOWC and the development of OIT, thereby improving the performance of data communication [5,6]. The study is conducted in four sections. Section 1 discusses numerous research works on SAOWC methods for OCOI. Section 2 is the model construction of OCOI’s SAOWC research. Section 3 is the experimental verification. Section 4 summarizes the content and points out the shortcomings.

2 Related works

With the growth of multi-core processors, efficient on-chip communication plays an essential role in improving the system performance of multiprocessors. The traditional on-chip interconnection (OCI) method not only fulfills the communication requirements but also results in a significant increase in energy consumption, thereby creating a challenge in achieving an equilibrium between latency, power consumption, and reliability. The emergence of optical networks on chip (ONOC) provides new ideas for solving this difficulty. ONOC has more advantages in transmission speed, communication bandwidth, energy efficiency, and is the most promising solution for chip level communication. To reduce the hardware overhead of ONOC and improve the performance of optical networks, Jinyang et al. proposed a 16-port passive H-tree optical interconnection network based on micro ring resonators. The passive optical H-tree network required only 72 micro ring resonators in a 16 × 16 array scale. This demonstrated its excellent performance compared to passive network structures such as Crossbar, λ-Router, GWOR, LACE, and Light. The average insertion loss of the network was only 1.49 dB, which was 21.5, 10.7, and 59.7% lower than that of λ-Router, GWOR, and TAONoC, respectively. This study has made significant breakthroughs by optimizing the use of micro ring resonators and organizing network structures [7]. The quantum and neural morphology computing platforms in integrated photonic circuits required next-generation optical capabilities. Usually, increasingly complex on-chip optical paths allowed for overlays that could not be achieved by planar techniques, which is crucial for artificial neural networks. Multifunctional 3D wave-guides could be achieved through micro printing based on two-photon polymerization (TPP). Adao et al. proposed a 3D shape prediction tool that considers experimental TPP parameters, achieving simulation of 3D wave-guide performance on chip. This simulation could reduce the cost intensive optimization process of system experiments. The optical transmission characteristics of the prepared 3D wave-guide were consistent with the simulation results, indicating that the developed morphology prediction method was conducive to the development of universal on-chip and inter-chip photon interconnection technologies [8]. Li et al. proposed the method of adding excessive methyl ammonium bromide (MABr) to its precursor. They studied self-passivating perovskite films with excessive doping of MABr by changing the stoichiometric ratio of perovskite precursors. As the MABr content increased, the interaction between free charges and longitudinal optical phonons increased, thereby reducing the exciton binding energy, leading to emission broadening and blue shift of the photoluminescence spectroscopy spectrum. When the molar ratio of PbBr2 to MABr was 1:1.315, PeLED achieved a good balance between brightness and stability, exhibiting a maximum brightness of 9,981 CD·m(−2), a lifespan of 14 min, and a maximum current efficiency of 6.59 CD·A(−1) [9]. Mode division multiplexing (MDM) for OCI has received widespread attention as a degree of freedom that can further expand communication ability. However, the selective loading of multi-mode optical carriers in MDM was not as easy as wavelength division multiplexing (WDM).

Hao and Xin designed a scalable mode selection device for OCI. It had two functional blocks. In one module, the carrier depletion plus silicon micro ring resonator was used to achieve simultaneous mode demultiplexing of multi-mode bus wave-guides and high-speed modulation functions. In another module, a mode multiplexer based on an asymmetric directional coupler restored the modulated signal from the fundamental mode to the original mode sequence. Through this, each channel was separated from the input port and could proceed separately. Researchers can selectively modulate any mode channel as needed, and this structure can be extended to multiple channels. As a concept proof, researchers designed and manufactured a device with 4 micro ring resonators and channel mode multiplexers. This method could provide greater operational flexibility for multi-mode optical interconnects [10]. Long wave infrared covers a rich range of molecular absorption fingerprints and has great potential in chemical sensing. Wave-guides provide an attractive chip level miniaturization solution for optical sensors. However, the exploration of wave-guide sensors within this wavelength range is limited. Weixin et al. developed a photon platform for fast and sensitive on-chip gas sensing utilizing suspended silicon wave-guides. It gave a feasible method for fully utilizing the transparent window of silicon. The subwavelength grating (SWG) presented a potential solution for designing patterns of interaction between strong light analytes. The field research covered the issues of transmission loss and bending loss in the wide wavelength range of 6.4–6.8 µm, and has achieved satisfactory results. In addition, the performance of functional devices such as grating couplers, Y-junctions, and directional couplers also proved their high performance. In view of vapor detection, the sensing function based on this platform was demonstrated. The test results showed that it could reach the detection limit of 75 ppm toluene, with a response time of 0.8 s and a recovery of 3.4 s for 75 ppm toluene, respectively. This powerful and efficient performance had huge potential for developing the SWG structures in on-site medical and environmental applications, making it a highly promising platform [11]. Huang et al. proposed a model for predicting human gait reaction forces and energy metabolism under a spring-loaded suspension backpack. The model was subsequently validated through experimentation, demonstrating that the discrepancy between the model-predicted energy cost and the experimental data was less than 7%. This outcome substantiated the efficacy of the model in optimizing the system performance [12]. Similarly, Sharma et al. developed a brain signal classification method based on multiple machine learning algorithms. The results showed that by combining multiple signal processing techniques, the performance of the model in complex signal processing was significantly improved, which further contributed to the optimization of complex systems. These works share some similarities with the study of complex signal processing challenges encountered in SAOWC design. These novel modeling techniques provide more scalability and robustness support in complex optical signal processing compared to traditional approaches [13].

In summary, ONOC has the advantages of fast transmission speed, large communication bandwidth, and high energy efficiency, and is considered a powerful solution to chip-level communication problems. However, the development and implementation of this technology also face many challenges, including device size, power consumption, reliability, and manufacturing complexity. Considering the complexity and cost of the manufacturing process, it is necessary to develop new manufacturing and packaging technologies to improve production efficiency and output quality. Despite these challenges, with the continuous development and optimization of photonic technology, it is expected that ONOC will occupy an increasingly vital position in future computing and communication systems. It is looked forward to achieving more efficient, reliable, and cost-effective ONOC solutions through continuous research and innovation.

3 Design of SAOWC method for OCOI

The research on the SAOWC method of OCOI mainly adopts the theories and methods of nonlinear optics, material science, and integrated circuit design to design and manufacture AOWC. This converter can achieve wavelength conversion of light without the involvement of electronic devices, thereby transmitting multiple signals of different wavelengths in the same optical fiber.

3.1 Optical waveguide construction of 3D interlayer silicon waveguide coupler

The construction of 3D interlayer silicon waveguide couplers began with precise etching techniques to prepare silicon waveguides. Then, through a specially designed interlayer coupling scheme, precise docking of silicon waveguides is achieved. In this process, the optimization of coupling efficiency is particularly concerned, by adjusting the geometric parameters and material properties of the waveguide to maximize coupling performance [14]. To ensure the implementation of final device performance, relying on advanced simulation tools for detailed simulation and optimization is an essential step. In this context, the propagation of electromagnetic waves along metal lines can be considered the construction of a minimalist waveguide structure, whereby the propagation of electromagnetic waves is restricted to this specific route. Similarly, visual light is a part of electromagnetic waves, and its wave nature can be utilized to confine light to a material with a specific shape, making it an optical waveguide. This is the basic construction method of optical waveguides in 3D interlayer silicon waveguide couplers. There are multiple commonly used methods for solving strip waveguides, including the Marcatili approximation and Kumar method for effective refractive index. The two methods are shown in Figure 1.

Figure 1

Schematic diagram of (a) Marcatili approximation method and (b) Kumar method.

In Figure 1(a), due to the relatively low influence of four specific regions R1, R2, R3, and R4 on the refractive index of the core layer, the Marcatili method chooses to ignore these regions. This strategy helps to reduce the number of boundary conditions that need to be processed, thereby reducing computational complexity. However, in Figure 1(b), although Kumar’s method is similar in core ideas to Marcatili, it takes into account the influence of these four corner regions on refractive index, which is ignored in Marcatili. Waveguides have two intrinsic polarization states, transverse electric (TE) and transverse magnetic (TM). Electromagnetic waves have three directions: electric field component, magnetic field component, and transmission, which are perpendicular to each other. If the direction of the electric field is parallel to the transmission plane and the direction of the magnetic field is perpendicular to the transmission plane, it is called the TE mode. If the magnetic field direction is parallel to the transmission plane and the electric field direction is perpendicular to the transmission plane, it is called the TM mode [15]. The eigenmode equations of TE and TM are shown in Eq. (1).

(1) E y = − ω μ 0 β H x ∂ E y ∂ x = − i ω μ 0 H z − i β H x − ∂ E z ∂ x = − i ω ε E y H y = − ω ε β E x − i ω ε ∂ H y ∂ x = E z − i β E x − ∂ E z ∂ x = − i ω μ 0 H y .

In Eq. (1), light propagates along the z direction in the waveguide. β is the propagation constant. On z the direction of the electric field changes. Taking TE mode as an example, after solving the wave equation, the field distribution in the waveguide can be derived from Eq. (2) after determining parameters such as waveguide size.

(2) K d = m π + tan − 1 p k + tan − 1 q k , m = 0 , 1 , 2 , 3 … ,

where K , p , and q are all functions of β . Assuming two light waves propagate in two waveguides, the coupled mode equation is given by Eq. (3).

(3) ∂ A 1 ( z ) ∂ z = i k 1 A 2 ( z ) e i ( β 1 − β 2 ) z ⋯ ∂ A 2 ( z ) ∂ z = i k 2 A 1 ( z ) e i ( β 1 − β 2 ) z ⋯ ,

where A 1 and A 2 are the electric field components, respectively. β 1 and β 2 are the propagation constants in the two waveguides, respectively. k 1 and k 1 are coupling coefficients. The coupling length is given by Eq. (4).

(4) k L = m π + π 2 ,

where L is the coupling length and m is any positive integer. The structural diagram of the MZI type optical switch is shown in Figure 2.

Figure 2

Schematic diagram of 2 × 2 optical switch structure.

In Figure 2, a beam of light enters from input port I₁, passes through the first 3-dB coupler, and is evenly divided into two beams with π/2 phase difference. These two beams of light continue to enter the modulation region. If no additional modulation is applied, the two beams of light maintain a π/2 phase difference and enter the second 3-dB coupler. The coupler is designed symmetrically and can compensate for the phase difference between two beams of light, so that the two beams are combined into one beam and output at the O₂ port. However, if modulation is applied, it will cause the π/2 phase difference between the two beams of light to become −π/2, resulting in phase inversion. Therefore, light will be output from the O₁ port. If the two states are light input from I₁ and output from O₂, it is known as the “Cross” state. On the contrary, if the two states are input from I₁ and output from O₁, it is named as the “Bar” state. This forms two states of a 2 × 2 optical switch. The distribution of guided modes in multi-mode waveguides is shown in Figure 3.

Figure 3

Distribution of guided modes in multi-mode waveguides.

In Figure 3, the self-imaging effect (SIE) plays a crucial role in the design of a multi-mode interferometer (MMI). SIE is a fundamental optical phenomenon in multi-mode waveguides, which is achieved through complex interactions between phase consistency and geometric shape. Essentially, it describes the process of how light replicates its incident mode in the propagation of matter, while also reproducing the shape and structure of the scene. According to SIE, by cleverly designing MMI, the research can achieve the generation of input mode images at the waveguide exit. This is why in the design of MMI, the target size parameters are determined based on SIE. Geometric size parameters are key elements that determine the generation of functional images, including the length and width of MMI, as well as the width of the designed multi-mode zone. Each parameter is greatly affected by the interferometer, including loss, phase error, crosstalk, and bandwidth. Based on the guided mode distribution in SIE, it is possible to design a MMI with specific characteristics and performance. Especially in optical communication and integrated photonics, it has extensive applications, such as being able to achieve complex optical signal processing functions, WDM devices, MDM devices, etc. [16,17]. The distribution mode of guided modes under SIE directly determines the output characteristics and application range of MMIs. The optical field intensity at the input of the multi-mode interference region is given by Eq. (5).

(5) F = ∑ m C m ϕ m ( x ) ,

where ϕ m ( x ) is the m -order eigenmode of the multi-mode interference region. C _m is the weight coefficient. At the position of the L multi-mode interference interval, the light field formula is given by Eq. (6).

(6) F ( x , L ) = ∑ m C m ϕ m ( x ) e − j β m z ,

where β m is the propagation constant. The position of the β m -fold self-image of the input light field generated in the multi-mode interference region is given by Eq. (7).

(7) L = 3 L π N ,

where L _π represents the beat length and β m represents the self-reflection position.

3.2 Building an AOWC model based on quantum dot semiconductor optical amplifier Mach Zehnder interferometer (QD-SOA-MZI)

This study explores the construction methods of AOWC. The converter is based on the QD-SOA-MZI [18,19]. First, the QD-SOA is designed to accurately control the parameters of quantum dots to optimize broadband wavelength conversion performance. On this basis, a MZI is built to provide an optical architecture for AOWC. In the process of designing and constructing this device, theoretical analysis and experimental verification are deeply integrated. On the one hand, detailed theoretical analysis provides accurate guidance for design, predicts experimental results, and provides theoretical basis for system optimization through multiple iterations. In addition, experimental verification provides theoretical analysis with real-world references. The blind spots and unknown problems can also be identified and solved during the experimental process. The working diagram of the converter is shown in Figure 4.

Figure 4

Working principle diagram of cross-phase modulation (XPM) type wavelength converter based on QD-SOA-MZI.

In Figure 4, two identical QD-SOAs (QD-SOA1 and QD-SOA2) are placed in the arms of MZI. In this design, the pump light is input into QD-SOA1, and the continuous detection light is decomposed by a splitter and then input into two QD-SOAs, respectively. Due to the fact that the detection light in QD-SOA1 is modulated by pump light, there is a phase difference in the detection light compared to the one that is not modulated by pump light. Furthermore, through interference, this phase difference is converted into an intensity change, allowing the detection light at the output port to have pump light information, achieving wavelength conversion. In modern optical communication systems, bit error rate (BER) is usually used as one of the parameters to measure the accuracy of signal transmission. Subsequently, through interference, this phase difference is converted into an intensity change, and the probe light carries pump light information in the output port, thus achieving wavelength conversion. In today’s optical communication systems, the BER is often regarded as one of the key parameters to measure the accuracy of signal transmission, and its formula is given in Eq. (8).

(8) BER = e − Q 2 2 2 π Q .

The transformation extinction ratio, as an important indicator of system robustness and sensitivity, is defined as the ratio of the average power of output signal “1” to “0”, as shown in Eq. (9).

(9) ER = 10 lg p 1 p 0 ,

where ER is the extinction ratio of the transformed light, and ω 0 is the pulse width. The principle of all-optical logic XOR gate (AoL-XORg) built on QD-SOA-MZI is shown in Figure 5.

Figure 5

AoL-XORg principle based on QD-SOA-MZI.

In Figure 5, two identical QD-SOAs are placed on both arms of the interferometer. The detection light is divided into two identical beams through the wavelength divider C1, and is input into QD-SOA through couplers C2 and C3 along with the signal light Data A and Data B. Due to the higher input power of Data A and Data B compared to the maximum linear input power of QD-SOA, the probe light is modulated by Data A and Data B in QD-SOA. These modulated lights undergo interference to achieve an AoL-XORg. To achieve logical XOR operation, the output optical power is given by Eq. (10).

(10) P XOR ( t ) = P probe { k 1 k 2 G 1 ( t ) + ( 1 − k 1 ) ( 1 − k 2 ) G 2 ( t ) − 2 k 1 k 2 ( 1 − k 1 ) ( 1 − k 2 ) G 1 ( t ) G 2 ( t ) × cos [ ϕ 1 ( t ) − ϕ 2 ( t ) ] } ,

where G 1 ( t ) and G 2 ( t ) are the gains of the detection light after passing through the interferometer QD-SOA1 and QD-SOA2, respectively. ϕ 1 ( t ) and ϕ 2 ( t ) are the phases of the detection light modulated by the signal light, respectively. k 1 and k 2 are the coupling coefficients of coupling [20,21]. Overall, the AOWC and AoL-XORg based on QD-SOA-MZI provide a unique method to explore and optimize the performance of optical communication systems.

4 Predictive analysis of OCOI’s SAOWC application

The application prospects of SAOWC are broad, especially in data communication and silicon photonics. In data communication, AOWC can achieve wavelength conversion of light without the involvement of electronic devices, thereby transmitting multiple signals of different wavelengths in the same optical fiber, greatly improving data transmission capacity. It can effectively reduce signal interference and improve signal quality and system stability. The method is highly scalable due to the maturity of silicon photonics technology, which allows for seamless integration with existing photonic systems, particularly in standard silicon-based chip fabrication processes without extensive process modifications. This makes it suitable for mass production. However, commercial implementation still faces a number of challenges, especially in the manufacturing and packaging processes. Although the cost of integrated silicon photonic systems has been gradually decreasing, the complex design and high-precision fabrication requirements of wavelength converters may increase the cost, especially in the early stages. Further cost reductions are expected with advances in manufacturing technology and the realization of mass production, but initial capital investment remains a major obstacle to commercial implementation.

4.1 Application analysis of AOWC in data communication

The application of AOWC in data communication will be of great significance. By using optical wavelength conversion without the involvement of electronic devices, the same optical fiber can transmit multiple signals of different wavelengths, greatly improving the ability of data communication. In addition, AOWC can effectively reduce signal interference and improve signal quality and system stability. In comparison to conventional electronic wavelength converters, SAOWC exhibits a notable advantage in terms of energy consumption. This is due to the fact that its operation is primarily based on optical signal processing, which eliminates the necessity for photovoltaic conversion. This significantly reduces the energy consumption issues that are commonly observed in electronic devices. However, despite the excellent energy-saving characteristics of SAOWC, its overall energy efficiency is still dependent on the use of light sources and amplifiers. In some application scenarios, it may be necessary to additionally optimize the energy-efficient design of the device to reduce system power consumption, thus preventing it from becoming a limiting factor in large-scale applications. In conducting experiments on the application of AOWC in data communication, the selection of parameters and equipment is crucial. Table 1 shows the details of the model.

Table 1

System parameters

Parameter name	Parameter value	Equipment selection
Fiber type	Multi-mode fiber	Fiber suitable for multi-wavelength signal transmission
Optical wavelength converter	AOWC	Device with high-performance optical wavelength conversion ability
Fiber coupler	High precision fiber coupler	Used to reduce signal interference and improve signal quality
Optical detector	High performance optical detector	Used to detect optical signals
Power supply	Stable power supply	Used to ensure system stability
Temperature control equipment	High precision temperature control equipment	Used to ensure the normal operation of the device

In Table 1, first, it is necessary to select optical fibers that can support multi-wavelength signal transmission, as well as AOWC with high-performance optical wavelength conversion capabilities. In addition, to reduce signal interference and improve signal quality, it is also necessary to choose high-precision fiber optic couplers and high-performance photodetectors. At the same time, to achieve the system operation stably, a stable power supply and high-precision temperature control equipment are also needed. The curve of the coupling efficiency of the inverted cone waveguide interlayer coupler with wavelength variation is shown in Figure 6.

Figure 6

Curve of coupling efficiency of inverted cone waveguide interlayer coupler changing with wavelength. (a) Wavelength change rate curve 1. (b) Wavelength change rate curve 2.

In Figure 6, through simulation calculations, there is a clear pattern in the rate of change of wavelength within a specific time interval using the finite-difference time-domain (FDTD) method. At t = 0 s, the wavelength is 1,550 nm, while at t1 seconds, the wavelength increases to 1,600 nm. During this time interval, the rate of change in wavelength (i.e., the amount of change in wavelength per second) remains within a certain range, demonstrating the stability of the device’s operation throughout the C-band. This high coupling efficiency not only demonstrates the excellent performance of the converter in OCOIs but also hints at its potential for applications in biomedical imaging. For example, the depth of photon penetration in biological tissues and the quality of imaging are directly related to the coupling efficiency. Higher coupling efficiency means more stable signal transmission, which can provide sharper biomedical images, especially within the infrared band, which penetrates deeper into biological tissues. The curve of the coupling efficiency of the rectangular waveguide interlayer coupler with wavelength variation is shown in Figure 7.

Figure 7

Curve of coupling efficiency of rectangular waveguide interlayer coupler with wavelength variation. (a) Curve 1 of the coupling efficiency of rectangular waveguide interlayer couplers with wavelength variation. (b) Curve 2 of the coupling efficiency of rectangular waveguide interlayer couplers with wavelength variation.

Figure 7 simulates the interlayer coupling efficiency at different wavelengths using the FDTD method. At a wavelength of 1,550 nm, the interlayer coupling efficiency can reach 98.14%, and the coupling efficiency range of over 90% corresponds to a bandwidth of greater than 100 nm, which can cover the entire C-band.

4.2 Application analysis of AOWC in silicon photonics

The application of AOWC in silicon photonics will also have a profound impact. Through SAOWC, more efficient photonic integrated circuits can be achieved, further improving the performance of optoelectronic devices. This will promote the development of silicon photonics and provide new possibilities for the design and manufacturing of optoelectronic devices. The research and application of AOWC will open up new research fields and application prospects for the development of silicon photonics. Table 2 shows the parameters of the model.

Table 2

System parameters of all-optical wavelength converters

Parameter name	Parameter value	Equipment selection
Optical wavelength converter	SAOWC	Device with high-efficiency wavelength conversion performance
Photonic integrated circuit	Photonic integrated circuit	Device for simulating and testing optoelectronic device performance
Light source device	Laser	Stable, moderate intensity light source device
Optical detector	Photodiode	High sensitivity, fast response optical detector
Power supply	Direct current power supply	Stable output, low current fluctuation power supply
Temperature control equipment	Digital temperature controller	High precision, fast response temperature control equipment

In Table 2, when choosing AOWC, SAOWC with high efficiency wavelength conversion performance is preferred. Meanwhile, to simulate and test the performance of optoelectronic devices, photonic integrated circuits have been selected as the preferred equipment. In terms of the selection of light source equipment, a stable and moderately strong laser is considered as the best option. In addition, due to its high sensitivity and fast response speed, photodiodes have been selected as the most ideal photodetectors. On the power supply, a DC power supply with stable output and small current fluctuations is selected. In addition, digital temperature controllers with high accuracy and fast response speed have been selected as temperature control devices to ensure the accuracy and effectiveness of the experiment. The relationship between error rate and pulse width is shown in Figure 8.

Figure 8

The relationship between BER and pulse width.

In Figure 8, when Λ = 0.1 ps, BER is 4.341310⁻¹⁷. When Λ = 1.0 ps, BER is 3.666410⁻⁴. In view of this, when Λ increases from 0.1 to 1.0 ps, the order of 10⁻¹⁷ is increased to the order of 10⁻⁴. An increase in BER can mean more interference or loss of data during the transmission of biological signals, similar to the loss of information in neural signaling. Therefore, the BER results underscore the necessity for additional optimization of the technique when processing biological data, aiming to ensure the maintenance of high signal accuracy in complex environments. The simulation results of XPM type wavelength converter based on QD-SOA-MZI are shown in Figure 9.

Figure 9

Simulation results of XPM wavelength converter based on QD-SOA-MZI. (a) Pump power before conversion. (b) Pump power after conversion. (c) Conversion power before conversion. (d) Conversion power after conversion.

In Figure 9, during the simulation of the all-optical wavelength converter, it is assumed that the pump light wavelength is 1,500 nm and the detection light wavelength is 1,550 nm. When the injection current 1I of QD-SOA1 is 40 mA and the injection current 2I of QD-SOA2 is 35 mA, in-phase conversion is achieved. The selection of these two types of injection currents is mainly based on the preliminary simulation analysis of the gain characteristics of the quantum dot semiconductor optical amplifier (QD-SOA) and the reference values in the literature. The specific reason is that the 40 mA injection current can provide sufficient optical gain for QD-SOA1, while the 35 mA injection current ensures the steady-state output of QD-SOA2. The difference between the two also helps to realize the phase modulation in the interferometer. The simulation under these environmental conditions reveals the performance of the AOWC under specific operating conditions. The relationship between pulse width and influence factor is shown in Figure 10.

Figure 10

Relationship between pulse width and influence factor.

In Figure 10, during the process of reducing the parameter factor from 2.0 to 0.5 ps, the error rate (ER) increases from 33.9738 to 33.9537 dB, while the impact factor increases from 5.6898 to 11.1202 dB. This process reveals a phenomenon of mutual hindrance, namely, the difficulty in attempting to increase both ER and influence factors by changing parameter factors. When determining the parameter factor, this factor should be considered to ensure that while improving the influence factor, the obtained ER value can also meet the requirements. This is a strategy to balance optimization indicators, aiming to maintain a high level of ER while meeting the improvement of impact factors. In addition, when optimizing the injection current, a dual-objective optimization method based on BER and coupling efficiency is used, and the test results for QD-SOA1 and QD-SOA2 are shown in Figure 11.

Figure 11

ER and coupling efficiency testing of different injected currents. (a) Error rate testing. (b) Coupling efficiency test.

Figure 11 shows the results of the ER and coupling efficiency at different injection currents. From Figure 11(a), the ER decreases significantly with the increase in QD-SOA1 injection current. When the injection current is increased from 35 to 45 mA, the BER decreases from 10 to 17 levels to nearly 10–18 levels, proving that the ER can be effectively reduced by adjusting the current. In Figure 11(b), the coupling efficiency gradually increases from 98.44% to more than 98.7% with the increase in injection current. This proves that the reasonable adjustment of injection current helps to improve the coupling performance of the optical path. In addition, the study conducts the temperature fluctuation test and optical signal intensity degradation test in the stress test, and the results are shown in Figure 12.

Figure 12

Temperature fluctuation test and light signal intensity degradation test. (a) Optical power loss under temperature and pressure. (b) Signal to noise ratio under temperature and pressure. (c) Optical power loss caused by degradation of optical signal intensity. (d) Signal to noise (SNR) ratio of degraded optical signal intensity.

Figure 12(a) and (b) show the test results of optical power loss and SNR under temperature stress. Figure 12(c) and (d) show the test results of optical power loss and SNR ratio with degradation of optical signal strength. From Figure 12(a) and (b), the optical power loss gradually increases from 0.5 to 3.0 dB as the temperature rises from 20 to 80°C, while the SNR gradually decreases from 40 to 22 dB. This indicates that the temperature rise negatively affects the transmission efficiency of the system, resulting in more signal loss. At the same time, the temperature fluctuation leads to the degradation of the signal quality, and the noise relative to the signal becomes more pronounced. This phenomenon can also be compared to temperature sensitivity in biological systems. For example, neural signaling and enzymatic reactions change in efficiency and accuracy at different temperatures. Increased molecular vibrations in biological systems at high temperatures can lead to signal distortion. Similarly, the effect of temperature on the photon transport system suggests that the performance of the converter in different biological environments may also need to be optimized for temperature changes. From Figure 12(c) and (d), the optical power loss increases from 0.5 to 3.2 dB and the SNR decreases from 40 to 20 dB as the signal strength decreases from 0 to 10 dBm. This indicates that the decreasing of the signal strength has a large impact on the loss in the transmission process. At the same time, the quality of the signal with an increase in the noise component is significantly reduced. The optical power loss increases with the temperature, while the SNR decreases with the temperature. In addition, the optical power loss increases with the weakening of the signal strength, while the SNR decreases significantly. It illustrates that temperature fluctuation needs to be controlled in practical applications to avoid system reliability degradation, while keeping the intensity of the optical signal within a certain range is the key to ensure the stability of the system.

5 Conclusion

In the current field of high-speed data transmission and communication, on-chip OIT has received widespread attention due to its characteristics such as low latency and energy consumption, and high bandwidth. Among them, efficient optical wavelength converters are a key component of smooth information transmission and play a crucial role in supporting the comprehensive operation of OCOIs. Therefore, the research on SAOWC for OCOIs has remarkable significance. The existing research mainly uses the FDTD method for simulation calculations. During this process, attempts are made to achieve in-phase conversion at different wavelengths by adjusting the injection current 1I of QD-SOA1 and the injection current 2I of QD-SOA2.

Through simulation and experimental verification, the study successfully demonstrated the efficient performance of SAOWCs in OCOIs. First, the FDTD simulation results showed that the interlayer coupling efficiency reached 98.44% at 1,550 nm wavelength, and this performance was consistent with the experimental data, proving the high efficiency of the proposed converter. Second, the BER test clearly showed that the increase in pulse width had a significant effect on the BER, which rose from 10⁻¹⁷ to 10⁻⁴ when the pulse width increased from 0.1 to 1.0 ps. This further illustrated the criticality of the pulse parameter on the system performance. In addition, the test results for temperature fluctuation and optical signal strength degradation showed the stability of the system in different environments, where the optical power loss increased with temperature rise and signal strength degradation, while the SNR decreased. This indicated that the converter performed well in complex environments with good operational stability as predicted by simulations. In the high wavelength range, existing silicon-based materials will experience increased losses and nonlinear effects, which will affect conversion efficiency and signal quality. The design of wavelength converters must be adapted to accommodate the varying wavelengths of light waves, encompassing the engineering of waveguide structures and resonant cavities. This is essential to enhance the coupling and transmission of light. The experimental equipment used for testing and validation also needs to be updated to enable accurate measurement and analysis in a higher wavelength range. The application of SAOWC in on-chip OIT has demonstrated extremely high interlayer coupling efficiency and reliable in-phase conversion performance. It is expected to have more optimization research works in the future to promote the deeper development of optical wavelength converters and on chip OITs.

Funding information: The research was supported by 2021 Shaanxi Provincial Department of Education Project, Research on silicon-based all-optical wavelength converters for on-chip optical interconnection Project No. “21JK0513.”
Author contributions: All the contributions in the article are attributed to Shuhong Liu.
Conflict of interest: The author declares that there is no conflict of interest.
Data availability statement: The data used to support the findings of the research are available from the corresponding author upon reasonable request.

References

[1] Huang X, Li Z. Multi-scale analyses on mechano-electric degradation of film interconnects in flexible electronics. Fatigue Fract Eng Mater Struct. 2023;46(1):259–70.10.1111/ffe.13861Suche in Google Scholar

[2] Bharat S, Chaudhari S, Patil S. Optimized designs of low loss non-blocking optical router for ONoC applications. Int J Inf Technol. 2020;12(1):91–6.10.1007/s41870-019-00298-7Suche in Google Scholar

[3] Cao XP, Zheng S, Zhou N. On-chip multidimensional 1 × 4 selective switch with simultaneous mode-/polarization-/wavelength-division multiplexing. IEEE J Quantum Electron. 2020;56(5):1–8.10.1109/JQE.2020.3004821Suche in Google Scholar

[4] Zhang Y, Zhang RH, Zhu QM. Architecture and devices for silicon photonic switching in wavelength, polarization and mode. J Lightw Technol. 2019;38(2):215–25.10.1109/JLT.2019.2946171Suche in Google Scholar

[5] Zhang J, Kuo BPP, Radic S. 64 Gb/s PAM4 and 160 Gb/s 16QAM modulation reception using a low-voltage Si-Ge waveguide-integrated APD. Opt Express. 2020;28(16):23266–73.10.1364/OE.396979Suche in Google Scholar PubMed

[6] Fang Y, Luo B, Zhao T, He D, Jiang B, Liu Q. ST-SIGMA: Spatio-temporal semantics and interaction graph aggregation for multi-agent perception and trajectory forecasting. CAAI Trans Intell Technol. 2022;7(4):744–57.10.1049/cit2.12145Suche in Google Scholar

[7] Jinyang H, Lin J, Yan Z. Design of passive H tree network for on-chip optical interconnection. J Optoelectron Laser. 2022;33(June):570–7.Suche in Google Scholar

[8] Adao RMR, Alves TL, Maibohm C. Two-photon polymerization simulation and fabrication of 3D microprinted suspended waveguides for on-chip optical interconnects. Opt Express. 2022;30(1):9623–42. Optica Publ Group.10.1364/OE.449641Suche in Google Scholar PubMed

[9] Li G, Li M, Zuo Y, Suo X, Su H, Wang Y. Self-passivated perovskite film by overdoping MABr to enhance the luminescence efficiency of MAPbBr3-based light-emitting diodes. Opt Eng. 2022;61(8):1–10.10.1117/1.OE.61.8.087104Suche in Google Scholar

[10] Hao J, Xin F. Mode-selective modulation by silicon microring resonators and mode multiplexers for on-chip optical interconnect. Opt Express. 2019;27(Jan):2915–25.10.1364/OE.27.002915Suche in Google Scholar PubMed

[11] Weixin L, Yiming M. Suspended silicon waveguide platform with subwavelength grating metamaterial cladding for long-wave infrared sensing applications. Nanophotonics. 2021;10(May):1861–70.10.1515/nanoph-2021-0029Suche in Google Scholar

[12] Huang L, Yang Z, Wang R, Xie L. A model for predicting ground reaction force and energetics of human locomotion with an elastically suspended backpack. Comput Methods Biomech Biomed Eng. 2022;25(14):1554–64.10.1080/10255842.2021.2023808Suche in Google Scholar PubMed

[13] Sharma H, Raj R, Juneja M. An empirical comparison of machine learning algorithms for the classification of brain signals to assess the impact of combined yoga and Sudarshan Kriya. Comput Methods Biomech Biomed Eng. 2022;25(7):721–8.10.1080/10255842.2021.1975682Suche in Google Scholar PubMed

[14] Huang L, Gu H, Tian Y, Zhao T. Universal method for constructing the on-chip optical router with wavelength routing technology. J Lightw Technol. 2020;38(15):3815–21.10.1109/JLT.2020.2986237Suche in Google Scholar

[15] Mere V, Dash A, Kallega R, Pratap R, Naik A, Selvaraja SK. On-chip silicon photonics based grating assisted vibration sensor. Opt Express. 2020;28(19):27495–505.10.1364/OE.394393Suche in Google Scholar PubMed

[16] Zhang K, Hu H. An experimental study on the transient runback characteristics of wind-driven film/rivulet flows. Phys Fluids. 2021;33(11):1–14.10.1063/5.0067672Suche in Google Scholar

[17] Dabaghyan G. Modification of a silicon surface visible light reflectivity by the carbon nanofilm. Int J Mod Phys B. 2021;35(26):1–5.10.1142/S0217979221502647Suche in Google Scholar

[18] Guo PX, Hou WG, Guo L. Fault tolerant routing mechanism in 3D optical network-on-chip based on node reuse. IEEE Trans Parallel Distrib Syst. 2020;31(3):547–64.10.1109/TPDS.2019.2939240Suche in Google Scholar

[19] Mahmood T, Ali Z. Prioritized Muirhead mean aggregation operators under the complex single-valued neutrosophic settings and their application in multi-attribute decision-making. J Comput Cogn Eng. 2022;1(2):56–73.10.47852/bonviewJCCE2022010104Suche in Google Scholar

[20] Muhiuddin G, Mahboob A, Elnair MEA. A new study based on fuzzy bi-Γ-ideals in ordered-Γ-semigroups. J Comput Cogn Eng. 2022;1(1):42–6.10.47852/bonviewJCCE19919205514Suche in Google Scholar

[21] Waziri TA, Yakasai BM. Assessment of some proposed replacement models involving moderate fix-up. J Comput Cogn Eng. 2023;2(1):28–37.10.47852/bonviewJCCE2202150Suche in Google Scholar

Received: 2024-07-26

Revised: 2024-10-08

Accepted: 2024-11-04

Published Online: 2025-02-10

This work is licensed under the Creative Commons Attribution 4.0 International License.

Artikel in diesem Heft

https://doi.org/10.1515/nleng-2024-0059

Schlagwörter für diesen Artikel

silicon-based all-optical wavelength converter; optical interconnection technology; data communication; nonlinear optics; integrated circuit design

Creative Commons

BY 4.0