Optimizing P2P streaming throughput under peer churning

High-throughput P2P streaming relies on peer selection, the strategy a peer uses to select other peer(s) as its parent(s) of streaming. Although this problem has been thoroughly investigated in the classical optimization framework under static settings, it still remains unaddressed as how to sustain throughput competitive to the optimum under highly dynamic peer churning. To accommodate such peer dynamics, we extend the classical optimization framework and propose a distributed online peer selection algorithm. This basic algorithm is further extended to a variety of settings commonly seen in operational P2P networks, such as multi-parent streaming, admission control, delay constraint, etc. We prove approximation bound of our algorithm to the optimal throughput. Through evaluation under different topological setups and peer churning sequences, we show that our solution can consistently deliver competitive throughput, which greatly outperforms its theoretical bound. This work was supported by NSF award 0643488, Vanderbilt Discovery grant, and a gift from Microsoft Research. Views and conclusions of this paper are those of authors, which should not be interpreted as representing the official policies, either expressed or implied, of the funding agencies.     

CollectCast: A peer-to-peer service for media streaming

We present CollectCast, a peer-to-peer (P2P) service for media streaming where a receiver peer is served by multiple sender peers. CollectCast operates at the application level but infers underlying network properties to correlate end-to-end connections between peers. The salient features of CollectCast include: (1) a novel multisender selection method that exploits the performance correlation and dependency among connections between different candidate senders and the receiver, (2) a customization of network tomography techniques and demonstration of improved practicality and efficiency, and (3) an aggregation-based P2P streaming mechanism that sustains receiver-side quality in the presence of sender/network dynamics and degradation. We have performed both real-world (on PlanetLab) and simulation evaluation of CollectCast. Our simulation results show that for a receiver, CollectCast makes better selection of multiple senders than other methods that do not infer underlying network properties. Our PlanetLab experiments are performed using a P2P media streaming application (called PROMISE) which we developed on top of CollectCast. Both packet-level and frame-level performance of MPEG-4 video streaming demonstrates the practicality and effectiveness of CollectCast.     

Transaction processing in a peer to peer database network

This paper investigates a transaction processing mechanism in a peer to peer database network. A peer to peer database network is a collection of autonomous data sources, called peers, where each peer augments a conventional database management system with an inter-operability layer (i.e. mappings) for sharing data. In this network, each peer independently manages its database and executes queries as well as updates over the related data in other peers. In this paper, we consider a peer to peer database network where mappings between peers are established through data-level mappings for sharing data and resolving data heterogeneity. With regards to transaction processing in a peer to peer database network, we mainly focus on how to maintain a consistent execution view of concurrent transactions in peers without a global transaction coordinator. Since there is no global transaction coordinator and each peer executes concurrent transactions independently, different peers may produce different execution views for the same set of transactions. For this purpose, we investigate potential problems that arise when maintaining a consistent execution of concurrent transactions. In order to guarantee consistent execution, we introduce a correctness criteria and propose two approaches, namely Merged Transactions and OTM based propagation. We assume that one single peer initiates the concurrent transactions. We also present a solution for ensuring the consistent execution view of concurrent transactions considering the failures of transactions.     

Multi-path continuous media streaming: what are the benefits?

Quality of service (QoS) in delivery of continuous media (CM) over the Internet is still relatively poor and inconsistent. Although many such applications can tolerate some degree of missing information, significant losses degrade an application’s QoS. In this paper, we investigate the potential benefits of mitigating this problem through the exploitation of multiple paths existing in the network between a set of senders and a receiver of CM. Our focus in this work is on providing a fundamental understanding of the benefits of using multiple paths to deliver CM over best-effort wide-area networks. Specifically, we consider pre-recorded CM applications and use the following metrics in evaluating the performance of multi-path streaming as compared to single-path streaming: (a) data loss rate, (b) conditional error burst length distribution, and (c) lag1-autocorrelation. The results of this work can be used in guiding the design of multi-path CM systems streaming data over best-effort wide-area networks.     

Performance evaluation of peering-agreements among autonomous systems subject to peer-to-peer traffic

The interconnection of thousands of Autonomous Systems (ASs) makes up the Internet. Each AS shares trade agreements with its neighbors that regulate the costs associated with traffic exchanged on the physical links. These agreements are local, i.e., are settled only between directly connected ASs, but have a global impact by influencing the paths allowed for the routing of network packets and the costs associated with these routes. Indeed, the costs and earnings of interconnected ASs is a function of many factors, such as size of the ASs, existing agreements, routing policy, traffic pattern and AS-level topology. In this paper we present an approach that takes these factors into account to assess peering and transit agreements. Here we focus on traffic generated from P2P activities, but the approach is general enough to be applied to different traffic classes. The P2P model we present is based on the use of the generating function, it allows to perform an analytical study of the traffic associated to file-sharing. The proposed P2P model is able to consider a large number of peers sharing several resources, spread along different ASs connected through a series of links. We validate the results of our P2P model against one of the most widely used P2P simulators, i.e. PeerSim. Using both the AS-level and P2P models we evaluate how the inter-AS P2P traffic influences the AS network cost and earning.     

Analysis of a CDN–P2P hybrid architecture for cost-effective streaming media distribution

To distribute video and audio data in real-time streaming mode, two different technologies – Content Distribution Network (CDN) and Peer-to-Peer (P2P) – have been proposed. However, both technologies have their own limitations: CDN servers are expensive to deploy and maintain, and consequently incur a cost for media providers and/or clients for server capacity reservation. On the other hand, a P2P-based architecture requires sufficient number of seed supplying peers to jumpstart the distribution process. Compared with a CDN server, a peer usually offers much lower out-bound streaming rate and hence multiple peers must jointly stream a media data to a requesting peer. Furthermore, it is not clear how to determine how much a peer should contribute back to the system after receiving the media data, in order to sustain the overall media distribution capacity.In this paper, we propose and analyze a novel hybrid architecture that integrates both CDN- and P2P-based streaming media distribution. The architecture is highly cost-effective: it significantly lowers the cost of CDN capacity reservation, without compromising the media quality delivered. In particular, we propose and compare different limited contribution policies for peers that request a media data, so that the streaming capacity of each peer can be exploited on a fair and limited basis. We present: (1) in-depth analysis of the proposed architecture under different contribution policies, and (2) extensive simulation results which validate the analysis. Our analytical and simulation results form a rigorous basis for the planning and dimensioning of the hybrid architecture.     

Performance analysis of structured peer-to-peer overlays for mobile networks

Distributed Hash Table (DHT) based Peer-to-Peer (P2P) overlays have been widely researched and deployed in many applications such as file sharing, IP telephony, content distribution and media streaming applications. However, their deployment has largely been restricted to fixed, wired networks. This is due to the fact that supporting P2P overlays on wireless networks such as the public mobile data network is more challenging due to constraints in terms of data transmissions on cellular networks, limited battery power of the handsets and increased levels of node churn. However, the proliferation of smartphones makes the use of P2P applications on mobile handsets very desirable. In this article, we have analysed and evaluated the performance and efficiency of five popular DHT based structured P2P overlays (Chord, Pastry, Kademlia, Broose and EpiChord) under conditions as commonly experienced in public mobile data networks. Our results show that the conditions in mobile networks, including a high churn rate and the relatively low bandwidth availability is best matched by Kademlia and EpiChord. These overlays exhibit a high lookup success ratio and low hop count while consuming a moderate amount of bandwidth. These characteristics make these two overlays suitable candidates for use in mobile networks.     

Robust control in cloud assisted peer to peer live streaming systems

Existing live video streaming systems can be classified as server (cloud) based or as peer-to-peer (P2P). The client–server approach promises stability and (Quality of Service) QoS by incurring expensive bandwidth provision cost on the server. On the other hand, P2P architecture is scalable with low bandwidth and maintenance cost. Here we propose a cloud assisted P2P live streaming architecture which is scalable and stable. In order to achieve this we have developed: (i) a scalable gossip protocol that monitors dynamically the total available bandwidth resources of the participating peers, (ii) a control strategy that dynamically allocates the bandwidth that is required.The first step towards this direction is to create a theoretical model that captures the dynamic relationship between the total bandwidth surplus/deficit and peers’ bandwidth utilization in order to be able to apply a control theoretical approach. Moreover, we quantify the impact of monitoring inaccuracies and peers’ dynamic bandwidth changes and we calculate analytically, as a function of them, the minimum amount of bandwidth overprovision that ensures the undisturbed distribution of the stream. System is evaluated through a detailed simulator of a complete P2P live streaming system and testified the uninterrupted and complete stream delivery even in very adverse bandwidth changes.     

Flow rate-modified streaming effects in heterogeneous microchannels

A general model based on the Onsager reciprocal relations is developed to study the streaming potential and streaming current in heterogeneous microchannels. The surface heterogeneities may be symmetrically or asymmetrically distributed parallel or perpendicular to the flow axis. Both streaming effects are modified by the flow rate through the heterogeneous channel, to eliminate the possible influence of electrokinetic flow on the streaming potential and streaming current measurements. Although they are still dependent on the distribution of surface heterogeneity, the flow rate-modified streaming effects are demonstrated to provide more consistent results with the traditional linear assumption than do the traditional ones, especially apparent in small microchannels.     

Chaining watermarks for detecting malicious modifications to streaming data

Authenticating streaming data is a very important research area due to its wide range of applications. Previous technologies mainly focused on authenticating data packets at the IP layer and ensuring the robustness of the verification. These schemes usually incur large communications overhead, which is not desirable in applications with limited bandwidth. In this paper, we propose a novel fragile watermarking algorithm which verifies the integrity of streaming data at the application layer. The data are divided into groups based on synchronization points, so each group can be synchronized and any modifications made to one group only affect up to two groups. A unique watermark is embedded directly into each group to save communications bandwidth. The embedded watermark can detect as well as locate any modifications made to a data stream. To ensure the completeness of the data stream, watermarks are chained across groups so that no matter how many data are deleted, these deletions can be correctly detected. Security analysis and experimental results show that the proposed scheme can efficiently detect and locate modifications and ensure the completeness of data streams.     

Peer to peer video streaming in Bluetooth overlays

As Bluetooth is available in most personal and portable terminals (eg, cellular phone, PDA, videocamera, laptop, etc) Peer-to-peer video streaming through Bluetooth networks is now a reality. Camera equipped Bluetooth phones capture video and broadcast it to other Bluetooth devices and to the infrastructure. Tra ditionally, large scale Bluetooth networks were designed using scatternet concepts. However, many Bluetooth devices do not support Scatternet connections and, even if they support it, they provide only very limited features suitable mostly for static environments. In high mobility situations, a traditional Scatternet design is not useful because of frequent disconnections and reconnections. To overcome these problems, we propose overlaid Bluetooth Piconets (OBP) and simplified overlaid Bluetooth Piconets (SOBP) that interconnect Piconets forming virtual Scatternets. In OBP, every Piconet dynamically reconfigures to collect metadata from neighboring Piconets. If metadata shows the existence of useful data to transfer, an inter-Piconet connection is made to carry out the transfer. SOBP can be used instead of OBP once neighbor Piconets have already discovered each other. In this paper, we compare via analysis and simulation the throughput and efficiency of OBP, SOBP and Scatternet for video applications. We demonstrate the feasibility of video over OBP and SOBP for a representative application.

Enhancing aggregate QoS for video streaming

The IETF's Differentiated Services architecture is designed to provide different types or levels of service for Internet traffic. One of its key features is that traffic flows are aggregated so that routers in the core network only need to distinguish a relatively small number of aggregate flows, even if those flows consist of hundreds or thousands of individual flows. However, network-level QoS differentiation may not satisfy the requirements of many QoS-sensitive applications such as live video streaming, where end-to-end delay and reliability must be guaranteed.

This paper investigates application-level service differentiation for MPEG video streaming in a Diff-Serv-aware MPLS network infrastructure, along with routing support at the network layer. Based on the fact that MPEG video has become one of the most popular formats for Internet (wired and wireless) users, our approach enhances aggregate QoS for video streaming by employing existing application-level knowledge of the MPEG video structure; therefore it requires neither a new video compression algorithm nor additional bandwidth. Several MPEG video dispersion models are proposed and analyzed. Simulation results show that two such strategies are superior to the other tested. In addition, a new routing scheme is proposed to support searching the ‘best’ paths for efficient multi-path video streaming. Through extensive simulations, we demonstrate that our approaches improve the aggregate QoS of MPEG video streams, as well as overall network efficiency.     

Density-based hierarchical clustering for streaming data

For streaming data that arrive continuously such as multimedia data and financial transactions, clustering algorithms are typically allowed to scan the data set only once. Existing research in this domain mainly focuses on improving the accuracy of clustering. In this paper, a novel density-based hierarchical clustering scheme for streaming data is proposed in order to improve both accuracy and effectiveness; it is based on the agglomerative clustering framework. Traditionally, clustering algorithms for streaming data often use the cluster center to represent the whole cluster when conducting cluster merging, which may lead to unsatisfactory results. We argue that even if the data set is accessed only once, some parameters, such as the variance within cluster, the intra-cluster density and the inter-cluster distance, can be calculated accurately. This may bring measurable benefits to the process of cluster merging. Furthermore, we employ a general framework that can incorporate different criteria and, given the same criteria, will produce similar clustering results for both streaming and non-streaming data. In experimental studies, the proposed method demonstrates promising results with reduced time and space complexity.     

基于P2P流媒体的在线直播系统模型研究

在传统网络电视直播系统的基础上,提出了一种基于P2P流媒体的在线直播系统架构,它针对流媒体传输要求带宽高、延迟小、实时性强等特点,采用P2P模型中的混合模型,实现了一种动态自适应的区域自治机制.对P2P网络中的信息源进行实时动态分配,极大地提高了各节点间信息传输速率,并且采用RTP、RTSP等流式传输协议,使整个P2P网络的信息传输稳定性、安全性都有一定的保障.     

Optimum piece selection strategies for a peer-to-peer video streaming platform

The client–server architecture is still popular due to its high predictable service and performance. However, it is not bandwidth scalable. An alternative setup for Internet video-streaming is offered by the peer-to-peer architecture, in which peers are servers as well as clients. Peers basically communicate in a three-level based policy. First, they meet other peers with common interests: this is called swarming. Then, each peer selects a small number of them for cooperation, called the peer selection strategy. In the last step peers cooperate sending pieces, defining the piece selection strategy.     

Evolving schemas for streaming XML

In this paper we model schema evolution for XML by defining formal language operators on Visibly Pushdown Languages (VPLs). Our goal is to provide a framework for efficient validation of streaming XML in the realistic setting where the schemas of the exchanging parties evolve and thus diverge from one another. We show that Visibly Pushdown Languages are closed under the defined language operators and this enables us to expand the schemas (for XML) in order to account for flexible or constrained evolution.     

Continuous subspace clustering in streaming time series

Performing data mining tasks in streaming data is considered a challenging research direction, due to the continuous data evolution. In this work, we focus on the problem of clustering streaming time series, based on the sliding window paradigm. More specifically, we use the concept of subspace α$\alpha$-clusters. A subspace α$\alpha$-cluster consists of a set of streams, whose value difference is less than α$\alpha$ in a consecutive number of time instances (dimensions). The clusters can be continuously and incrementally updated as the streaming time series evolve with time. The proposed technique is based on a careful examination of pair-wise stream similarities for a subset of dimensions and then it is generalized for more streams per cluster. Additionally, we extend our technique in order to find maximal pClusters in consecutive dimensions that have been used in previously proposed clustering methods. Performance evaluation results, based on real-life and synthetic data sets, show that the proposed method is more efficient than existing techniques. Moreover, it is shown that the proposed pruning criteria are very important for search space reduction, and that the cost of incremental cluster monitoring is more computationally efficient that the re-clustering process.     

Adaptive receiver-driven streaming from multiple senders

This paper presents the design and evaluation of an adaptive streaming mechanism from multiple senders to a single receiver in Peer-to-Peer (P2P) networks, called P2P Adaptive Layered Streaming, or PALS. PALS is a receiver-driven mechanism. It enables a receiver peer to orchestrate quality adaptive streaming of a single, layer-encoded video stream from multiple congestion-controlled senders, and is able to support a spectrum of noninteractive streaming applications. The primary challenge in the design of a streaming mechanism from multiple senders is that available bandwidth from individual peers is not known a priori, and could significantly change during delivery. In PALS, the receiver periodically performs quality adaptation based on the aggregate bandwidth from all senders to determine: (i) the overall quality (i.e number of layers) that can be collectively delivered by all senders, and more importantly (ii) the specific subset of packets that should be delivered by individual senders in order to gracefully cope with any sudden change in their bandwidth. Our detailed simulation-based evaluations illustrate that PALS can effectively cope with several angles of dynamics in the system including: bandwidth variations, peer participation, and partially available content at different peers. We also demonstrate the importance of coordination among senders and examine key design tradeoffs for the PALS mechanism. Nazanin Magharei is currently a PhD student in the Computer Science Department at the University of Oregon. She received her BSc degree in Electrical Engineering from Sharif University of Technology, Iran in 2002. Her research interests include Peer-to-Peer streaming and multimedia caching. Reza Rejaie is currently an Assistant Professor at the Department of Computer and Information Science at the University of Oregon. From October 1999 to March 2002, he was a Senior Technical Staff member at AT&T Labs-Research in Menlo Park, California. He received a NSF CAREER Award for his work on P2P streaming in 2005. Reza has served on the editorial board of IEEE Communications Surveys & Tutorials, as well as the program committee of major networking conferences including INFOCOM, ICNP, Global Internet, ACM Multimedia, IEEE Multimedia, NOSSDAV, ICDCS, and MMCN. Reza received his MS and PhD degrees in Computer Science from the University of Southern California (USC) in 1996 and 1999, and his BS degree in Electrical Engineering from the Sharif University of Technology (Tehran, Iran) in 1991, respectively. Reza has been a member of both the ACM and IEEE since 1997.     

Intractability of min- and max-cut in streaming graphs

We show that the exact computation of a minimum or a maximum cut of a given graph G is out of reach for any one-pass streaming algorithm, that is, for any algorithm that runs over the input stream of G's edges only once and has a working memory of o(n²) bits. This holds even if randomization is allowed.