Pub Date : 2021-10-01DOI: 10.1109/NaNA53684.2021.00028
Ahmed Salem, Huihui Wu, Xiaohong Jiang
This paper investigates the parameters of covert communications in a finite blocklength regime. The optimal hypothesis test in covert communications is directly related to the total variation distance (TVD), where obtaining it becomes crucial when practical covert communication is considered. The current literature provides different TVD calculation methods that are inaccurate, especially when the blocklength n is large. We provide an exact evaluation for TVD theoretically based on the incomplete Gamma function. We also provide an accurate approach to calculate TVD when n is small (n ≤ 140). Extensive numerical results were provided to verify the accuracy of our proposed expressions.
{"title":"Exact Evaluation of Total Variation Distance in Covert Communications","authors":"Ahmed Salem, Huihui Wu, Xiaohong Jiang","doi":"10.1109/NaNA53684.2021.00028","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00028","url":null,"abstract":"This paper investigates the parameters of covert communications in a finite blocklength regime. The optimal hypothesis test in covert communications is directly related to the total variation distance (TVD), where obtaining it becomes crucial when practical covert communication is considered. The current literature provides different TVD calculation methods that are inaccurate, especially when the blocklength n is large. We provide an exact evaluation for TVD theoretically based on the incomplete Gamma function. We also provide an accurate approach to calculate TVD when n is small (n ≤ 140). Extensive numerical results were provided to verify the accuracy of our proposed expressions.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123469593","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00069
Xue Wang, Hao Zhang, Kaijun Wu
In attribute-based access control services, there are problems such as cumbersome policy control, prone to authorization conflicts, and conflicts caused by complex attribute structures are not easy to identify. Due to the difficulty of the conflict detection problem, most of the detection methods have strict requirements for policy structure. Normally, the priority strategy are chosen uniformly when using the directed acyclic graph approach to disambiguate conflict rule pairs. The present methods are not thorough, flexible and user-friendly for the policy design in practical applications. To address these problems, an access control policy conflict detection algorithm based on intersection of target expression trees under the XACML (eXtensible Access Control Markup Language) specification is proposed. The method efficiently locates the conflict rule pairs based on the index structure through policy tree and rule effects, determines the conflict by expression comparison n, and marks the possible causes of the conflict, provides analysis of the disambiguation scheme, and achieves access control with fine granularity.
在基于属性的访问控制服务中，存在策略控制繁琐、容易发生授权冲突、属性结构复杂导致的冲突不易识别等问题。由于冲突检测问题的难度，大多数检测方法对策略结构都有严格的要求。通常使用有向无环图方法对冲突规则对进行消歧时，优先级策略是统一选择的。目前的方法对于实际应用中的政策设计不够彻底、灵活和人性化。针对这些问题，提出了一种基于XACML (eXtensible access control Markup Language，可扩展访问控制标记语言)规范下目标表达式树交集的访问控制策略冲突检测算法。该方法通过策略树和规则效果有效地定位基于索引结构的冲突规则对，通过表达式比较n确定冲突，并标记冲突的可能原因，提供消歧方案分析，实现细粒度访问控制。
{"title":"Expression Tree-based Policy Conflict Detection Algorithm","authors":"Xue Wang, Hao Zhang, Kaijun Wu","doi":"10.1109/NaNA53684.2021.00069","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00069","url":null,"abstract":"In attribute-based access control services, there are problems such as cumbersome policy control, prone to authorization conflicts, and conflicts caused by complex attribute structures are not easy to identify. Due to the difficulty of the conflict detection problem, most of the detection methods have strict requirements for policy structure. Normally, the priority strategy are chosen uniformly when using the directed acyclic graph approach to disambiguate conflict rule pairs. The present methods are not thorough, flexible and user-friendly for the policy design in practical applications. To address these problems, an access control policy conflict detection algorithm based on intersection of target expression trees under the XACML (eXtensible Access Control Markup Language) specification is proposed. The method efficiently locates the conflict rule pairs based on the index structure through policy tree and rule effects, determines the conflict by expression comparison n, and marks the possible causes of the conflict, provides analysis of the disambiguation scheme, and achieves access control with fine granularity.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124250124","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00017
Ze Yang, Youliang Tian
At present, the research on shares in the t-out-of-n secret sharing scheme mainly focuses on the delimitation of share length, and there is no specific quantitative method for the security of authorized shares t in secret sharing without considering any application scenarios. In this paper, we presents a method to quantity the capabilities of the attacker and the defender under the t-out-of-n secret sharing scheme. More specifically, we present a new general notion of limitation to provide a precise bound of attack and defense capability under the $(t, n)$-threshold sharing scheme. Furthermore, we discuss the best selection of authorized shares t. Firstly, we transform the attack and defense under the $(t, n)$-threshold secret sharing scheme into the communication problem in information theory, and establish the attack channel and defense channel from the point of view of the attacker and the defender. Moreover, we describe the capability of both attacking and defending by introducing average mutual information, and analyze the limitation of the capability of both sides, as well as the maximum value of the average mutual information is the channel capacity. Finally, according to the average amount of mutual information of the change curve under different scenarios of secret sharing, we compare and analyze the influence of the capability limitation of the attacker and the defender on the value of t, and get the best selection of t.
{"title":"Towards Attack and Defense Views to (t, n)-threshold Secret Sharing Scheme Using Information Theory","authors":"Ze Yang, Youliang Tian","doi":"10.1109/NaNA53684.2021.00017","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00017","url":null,"abstract":"At present, the research on shares in the t-out-of-n secret sharing scheme mainly focuses on the delimitation of share length, and there is no specific quantitative method for the security of authorized shares t in secret sharing without considering any application scenarios. In this paper, we presents a method to quantity the capabilities of the attacker and the defender under the t-out-of-n secret sharing scheme. More specifically, we present a new general notion of limitation to provide a precise bound of attack and defense capability under the $(t, n)$-threshold sharing scheme. Furthermore, we discuss the best selection of authorized shares t. Firstly, we transform the attack and defense under the $(t, n)$-threshold secret sharing scheme into the communication problem in information theory, and establish the attack channel and defense channel from the point of view of the attacker and the defender. Moreover, we describe the capability of both attacking and defending by introducing average mutual information, and analyze the limitation of the capability of both sides, as well as the maximum value of the average mutual information is the channel capacity. Finally, according to the average amount of mutual information of the change curve under different scenarios of secret sharing, we compare and analyze the influence of the capability limitation of the attacker and the defender on the value of t, and get the best selection of t.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"30 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125881465","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00037
Gaofeng Hong, Qili Wen, Wei Su
The 5G non-standalone (NSA) network deployment based on the legacy Evolved Packet Core (EPC) emerges a denser access cells scenario. Due to the complex mobility characteristics of vehicles and the diversified vehicular service requirements, the traditional cellular handover mechanism may not maintain high-quality network service for various connected vehicles. This paper concentrates on satisfying the network requirements of different vehicular services, keeping network load balance and avoiding unnecessary handover under the EPC-based LTE-5G RAN-level network architecture. We first develop the modified heterogeneous cellular network architecture with the assistance of the multi-access edge computing (MEC) technology, the MEC server works as a coordinator which is responsible for handover state information management and executing network selection algorithm to help vehicles access the most suitable candidate network. The network selection algorithm uses a quality of service (QoS) coverage conversion methods which calculate the QoS boundary of candidate networks to satisfy the specific service requirement and balance the network load. A long short-term memory (LSTM)-based trajectory prediction method is designed to obtain the sojourn time of a vehicle staying in the QoS boundary of candidate networks. The sojourn time acts as a vital reference for the handover decision. Further, we simplify handover signaling interaction to reduce the handover delay. Simulation results show that the proposed scheme is validated in improving network handover performance and the QoS of users under several metrics.
{"title":"A Modified Vehicular Handover Scheme in Non-standalone 5G Networks With the Assistance of Multi-access Edge Computing","authors":"Gaofeng Hong, Qili Wen, Wei Su","doi":"10.1109/NaNA53684.2021.00037","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00037","url":null,"abstract":"The 5G non-standalone (NSA) network deployment based on the legacy Evolved Packet Core (EPC) emerges a denser access cells scenario. Due to the complex mobility characteristics of vehicles and the diversified vehicular service requirements, the traditional cellular handover mechanism may not maintain high-quality network service for various connected vehicles. This paper concentrates on satisfying the network requirements of different vehicular services, keeping network load balance and avoiding unnecessary handover under the EPC-based LTE-5G RAN-level network architecture. We first develop the modified heterogeneous cellular network architecture with the assistance of the multi-access edge computing (MEC) technology, the MEC server works as a coordinator which is responsible for handover state information management and executing network selection algorithm to help vehicles access the most suitable candidate network. The network selection algorithm uses a quality of service (QoS) coverage conversion methods which calculate the QoS boundary of candidate networks to satisfy the specific service requirement and balance the network load. A long short-term memory (LSTM)-based trajectory prediction method is designed to obtain the sojourn time of a vehicle staying in the QoS boundary of candidate networks. The sojourn time acts as a vital reference for the handover decision. Further, we simplify handover signaling interaction to reduce the handover delay. Simulation results show that the proposed scheme is validated in improving network handover performance and the QoS of users under several metrics.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121701103","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}
The widespread application of embedded devices has attracted great concern about device security issues. Silicon physical unclonable functions (PUFs) have been proven to be a low-cost and effective hardware-based solution to ensure the security of embedded devices. Among many schemes, DRAM-based PUF is an attracted option for the reason that DRAM is ubiquitous in embedded devices and has a large address space. However, the existing DRAM PUF schemes have some defects, such as low PUF response reliability and complex key post-processing operations. Therefore, this paper presents a hybrid DRAM PUF composed of PicoPUF and DRAM PUF to address these problems. We implement the proposed hybrid DRAM PUFs on Xilinx Kintex 7 FPGA board and validate the effectiveness of our scheme. The experimental results show that compared to current DRAM PUFs, the proposed scheme can generate PUF responses with improved reliability and reduce the key post-process procedure.
{"title":"Implementation and Analysis of Hybrid DRAM PUFs on FPGA","authors":"Yu Zheng, Zhao Huang, Liang Li, Changjian Xie, Quan Wang, Zili Wu","doi":"10.1109/NaNA53684.2021.00074","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00074","url":null,"abstract":"The widespread application of embedded devices has attracted great concern about device security issues. Silicon physical unclonable functions (PUFs) have been proven to be a low-cost and effective hardware-based solution to ensure the security of embedded devices. Among many schemes, DRAM-based PUF is an attracted option for the reason that DRAM is ubiquitous in embedded devices and has a large address space. However, the existing DRAM PUF schemes have some defects, such as low PUF response reliability and complex key post-processing operations. Therefore, this paper presents a hybrid DRAM PUF composed of PicoPUF and DRAM PUF to address these problems. We implement the proposed hybrid DRAM PUFs on Xilinx Kintex 7 FPGA board and validate the effectiveness of our scheme. The experimental results show that compared to current DRAM PUFs, the proposed scheme can generate PUF responses with improved reliability and reduce the key post-process procedure.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132792528","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00032
Zhonghua Xie, Tao Tao, Lisheng Ma
Service provided by virtual machine hosted at cloud data center (DC) has become the mainstream mode. However, cloud data centers are facing the threat from natural disasters. For disaster-affected data centers, to ensure the sustainability of services, virtual machine hosted at such data centers should be migrated quickly to safe data centers. This paper investigates an emergency virtual machine online migration scheme where the migration operation is carried out within the early warning of disaster. In the proposed scheme, a virtual machine can be migrated through multiple paths and the migration data center need to meet the required bandwidth and delay of connection requests supported by the virtual machine. An Integer Linear Program (ILP) model is established to get the solutions for the proposed scheme. The objective of the ILP model is to maximize the number of virtual machines successfully migrated within the given early warning time and the constraints are the available network resources. Numerical results show that the proposed model can effectively implement emergency virtual machine online migration.
{"title":"Emergency Virtual Machine Online Migration in Cloud Data Centers","authors":"Zhonghua Xie, Tao Tao, Lisheng Ma","doi":"10.1109/NaNA53684.2021.00032","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00032","url":null,"abstract":"Service provided by virtual machine hosted at cloud data center (DC) has become the mainstream mode. However, cloud data centers are facing the threat from natural disasters. For disaster-affected data centers, to ensure the sustainability of services, virtual machine hosted at such data centers should be migrated quickly to safe data centers. This paper investigates an emergency virtual machine online migration scheme where the migration operation is carried out within the early warning of disaster. In the proposed scheme, a virtual machine can be migrated through multiple paths and the migration data center need to meet the required bandwidth and delay of connection requests supported by the virtual machine. An Integer Linear Program (ILP) model is established to get the solutions for the proposed scheme. The objective of the ILP model is to maximize the number of virtual machines successfully migrated within the given early warning time and the constraints are the available network resources. Numerical results show that the proposed model can effectively implement emergency virtual machine online migration.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129072892","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00081
Chun Jen Lin, Yan Luo, Liang-Min Wang
Deep Reinforcement Learning (DRL) has yielded proficient controllers for complex tasks. DRL trains machine learning models for decision making to maximize rewards in uncertain environments such as network function virtualization (NFV). However, when facing limited information, agents often have difficulties making decisions at some decision point. In a real-world NFV environment, we may have incomplete information about network flow patterns. Compared with complete information feedback, it increases the difficulty to predict an optimal policy since important state information is missing. In this paper, we formulate a Partially Observable Markov Decision Process (POMDP) with a partially unknown NFV system. To address the shortcomings in real-world NFV, we conduct an extensive simulation to investigate the effects of adding recurrency to a Proximal Policy optimization (PPO2) by replacing the first post-convolutional fully-connected layer with a recurrent LSTM or adding stacked frames as input. The results show that RL based schedulers using stacking a history of frames in the PPO2’s input layer can easily adapt at evaluation time if the quality of observations changes.
{"title":"Heterogeneous Flow Scheduling using Deep Reinforcement Learning in Partially Observable NFV Environment","authors":"Chun Jen Lin, Yan Luo, Liang-Min Wang","doi":"10.1109/NaNA53684.2021.00081","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00081","url":null,"abstract":"Deep Reinforcement Learning (DRL) has yielded proficient controllers for complex tasks. DRL trains machine learning models for decision making to maximize rewards in uncertain environments such as network function virtualization (NFV). However, when facing limited information, agents often have difficulties making decisions at some decision point. In a real-world NFV environment, we may have incomplete information about network flow patterns. Compared with complete information feedback, it increases the difficulty to predict an optimal policy since important state information is missing. In this paper, we formulate a Partially Observable Markov Decision Process (POMDP) with a partially unknown NFV system. To address the shortcomings in real-world NFV, we conduct an extensive simulation to investigate the effects of adding recurrency to a Proximal Policy optimization (PPO2) by replacing the first post-convolutional fully-connected layer with a recurrent LSTM or adding stacked frames as input. The results show that RL based schedulers using stacking a history of frames in the PPO2’s input layer can easily adapt at evaluation time if the quality of observations changes.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"393 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132671308","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00082
Xiangyi Lu, Qing Ren, Feng Tian
With the development of machine learning, the issue of privacy leakage has attracted much attention. Member inference attack is an attack method that threatens the privacy of training datasets. It uses the model’s behavior to infer whether the input user record belongs to the training datasets, and then get the user’s private information according to the purpose of the model. This paper studies the member inference attack under the black box model. We design a defense mechanism to make the learning model and the inference attack model learn from each other, and use the gains from the attack model to fine-tune the last layer’s parameters of the learning model. The fine-tuned learning model can reduce the gains from the membership inference attack with less loss of prediction accuracy. We use different datasets to evaluate the defense mechanism on deep neural networks. The results show that when the training accuracy and test accuracy of the learning model convergence are similar, the learning model only losses about 1% of the prediction accuracy, which the accuracy of the member inference attack drops by a maximum of around 20%.
{"title":"A Fine-tuning-based Adversarial Network for Member Privacy Preserving","authors":"Xiangyi Lu, Qing Ren, Feng Tian","doi":"10.1109/NaNA53684.2021.00082","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00082","url":null,"abstract":"With the development of machine learning, the issue of privacy leakage has attracted much attention. Member inference attack is an attack method that threatens the privacy of training datasets. It uses the model’s behavior to infer whether the input user record belongs to the training datasets, and then get the user’s private information according to the purpose of the model. This paper studies the member inference attack under the black box model. We design a defense mechanism to make the learning model and the inference attack model learn from each other, and use the gains from the attack model to fine-tune the last layer’s parameters of the learning model. The fine-tuned learning model can reduce the gains from the membership inference attack with less loss of prediction accuracy. We use different datasets to evaluate the defense mechanism on deep neural networks. The results show that when the training accuracy and test accuracy of the learning model convergence are similar, the learning model only losses about 1% of the prediction accuracy, which the accuracy of the member inference attack drops by a maximum of around 20%.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133454050","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00087
Lin An, Zhuo Wang, Jiahao Yue, Xiaoliang Ma
Various innovative applications of emerging mobile Internet have exploded in recent years, which brings huge challenges to terminal devices with limited CPU computing ability and battery capacity. The realization of high-performance computing offloading based on different optimization indicators (e.g., task delay and energy consumption) is currently a research hotspot in the field of mobile edge computing (MEC). This paper proposes a joint task offloading and resource allocation algorithm via proximal policy optimization for multiple terminal users and multiple MEC servers. The proposed algorithm designs the local task butter queues for terminal users and edge task butter queues for MEC servers, which allows the tasks to be executed on butter queues in a first-in-first-out way, leading to a precise calculation of waiting delays of tasks. Moreover, it formulates the objective optimization problem as the Markov decision process and employs the proximal policy optimization algorithm to minimize the weighted sum of the task delay and energy consumption. Simulation results show the proposed algorithm outperforms the baselines with better performance.
{"title":"Joint Task Offloading and Resource Allocation via Proximal Policy Optimization for Mobile Edge Computing Network","authors":"Lin An, Zhuo Wang, Jiahao Yue, Xiaoliang Ma","doi":"10.1109/NaNA53684.2021.00087","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00087","url":null,"abstract":"Various innovative applications of emerging mobile Internet have exploded in recent years, which brings huge challenges to terminal devices with limited CPU computing ability and battery capacity. The realization of high-performance computing offloading based on different optimization indicators (e.g., task delay and energy consumption) is currently a research hotspot in the field of mobile edge computing (MEC). This paper proposes a joint task offloading and resource allocation algorithm via proximal policy optimization for multiple terminal users and multiple MEC servers. The proposed algorithm designs the local task butter queues for terminal users and edge task butter queues for MEC servers, which allows the tasks to be executed on butter queues in a first-in-first-out way, leading to a precise calculation of waiting delays of tasks. Moreover, it formulates the objective optimization problem as the Markov decision process and employs the proximal policy optimization algorithm to minimize the weighted sum of the task delay and energy consumption. Simulation results show the proposed algorithm outperforms the baselines with better performance.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"146 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133073885","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00044
Yanchun Kong, Weibin Su, Gang Xu
The traditional carbon monitoring system uses eddy covariance, remote sensing and geographic information system combined with artificial ground survey, which is difficult to achieve long-term, large-scale forest carbon measurement, and has uncertainty. Through statistics of carbon emission and absorption data, we design algorithms on edge computing nodes with intelligent analysis, real-time collection of forest resources, the type of ground vegetation, soil nutrients, carbon dioxide and meteorological data, the establishment of mountain forest carbon dioxide fuzzy measurement monitoring model, to estimate forest absorption. This project aims to improve the measurement accuracy of carbon neutralization in high mountain forest land, further densify the data acquisition from grid sample plots of forest, increase the number of sample plots, and reveal the spatiotemporal evolution law through the implementation of scientific and effective comprehensive real-time monitoring of carbon dioxide absorption of forest resources, so as to provide theoretical and technical support for the study of regional carbon peak estimation.
{"title":"Monitoring System of Carbon Neutralization Forestland in Plateau based on Edge Computing","authors":"Yanchun Kong, Weibin Su, Gang Xu","doi":"10.1109/NaNA53684.2021.00044","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00044","url":null,"abstract":"The traditional carbon monitoring system uses eddy covariance, remote sensing and geographic information system combined with artificial ground survey, which is difficult to achieve long-term, large-scale forest carbon measurement, and has uncertainty. Through statistics of carbon emission and absorption data, we design algorithms on edge computing nodes with intelligent analysis, real-time collection of forest resources, the type of ground vegetation, soil nutrients, carbon dioxide and meteorological data, the establishment of mountain forest carbon dioxide fuzzy measurement monitoring model, to estimate forest absorption. This project aims to improve the measurement accuracy of carbon neutralization in high mountain forest land, further densify the data acquisition from grid sample plots of forest, increase the number of sample plots, and reveal the spatiotemporal evolution law through the implementation of scientific and effective comprehensive real-time monitoring of carbon dioxide absorption of forest resources, so as to provide theoretical and technical support for the study of regional carbon peak estimation.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121124003","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：