P2P Flows Research Articles

Efficient Training Support Vector Clustering With Appropriate Boundary Information

Due to the remarkable capability in handling arbitrary cluster shapes, support vector clustering (SVC) benefits data analysis in terms of data description. However, large-scale data such as network traffic frequently makes it suffer from highly intensive pricey computation and storage for solving the dual problem and storing the kernel matrix, respectively. Fortunately, support vectors which describe the clusters, in a sense, are expected in the boundaries. To tackle this issue, we propose an efficient training SVC with appropriate boundary information (ETSVC), which features excellent flexibility and scalability. In ETSVC, we first give a shrinkable boundary selection (SBS) method which collects appropriate boundaries while reducing redundancy and noise. Based on the boundary information, a redefined dual problem is then designed without scarifying the principle of SVC. Finally, we design a reformative solver (RSolver) to reformulate the training phase, which estimates the support vector function by solving the dual problem. Since only a subset of boundaries is employed for model training, theoretical analysis suggests that ETSVC reaches efficiency improvement and consumes much less memory if sacrificing efficiency to reduce storage consumption. Towards grouping P2P flows and large-scale intrusion traffic, as well as other non-traffic data, experimental results confirm that ETSVC could be applied to resources constrained platform while achieving comparable accuracies with the state-of-the-art methods.

Multilevel Bloom Filters for P2P Flows Identification Based on Cluster Analysis in Wireless Mesh Network

With the development of wireless mesh networks and distributed computing, lots of new P2P services have been deployed and enrich the Internet contents and applications. The rapid growth of P2P flows brings great pressure to the regular network operation. So the effective flow identification and management of P2P applications become increasingly urgent. In this paper, we build a multilevel bloom filters data structure to identify the P2P flows through researches on the locality characteristics of P2P flows. Different level structure stores different numbers of P2P flow rules. According to the characteristics values of the P2P flows, we adjust the parameters of the data structure of bloom filters. The searching steps of the scheme traverse from the first level to the last level. Compared with the traditional algorithms, our method solves the drawbacks of previous schemes. The simulation results demonstrate that our algorithm effectively enhances the performance of P2P flows identification. Then we deploy our flow identification algorithm in the traffic monitoring sensors which belong to the network traffic monitoring system at the export link in the campus network. In the real environment, the experiment results demonstrate that our algorithm has a fast speed and high accuracy to identify the P2P flows; therefore, it is suitable for actual deployment.

A NOVEL TOKEN BASED APPROACH TOWARDS PACKET LOSS CONTROL

Due to the advent of technologies like Web 2.0, the Internet applications are able to support transmis sion of multimedia content to end users. In such applications the transmission might result in packet loss as well. In this context, it is essential to have packet loss control mechanisms that can avoid deterioration of quality of services while rendering media rich cont ent. The quality of service in this case depends on congestion control. Many proto cols have been introduced in order to supplement th e standard TCP protocol in order to control network congestion. The CSFQ which was built for fair service with open ‐ loop contro ller has started deterioration in quality as P2P flows dominated Internet traffic of late. One of the closed loop congestion control known as Token-Based Congestion Control (TBCC) was able to restrict consuming resou rces and provide best service to end users. It moni tors inter-domain traffic for trust relationships. Recently, Shi et al. presented a new mechanism known as Stable Token-Limited Cong estion Control (STLCC) for controlling inter-domain congestion and improve net work performance. In this paper we implement the ST LCC mechanism. We built a prototype application that demonstrates the proof o f concept. The experimental results revealed that t he proposed application is able to control network congestion by controlling packet loss thus improving performance of network.

EPFIA: Extensible P2P Flows Identification Architecture

The fundament of managing P2P traffic is identifying various P2P flows accurately. Although many P2P flows identification methods are presented nowadays, there are no ideas for either integrating these independent methods together or being extended fast to support new method. In this work, an extensible P2P flows identification architecture (EPFIA for short) is proposed. In order to identify many specific P2P flows, EPFIA uses many different identification methods simultaneously, and obtains the highest efficiency via adjusting their identification sequence. An online mechanism of renewing identification methods is designed, which can extend new identification method without compiling the whole program. Applying policy mechanism, identification methods can be updated, started and halted remotely. The experiment results of running the prototype system show us that EPFIA could effectively promote the performance of system and support online renew P2P identification methods and manage them remotely.

Hidden Markov model-based P2P flow identification: a hidden Markov model-based P2P flow identification method

It is critical to identify various P2P flows accurately for managing P2P traffic as well as for ensuring network security. In this study, a hidden Markov model -based P2P flow identification (HMM-PFI) method is put forward, which makes use of packet size, the inter-arrival time and the arrival order of packets to build flow identification model, the packets in sample flow are associated with the states of hidden Markov model (HMM). Moreover, some discrete random variables are introduced to depict the characteristics of HMM state. The method dramatically decreases the time needed for building the model and improves the real-time and accuracy in identifying unknown flows. In HMM-PFI, various P2P flows can be identified simultaneously. Meanwhile, the algorithm for selecting the number of HMM state is designed. In a controllable experimental circumstance such as our campus network, HMM-PFI is utilised to identify P2P flows whereas several other identification methods are chosen as benchmark. The results show that HMM-PFI can correctly identify the packet flows produced by various P2P protocols and it is preferably adaptive to different network circumstance.

Packet Loss Control Using Tokens at the Network Edge

Presently, the Internet accommodates simultaneous audio, video, and data traffic. This requires the Internet to guarantee the packet loss which at its turn depends very much on congestion control. A series of protocols have been introduced to supplement the insufficient TCP mechanism controlling the network congestion. CSFQ was designed as an open-loop controller to provide the fair best effort service for supervising the per-flow bandwidth consumption and has become helpless when the P2P flows started to dominate the traffic of the Internet. Token-Based Congestion Control (TBCC) is based on a closed-loop congestion control principle, which restricts token resources consumed by an end-user and provides the fair best effort service with O(1) complexity. As Self-Verifying CSFQ and Re-feedback, it experiences a heavy load by policing inter-domain traffic for lack of trust. In this paper, Stable Token-Limited Congestion Control (STLCC) is introduced as new protocols which appends inter-domain congestion control to TBCC and make the congestion control system to be stable. STLCC is able to shape output and input traffic at the inter-domain link with O(1) complexity. STLCC produces a congestion index, pushes the packet loss to the network edge and improves the network performance. Finally, the simple version of STLCC is introduced. This version is deployable in the Internet without any IP protocols modifications and preserves also the packet datagram.

Host-Based P2P Flow Identification and Use in Real-Time

Data identification and classification is a key task for any Internet Service Provider (ISP) or network administrator. As port fluctuation and encryption become more common in P2P applications wishing to avoid identification, new strategies must be developed to detect and classify their flows. This article introduces a method of separating P2P and standard web traffic that can be applied as part of an offline data analysis process, based on the activity of the hosts on the network. Heuristics are analyzed and a classification system proposed that focuses on classifying those “long” flows that transfer most of the bytes across a network. The accuracy of the system is then tested using real network traffic from a core Internet router showing misclassification rates as low as 0.54% of flows in some cases. We expand on this proposed strategy to investigate its relevance to real-time, early classification problems. New proposals are made and the results of real-time experiments are compared to those obtained in the offline analysis. It is shown that classification accuracies in the real-time strategy are similar to those achieved in offline analysis with a large portion of the total web and P2P flows correctly identified.

Graption: A graph-based P2P traffic classification framework for the internet backbone

Monitoring network traffic and classifying applications are essential functions for network administrators. Current traffic classification methods can be grouped in three categories: (a) flow-based (e.g., packet sizing/timing features), (b) payload-based, and (c) host-based. Methods from all three categories have limitations, especially when it comes to detecting new applications, and classifying traffic at the backbone. In this paper, we propose the use of Traffic Dispersion Graphs (TDGs) to remedy these limitations. Given a set of flows, a TDG is a graph with an edge between any two IP addresses that communicate; thus TDGs capture network-wide interactions. Using TDGs, we develop an application classification framework dubbed Graption ( Graph-based classifica tion). Our framework provides a systematic way to classify traffic by using information from the network-wide behavior and flow-level characteristics of Internet applications. As a proof of concept, we instantiate our framework to detect P2P traffic, and show that it can identify 90% of P2P flows with 95% accuracy in backbone traces, which are particularly challenging for other methods.

Optimally designing caches to reduce P2P traffic

Traffic caused by P2P services dominates a large part of traffic on the Internet and imposes significant loads on the Internet, so reducing P2P traffic within networks is an important issue for ISPs. In particular, a huge amount of traffic is transferred within backbone networks; therefore reducing P2P traffic is important for transit ISPs to improve the efficiency of network resource usage and reduce network capital cost. To reduce P2P traffic, it is effective for ISPs to implement cache devices at some router ports and reduce the hop length of P2P flows by delivering the required content from caches. However, the design problem of cache locations and capacities has not been well investigated, although the effect of caches strongly depends on the cache locations and capacities. We propose an optimum design method of cache capacity and location for minimizing the total amount of P2P traffic based on dynamic programming, assuming that transit ISPs provide caches at transit links to access ISP networks. We apply the proposed design method to 31 actual ISP backbone networks and investigate the main factors determining cache efficiency. We also analyze the property of network structure in which deploying caches are effective in reducing P2P traffic for transit ISPs. We show that transit ISPs can reduce the P2P traffic within their networks by about 50–85% by optimally designing caches at the transit links to the lower networks.

Comparison of Evolutionary Algorithm and Heuristics for Flow Optimization in P2P Systems

Comparison of Evolutionary Algorithm and Heuristics for Flow Optimization in P2P SystemsNowadays, many Internet users make use of Peer-to-Peer (P2P) systems to download electronic content including music, movies, software, etc. Growing popularity in P2P based protocol implementations for file sharing purposes caused that the P2P traffic exceeds Web traffic and in accordance with to many statistics, P2P systems produce a more than 50% of the whole Internet traffic. Therefore, P2P systems provide remarkable income for Internet Service Providers (ISP). However, at the same time P2P systems generates many problems related to traffic engineering, optimization, network congestion. In this paper we focus on the problem of flow optimization in P2P file sharing systems. Corresponding to BitTorrent-based systems behaviour, the optimization of P2P flows is very complex and in this work we consider different heuristic strategies for content distribution and moreover we propose a new evolutionary algorithm (EA) for this problem. We compare results of the algorithms against optimal results yielded by CPLEX solver for networks including 10 peers and relation to random algorithm for 100-node systems. According to numerical experiments, the EA provides solutions close to optimal for small instances and all of the heuristics exhibit a superior performance over random search.

Identify P2P traffic by inspecting data transfer behavior

Classifying network traffic according to its applications is important to a broad range of network areas. Since new applications, especially P2P applications, no longer use well-known fixed port numbers, the native port-based traffic classification technique has become much less effective. In this paper, we propose a novel approach to identify P2P traffic by leveraging the data transfer behavior of P2P applications. The behavior investigated in the paper is that downloaded data from a P2P host will be uploaded to other hosts later. To find the shared data of downloading flows and uploading flows online, the content-based partitioning scheme is proposed to partition the flows into data blocks. Flows sharing the same data blocks are identified as P2P flows. Theoretical analysis proves that the content-based partitioning scheme is stable and effective. Experiments on various P2P applications demonstrate that the method is generic and can be applied to most P2P applications. Experimental results show that the algorithm can identify P2P applications accurately while only keeping a small set of data blocks.

Analyzing the Number of Varieties in Frequently Found Flows

Abnormal traffic that causes various problems on the Internet, such as P2P flows, DDoS attacks, and Internet worms, is increasing; therefore, the importance of methods that identify and control abnormal traffic is also increasing. Though the application of frequent-itemset-mining techniques is a promising way to analyze Internet traffic, the huge amount of data on the Internet prevents such techniques from being effective. To overcome this problem, we have developed a simple frequent-itemset-mining method that uses only a small amount of memory but is effective even with the large volumes of data associated with broadband Internet traffic. Using our method also involves analyzing the number of distinct elements in the itemsets found, which helps identify abnormal traffic. We used a cache-based implementation of our method to analyze actual data on the Internet and demonstrated that such an implementation can be used to provide on-line analysis of data while using only a small amount of memory.