Employing QoS Driven Neighbor Selection for Heterogeneous Peer-to-Peer Streaming

Nowadays, customers spend much time and effort in finding the best suitable goods since more and more information is placed on-line. To save their time and effort in searching the goods they want, a customized recommender system is required. In this paper we present an improved neighbor selection algorithm that exploits a graph approach. The graph approach allows us to exploit the transitivity of similarities. The algorithm searches more efficiently for set of influential customers with respect to a given customer. We compare the proposed recommendation algorithm with other neighbor selection methods. The experimental results show that the proposed algorithm outperforms other methods.

The neighbor selection that determines which users are exploited to estimate a target user's ratings has an important influence on the accuracy of recommendations of collaborative filtering based recommender system. Two kinds of ways for neighbor selection: KNN and cluster-based, are lack of specificity which refers to selecting different appropriate neighbors for different given target users, and thus limit the accuracy of recommendation. Therefore, in this paper, firstly, we propose a method that employs the evolutionary algorithm to optimize neighbors for all target users. Secondly, overcoming the high time complexity of the first one, we present another approach in which community detection algorithm is utilized as a preprocessing, and then the evolution algorithm is employed to optimize the neighborhood size for every community. We present experiments on a standard benchmark data-set, and the results show that the two methods both realize the specificity in neighbor selection, and accordingly lead to a higher accuracy of recommendations. Besides, the second one makes a good compromise between the specificity and time complexity.

Because existing simulator can't meet the need of the P2P streaming system, we research and design a P2P streaming simulator called P2PMedSim. Our simulator comes with a rich set of join, duration, and bandwidth distributions as well as overlay metrics. It has good visual interface and supports C/S model, tree model, and mesh model. The simulator design is introduced in details. Furthermore, we compare the original system and the improved syatem in testing part, which are both based on the mesh network topology.

Due to the recent spread of different styles of watching movies, streaming using Peer-to-Peer (P2P) technology has attracted great attention. In P2P streaming systems, to distribute the network load, since peers from which the user receives data are selected at random, clients have to wait until their desired data are delivered. Therefore, many researches are attempting to reduce the waiting time. However, due to the complexity of implementation, they usually evaluate these methods using machine simulations. In actual environments, interruption time is not always reduced by increasing the number of clients who deliver data. To evaluate the availability of P2P streaming systems, implementing a P2P streaming system is crucial. In this paper, we design and implement a P2P streaming system. With our implemented system, we consider situations in which the proposed system is effective.

인터넷 스트리밍 시스템은 스트리밍 소스를 공급하는 미디어 서버와 이로부터 미디어 스트림을 받아 분배하는 스트리밍 서버, 그리고 스트리밍 단말기들로 구성되며, 기존에는 하나의 스트리밍 서버에 다수의 단말기들이 직접적으로 연결되는 방식을 취하고 있다. 이러한 클라이언트-서버 형태의 중앙 구조에서는 스트리밍 서버에 대한 트래픽 집중으로 병목 현상이 발생하며, 한 스트리밍 서버의 용량에 따라 스트리밍 단말기의 개수가 제한되는 등 확장성 및 수용 능력의 한계라는 단점을 근본적으로 가지고 있다. 이러한 문제를 극복하여 확장성을 제공하기 위해 P2P 분산 아키텍처를 이용하는 P2P 스트리밍 시스템에 대한 연구가 최근 들어 진행되고 있다. 그러나 P2P 방식을 이용한 인터넷 스트리밍 시스템을 설계, 구현, 테스트하기 위해서는 실제 많은 컴퓨터들이 필요하며, 네트워크의 다양한 구성이나 트래픽 변화에 따른 실험을 하기에는 현실적인 어려움이 있다. 그러므로 본 논문에서는 P2P 스트리밍 시스템에 대한 다양한 연구 및 실험을 지원하기 위해, P2P 스트리밍 시스템에 대한 구조적 모델 및 동작 모델, 시간 모델, 성능 지수들을 정의하고, 네트워크 시뮬레이터를 지원하는 NS2 시뮬레이션 라이브러리를 이용하여 P2P 스트리밍 시스템 시뮬레이터 P2PStreamSim을 설계 및 구현하였다. 또한 테스트 P2P 스트리밍 시스템을 사례로 적용하여 P2PStreamSim 의 동작을 검증하고 성능을 평가하였다. Internet streaming systems consist of a media server, a streaming sewer, and terminals. The media server delivers multimedia contents such as video and/or audio to the streaming server, which distributes the contents to terminals as well. Existing Internet streaming systems have a bottleneck problem in the streaming server because of the limit of the processing capacity of the streaming server and therefore a streaming server can not accomodate more terminals than the limit. As a solution to this problem, P2P streaming systems have been lately proposed and investigated, using P2P distributed architectures. Actually, however, there exist many difficulties in the design and implementation of P2P streaming systems, because it needs many real computers and various network constructions. In this paper, we have proposed and defined a P2P streaming system model such as the architectural model, the timing model, the behavior model, and the performance metrics. And also we have implemented an NS2 based P2P streaming system simulator called P2PStreamSim. Finally, we have verified it through test simulations and analyzed the results.

Layered streaming is being considered as the most promising approach to adapt to bandwidth variations and heterogeneous end users in streaming applications. The goal of a layered streaming protocol is not only to optimize the average playback skip rate as in single-layer streaming, but also to maximize possible quality level (quality satisfaction) based on the available bandwidth capacity at the end user. In unstructured layered peer-to-peer streaming, however, achieving high quality satisfaction is challenging due to content and bandwidth bottlenecks. With experiments, in this paper, we demonstrate the importance and identify unique challenges of neighbor selection to the system performance in terms of the average skip rate and quality satisfaction. Then, we propose a new neighbor selection technique that can offer good performance while keeping the scalability of the mesh overlay under network fluctuations. The core of the technique is a preemption rule that allows a higher capacity peer to replace a lower capacity peer to be a neighbor of another peer with a certain probability. This preemption rule gears high capacity peers to good locations in the overlay to maximize the use of their bandwidth capacity and available layers. Simulation results demonstrate the efficiency of the method.

Formal verification of a peer-to-peer streaming protocol

Collaborative filtering methods are utilized to provide personalized recommendations for users in order to alleviate information overload problem in different domains. Traditional collaborative filtering methods operate on a user-item matrix in which each user reveal her admiration about an item based on a single criterion. However, recent studies indicate that recommender systems depending on multi-criteria can improve accuracy level of referrals. Since multi-criteria rating-based collaborative filtering systems consider users in multi-aspects of items, they are more successful at forming correlation-based user neighborhoods. Although, proposed multi-criteria user-based collaborative filtering algorithms' accuracy results are very promising, they have online scalability issues. In this paper, we propose an item-based multi-criteria collaborative filtering framework. In order to determine appropriate neighbor selection method, we compare traditional correlation approaches with multi-dimensional distance metrics. Also, we investigate accuracy performance of statistical regression-based predictions. According to real data-based experiments, it is possible to produce more accurate recommendations by utilizing multi-criteria item-based collaborative filtering algorithm instead of a single criterion rating-based algorithm.

The system throughput, outage probability and packet delay are analyzed for a feedback channel state information (FCSI) based soft decision MAC protocol in a centralized asynchronous CDMA packet radio system with finite population. The relative system performance loss due to effects of the feedback delay, access delay and imperfect channel sensing is investigated in terms of sensitivity functions. The motivation behind this protocol is to control the access of data transmission by dynamically changing permission probability depending on FCSI. The channel state is defined as a number of packets currently being transmitted in up-link (UL) multiple access channel. The FCSI is sensed and emitted periodically by a central base station (BS). The UL interfering load is therefore softly controlled under a certain threshold for a QoS requirement. The system performance can be improved significantly compared with conventional ALOHA or statically hard decision MAC packet radio system. The system is modeled with a modified Erlang loss queuing system, which has a variable number of servers and Erlang distributed service function. Tradeoff of the system parameters and performance characteristics is discussed based on number of numerical results.

This study discusses dispersion as a key issue influencing the performance of optical fiber communication technology. Poor bit rate, pulse broadening, and transmission distance restrictions are the outcomes. The dispersion correction method frequently employs fiber with dispersion compensation (DCF) and fiber Bragg grating (FBG). A model transmission system has been analyzed in light of numerous parameters using the Opti System 20.0 simulator. The DCF and FBG dispersion compensation methods are two that have been proposed. The Q-factor, BER, and eye height measured at the receiver end will be utilized as three metrics to describe the results of the two dispersion-correcting strategies. Additionally, the system's transmission performance will vary as a result of varied compensatory approaches. The DCF approach does not significantly increase system performance compared to the compensation technique. In addition, the properties of the optical fiber used for transmission in the system affect the characteristics of the FBG and the fiber used for the compensating process. It has been demonstrated that post-compensation, followed by symmetrical compensation, produces the best result. In addition, the transmission performance of the system with different compensation methods will be different, and the compensation technique will improve the system performance much better than the DCF technique. Moreover, the characteristics of the FBG and the fiber that is used for the compensation process depend on the characteristics of the optical fiber that is used for the transmission in system. It is shown that the symmetrical compensation performance is the best. Due to the nonlinear nature of propagation, system performance depends on power levels, followed by post-compensation.

The fundamental advantage of peer-to-peer (P2P) multimedia streaming applications is to leverage peer upload capacities to minimize bandwidth costs on dedicated streaming servers. The available bandwidth among peers is of pivotal importance to P2P streaming applications, especially as the number of peers in the streaming session reaches a very large scale. In this paper, we utilize more than 230 GB of traces collected from a commercial P2P streaming system, UUSee, over a four-month period of time. With such traces, we seek to thoroughly understand and characterize the achievable bandwidth of streaming flows among peers in large-scale real-world P2P live streaming sessions, in order to derive useful insights towards the improvement of current-generation P2P streaming protocols, such as peer selection. Using continuous traces over a long period of time, we explore evolutionary properties of inter-peer bandwidth. Focusing on representative snapshots of the entire topology at specific times, we investigate distributions of inter-peer bandwidth in various peer ISP/area/type categories, and statistically test and model the deciding factors that cause the variance of such inter-peer bandwidth. Our original discoveries in this study include: (1) The ISPs that peers belong to are more correlated to inter-peer bandwidth than their geographic locations; (2) There exist excellent linear correlations between peer last-mile bandwidth availability and inter-peer bandwidth within the same ISP, and between a subset of ISPs as well; and (3) The evolution of inter-peer bandwidth between two ISPs exhibits daily variation patterns. Based on these insights, we design a throughput expectation index that facilitates high-bandwidth peer selection without performing any measurements.

Consideration is given to a multiaccess communications channel which is shared on an integrated circuit-switched and packet-switched basis. Applications include time-division multiplexing and demand-assigned/TDMA (time-division multiaccess) communication channels which are shared by circuit-switched services (such as isochronous voice) and packet-switched services (such as data). Frames of fixed duration are established and divided into two parts: one part is used for voice transmissions while the other part serves to accommodate packetized data traffic. These two parts are separated by a movable boundary, so that data traffic can use the frame capacity which is temporarily unoccupied by voice transmissions. An exact expression is derived for the mean packet system size and delay. To reduce the computational complexity involved in using this expression, simpler expressions for upper and lower bounds are derived for the mean packet system size and packet delay. Examples that demonstrate the tightness of these numerically efficient bounds are presented.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

While recent measurement studies have shown the effectiveness of P2P network in media streaming, there have been questions raised about the quality of service (QoS), reliability of streaming services and sub optimal uplink utilization in particular. P2P streaming systems are inherently less reliable because of churn, internet dynamics, node heterogeneity and randomness in the swarm. We present a new model for P2P media streaming based on clustering of peers, called alliances. We show that alliance formation is a loosely coupled and an effective way to organize the peers. We show that our model maps to a ldquosmall-worldrdquo network, which form efficient overlay structures and are robust to network perturbations such as churn. We present a comparative simulation based study of our model with CoolStreaming/DONet and present a quantitative performance evaluation. Simulation results are promising and show that our model scales well under varying workloads and conditions, delivers near optimal levels of QoS, and for most cases, performs at par or even better than CoolStreaming/DONet.

Peer-to-peer (P2P) streaming systems rely on that peers voluntarily share their bandwidth to maintain high performance. Credit-based incentive mechanisms are widely used to encourage peers to share more bandwidth. However, in credit-based mechanisms, how to set a suitable bandwidth price is a critical issue. In this paper, we propose a new dynamic bandwidth pricing mechanism based on Stackelberg game and trust management. This mechanism can dynamically adjust the bandwidth price based on peers' trust values: the lower trust a peer has, the higher price the peer pays. A Stackelberg game is formulated to obtain peers' optimal bandwidth pricing and purchasing strategies. Through mathematical analysis, we derive the equilibrium of our proposed game and an algorithm to obtain equilibrium is also proposed. Extensive simulations show that our proposed mechanism can effectively induce peers to share more bandwidth, maintain system fairness and defend typical data pollution attack to guarantee the robustness of the system.

In this paper, we measure and study the peer startup process in PPLive, a popular commercial P2P streaming system, and focus on a fundamental issue in this aspect: how a peer initializes its buffer when it joins a channel, i.e., initial offset placement of peers' buffers in the startup stage. We build a general model of peer startup process in chunk-based P2P streaming systems and present an initial offset placement scheme we inferred from the measurement results, i.e., proportional placement (PP) scheme. With FP scheme, the initial buffer offset is set to the offset of the reference neighbor peer plus an advance proportional to the reference neighbor peer's offset lag or buffer width. We evaluate the performance of PP scheme and find it is stable when the placement is based on offset lag, but will be unstable when it is based on buffer width if the chunk fetching strategy and neighbor peer selection mechanism are not properly designed. We finally report our detailed measurement results of the peer startup process and initial offset placement algorithms used in PPLive. Our models and measurement results could be useful for guiding the analysis and design of buffering protocols for a real P2P live streaming system.