Volume 14,Issue 3,2018 Table of Contents
2018, 14(3):161-163.
DOI: https://doi.org/10.1007/s11801-018-7249-0
Abstract:
A self-mixing interferometer (SMI) with resolution twenty times higher than that of a conventional interferometer is developed by multiple reflections. Only by employing a simple external reflecting mirror, the multiple-pass optical configuration can be constructed. The advantage of the configuration is simple and easy to make the light re-injected back into the laser cavity. Theoretical analysis shows that the resolution of measurement is scalable by adjusting the number of reflections. The experiment shows that the proposed method has the optical resolution of approximate λ/40. The influence of displacement sensitivity gain (G) is further analyzed and discussed in practical experiments.
A self-mixing interferometer (SMI) with resolution twenty times higher than that of a conventional interferometer is developed by multiple reflections. Only by employing a simple external reflecting mirror, the multiple-pass optical configuration can be constructed. The advantage of the configuration is simple and easy to make the light re-injected back into the laser cavity. Theoretical analysis shows that the resolution of measurement is scalable by adjusting the number of reflections. The experiment shows that the proposed method has the optical resolution of approximate λ/40. The influence of displacement sensitivity gain (G) is further analyzed and discussed in practical experiments.
2018, 14(3):164-169.
DOI: https://doi.org/10.1007/s11801-018-7266-z
Abstract:
In this study, we design and fabricate phosphor-doped host-guest structure organic light-emitting diodes (OLEDs), where the blue-ray iridium complex electrophosphorescent material FIrpic acts as object material. Properties of the device can be accommodated by changing the host materials, dopant concentration and thickness of the light-emitting layer. The study shows that the host material N,N'-dicarbazolyl-3,5-benzene (mCP) has a higher triplet excited state energy level, which can effectively prevent FIrpic triplet excited state energy backtracking to host material, thus the luminous efficiency is improved. When mCP is selected as the host material, the thickness of the light-emitting layer is 30 nm and the dopant concentration is 8 wt%, the excitons can be effectively confined in the light-emitting region. As a result, the maximum current efficiency and the maximum brightness of the blue device can reach 15.5 cd/A and 7 196.3 cd/m2, respectively.
In this study, we design and fabricate phosphor-doped host-guest structure organic light-emitting diodes (OLEDs), where the blue-ray iridium complex electrophosphorescent material FIrpic acts as object material. Properties of the device can be accommodated by changing the host materials, dopant concentration and thickness of the light-emitting layer. The study shows that the host material N,N'-dicarbazolyl-3,5-benzene (mCP) has a higher triplet excited state energy level, which can effectively prevent FIrpic triplet excited state energy backtracking to host material, thus the luminous efficiency is improved. When mCP is selected as the host material, the thickness of the light-emitting layer is 30 nm and the dopant concentration is 8 wt%, the excitons can be effectively confined in the light-emitting region. As a result, the maximum current efficiency and the maximum brightness of the blue device can reach 15.5 cd/A and 7 196.3 cd/m2, respectively.
REN Mei-zhen
,
ZHANGJia-shun
,
AN Jun-ming
,
WANG Yue
,
WANG Liang-liang
,
LI Jian-guang
,
WU Yuan-da
,
YIN Xiao-jie
,
HU Xiong-wei
2018, 14(3):170-174.
DOI: https://doi.org/10.1007/s11801-018-8008-y
Abstract:
We report investigations of three types of silica-based thermo-optic modulating Mach-Zehnder interferometers (MZIs). They are widely used in optical communication and quantum photonics. Three types of MZIs are fabricated. The waveguide structure and fabrication process are paid special attention. The power consumption is less than 250 mW for all MZIs. The polarization dependent loss (PDL) at the same attenuation using the upper heater is less than that using the lower heater for the three types of MZIs. In addition, it is found that the PDL at the same attenuation increases gradually for π, 2π and 0 phase differences. The measured response time of the three types of MZIs is less than 1.8 ms.
We report investigations of three types of silica-based thermo-optic modulating Mach-Zehnder interferometers (MZIs). They are widely used in optical communication and quantum photonics. Three types of MZIs are fabricated. The waveguide structure and fabrication process are paid special attention. The power consumption is less than 250 mW for all MZIs. The polarization dependent loss (PDL) at the same attenuation using the upper heater is less than that using the lower heater for the three types of MZIs. In addition, it is found that the PDL at the same attenuation increases gradually for π, 2π and 0 phase differences. The measured response time of the three types of MZIs is less than 1.8 ms.
2018, 14(3):175-179.
DOI: https://doi.org/10.1007/s11801-018-7214-y
Abstract:
An N×N optical switch based on cascaded microring resonators on chip is proposed. As an example, the 4×4 optical switch is further investigated. It is successfully demonstrated that its insertion loss is relatively low as 2.2 dB, the crosstalk is negligible, and the extinction ratio (ER) is as large as 130 dB. Thermal tuning is employed to make the microrings be in resonance or not, which leads to a response time of several hundred microseconds. Alternatively, doping the desired waveguide regions with p-type or n-type dopants is able to achieve a better response time of several nanoseconds. The proposed design is easily integrated to a large scale with less microring resonators, which ensures the compact size and the low power consumption.
An N×N optical switch based on cascaded microring resonators on chip is proposed. As an example, the 4×4 optical switch is further investigated. It is successfully demonstrated that its insertion loss is relatively low as 2.2 dB, the crosstalk is negligible, and the extinction ratio (ER) is as large as 130 dB. Thermal tuning is employed to make the microrings be in resonance or not, which leads to a response time of several hundred microseconds. Alternatively, doping the desired waveguide regions with p-type or n-type dopants is able to achieve a better response time of several nanoseconds. The proposed design is easily integrated to a large scale with less microring resonators, which ensures the compact size and the low power consumption.
2018, 14(3):180-184.
DOI: https://doi.org/10.1007/s11801-018-8001-5
Abstract:
A TM-pass polarizer based on multilayer graphene polymer waveguide is proposed and theoretically analyzed. The mode properties, the extinction ratio, the insertion loss and the bandwidth are also discussed. The results show that a TM-pass polarizer, which only guides the TM mode, can be achieved by multilayer graphene polymer waveguide. With length of 150 μm, the proposed polarizer can achieve extinction ratio of 33 dB and insertion loss of 0.5 dB at optical wavelength of 1.55 μm. This device has an excellent performance, including large extinction ratio and low insertion loss within the spectral range from 1.45 μm to 1.6 μm.
A TM-pass polarizer based on multilayer graphene polymer waveguide is proposed and theoretically analyzed. The mode properties, the extinction ratio, the insertion loss and the bandwidth are also discussed. The results show that a TM-pass polarizer, which only guides the TM mode, can be achieved by multilayer graphene polymer waveguide. With length of 150 μm, the proposed polarizer can achieve extinction ratio of 33 dB and insertion loss of 0.5 dB at optical wavelength of 1.55 μm. This device has an excellent performance, including large extinction ratio and low insertion loss within the spectral range from 1.45 μm to 1.6 μm.
2018, 14(3):185-188.
DOI: https://doi.org/10.1007/s11801-018-7246-3
Abstract:
A novel technique is proposed for the generation of ultra-wide and flat optical frequency comb (OFC) based on serially cascading three stages of electro absorption modulators (EAMs) through sinusoidal radio frequency (RF) signals by setting frequencies at f GHz, f/2 GHz and f/4 GHz. Here, the first stage acts as subcarrier generator, the second stage acts as subcarrier doubler, and the third stage acts as subcarrier quadrupler. In addition, a higher number of subcarriers can easily be generated by adjusting the driving sinusoidal RF signal. In this paper, cascading three stages of EAMs driven by 50 GHz, 25 GHz and 12.5 GHz clock sources, we obtain 272 subcarriers with spacing of 2.5 GHz and power deviation within 1 dB. Theoretical analysis of serially cascaded EAMs for subcarrier generation is also investigated. Principal analysis and simulation of this technique are demonstrated.
A novel technique is proposed for the generation of ultra-wide and flat optical frequency comb (OFC) based on serially cascading three stages of electro absorption modulators (EAMs) through sinusoidal radio frequency (RF) signals by setting frequencies at f GHz, f/2 GHz and f/4 GHz. Here, the first stage acts as subcarrier generator, the second stage acts as subcarrier doubler, and the third stage acts as subcarrier quadrupler. In addition, a higher number of subcarriers can easily be generated by adjusting the driving sinusoidal RF signal. In this paper, cascading three stages of EAMs driven by 50 GHz, 25 GHz and 12.5 GHz clock sources, we obtain 272 subcarriers with spacing of 2.5 GHz and power deviation within 1 dB. Theoretical analysis of serially cascaded EAMs for subcarrier generation is also investigated. Principal analysis and simulation of this technique are demonstrated.
2018, 14(3):189-194.
DOI: https://doi.org/10.1007/s11801-018-7164-4
Abstract:
The inverted bottom-emitting organic light-emitting devices (IBOLEDs) were prepared, with the structure of ITO/Al (x nm)/LiF (1 nm)/Bphen (40 nm)/CBP:GIr1 (14%):R-4b (2%) (10 nm)/BCP (3 nm)/CBP:GIr1 (14%):R-4b (2%) (20 nm)/TCTA (10 nm)/NPB (40 nm)/MoO3 (40 nm)/Al (100 nm), where the thickness of electron injection layer Al (x) are 0 nm, 2 nm, 3 nm, 4 nm and 5 nm, respectively. In this paper, the electron injection condition and luminance properties of inverted devices were investigated by changing the thickness of Al layer in Al/LiF compound thin film. It turns out that the introduction of Al layer can improve electron injection of the devices dramatically. Furthermore, the device exerts lower driving voltage and higher current efficiency when the thickness of electron injection Al layer is 3 nm. For example, the current efficiency of the device with 3-nm-thick Al layer reaches 19.75 cd.A-1 when driving voltage is 7 V, which is 1.24, 1.17 and 17.03 times larger than those of the devices with 2 nm, 4 nm and 5 nm Al layer, respectively. The device property reaches up to the level of corresponding conventional device. In addition, all inverted devices with electron injection Al layer show superior stability of color coordinate due to the adoption of co-evaporation emitting layer and BCP spacer-layer, and the color coordinate of the inverted device with 3-nm-thick Al layer only changes from (0.580 6, 0.405 6) to (0.532 8, 0.436 3) when driving voltage increases from 6 V to 10 V.
The inverted bottom-emitting organic light-emitting devices (IBOLEDs) were prepared, with the structure of ITO/Al (x nm)/LiF (1 nm)/Bphen (40 nm)/CBP:GIr1 (14%):R-4b (2%) (10 nm)/BCP (3 nm)/CBP:GIr1 (14%):R-4b (2%) (20 nm)/TCTA (10 nm)/NPB (40 nm)/MoO3 (40 nm)/Al (100 nm), where the thickness of electron injection layer Al (x) are 0 nm, 2 nm, 3 nm, 4 nm and 5 nm, respectively. In this paper, the electron injection condition and luminance properties of inverted devices were investigated by changing the thickness of Al layer in Al/LiF compound thin film. It turns out that the introduction of Al layer can improve electron injection of the devices dramatically. Furthermore, the device exerts lower driving voltage and higher current efficiency when the thickness of electron injection Al layer is 3 nm. For example, the current efficiency of the device with 3-nm-thick Al layer reaches 19.75 cd.A-1 when driving voltage is 7 V, which is 1.24, 1.17 and 17.03 times larger than those of the devices with 2 nm, 4 nm and 5 nm Al layer, respectively. The device property reaches up to the level of corresponding conventional device. In addition, all inverted devices with electron injection Al layer show superior stability of color coordinate due to the adoption of co-evaporation emitting layer and BCP spacer-layer, and the color coordinate of the inverted device with 3-nm-thick Al layer only changes from (0.580 6, 0.405 6) to (0.532 8, 0.436 3) when driving voltage increases from 6 V to 10 V.
2018, 14(3):195-199.
DOI: https://doi.org/10.1007/s11801-018-7267-y
Abstract:
Silver nanowires (AgNWs) with diameter of 90—150 nm and length of 20—50 μm were successfully synthesized by a polyol process. Graphene oxide (GO) was prepared by Hummers method, and was reduced with strong hydrazine hy-drate at room temperature. The flexible transparent conductive films (TCFs) were fabricated using the mixed cellulose eater (MCE) as matrix and AgNWs and reduced graphene oxide (rGO) as conductive fillers by the improved vacuum fil-tration process. Then, the optical, electrical and mechanical properties of the AgNWs-rGO films were investigated. The results show that for the AgNWs-rGO film produced with the deposition densities of AgNWs and rGO as 110 mg·m-2 and 55 mg·m-2, the optical transmission at 550 nm is 88.4% with Rs around 891 Ω·sq-1, whereas the optical transmission for the AgNWs-rGO film with deposition densities of AgNWs and rGO of 385 mg·m-2 and 55 mg·m-2 is 79.0% at 550 nm with Rs around 9.6 Ω·sq-1. There is little overt increase in Rs of the AgNWS-rGO film after tape tests for 200 times. The bending test results indicate that the change in Rs of AgNWs-MCE film is less than 2% even after 200 cycles of compressive or tensile bending. The excellent mechanical properties of the AgNWs-rGO film can be attributed to the burying of AgNWs and rGO at the surface of MCE
Silver nanowires (AgNWs) with diameter of 90—150 nm and length of 20—50 μm were successfully synthesized by a polyol process. Graphene oxide (GO) was prepared by Hummers method, and was reduced with strong hydrazine hy-drate at room temperature. The flexible transparent conductive films (TCFs) were fabricated using the mixed cellulose eater (MCE) as matrix and AgNWs and reduced graphene oxide (rGO) as conductive fillers by the improved vacuum fil-tration process. Then, the optical, electrical and mechanical properties of the AgNWs-rGO films were investigated. The results show that for the AgNWs-rGO film produced with the deposition densities of AgNWs and rGO as 110 mg·m-2 and 55 mg·m-2, the optical transmission at 550 nm is 88.4% with Rs around 891 Ω·sq-1, whereas the optical transmission for the AgNWs-rGO film with deposition densities of AgNWs and rGO of 385 mg·m-2 and 55 mg·m-2 is 79.0% at 550 nm with Rs around 9.6 Ω·sq-1. There is little overt increase in Rs of the AgNWS-rGO film after tape tests for 200 times. The bending test results indicate that the change in Rs of AgNWs-MCE film is less than 2% even after 200 cycles of compressive or tensile bending. The excellent mechanical properties of the AgNWs-rGO film can be attributed to the burying of AgNWs and rGO at the surface of MCE
2018, 14(3):200-203.
DOI: https://doi.org/10.1007/s11801-018-8014-0
Abstract:
We theoretically propose the structure of all-solid photonic band gap fiber to support LP01 mode at 980 nm and four modes of LP01, LP11, LP02 and LP21 at C bands. We research characteristics of the photonic band gab and apply photonic fiber as gain medium to fiber lasers, finally realizing separate amplification of LP11 and LP21 high-order modes with single-mode stable pump light of LP01 mode at 980 nm. Besides, we realize LP11 mode and LP21 mode laser output of dual-wavelength.
We theoretically propose the structure of all-solid photonic band gap fiber to support LP01 mode at 980 nm and four modes of LP01, LP11, LP02 and LP21 at C bands. We research characteristics of the photonic band gab and apply photonic fiber as gain medium to fiber lasers, finally realizing separate amplification of LP11 and LP21 high-order modes with single-mode stable pump light of LP01 mode at 980 nm. Besides, we realize LP11 mode and LP21 mode laser output of dual-wavelength.
2018, 14(3):204-208.
DOI: https://doi.org/10.1007/s11801-018-7251-6
Abstract:
Based on the optical fiber mode theory and employing the expertized software COMSOL, we study the effect of ellipticity and misalignment on the effective refractive indices, walk-off and intensity distribution of the even and odd eigenmodes that form the basis of the orbital angular momentum (OAM) modes in a ring fiber. Our results show that the effective refractive index difference and the walk-off increase with the ellipticity and misalignment, thus reducing the stability of the OAM modes. We find that the misalignment has a greater impact on the OAM modes than the ellipticity, and both the misalignment and ellipticity affect the lower-order OAM modes more significantly, suggesting that the higher-order OAM modes are more stable during propagation.
Based on the optical fiber mode theory and employing the expertized software COMSOL, we study the effect of ellipticity and misalignment on the effective refractive indices, walk-off and intensity distribution of the even and odd eigenmodes that form the basis of the orbital angular momentum (OAM) modes in a ring fiber. Our results show that the effective refractive index difference and the walk-off increase with the ellipticity and misalignment, thus reducing the stability of the OAM modes. We find that the misalignment has a greater impact on the OAM modes than the ellipticity, and both the misalignment and ellipticity affect the lower-order OAM modes more significantly, suggesting that the higher-order OAM modes are more stable during propagation.
2018, 14(3):209-211.
DOI: https://doi.org/10.1007/s11801-018-7210-2
Abstract:
Aiming at the problem of the peak to average power ratio (PAPR) in coherent optical orthogonal frequency division multiplexing (CO-OFDM), a hybrid PAPR reduction technique of the CO-OFDM system by combining iterative partial transmit sequence (IPTS) scheme with modified clipping and filtering (MCF) is proposed. The simulation results show that at the complementary cumulative distribution function (CCDF) of 10-4, the PAPR of proposed scheme is optimized by 1.86 dB and 2.13 dB compared with those of IPTS and CF schemes, respectively. Meanwhile, when the bit error rate (BER) is 10-3, the optical signal to noise ratio (OSNR) are optimized by 1.57 dB and 0.66 dB compared with those of CF and IPTS-CF schemes, respectively.
Aiming at the problem of the peak to average power ratio (PAPR) in coherent optical orthogonal frequency division multiplexing (CO-OFDM), a hybrid PAPR reduction technique of the CO-OFDM system by combining iterative partial transmit sequence (IPTS) scheme with modified clipping and filtering (MCF) is proposed. The simulation results show that at the complementary cumulative distribution function (CCDF) of 10-4, the PAPR of proposed scheme is optimized by 1.86 dB and 2.13 dB compared with those of IPTS and CF schemes, respectively. Meanwhile, when the bit error rate (BER) is 10-3, the optical signal to noise ratio (OSNR) are optimized by 1.57 dB and 0.66 dB compared with those of CF and IPTS-CF schemes, respectively.
2018, 14(3):212-215.
DOI: https://doi.org/10.1007/s11801-018-7262-3
Abstract:
A novel optical fiber sensor with a U-shaped micro-groove structure ablated by femtosecond laser on single-mode fiber for measuring air relative humidity (RH) is reported in this paper. In order to improve the accuracy of sensor, a graphene oxide (GO)/polyvinyl alcohol (PVA) composite film is coated on the surface of micro-groove structure. In the U-shaped micro-groove structure, the remaining core and micro-cavity in the micro-groove make up two major optical propagation paths, forming a Mach-Zehnder interferometer (MZI). The sensor has a good linear response within the RH range of 30%—85%, and the maximum sensitivity can reach 0.638 1 nm/%RH. The effect of temperature on the overall performance of the humidity sensor is also investigated. As a new type of all-fiber device, the sensor shows excellent sensitivity and stability.
A novel optical fiber sensor with a U-shaped micro-groove structure ablated by femtosecond laser on single-mode fiber for measuring air relative humidity (RH) is reported in this paper. In order to improve the accuracy of sensor, a graphene oxide (GO)/polyvinyl alcohol (PVA) composite film is coated on the surface of micro-groove structure. In the U-shaped micro-groove structure, the remaining core and micro-cavity in the micro-groove make up two major optical propagation paths, forming a Mach-Zehnder interferometer (MZI). The sensor has a good linear response within the RH range of 30%—85%, and the maximum sensitivity can reach 0.638 1 nm/%RH. The effect of temperature on the overall performance of the humidity sensor is also investigated. As a new type of all-fiber device, the sensor shows excellent sensitivity and stability.
2018, 14(3):216-219.
DOI: https://doi.org/10.1007/s11801-018-7250-7
Abstract:
In the air-water quantum key distribution (QKD), the irregular sea surface has some influence on the photon polarization state. The wind is considered as the main factor causing the irregularity, so the model of irregular sea surface based on the wind speed is adopted. The relationships of the quantum bit error rate with the wind speed and the initial incident angle are simulated. Therefore, the maximum secure transmission depth of QKD is confirmed, and the limitation of the wind speed and the initial incident angle is determined. The simulation results show that when the wind speed and the initial incident angle increase, the performance of QKD will fall down. Under the intercept-resend attack condition, the maximum safe transmission depth of QKD is up to 105 m. To realize safe communications in the safe diving depth of submarines (100 m), the initial incident angle is requested to be not exceeding 26°, and with the initial incident angle increased, the limitation of wind speed is decreased.
In the air-water quantum key distribution (QKD), the irregular sea surface has some influence on the photon polarization state. The wind is considered as the main factor causing the irregularity, so the model of irregular sea surface based on the wind speed is adopted. The relationships of the quantum bit error rate with the wind speed and the initial incident angle are simulated. Therefore, the maximum secure transmission depth of QKD is confirmed, and the limitation of the wind speed and the initial incident angle is determined. The simulation results show that when the wind speed and the initial incident angle increase, the performance of QKD will fall down. Under the intercept-resend attack condition, the maximum safe transmission depth of QKD is up to 105 m. To realize safe communications in the safe diving depth of submarines (100 m), the initial incident angle is requested to be not exceeding 26°, and with the initial incident angle increased, the limitation of wind speed is decreased.
14 Study on polarized optical flow algorithm for imaging bionic polarization navigation micro sensor
2018, 14(3):220-225.
DOI: https://doi.org/10.1007/s11801-018-8004-2
Abstract:
At present, both the point source and the imaging polarization navigation devices only can output the angle information, which means that the velocity information of the carrier cannot be extracted from the polarization field pattern directly. Optical flow is an image-based method for calculating the velocity of pixel point movement in an image. However, for ordinary optical flow, the difference in pixel value as well as the calculation accuracy can be reduced in weak light. Polarization imaging technology has the ability to improve both the detection accuracy and the recognition probability of the target because it can acquire the extra polarization multi-dimensional information of target radiation or reflection. In this paper, combining the polarization imaging technique with the traditional optical flow algorithm, a polarization optical flow algorithm is proposed, and it is verified that the polarized optical flow algorithm has good adaptation in weak light and can improve the application range of polarization navigation sensors. This research lays the foundation for day and night all-weather polarization navigation applications in future.
At present, both the point source and the imaging polarization navigation devices only can output the angle information, which means that the velocity information of the carrier cannot be extracted from the polarization field pattern directly. Optical flow is an image-based method for calculating the velocity of pixel point movement in an image. However, for ordinary optical flow, the difference in pixel value as well as the calculation accuracy can be reduced in weak light. Polarization imaging technology has the ability to improve both the detection accuracy and the recognition probability of the target because it can acquire the extra polarization multi-dimensional information of target radiation or reflection. In this paper, combining the polarization imaging technique with the traditional optical flow algorithm, a polarization optical flow algorithm is proposed, and it is verified that the polarized optical flow algorithm has good adaptation in weak light and can improve the application range of polarization navigation sensors. This research lays the foundation for day and night all-weather polarization navigation applications in future.
2018, 14(3):226-231.
DOI: https://doi.org/10.1007/s11801-018-7268-x
Abstract:
In order to suppress complex mixing noise in low-illumination images for wide-area search of nighttime sea surface, a model based on total variation (TV) and split Bregman is proposed in this paper. A fidelity term based on L1 norm and a fidelity term based on L2 norm are designed considering the difference between various noise types, and the regularization mixed first-order TV and second-order TV are designed to balance the influence of details information such as texture and edge for sea surface image. The final detection result is obtained by using the high-frequency component solved from L1 norm and the low-frequency component solved from L2 norm through wavelet transform. The experimental results show that the proposed denoising model has perfect denoising performance for artificially degraded and low-illumination images, and the result of image quality assessment index for the denoising image is superior to that of the contrastive models.
In order to suppress complex mixing noise in low-illumination images for wide-area search of nighttime sea surface, a model based on total variation (TV) and split Bregman is proposed in this paper. A fidelity term based on L1 norm and a fidelity term based on L2 norm are designed considering the difference between various noise types, and the regularization mixed first-order TV and second-order TV are designed to balance the influence of details information such as texture and edge for sea surface image. The final detection result is obtained by using the high-frequency component solved from L1 norm and the low-frequency component solved from L2 norm through wavelet transform. The experimental results show that the proposed denoising model has perfect denoising performance for artificially degraded and low-illumination images, and the result of image quality assessment index for the denoising image is superior to that of the contrastive models.
2018, 14(3):232-235.
DOI: https://doi.org/10.1007/s11801-018-7254-3
Abstract:
According to the classic Karush-Kuhn-Tucker (KKT) theorem, at every step of incremental support vector machine (SVM) learning, the newly adding sample which violates the KKT conditions will be a new support vector (SV) and migrate the old samples between SV set and non-support vector (NSV) set, and at the same time the learning model should be updated based on the SVs. However, it is not exactly clear at this moment that which of the old samples would change between SVs and NSVs. Additionally, the learning model will be unnecessarily updated, which will not greatly increase its accuracy but decrease the training speed. Therefore, how to choose the new SVs from old sets during the incremental stages and when to process incremental steps will greatly influence the accuracy and efficiency of incremental SVM learning. In this work, a new algorithm is proposed to select candidate SVs and use the wrongly predicted sample to trigger the incremental processing simultaneously. Experimental results show that the proposed algorithm can achieve good performance with high efficiency, high speed and good accuracy.
According to the classic Karush-Kuhn-Tucker (KKT) theorem, at every step of incremental support vector machine (SVM) learning, the newly adding sample which violates the KKT conditions will be a new support vector (SV) and migrate the old samples between SV set and non-support vector (NSV) set, and at the same time the learning model should be updated based on the SVs. However, it is not exactly clear at this moment that which of the old samples would change between SVs and NSVs. Additionally, the learning model will be unnecessarily updated, which will not greatly increase its accuracy but decrease the training speed. Therefore, how to choose the new SVs from old sets during the incremental stages and when to process incremental steps will greatly influence the accuracy and efficiency of incremental SVM learning. In this work, a new algorithm is proposed to select candidate SVs and use the wrongly predicted sample to trigger the incremental processing simultaneously. Experimental results show that the proposed algorithm can achieve good performance with high efficiency, high speed and good accuracy.
2018, 14(3):236-240.
DOI: https://doi.org/10.1007/s11801-018-7261-4
Abstract:
The negativity of Wigner function (WF) is one of the important symbols of non-classical properties of light field. Therefore, it is of great significance to study the evolution of WF in dissipative process. The evolution formula of WF in laser process under the action of linear resonance force is given by virtue of thermo entangled state representation and the technique of integration within an ordered product of operator. As its application, the evolution of WF of thermal field and that of single-photon-added coherent state are discussed. The results show that the WF of thermal field maintains its original character. On the other hand, the negative region size and the depth of negativity of WF of single-photon-added coherent state decrease until it vanishes with dissipation. This shows that the non-classical property of single-photon-added coherent state is weakened, until it disappears with dissipation time increasing.
The negativity of Wigner function (WF) is one of the important symbols of non-classical properties of light field. Therefore, it is of great significance to study the evolution of WF in dissipative process. The evolution formula of WF in laser process under the action of linear resonance force is given by virtue of thermo entangled state representation and the technique of integration within an ordered product of operator. As its application, the evolution of WF of thermal field and that of single-photon-added coherent state are discussed. The results show that the WF of thermal field maintains its original character. On the other hand, the negative region size and the depth of negativity of WF of single-photon-added coherent state decrease until it vanishes with dissipation. This shows that the non-classical property of single-photon-added coherent state is weakened, until it disappears with dissipation time increasing.
External Links