Pub Date : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996136
B. Bhaskar, S. Veluchamy
Biometrics has wide applications in the fields of security and privacy. Since unimodal biometrics are subjected to various problems regarding recognition and security, multimodal biometrics have been used extensively nowadays for personal authentication. In this paper we have proposed an efficient personal identification system using two biometric identifiers, palm print and Inner knuckle print. In the recent years, palm prints and knuckle prints have overruled other biometric identifiers because of their unique, stable and novelty feature. The proposed feature extraction method for palm print is Monogenic Binary Coding (MBC), which is an efficient approach for extracting palm print features. Then for inner knuckle print recognition we have tried two algorithms named Ridgelet Transform and Scale Invariant Feature Transform (SIFT). Also we have compared their results in terms of recognition rate. We then adopt Support Vector Machine (SVM) for classifying the extracted feature vectors. Combining both knuckle print and palm print for personal identification will give better security and accuracy.
{"title":"Hand based multibiometric authentication using local feature extraction","authors":"B. Bhaskar, S. Veluchamy","doi":"10.1109/ICRTIT.2014.6996136","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996136","url":null,"abstract":"Biometrics has wide applications in the fields of security and privacy. Since unimodal biometrics are subjected to various problems regarding recognition and security, multimodal biometrics have been used extensively nowadays for personal authentication. In this paper we have proposed an efficient personal identification system using two biometric identifiers, palm print and Inner knuckle print. In the recent years, palm prints and knuckle prints have overruled other biometric identifiers because of their unique, stable and novelty feature. The proposed feature extraction method for palm print is Monogenic Binary Coding (MBC), which is an efficient approach for extracting palm print features. Then for inner knuckle print recognition we have tried two algorithms named Ridgelet Transform and Scale Invariant Feature Transform (SIFT). Also we have compared their results in terms of recognition rate. We then adopt Support Vector Machine (SVM) for classifying the extracted feature vectors. Combining both knuckle print and palm print for personal identification will give better security and accuracy.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127039129","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996159
M. Pallikonda Rajasekaran, M. Suresh, U. Dhanasekaran
Biometrics is the ID of humans utilizing intrinsic physical, biological, otherwise activity features, traits, or habits. Biometrics has the potential to provide this desired ability to clearly and discretely determine a person's identity with additional accuracy and security. Biometric systems primarily based on individual antecedent of advice which is referred as unimodal frameworks. Even though some unimodal frameworks (e.g. Palm, Finger impression, Face, Iris), have got significant change in consistency plus precision yet has experienced selection issues attributable to non-all-inclusiveness of biometrics attributes, vulnerability to biometric mocking and insufficient exactness created by boisterous information as their inconveniences. In future, single biometric framework might not be in a position to accomplish the wanted execution prerequisite in genuine world provisions. To defeat these issues, we have to utilize multimodal biometric confirmation frameworks which blend data from various modalities to make a choice. Multimodal biometric confirmation framework utilize use more than one human modalities such as face, iris, retina, sclera and fingerprint etc. to improve their security of the method. In this approach, combined the biometric traits of sclera and fingerprint for addressing authentication issues, which has not discussed and implemented earlier. The fusion of multimodal biometric system helps to reduce the system error rates. The ANFIS model consolidated the neural system versatile capacities and the fluffy rationale qualitative strategy will have low false dismissal degree contrasted with neural network and fluffy rationale qualitative frame work. The combination of multimodal biometric security conspires in the ANFIS will show higher accuracy come close with Neural Network and Fuzzy Inference System.
{"title":"Multimodal biometric recognition using sclera and fingerprint based on ANFIS","authors":"M. Pallikonda Rajasekaran, M. Suresh, U. Dhanasekaran","doi":"10.1109/ICRTIT.2014.6996159","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996159","url":null,"abstract":"Biometrics is the ID of humans utilizing intrinsic physical, biological, otherwise activity features, traits, or habits. Biometrics has the potential to provide this desired ability to clearly and discretely determine a person's identity with additional accuracy and security. Biometric systems primarily based on individual antecedent of advice which is referred as unimodal frameworks. Even though some unimodal frameworks (e.g. Palm, Finger impression, Face, Iris), have got significant change in consistency plus precision yet has experienced selection issues attributable to non-all-inclusiveness of biometrics attributes, vulnerability to biometric mocking and insufficient exactness created by boisterous information as their inconveniences. In future, single biometric framework might not be in a position to accomplish the wanted execution prerequisite in genuine world provisions. To defeat these issues, we have to utilize multimodal biometric confirmation frameworks which blend data from various modalities to make a choice. Multimodal biometric confirmation framework utilize use more than one human modalities such as face, iris, retina, sclera and fingerprint etc. to improve their security of the method. In this approach, combined the biometric traits of sclera and fingerprint for addressing authentication issues, which has not discussed and implemented earlier. The fusion of multimodal biometric system helps to reduce the system error rates. The ANFIS model consolidated the neural system versatile capacities and the fluffy rationale qualitative strategy will have low false dismissal degree contrasted with neural network and fluffy rationale qualitative frame work. The combination of multimodal biometric security conspires in the ANFIS will show higher accuracy come close with Neural Network and Fuzzy Inference System.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"183 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127040851","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996170
D. Jagadish
Collaborative learning is an online classroom can take the form of conversation between the whole classes or within smaller groups. Moodle (Modular Object-Oriented Dynamic Learning Environment) is a free and open source e-learning software platform, also known as a Learning Management System, or Virtual Learning Environment (VLE). As a web-based tool, Moodle offers the possible way to deliver courses which include an enormous variety of information sources - links to multimedia, websites and image - which are hard to deliver in a traditional teaching atmosphere. The converse (chat) activity module in moodle allows participants to encompass a realtime synchronous discussion in a moodle course. A teacher can organize users into groups within the course or within particular activities. This paper aims in efficient group formation of learners in a collaborative learning environment so that every individual in the group is benefitted. As a testing platform tenth standard Tamil text book is incorporated in to moodle. In this paper K-NN clustering algorithm is used to improve the group performance. This algorithm achieves good performance in terms of balancing the knowledge level among all the students.
{"title":"Grouping in collaborative e-learning environment based on interaction among students","authors":"D. Jagadish","doi":"10.1109/ICRTIT.2014.6996170","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996170","url":null,"abstract":"Collaborative learning is an online classroom can take the form of conversation between the whole classes or within smaller groups. Moodle (Modular Object-Oriented Dynamic Learning Environment) is a free and open source e-learning software platform, also known as a Learning Management System, or Virtual Learning Environment (VLE). As a web-based tool, Moodle offers the possible way to deliver courses which include an enormous variety of information sources - links to multimedia, websites and image - which are hard to deliver in a traditional teaching atmosphere. The converse (chat) activity module in moodle allows participants to encompass a realtime synchronous discussion in a moodle course. A teacher can organize users into groups within the course or within particular activities. This paper aims in efficient group formation of learners in a collaborative learning environment so that every individual in the group is benefitted. As a testing platform tenth standard Tamil text book is incorporated in to moodle. In this paper K-NN clustering algorithm is used to improve the group performance. This algorithm achieves good performance in terms of balancing the knowledge level among all the students.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122796318","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996121
Abinash Tripathy, S. Rath
Software Development Life Cycle (SDLC) starts with eliciting requirement of user as a document called Software Requirement Specification (SRS). SRS document is mostly written in the form of any natural language (NL) that is convenient for the client. In order to develop a right software based on user's requirements, the objects, methods and attributes needs to be identified from SRS document. In this paper, an attempt is made to develop a methodology, using the concept of Natural Language Processing (NLP) for Object Oriented (OO) Programming System analysis concept, by finding out the class name and its details directly form SRS.
{"title":"Application of Natural Language Processing in Object Oriented Software Development","authors":"Abinash Tripathy, S. Rath","doi":"10.1109/ICRTIT.2014.6996121","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996121","url":null,"abstract":"Software Development Life Cycle (SDLC) starts with eliciting requirement of user as a document called Software Requirement Specification (SRS). SRS document is mostly written in the form of any natural language (NL) that is convenient for the client. In order to develop a right software based on user's requirements, the objects, methods and attributes needs to be identified from SRS document. In this paper, an attempt is made to develop a methodology, using the concept of Natural Language Processing (NLP) for Object Oriented (OO) Programming System analysis concept, by finding out the class name and its details directly form SRS.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122868152","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996166
S. Kiruthiga, P. Kola Sujatha, A. Kannan
Social Networks (SN) are popular among the people to interact with their friends through the internet. Users spending their time in popular social networking sites like facebook, Myspace and twitter to share the personal information. Cloning attack is one of the insidious attacks in facebook. Usually attackers stole the images and personal information about a person and create the fake profile pages. Once the profile gets cloned they started to send a friend request using the cloned profile. Incase if the real users account gets blocked, they used to send a new friend request to their friends. At the same time cloned one also sending the request to the person. At that time it was hard to identify the real one for users. In the proposed system the clone attack is detected based on user action time period and users click pattern to find the similarity between the cloned profile and real one in facebook. Using Cosine similarity and Jaccard index the performance of the similarity between the users is improved.
{"title":"Detecting cloning attack in Social Networks using classification and clustering techniques","authors":"S. Kiruthiga, P. Kola Sujatha, A. Kannan","doi":"10.1109/ICRTIT.2014.6996166","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996166","url":null,"abstract":"Social Networks (SN) are popular among the people to interact with their friends through the internet. Users spending their time in popular social networking sites like facebook, Myspace and twitter to share the personal information. Cloning attack is one of the insidious attacks in facebook. Usually attackers stole the images and personal information about a person and create the fake profile pages. Once the profile gets cloned they started to send a friend request using the cloned profile. Incase if the real users account gets blocked, they used to send a new friend request to their friends. At the same time cloned one also sending the request to the person. At that time it was hard to identify the real one for users. In the proposed system the clone attack is detected based on user action time period and users click pattern to find the similarity between the cloned profile and real one in facebook. Using Cosine similarity and Jaccard index the performance of the similarity between the users is improved.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130959535","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996151
S. Anjanadevi, D. Vijayakumar, K. .. Srinivasagan
Cloud computing is an emerging, computing model wherein the tasks are allocated to software, combination of connections, and services accessed over a network. This connections and network of servers is collectively known as the cloud. In place of operating their own data centers, users might rent computing power and storage capacity from a service provider and pays only for what they use. Cloud storage is delivering the data storage as service. If the data is stored in cloud, it must provide the data access and heterogeneity. With the advances in cloud computing it allows storing of large number of images and data throughout the world. This paper proposes the indexing and metadata management which helps to access the distributed data with reduced latency. The metadata management can be enhanced for large scale file system applications. When designing the metadata, the storage location of the metadata and attributes is important for the efficient retrieval of the data. Indexes are used to quickly locate data without having to search over every location in storage. Based on these two models, the data can be easily fetched and the search time was reduced to retrieve the appropriate data.
{"title":"An efficient dynamic indexing and metadata based storage in cloud environment","authors":"S. Anjanadevi, D. Vijayakumar, K. .. Srinivasagan","doi":"10.1109/ICRTIT.2014.6996151","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996151","url":null,"abstract":"Cloud computing is an emerging, computing model wherein the tasks are allocated to software, combination of connections, and services accessed over a network. This connections and network of servers is collectively known as the cloud. In place of operating their own data centers, users might rent computing power and storage capacity from a service provider and pays only for what they use. Cloud storage is delivering the data storage as service. If the data is stored in cloud, it must provide the data access and heterogeneity. With the advances in cloud computing it allows storing of large number of images and data throughout the world. This paper proposes the indexing and metadata management which helps to access the distributed data with reduced latency. The metadata management can be enhanced for large scale file system applications. When designing the metadata, the storage location of the metadata and attributes is important for the efficient retrieval of the data. Indexes are used to quickly locate data without having to search over every location in storage. Based on these two models, the data can be easily fetched and the search time was reduced to retrieve the appropriate data.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133641606","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996200
M. Shankar, R. Senthilkumar
In the field of Data retrieval, accessing web resources is frequent task. This domain is shifting radically from the amplified data growth to the way in which it is structured and retrieved across web. This explosive growth of data is the result of billions of people using the Internet and mobile devices for commerce, entertainment, social interactions and as well as the Internet of things that constantly share machine-generated data. Even with lot of research, the task of analyzing this data to extract its business values with precision still remains as a trivial issue. To address this issue, the paper presents a novel Semantic Based Lesk Algorithm (SBLA), which traces the meaning of user defined tags and categorizes the web data by means of Support Vector Machine (SVM) classifier. On comparing with existing methods, the proposed method performs well in extraction of admissible data with the better accuracy and precision as discussed in result analysis.
{"title":"Harnessing the semantic analysis of tag using Semantic Based Lesk Algorithm","authors":"M. Shankar, R. Senthilkumar","doi":"10.1109/ICRTIT.2014.6996200","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996200","url":null,"abstract":"In the field of Data retrieval, accessing web resources is frequent task. This domain is shifting radically from the amplified data growth to the way in which it is structured and retrieved across web. This explosive growth of data is the result of billions of people using the Internet and mobile devices for commerce, entertainment, social interactions and as well as the Internet of things that constantly share machine-generated data. Even with lot of research, the task of analyzing this data to extract its business values with precision still remains as a trivial issue. To address this issue, the paper presents a novel Semantic Based Lesk Algorithm (SBLA), which traces the meaning of user defined tags and categorizes the web data by means of Support Vector Machine (SVM) classifier. On comparing with existing methods, the proposed method performs well in extraction of admissible data with the better accuracy and precision as discussed in result analysis.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"30 5","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132869878","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996174
M. Sumalatha, C. Selvakumar, T. Priya, R. T. Azariah, P. Manohar
In cloud computing, remote based massive data storage and dynamic computation services are provided to the users. The cloud enables the user to complete their tasks using pay-as-you-go cost model which typically works on the incurred virtual machine hours, so reducing the execution time will minimize the computational cost. Therefore the scheduler should bring maximum throughput in order to achieve effective resource allocation in cloud. Hence, in this work, DBPS (Deadline Based Pre-emptive Scheduling) and a TLBC (Throttled Load Balancing for Cloud) load balancing model based on cloud partitioning using virtual machine has been proposed. Workload prediction is done using statistics and training set, so that error tolerance can be achieved in TLBC. The preliminary results obtained when measuring performance based on the computational cost of the task set and the number of tasks executed in a particular time shows the proposed TLBC outperforms compared with existing systems. OpenNebula has been used as the cloud management tool for doing real time analysis and improving performance.
在云计算中，为用户提供基于远程的海量数据存储和动态计算服务。云允许用户使用按需付费的成本模型来完成他们的任务，这种模型通常在产生的虚拟机小时上工作，因此减少执行时间将使计算成本最小化。因此，调度器应该带来最大的吞吐量，以便在云中实现有效的资源分配。因此，本文提出了一种基于虚拟机的基于云分区的DBPS (Deadline Based preemptive Scheduling)和TLBC (throttledloadbalancing for Cloud)负载均衡模型。利用统计数据和训练集对工作负载进行预测，从而实现TLBC的容错性。根据任务集的计算成本和在特定时间内执行的任务数量对性能进行测量时获得的初步结果表明，与现有系统相比，建议的TLBC性能更好。OpenNebula被用作云管理工具，用于进行实时分析和提高性能。
{"title":"CLBC - Cost effective load balanced resource allocation for partitioned cloud system","authors":"M. Sumalatha, C. Selvakumar, T. Priya, R. T. Azariah, P. Manohar","doi":"10.1109/ICRTIT.2014.6996174","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996174","url":null,"abstract":"In cloud computing, remote based massive data storage and dynamic computation services are provided to the users. The cloud enables the user to complete their tasks using pay-as-you-go cost model which typically works on the incurred virtual machine hours, so reducing the execution time will minimize the computational cost. Therefore the scheduler should bring maximum throughput in order to achieve effective resource allocation in cloud. Hence, in this work, DBPS (Deadline Based Pre-emptive Scheduling) and a TLBC (Throttled Load Balancing for Cloud) load balancing model based on cloud partitioning using virtual machine has been proposed. Workload prediction is done using statistics and training set, so that error tolerance can be achieved in TLBC. The preliminary results obtained when measuring performance based on the computational cost of the task set and the number of tasks executed in a particular time shows the proposed TLBC outperforms compared with existing systems. OpenNebula has been used as the cloud management tool for doing real time analysis and improving performance.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130663596","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996189
P. Selvaperumal, A. Suruliandi
Twitter users tweet their views in the form of short text messages. Twitter topic classification is classifying the tweets in to a set of predefined classes. In this work, a new tweet classification Method that makes use of tweet features like URL's in the tweet, retweeted tweets and influential users tweet is proposed. Experiments were carried out with extensive tweet data set. The performance of the proposed algorithm in classifying the tweets was compared with the text classification algorithms like SVM, Naïve Bayes, KNN etc. It is observed that the proposed method outclasses the conventional text classification algorithms in classifying the tweets.
{"title":"A short message classification algorithm for tweet classification","authors":"P. Selvaperumal, A. Suruliandi","doi":"10.1109/ICRTIT.2014.6996189","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996189","url":null,"abstract":"Twitter users tweet their views in the form of short text messages. Twitter topic classification is classifying the tweets in to a set of predefined classes. In this work, a new tweet classification Method that makes use of tweet features like URL's in the tweet, retweeted tweets and influential users tweet is proposed. Experiments were carried out with extensive tweet data set. The performance of the proposed algorithm in classifying the tweets was compared with the text classification algorithms like SVM, Naïve Bayes, KNN etc. It is observed that the proposed method outclasses the conventional text classification algorithms in classifying the tweets.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114713257","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 : 2014-04-10DOI: 10.1109/ICRTIT.2014.6996178
Arijit Mallik, S. Hegde
Modern data center networks (DCNs) often use multi-rooted topologies, which offer multipath capability, for increased bandwidth and fault tolerance. However, traditional routing algorithms for the Internet have no or limited support for multipath routing, and cannot fully utilize available bandwidth in such DCNs. As a result, they route all the traffic through a single path, and thus form congestion. Multipath (MP) routing might be a good alternative, but is not sufficient alone to handle congestion that comes from the contention of end stations. Dynamic load balancing, on the other hand, protects the network from sudden congestions which could be caused by load spikes or link failures. However, little work has been done to incorporate all these features in a single and comprehensive solution for Data Center Ethernet (DCE). In this paper, we propose a novel method that attempts to integrate dynamic load balancing, multi-path scheme with congestion control (CC), with the use of pure Software-Defined-Networking (SDN) approach. SDN decouples control plane from the data forwarding plane, which reduces the overheads of the network switches. The major objectives that our solution attempts to achieve are, efficient utilization of network resources, high throughput and minimal frame loss.
{"title":"A novel proposal to effectively combine multipath data forwarding for data center networks with congestion control and load balancing using Software-Defined Networking Approach","authors":"Arijit Mallik, S. Hegde","doi":"10.1109/ICRTIT.2014.6996178","DOIUrl":"https://doi.org/10.1109/ICRTIT.2014.6996178","url":null,"abstract":"Modern data center networks (DCNs) often use multi-rooted topologies, which offer multipath capability, for increased bandwidth and fault tolerance. However, traditional routing algorithms for the Internet have no or limited support for multipath routing, and cannot fully utilize available bandwidth in such DCNs. As a result, they route all the traffic through a single path, and thus form congestion. Multipath (MP) routing might be a good alternative, but is not sufficient alone to handle congestion that comes from the contention of end stations. Dynamic load balancing, on the other hand, protects the network from sudden congestions which could be caused by load spikes or link failures. However, little work has been done to incorporate all these features in a single and comprehensive solution for Data Center Ethernet (DCE). In this paper, we propose a novel method that attempts to integrate dynamic load balancing, multi-path scheme with congestion control (CC), with the use of pure Software-Defined-Networking (SDN) approach. SDN decouples control plane from the data forwarding plane, which reduces the overheads of the network switches. The major objectives that our solution attempts to achieve are, efficient utilization of network resources, high throughput and minimal frame loss.","PeriodicalId":422275,"journal":{"name":"2014 International Conference on Recent Trends in Information Technology","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115912833","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：