Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864905
Orhun Koc, Alp Demirel, E. Aydin, Fatmagul Ibisoglu, Sevval Izel Solmaz, Kaan Ari, Ayse Idman, Ismail Uyanik
Animals successfully perform many behavioral tasks within the framework of a closed-loop sensorimotor control system during their daily lives. To achieve this, animals receive sensory signals from their environment through various sensory receptors and process these signals in their central nervous systems (CNS). Then, using this sensory feedback, animals produce necessary motor signals and transmit them to their muscles to perform the desired behavior. During this process, animals integrate sensory information perceived by different sensory receptors and they simultaneously stimulate multiple muscle combinations. The goal of this study is to identify the closed-loop sensorimotor control processes of animals during their unconstrained behaviors. To achieve this, we built a novel experimental setup that allows data-driven system identification of the target tracking behavior of zebrafish during rheotaxis. In that, a stimulus target oscillating in the frequency range of 0-2 Hz was presented to the zebrafish. Then, frequency response of the target tracking performance for N=5 fish were estimated.
{"title":"System Identification of the Target Tracking Behavior of Zebrafish During Rheotaxis","authors":"Orhun Koc, Alp Demirel, E. Aydin, Fatmagul Ibisoglu, Sevval Izel Solmaz, Kaan Ari, Ayse Idman, Ismail Uyanik","doi":"10.1109/SIU55565.2022.9864905","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864905","url":null,"abstract":"Animals successfully perform many behavioral tasks within the framework of a closed-loop sensorimotor control system during their daily lives. To achieve this, animals receive sensory signals from their environment through various sensory receptors and process these signals in their central nervous systems (CNS). Then, using this sensory feedback, animals produce necessary motor signals and transmit them to their muscles to perform the desired behavior. During this process, animals integrate sensory information perceived by different sensory receptors and they simultaneously stimulate multiple muscle combinations. The goal of this study is to identify the closed-loop sensorimotor control processes of animals during their unconstrained behaviors. To achieve this, we built a novel experimental setup that allows data-driven system identification of the target tracking behavior of zebrafish during rheotaxis. In that, a stimulus target oscillating in the frequency range of 0-2 Hz was presented to the zebrafish. Then, frequency response of the target tracking performance for N=5 fish were estimated.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124064673","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864756
Bilgeçağ Aydoğdu, Betül Ahat, A. Salah, Tuba Bircan
Measuring and monitoring immigrant integration is important for evidence based policy making. Mobile data proposes a valuable opportunity for developing fine grained indicators of segregation and integration of sub-populations. Due to privacy and ethical reasons, mobile phone data is rarely shared with researchers. In this study, we use anonymous and aggregated data sets prepared using mobile positioning data (XDR) collected by Turkcell throughout 2020 with the purpose of calculating different indicators of segregation for refugees in order to understand the temporal and spatial variation of these indicators. In this study, isolation index and dissimilarity index were calculated and compared in terms of their measure of migrant segregation.
{"title":"Temporal and spatial analysis of indicators on segregation of Syrian refugees in Turkey with mobile phone data","authors":"Bilgeçağ Aydoğdu, Betül Ahat, A. Salah, Tuba Bircan","doi":"10.1109/SIU55565.2022.9864756","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864756","url":null,"abstract":"Measuring and monitoring immigrant integration is important for evidence based policy making. Mobile data proposes a valuable opportunity for developing fine grained indicators of segregation and integration of sub-populations. Due to privacy and ethical reasons, mobile phone data is rarely shared with researchers. In this study, we use anonymous and aggregated data sets prepared using mobile positioning data (XDR) collected by Turkcell throughout 2020 with the purpose of calculating different indicators of segregation for refugees in order to understand the temporal and spatial variation of these indicators. In this study, isolation index and dissimilarity index were calculated and compared in terms of their measure of migrant segregation.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132537803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864901
M. Elamassie, M. Uysal
In an outdoor environment, the transmitted optical signal is affected by the atmospheric particles that may scatter or absorb light such as drizzle, rain, snow, fog, and aerosols. Various earlier studies have attempted to provide models that accurately describe the attenuation caused by the propagation of light waves through different atmospheric conditions. While the rain attenuation models have received significant research efforts, the modeling of drizzle attenuation was not yet fully studied. In this paper, we conduct extensive simulations in MODTRAN to model drizzle attenuation over the wavelength from 350 nm to 1550 nm. Based on data fitting to simulation results, we further propose a closed-form expression for the drizzle extinction coefficient.
{"title":"Drizzle Attenuation Model for Visible and Infrared Wavelengths","authors":"M. Elamassie, M. Uysal","doi":"10.1109/SIU55565.2022.9864901","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864901","url":null,"abstract":"In an outdoor environment, the transmitted optical signal is affected by the atmospheric particles that may scatter or absorb light such as drizzle, rain, snow, fog, and aerosols. Various earlier studies have attempted to provide models that accurately describe the attenuation caused by the propagation of light waves through different atmospheric conditions. While the rain attenuation models have received significant research efforts, the modeling of drizzle attenuation was not yet fully studied. In this paper, we conduct extensive simulations in MODTRAN to model drizzle attenuation over the wavelength from 350 nm to 1550 nm. Based on data fitting to simulation results, we further propose a closed-form expression for the drizzle extinction coefficient.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134150484","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864784
Besra Cetindere, Cenk Albayrak, K. Türk
Visible light communication (VLC) is a short-range wireless optical communication that can transmit data by switching lighting elements at high speeds in indoor areas. In common areas, VLC can provide data security at every layer of communication by using physical layer security (PLS) techniques as well as existing cryptography-based techniques. In the literature, PLS techniques have generally been studied for monochrome VLC systems, and multicolor VLC studies are quite limited. In this study, to the best of authors’ knowledge, null steering (NS) and artificial noise (AN), which are widely used PLS methods, have been applied to multi-colored LED-based VLC systems for the first time in the literature and the achievable secrecy rate has been calculated.
{"title":"Physical Layer Security For Indoor Multicolor Visible Light Communication","authors":"Besra Cetindere, Cenk Albayrak, K. Türk","doi":"10.1109/SIU55565.2022.9864784","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864784","url":null,"abstract":"Visible light communication (VLC) is a short-range wireless optical communication that can transmit data by switching lighting elements at high speeds in indoor areas. In common areas, VLC can provide data security at every layer of communication by using physical layer security (PLS) techniques as well as existing cryptography-based techniques. In the literature, PLS techniques have generally been studied for monochrome VLC systems, and multicolor VLC studies are quite limited. In this study, to the best of authors’ knowledge, null steering (NS) and artificial noise (AN), which are widely used PLS methods, have been applied to multi-colored LED-based VLC systems for the first time in the literature and the achievable secrecy rate has been calculated.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133885101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864992
Fatma Ceren Yücel, Meryem Maras, Talat Kepezkaya, Elif Nur Ayvaz, A. Özen
In this study, it is suggested to use artificial intelligence assisted fuzzy logic based LMS (F-LMS) algorithm to improve the performance of single carrier (SC) underwater acoustic communication (UWAC) systems in multipath underwater acoustic channel environments. Numerical simulation studies are carried out to compare the proposed F-LMS algorithm with the bit error rate (BER) and mean square error (MSE) performance measures over decision feedback equalizer (DFE) and channel matched filter DFE (CMF-DFE). From the produced numerical results, it is understood that the best gains are achieved in both MSE and BER simulations with the proposed F-LMS algorithm. In addition to these, the most important contribution of the proposed study is to eliminate the error floor that occurs in DFE equalizers in BER simulations with CMF-DFE equalizers.
{"title":"Artificial Intelligence Based Under Water Acoustic Channel Equalizer Design","authors":"Fatma Ceren Yücel, Meryem Maras, Talat Kepezkaya, Elif Nur Ayvaz, A. Özen","doi":"10.1109/SIU55565.2022.9864992","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864992","url":null,"abstract":"In this study, it is suggested to use artificial intelligence assisted fuzzy logic based LMS (F-LMS) algorithm to improve the performance of single carrier (SC) underwater acoustic communication (UWAC) systems in multipath underwater acoustic channel environments. Numerical simulation studies are carried out to compare the proposed F-LMS algorithm with the bit error rate (BER) and mean square error (MSE) performance measures over decision feedback equalizer (DFE) and channel matched filter DFE (CMF-DFE). From the produced numerical results, it is understood that the best gains are achieved in both MSE and BER simulations with the proposed F-LMS algorithm. In addition to these, the most important contribution of the proposed study is to eliminate the error floor that occurs in DFE equalizers in BER simulations with CMF-DFE equalizers.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128936494","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864882
Okan Özdemir
FPGAs are frequently used in embedded systems thanks to their flexibility and parallel processing power. One usage of this flexibility is to design the TCP/IP stack elements inside the FPGA for their speed benefits. In this work, an IP block which provides communication between FPGA and user via "User Datagram Protocol (UDP)" is designed. This block can send and receive datagrams from a remote UDP socket. The analysis shows that up to 966 Mbit/s throughput can be reached using the design. This way, the Gigabit Ethernet interface of the embedded system can be used without consuming any microcontroller resource.
{"title":"UDP Usage in FPGA","authors":"Okan Özdemir","doi":"10.1109/SIU55565.2022.9864882","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864882","url":null,"abstract":"FPGAs are frequently used in embedded systems thanks to their flexibility and parallel processing power. One usage of this flexibility is to design the TCP/IP stack elements inside the FPGA for their speed benefits. In this work, an IP block which provides communication between FPGA and user via \"User Datagram Protocol (UDP)\" is designed. This block can send and receive datagrams from a remote UDP socket. The analysis shows that up to 966 Mbit/s throughput can be reached using the design. This way, the Gigabit Ethernet interface of the embedded system can be used without consuming any microcontroller resource.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133095919","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864828
Safak Ozturk, Yunus Emre Esin, O. Ozdil, Yunus Can Bilge
Increasing air pollution, greenhouse gases and the emission of industrial toxic gases can cause serious pollution and climate changes. For this reason, the detection of gases is one of the most important remote sensing topics today. Gas detection and classification using hyperspectral sensor data is one of the methods that can be used for this purpose. In this study, an exemplary method for the detection of butane gas using hyperspectral LWIR data is described..
{"title":"Butane Gas Detection in Hyperspectral LWIR Data","authors":"Safak Ozturk, Yunus Emre Esin, O. Ozdil, Yunus Can Bilge","doi":"10.1109/SIU55565.2022.9864828","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864828","url":null,"abstract":"Increasing air pollution, greenhouse gases and the emission of industrial toxic gases can cause serious pollution and climate changes. For this reason, the detection of gases is one of the most important remote sensing topics today. Gas detection and classification using hyperspectral sensor data is one of the methods that can be used for this purpose. In this study, an exemplary method for the detection of butane gas using hyperspectral LWIR data is described..","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133715533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864800
Onur Kaymak, Furkan Sen, Furkan Yüksel, Ismail Uyanik, S. K. Yildiz
The main objective of this study is to manufacture a soft robotic antenna with bending and strain sensing properties. The ultimate goal is to achieve wall following using the feedback obtained from this antenna. The design of the soft robotic antenna is inspired from the movement of the cockroach and the interaction of its antenna with its environment. To that end, a conductive carbon grease is injected in a silicon rubber substrate to obtain a structure with bending and strain sensing properties. It is aimed that the resulting sensor performs its measuring capabilities even under various conditions such as under water or debris where commonly used sensors based on infrared and ultrasonic principles usually fail to function properly.
{"title":"A Biologically Inspired Soft Robotic Antenna for Insect-like Robots","authors":"Onur Kaymak, Furkan Sen, Furkan Yüksel, Ismail Uyanik, S. K. Yildiz","doi":"10.1109/SIU55565.2022.9864800","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864800","url":null,"abstract":"The main objective of this study is to manufacture a soft robotic antenna with bending and strain sensing properties. The ultimate goal is to achieve wall following using the feedback obtained from this antenna. The design of the soft robotic antenna is inspired from the movement of the cockroach and the interaction of its antenna with its environment. To that end, a conductive carbon grease is injected in a silicon rubber substrate to obtain a structure with bending and strain sensing properties. It is aimed that the resulting sensor performs its measuring capabilities even under various conditions such as under water or debris where commonly used sensors based on infrared and ultrasonic principles usually fail to function properly.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124592514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864975
Elif Değirmenci, Ilker Özçelik, A. Yazıcı
Adversarial attacks aim to deceive the target system. Recently, deep learning methods have become a target of adversarial attacks. Even small perturbations could lead to classification errors in deep learning models. In an intrusion detection system using deep learning method, adversarial attack can cause classification error and malicious traffic can be classified as benign. In this study, the effects of adversarial attacks on the accuracy of deep learning-based intrusion detection systems were examined. CICIDS2017 dataset was used to test the detection systems . At first, DDoS attacks weredetected using Autoencoder, MLP, AEMLP, DNN, AEDNN, CNN and AECNN methods. Then, the Fast Gradient Sign Method (FGSM) is used to perform adversarial attacks. Finally, the sensitivity of the methods against the adversarial attacks were examined. Our results showed that the classification performance of deep learning based detection methods decreased up to %17 after the adversarial attacks. The results obtained in this study form the basis for the validation and validation studies of learning-based intrusion detection systems.
{"title":"The Effects of Autoencoders on the Robustness of Deep Learning Models","authors":"Elif Değirmenci, Ilker Özçelik, A. Yazıcı","doi":"10.1109/SIU55565.2022.9864975","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864975","url":null,"abstract":"Adversarial attacks aim to deceive the target system. Recently, deep learning methods have become a target of adversarial attacks. Even small perturbations could lead to classification errors in deep learning models. In an intrusion detection system using deep learning method, adversarial attack can cause classification error and malicious traffic can be classified as benign. In this study, the effects of adversarial attacks on the accuracy of deep learning-based intrusion detection systems were examined. CICIDS2017 dataset was used to test the detection systems . At first, DDoS attacks weredetected using Autoencoder, MLP, AEMLP, DNN, AEDNN, CNN and AECNN methods. Then, the Fast Gradient Sign Method (FGSM) is used to perform adversarial attacks. Finally, the sensitivity of the methods against the adversarial attacks were examined. Our results showed that the classification performance of deep learning based detection methods decreased up to %17 after the adversarial attacks. The results obtained in this study form the basis for the validation and validation studies of learning-based intrusion detection systems.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116946541","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-15DOI: 10.1109/SIU55565.2022.9864685
Yusuf Eren, Buket Çolak Güvenç, Engin Cemal Menguc
This study introduces a novel online censoring-least mean square (OC-LMS) based acoustic feedback cancellation (AFC) system to eliminate undesired acoustic feedback that causes instability in hearing aids. The performance of the proposed OC-LMS based AFC system is tested on a sample feedback path measured from a behind-the-ear hearing aid and compared with its conventional LMS based version in terms of mean square deviation, misalignment, and added stable gain metrics. Simulation results shows that the proposed AFC system works at higher performance than its LMS based version with fewer data by censoring noninformative data thanks to its censoring mechanism. The proposed novel system will pave the way for the design of wearable hearing aids that work with both high performance and long battery life.
{"title":"Online Censoring Based Acoustic Feedback Cancellation for Wearable Hearing Aids","authors":"Yusuf Eren, Buket Çolak Güvenç, Engin Cemal Menguc","doi":"10.1109/SIU55565.2022.9864685","DOIUrl":"https://doi.org/10.1109/SIU55565.2022.9864685","url":null,"abstract":"This study introduces a novel online censoring-least mean square (OC-LMS) based acoustic feedback cancellation (AFC) system to eliminate undesired acoustic feedback that causes instability in hearing aids. The performance of the proposed OC-LMS based AFC system is tested on a sample feedback path measured from a behind-the-ear hearing aid and compared with its conventional LMS based version in terms of mean square deviation, misalignment, and added stable gain metrics. Simulation results shows that the proposed AFC system works at higher performance than its LMS based version with fewer data by censoring noninformative data thanks to its censoring mechanism. The proposed novel system will pave the way for the design of wearable hearing aids that work with both high performance and long battery life.","PeriodicalId":115446,"journal":{"name":"2022 30th Signal Processing and Communications Applications Conference (SIU)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117333492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：