Pub Date : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333883
Hodong Kim, Changhee Hahn, Junbeom Hur
Cache side-channel attack is a class of attacks to retrieve sensitive information from a system by exploiting shared resource in CPUs. As the attacks are delivered to wide range of environments from mobile systems to cloud recently, many detection strategies have been proposed. Since the conventional cache side-channel are likely to incur tremendous number of cache events, most of the previous detection mechanisms were designed to carefully monitor cache events. However, recently proposed attacks tend to incur less cache events during the attack. PRIME+ABORT attack, for example, leverages the Intel TSX instead of accessing cache to measure access time. Because of the characteristic, cache event based detection mechanisms may hardly distinguish the attack. In this paper, we conduct an in-depth analysis of the PRIME+ABORT attack to identify the other useful hardware events for detection rather than cache events. Based on our finding, we present a novel mechanism called PRIME+ABORT Detector to detect the PRIME+ABORT attack and demonstrate that the detection mechanism can achieve 99.5% success rates with 0.3% performance overhead.
{"title":"Real-time Detection of Cache Side-channel Attack Using Non-cache Hardware Events","authors":"Hodong Kim, Changhee Hahn, Junbeom Hur","doi":"10.1109/ICOIN50884.2021.9333883","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333883","url":null,"abstract":"Cache side-channel attack is a class of attacks to retrieve sensitive information from a system by exploiting shared resource in CPUs. As the attacks are delivered to wide range of environments from mobile systems to cloud recently, many detection strategies have been proposed. Since the conventional cache side-channel are likely to incur tremendous number of cache events, most of the previous detection mechanisms were designed to carefully monitor cache events. However, recently proposed attacks tend to incur less cache events during the attack. PRIME+ABORT attack, for example, leverages the Intel TSX instead of accessing cache to measure access time. Because of the characteristic, cache event based detection mechanisms may hardly distinguish the attack. In this paper, we conduct an in-depth analysis of the PRIME+ABORT attack to identify the other useful hardware events for detection rather than cache events. Based on our finding, we present a novel mechanism called PRIME+ABORT Detector to detect the PRIME+ABORT attack and demonstrate that the detection mechanism can achieve 99.5% success rates with 0.3% performance overhead.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"14 1","pages":"28-31"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87389574","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9334004
Jaehwan Kim, B. Derbel, Byung-Woo Hong
We present a potential application of the conventional scale-space theory to the estimation of optical flow in the deep learning framework. An unsupervised learning scheme for the computation of optical flow is integrated with a Gaussian scale space. The hierarchical propagation of intermediate estimations via a consecutive scales demonstrates a potential in the course of optimization leading to a better local minimum. The landscape of loss function associated with an optical flow problem in a neural network framework is highly complex and non-convex, which requires to guild the optimization path in such a way that a solution at a plateau region. The qualitative comparison of the optical flow solutions via a Gaussian scale-space provides the characteristics of solutions at different scales, thus provides a way to take into consideration of scales in further improving accuracy.
{"title":"Motion Estimation via Scale-Space in Unsupervised Deep Learning","authors":"Jaehwan Kim, B. Derbel, Byung-Woo Hong","doi":"10.1109/ICOIN50884.2021.9334004","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9334004","url":null,"abstract":"We present a potential application of the conventional scale-space theory to the estimation of optical flow in the deep learning framework. An unsupervised learning scheme for the computation of optical flow is integrated with a Gaussian scale space. The hierarchical propagation of intermediate estimations via a consecutive scales demonstrates a potential in the course of optimization leading to a better local minimum. The landscape of loss function associated with an optical flow problem in a neural network framework is highly complex and non-convex, which requires to guild the optimization path in such a way that a solution at a plateau region. The qualitative comparison of the optical flow solutions via a Gaussian scale-space provides the characteristics of solutions at different scales, thus provides a way to take into consideration of scales in further improving accuracy.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"4 1","pages":"730-731"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87525646","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333941
H. Otsuki, Eiji Kawai, Katsuyoshi Setoyama, H. Kimiyama, Katsuhiro Sebayashi, M. Maruyama
In this study, we propose an architecture for monitoring packets coming from a high-speed optical Ethernet network. Moreover, we implement a packet monitoring system adopting our proposed architecture using general PC-based equipment with a field-programmable gate array (FPGA)based network interface card (NIC). We also experimentally achieve a full line-rate processing capability for 100-Gbps Ethernet and examine its feasibility on 400-Gbps Ethernet.
{"title":"Parallel Monitoring Architecture for 100 Gbps and Beyond Optical Ethernet","authors":"H. Otsuki, Eiji Kawai, Katsuyoshi Setoyama, H. Kimiyama, Katsuhiro Sebayashi, M. Maruyama","doi":"10.1109/ICOIN50884.2021.9333941","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333941","url":null,"abstract":"In this study, we propose an architecture for monitoring packets coming from a high-speed optical Ethernet network. Moreover, we implement a packet monitoring system adopting our proposed architecture using general PC-based equipment with a field-programmable gate array (FPGA)based network interface card (NIC). We also experimentally achieve a full line-rate processing capability for 100-Gbps Ethernet and examine its feasibility on 400-Gbps Ethernet.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"18 1","pages":"358-360"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83593616","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333967
Jae-Woo Kim, C. I. Nwakanma, Dong-Seong Kim, Jae-Min Lee
This paper discusses intelligent edge computing technology using neuromorphic technology. Neuromorphic is a technology that uses pure hardware to implement intelligent systems, unlike traditional methods of implementing intelligent systems in a software manner using CPU or GPU hardware. In this paper, intelligent edge computing technology was introduced using NeuroEdge, one of the devices using Neurologic technology, and the performance was verified through a face recognition test. Results showed that using neuromorphic technology such as the NM500 chip saves the time needed for training systems and does not impose the burden of requiring many datasets for effective training.
{"title":"Intelligent Face Recognition on the Edge Computing using Neuromorphic Technology","authors":"Jae-Woo Kim, C. I. Nwakanma, Dong-Seong Kim, Jae-Min Lee","doi":"10.1109/ICOIN50884.2021.9333967","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333967","url":null,"abstract":"This paper discusses intelligent edge computing technology using neuromorphic technology. Neuromorphic is a technology that uses pure hardware to implement intelligent systems, unlike traditional methods of implementing intelligent systems in a software manner using CPU or GPU hardware. In this paper, intelligent edge computing technology was introduced using NeuroEdge, one of the devices using Neurologic technology, and the performance was verified through a face recognition test. Results showed that using neuromorphic technology such as the NM500 chip saves the time needed for training systems and does not impose the burden of requiring many datasets for effective training.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"35 1","pages":"514-516"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76443873","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333875
Hojoon Won, Younghan Kim
As the cloud infrastructure becomes more complex, the importance of fault detection technology is increasing. A machine learning-based fault detection technology is being used to overcome the limitations of the existing fault detection method through log analysis and threshold-based fault detection method. Machine learning-based fault detection methods are greatly influenced by features. In this paper, we introduce feature engineering techniques that can affect accuracy, and propose a method to improve the performance of fault detection models in cloud infrastructure through comparative analysis and verification of various feature analysis models.
{"title":"Performance Analysis of Machine Learning Based Fault Detection for Cloud Infrastructure","authors":"Hojoon Won, Younghan Kim","doi":"10.1109/ICOIN50884.2021.9333875","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333875","url":null,"abstract":"As the cloud infrastructure becomes more complex, the importance of fault detection technology is increasing. A machine learning-based fault detection technology is being used to overcome the limitations of the existing fault detection method through log analysis and threshold-based fault detection method. Machine learning-based fault detection methods are greatly influenced by features. In this paper, we introduce feature engineering techniques that can affect accuracy, and propose a method to improve the performance of fault detection models in cloud infrastructure through comparative analysis and verification of various feature analysis models.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"1 1","pages":"877-880"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83521698","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333874
Anwar Alyatama
This work extends the adaptive RSA algorithm that is based on the relative cost for solving the routing, modulation level and spectrum assignment (RMSA) for elastic optical network (EON). In planning and execution of EONs, RMSA is a viral aspect. Our proposed RMSA is rooted in the relative cost concept and evaluates the average effect upon the admission of a connection request at a given network state on a given set of resources. Besides, only if the minimal relative cost is less than the request’s value, a connection request is admitted. Simulation has been used to display the effectiveness of using relative cost RMSA for attaining higher fairness in EONs and lower lost revenue.
{"title":"Relative Cost Routing, Modulation and Spectrum Allocation in Elastic Optical Networks","authors":"Anwar Alyatama","doi":"10.1109/ICOIN50884.2021.9333874","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333874","url":null,"abstract":"This work extends the adaptive RSA algorithm that is based on the relative cost for solving the routing, modulation level and spectrum assignment (RMSA) for elastic optical network (EON). In planning and execution of EONs, RMSA is a viral aspect. Our proposed RMSA is rooted in the relative cost concept and evaluates the average effect upon the admission of a connection request at a given network state on a given set of resources. Besides, only if the minimal relative cost is less than the request’s value, a connection request is admitted. Simulation has been used to display the effectiveness of using relative cost RMSA for attaining higher fairness in EONs and lower lost revenue.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"24 1","pages":"127-131"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91153716","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333859
Seunghyun Lee, Changhee Joo
The Content Delivery Network (CDN) appears as a solution for the rapidly growing demand of Internet traffic. Through distributed surrogate servers, the CDN can manage higher traffic demand and improve the overall efficiency. Since the CDN is included in the conventional Internet traffic delivery chain, and becomes popular as a new passage between users and content providers, it starts playing a significant role in the market. In this paper, we investigate the strategies of the players in the Internet market with the CDN, taking into consideration the network factors that impacts on the players’ decision as well as the objective and the regulations.
{"title":"On the CDN Pricing Strategies in the Internet Traffic Delivery Chain","authors":"Seunghyun Lee, Changhee Joo","doi":"10.1109/ICOIN50884.2021.9333859","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333859","url":null,"abstract":"The Content Delivery Network (CDN) appears as a solution for the rapidly growing demand of Internet traffic. Through distributed surrogate servers, the CDN can manage higher traffic demand and improve the overall efficiency. Since the CDN is included in the conventional Internet traffic delivery chain, and becomes popular as a new passage between users and content providers, it starts playing a significant role in the market. In this paper, we investigate the strategies of the players in the Internet market with the CDN, taking into consideration the network factors that impacts on the players’ decision as well as the objective and the regulations.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"39 1","pages":"680-682"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73117682","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333984
Muhammad Diyan, Murad Khan, Zhenbo Cao, Bhagya Nathali Silva, Jihun Han, K. Han
The dynamic nature of the electricity market need an efficient energy management and control system to take perfect decisions accordingly. House hold appliances is the contemporary study being adopted to improve the performance and balance the fluctuation between power system and smart home. This article proposes an intelligent home energy management system (IHEMS) incorporated with a prediction model and optimization model. To address the uncertainty of future energy load and its cost, a suitable prediction model based on Bi-directional long short Term memory (Bi-LSTM) is contributed. In collaboration with the prediction model, an optimization model based on reinforcement learning is presented to schedule the home appliances by taking optimal decisions. To validate the performance of the proposed scheme, Intensive simulation is performed with adoptable, un-adoptable and manageable loads of household appliances. The results confirm that the proposed scheme address the problem of energy management for numerous appliances, reduce the total energy consumption with total energy bill and minimize the user comfort level.
{"title":"Intelligent Home Energy Management System based on Bi-directional Long-short Term Memory and Reinforcement Learning","authors":"Muhammad Diyan, Murad Khan, Zhenbo Cao, Bhagya Nathali Silva, Jihun Han, K. Han","doi":"10.1109/ICOIN50884.2021.9333984","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333984","url":null,"abstract":"The dynamic nature of the electricity market need an efficient energy management and control system to take perfect decisions accordingly. House hold appliances is the contemporary study being adopted to improve the performance and balance the fluctuation between power system and smart home. This article proposes an intelligent home energy management system (IHEMS) incorporated with a prediction model and optimization model. To address the uncertainty of future energy load and its cost, a suitable prediction model based on Bi-directional long short Term memory (Bi-LSTM) is contributed. In collaboration with the prediction model, an optimization model based on reinforcement learning is presented to schedule the home appliances by taking optimal decisions. To validate the performance of the proposed scheme, Intensive simulation is performed with adoptable, un-adoptable and manageable loads of household appliances. The results confirm that the proposed scheme address the problem of energy management for numerous appliances, reduce the total energy consumption with total energy bill and minimize the user comfort level.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"101 1","pages":"782-787"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73874162","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9333881
Tsuyoshi Arai, Y. Okabe, Yoshinori Matsumoto
CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) is commonly utilized as a technology for avoiding attacks to Web sites by bots. State-of-the-art CAPTCHAs vary in difficulty based on the client’s behavior, allowing for efficient bot detection without sacrificing simplicity. In this research, we focus on detecting bots by supervised machine learning from access-log time series in the past. We have analysed access logs to several Web services which are using a commercial cloud-based CAPTCHA service, Capy Puzzle CAPTCHA. Experiments show that bot detection in attacks over a month can be performed with high accuracy by precursory analysis of the access log in only the first day as training data. In addition, we have manually analyzed the data that are found to be False Positive in the discrimination results, and it is found that the proposed model actually detects access by bots, which had been overlooked in the first-stage manual discrimination of flags in preparation of training data.
{"title":"Precursory Analysis of Attack-Log Time Series by Machine Learning for Detecting Bots in CAPTCHA","authors":"Tsuyoshi Arai, Y. Okabe, Yoshinori Matsumoto","doi":"10.1109/ICOIN50884.2021.9333881","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9333881","url":null,"abstract":"CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) is commonly utilized as a technology for avoiding attacks to Web sites by bots. State-of-the-art CAPTCHAs vary in difficulty based on the client’s behavior, allowing for efficient bot detection without sacrificing simplicity. In this research, we focus on detecting bots by supervised machine learning from access-log time series in the past. We have analysed access logs to several Web services which are using a commercial cloud-based CAPTCHA service, Capy Puzzle CAPTCHA. Experiments show that bot detection in attacks over a month can be performed with high accuracy by precursory analysis of the access log in only the first day as training data. In addition, we have manually analyzed the data that are found to be False Positive in the discrimination results, and it is found that the proposed model actually detects access by bots, which had been overlooked in the first-stage manual discrimination of flags in preparation of training data.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"14 1","pages":"295-300"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73261793","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 : 2021-01-13DOI: 10.1109/ICOIN50884.2021.9334026
P. Aravinda, S. Sooriyaarachchi, C. Gamage, N. Kottege
This paper proposes a method for RSSI based indoor localization and tracking in cluttered environments using Deep Neural Networks. We implemented a real-time system to localize people using wearable active RF tags and RF receivers fixed in an industrial environment with high RF noise. The proposed solution is advantageous in analysing RSSI data in cluttered-indoor environments with the presence of human body attenuation, signal distortion, and environmental noise. Simulations and experiments on a hardware testbed demonstrated that receiver arrangement, number of receivers and amount of line of sight signals captured by receivers are important parameters for improving localization and tracking accuracy. The effect of RF signal attenuation through the person who carries the tag was combined with two neural network models trained with RSSI data pertaining to two walking directions. This method was successful in predicting the walking direction of the person.
{"title":"Optimization of RSSI based indoor localization and tracking to monitor workers in a hazardous working zone using Machine Learning techniques","authors":"P. Aravinda, S. Sooriyaarachchi, C. Gamage, N. Kottege","doi":"10.1109/ICOIN50884.2021.9334026","DOIUrl":"https://doi.org/10.1109/ICOIN50884.2021.9334026","url":null,"abstract":"This paper proposes a method for RSSI based indoor localization and tracking in cluttered environments using Deep Neural Networks. We implemented a real-time system to localize people using wearable active RF tags and RF receivers fixed in an industrial environment with high RF noise. The proposed solution is advantageous in analysing RSSI data in cluttered-indoor environments with the presence of human body attenuation, signal distortion, and environmental noise. Simulations and experiments on a hardware testbed demonstrated that receiver arrangement, number of receivers and amount of line of sight signals captured by receivers are important parameters for improving localization and tracking accuracy. The effect of RF signal attenuation through the person who carries the tag was combined with two neural network models trained with RSSI data pertaining to two walking directions. This method was successful in predicting the walking direction of the person.","PeriodicalId":6741,"journal":{"name":"2021 International Conference on Information Networking (ICOIN)","volume":"36 1","pages":"305-310"},"PeriodicalIF":0.0,"publicationDate":"2021-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73856849","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：