POSTER

Content Delivery Networks (CDNs) deliver much of the world's web, video, and application content on the Internet today. A key component of a CDN is the mapping system that uses the DNS protocol to route each client's request to a ``proximal'' server that serves the requested content. While traditional mapping systems identify a client using the IP of its name server, we describe our experience in building and rolling-out a novel system called end-user mapping that identifies the client directly by using a prefix of the client's IP address. Using measurements from Akamai's production network during the roll-out, we show that end-user mapping provides significant performance benefits for clients who use public resolvers, including an eight-fold decrease in mapping distance, a two-fold decrease in RTT and content download time, and a 30% improvement in the time-to-first byte. We also quantify the scaling challenges in implementing end-user mapping such as the 8-fold increase in DNS queries. Finally, we show that a CDN with a larger number of deployment locations is likely to benefit more from end-user mapping than a CDN with a smaller number of deployments.

Chapter 14 - Proxy Functions of Microsoft Internet Security and Acceleration Server 2006

A web server log is a data file created by a server. Many servers collect data about the information requested for the web server, examples are date, time, client IP address, recommendation, user agent, and so on. This data will help identify specific issues on the web. Log files are stored in the web server, web proxy server, and the client browser. This paper provides a detailed overview of weblog files, analytics tools, and reviews of existing research.KeywordsWeb serverWeb analytics toolsUser behavior

Port knocking is a form of host-to-host communication which relies on deliberately failed TCP connection attempts. The information is encoded into a port sequence. The client attempts to initiate several three-way-handshakes and receives no reply. These connection attempts are monitored by a daemon which interprets their destination port numbers as data. This mechanism has vulnerabilities that can be exploited by hackers with the help of data sniffed off the network. Through synchronization, these vulnerabilities can be minimized. A knock sequence is less vulnerable to replay and brute force attacks if its lifespan is shorter. All the entities involved in the knock sequence need to be aware all the time of the knock sequence that can be used. For this, it is required that clients and server share the same time. Both possess the means of the determining the sequence, which consists of a hash based on a preshared key, time value, client IP address and destination port. We present the means to achieve the client-server synchronization and describe an application that implements this.

Natural Language Processing (NLP) and related technologies are critical for the success of many Internet applications such as digital libraries, e-commerce and customer service. This talk presents some recent research efforts and trends of four sets of NLP technologies for Internet applications. First, neural language model has been a very popular research direction in the last a few years that serves as the foundation of many NLP technologies and has significantly improved the performance of many applications; Second, machine translation techniques have been substantially advanced to better bridge the language barriers for many Internet applications; Third, the identification of inappropriate Internet text information (e.g., pornographic content) is a challenging research topic due to the diversified text representation; Fourth, machine reading comprehension has been become an important question and answering technology to directly satisfy information needs of many Internet users. These technologies will be discussed with examples from large-scale real-world applications.

Behavior-based tracking is an unobtrusive technique that allows observers to monitor user activities on the Internet over long periods of time -- in spite of changing IP addresses. Previous work has employed supervised classifiers in order to link the sessions of individual users. However, classifiers need labeled training sessions, which are difficult to obtain for observers. In this paper we show how this limitation can be overcome with an unsupervised learning technique. We present a modified k-means algorithm and evaluate it on a realistic dataset that contains the Domain Name System (DNS) queries of 3,862 users. For this purpose, we simulate an observer that tries to track all users, and an Internet Service Provider that assigns a different IP address to every user on every day. The highest tracking accuracy is achieved within the subgroup of highly active users. Almost all sessions of 73% of the users in this subgroup can be linked over a period of 56 days. 19% of the highly active users can be traced completely, i.e., all their sessions are assigned to a single cluster. This fraction increases to 40% for shorter periods of seven days. As service providers may engage in behavior-based tracking to complement their existing profiling efforts, it constitutes a severe privacy threat for users of online services. Users can defend against behavior-based tracking by changing their IP address frequently, but this is cumbersome at the moment.

As the most popular application in Internet, the performance of web applications are drawing more attention of users and Internet service providers. In order to monitor and analyze user perceived web performance in real network of China, a group of web performance parameters are used which can reflect high layer protocol performance directly impacting user feeling when visiting webpages. A distributed active monitoring system was used to monitor web performance on real network of main ISPs in China. Based on monitoring data, different aspects of user perceived web performance are analyzed in comparison with users in different cities, target websites in different distance, idle and busy times, weekdays and weekend, different steps in web process. Some practical valuable conclusions are concluded.

Visualizing and characterizing DNS lookup behaviors via log-mining

Hadoop Distributed File System (HDFS) has been widely adopted to support Internet applications because of its reliable, scalable and low-cost storage capability. Blue Sky, one of the most popular e-Learning resource sharing systems in China, is utilizing HDFS to store massive courseware. However, due to the inefficient access mechanism of HDFS, access latency of reading files from HDFS significantly impacts the performance of processing user requests. This paper introduces a two-level correlation based file prefetching approach, taking the characteristics of HDFS into consideration, to improve performance by reducing access latency. Four placement patterns to store prefetched data are presented, with policies to achieve trade-off between performance and efficiency of HDFS prefetching. Moreover, a dynamic replica selection algorithm is investigated to improve the efficiency of HDFS prefetching. The proposed prefetching approach has been implemented in Blue Sky, and experimental results prove that correlation based file prefetching can significantly reduce access latency therefore improve performance of Hadoop-based Internet applications.

Several medium-to-large companies are currently in the process of using external hosting to deploy their Internet applications. External hosting by such application service providers (ASPs) as IBM provides a low-cost, secure and reliable way for companies to deploy e-commerce enterprise applications without the need to purchase costly infrastructure. However, while designing software for applications that run on the servers of external hosts, performance may suffer if the data and the application reside at two different locations. In this paper, we provide a technique for improving the performance in such applications. This technique is implemented in large industrial software systems. We provide the software architecture and preliminary performance results.

Network distance, measured as round-trip latency between hosts, is important for the performance of many Internet applications. For example, nearest server selection and proximity routing in peer-to-peer networks rely on the ability to select nodes based on inter-host latencies. This paper presents PCoord, a decentralized network coordinate system for Internet distance prediction. In PCoord, the network is modeled as a D-dimensional geometric space; each host computes its coordinates in this geometric space to characterize its network location based on a small number of peer-to-peer network measurements. The goal is to embed hosts in the geometric space so that the Euclidean distance between two hosts’ coordinates accurately predicts their actual inter-host network latency. PCoord constructs network coordinates in a fully decentralized fashion. We present several mechanisms in PCoord to stabilize the system convergence. Our simulation results using real Internet measurements suggest that, even under an extremely challenging flash-crowd scenario where 1740 hosts simultaneously join the system, PCoord with a 5-dimensional Euclidean model is able to converge to 11% median prediction error in 10 coordinate updates per host on average.

Many recent developments justify how social and political participation through new media and information and communication technology is an urgent matter for many developed countries [...]

For Web-based applications in Internet of Things (IoT) environment, the IETF CoRE Working Group developed the Constrained Application Protocol (CoAP). A limitation of CoAP is that it does not address a mobility management for CoAP sensor nodes, which may move between different IP domains of a wireless sensor network. In this paper, we propose the CoAP-based Mobility Management Protocol (CoMP), which can provide mobility management for mobile CoAP sensor nodes. CoMP keeps track of the current IP addresses of mobile sensor and enables sensing data to be reliably delivered to the Web clients, using both HTTP and CoAP. The idea is that CoMP employs a separate location management function to support CoAP service mobility, and two operating modes, CoMP holding and binding modes, are used for a reliable data transmission during handover. To do this, we have designed the architecture of CoMP. Furthermore, we have designed the detailed signaling flow, the message format and algorithm of CoMP for the mobility support. Finally, the numerical analysis and simulation results of the handover latency and packet loss show the benefits of the proposed protocol.

In the last fifteen years, services deployed via the Web have become a part of our everyday lives. Supporting Web technologies have also evolved very quickly to respond to the new services and features demanded by mobile and always connected users. Currently, the demand for services on the Web is still growing at a very fast pace as well as new techniques, methods and cutting-edge technologies that make the Web an increasingly complex environment.

Although the mobile RFID system has various applications in Internet of Things(IoT), it has a limited communication range and raises some serious privacy and security problems. To improve the issuers above, we propose a new mobile RFID network framework based on security and privacy protection. In the framework, we design a Reader Network using ad hoc mechanism, that readers with short communication range could extend their communication distance for meeting more application service requirement. Meanwhile, we introduce TPM(Trusted Platform Module) to reader node to ensure the security and privacy of node and give the detailed trusted and privacy mechanisms based on TPM. These mechanisms ensure the node secure and privacy.