Packet Size Distribution Research Articles

Non-Stationary Characteristics of AQM Based on the Queue Length.

We performed a non-stationary analysis of a class of buffer management schemes for TCP/IP networks, in which the arriving packets were rejected randomly, with probability depending on the queue length. In particular, we derived formulas for the packet waiting time (queuing delay) and the intensity of packet losses as functions of time. These results allow us to observe how the evolution of the waiting time and losses depend on initial conditions (e.g., the full buffer) and system parameters (e.g., dropping probabilities, load, packet size distribution). As side results, the stationary waiting time and packet loss probability were obtained. Numerical examples demonstrate applicability of the theoretical results.

Cloud-gaming: Analysis of Google Stadia traffic

Interactive, real-time, and high-quality cloud video games pose a serious challenge to the Internet due to simultaneous high-throughput and low round trip delay requirements. In this paper, we investigate the traffic characteristics of Stadia, the cloud-gaming solution from Google, which is likely to become one of the dominant players in the gaming sector. To do that, we design several experiments, and perform an extensive traffic measurement campaign to obtain all required data. Our first goal is to gather a deep understanding of Stadia traffic characteristics by identifying the different protocols involved for both signaling and video/audio contents, the traffic generation patterns, and the packet size and inter-packet time probability distributions. Then, our second goal is to understand how different Stadia games and configurations, such as the video codec and the video resolution selected, impact on the characteristics of the generated traffic. We also evaluate the ability of Stadia to adapt to different link capacity conditions, including cases where the capacity drops suddenly, as well as sudden increases in the network latency. Our results and findings, besides illustrating the characteristics of Stadia traffic, are also valuable for planning and dimensioning future networks, as well as for designing new resource management strategies. Finally, we compare Stadia traffic to other video streaming applications, showcasing the main differences between them, and introduce a traffic model using our captures. We show that this model can be used in simulations to further investigate the network performance in presence of Stadia traffic.

Towards threshold‐agnostic heavy‐hitter classification

SummaryA heavy‐hitter (HH) network traffic flow consumes considerably more network resources than other flows combined. The classification of HHs is critical to provide, among others, the required level of Quality of Service and reliability in both conventional and data center networks. HH classification is typically threshold‐based. However, there is no consistent and accepted threshold or set of thresholds that would reliably classify flows. Furthermore, existing threshold‐driven approaches use counters (e.g., duration, packets, and bytes); thus, their accuracy depends on how complete the flow information is. This paper paves the way to threshold‐agnostic HH identification by proposing an approach that performs HH classification based on per‐flow packet size distribution (PSD) and template matching (TM). PSD allows capturing the behavior and dynamism of network traffic flows (even from their first few packets). TM enables to classify HHs by measuring the similarity between the PSD of observed flows and a set of master templates representing the flow size behavior of HH classes. We evaluated the PSD‐ and TM‐based approach using flows extracted from real traffic traces. Results show that our approach classifies HHs accurately and timely, corroborating that the threshold‐less perspective is feasible for HH identification.

Modeling Queuing Delay of 5G NR With NOMA Under SINR Outage Constraint

In this paper, we present an analytical model to characterize the average queuing delay for non-orthogonal multiple access (NOMA) downlink system by taking the outage event into account such that the user fails either decoding its own signal or performing SIC for the signals of other users at the receiver when the SINR is lower than a predefined outage threshold. The departure process of the queuing model is characterized by obtaining the first and second moment statistics of the service time that depends on the resource allocation strategy, the packet size and channel distributions. The proposed model is utilized to obtain the optimum power allocation that minimizes the maximum of the average queuing delay (MAQD) for a two-user network scenario. The Monte Carlo simulation experiments are performed to numerically validate the model by providing MAQD results for both NOMA and orthogonal multiple access (OMA) schemes. The results demonstrate that the NOMA achieves lower latency for low SINR outage thresholds while its performance is degraded faster than OMA as the SINR outage threshold increases such that OMA outperforms NOMA beyond a certain threshold. Another important result is that the latency performance of NOMA is significantly degraded when the 5G NR frame types having wider bandwidth are utilized. The results provide powerful insights for queuing delay analysis of 5G services.

Optimal PHY Configuration in Wireless Networks

In this work, we study the optimal configuration of the physical layer in wireless networks by means of Semi-Markov Decision Process (SMDP) modeling. In particular, assume the physical layer is characterized by a set of potential operating points, with each point corresponding to a rate and reliability pair; for example, these pairs might be obtained through a now-standard diversity-multiplexing tradeoff characterization. Given the current network state (e.g., buffer occupancies), a Decision Maker (DM) needs to dynamically decide which operating point to use. The SMDP problem formulation allows us to choose from these points. A solution to the SMDP problem is an optimal selection of operating points, which is expressed by a decision rule as a function of the number of packets in the source’s finite queue, the channel state, and the size of the packet to be transmitted. We derive a general solution to the SMDP which covers various model configurations, packet size distributions and channel dynamics. For the specific case of exponential transmission times, we analytically prove the optimal policy has a threshold structure. Numerical results validate this finding, as well as depict muti-threshold policies for time varying channels such as the Gilbert-Elliott channel.

Combination between DASH and SVM machine learning in the field of video delivery

With the development of network and streaming media technology, network video traffic is growing rapidly. In order to better control and manage network traffic and guarantee the quality of service of network video, it is necessary to classify network video services effectively. In traffic identification and classification, feature analysis and acquisition of better features are the key points to achieve efficient classification. Starting with the characteristics of packet size distribution, rate, IP alternation, byte number ratio between downstream and upstream, number of sub-stream fragments and average packet arrival time interval, this paper uses Support Vector Machine (SVM) to verify the classification effect of this feature, and achieves a high classification accuracy

Selective void creation/filling for variable size packets and multiple wavelengths

With ever-increasing demand for bandwidth, both optical packet switching and optical burst switching are proposed as alternatives to increase the capacity of optical networks in the future. In these packet-based switching techniques, Fiber Delay Lines (for delay assignments) and wavelength conversion (for channel assignments) are used to avoid contention between contending packets. The involved scheduling algorithms decide on which Fiber Delay Line and wavelength each packet is scheduled in order to maximize performance. For the setting without wavelength conversion we proposed a scheduling algorithm for assigning delays called void-creating algorithm that outperforms existing void filling algorithms for a variety of packet size distributions. This is achieved by selectively delaying packets longer than strictly necessary based on a numerical procedure that assigns a theoretical value to each void based on how likely the void will eventually be filled and thus prove useful. This contribution extends the concept of void-creation to the important case with multiple wavelengths, where also the channel has to be assigned. Results obtained by Monte Carlo simulation show that with our void-creating algorithm the obtainable improvement in various performance measures highly depends on the number of wavelengths present.

Impact toughness scattering of bainitic steel in the ductile-brittle transition temperature region

The impact toughness scattering in the ductile-brittle transition temperature (DBTT) region was experimentally examined on mixed and homogeneous grains of low alloy high strength bainitic steel under dynamic loading conditions. The results revealed that the mixed grain microstructure had larger impact toughness scattering than the homogeneous one, and the impact toughness scattering was mainly caused by the scattering in the cleavage fracture stress σf. The value of σf is related to the size of the microcrack formed in the bainitic packet. When a bainitic packet-sized microcrack propagates from one bainitic packet into the adjacent packet, cleavage fracture occurs. The cleavage fracture is controlled by the few coarse packets in the microstructures, and the σf scattering is influenced by the varied distances/relative locations between these coarse packets, and homogenizing the distribution of fine bainitic packet sizes is an effective way to reduce the impact toughness scattering in the DBTT region.

High utilization dynamic bandwidth allocation algorithm based on sorting report messages with additive-polling thresholds in EPONs

High utilization is one of the design goals for MPCP-based dynamic bandwidth allocation (DBA) algorithms in Ethernet passive optical networks (EPON). However, utilization is sacrificed in order to meet the delay limits of the applications in most DBA design schemes. This paper proposes a dynamic bandwidth allocation algorithm based on sorting report messages with additive-polling thresholds (AP-Sort DBA). This has the characteristics of high utilization and low delay during medium network loading. AP-Sort DBA is an extension of the adaptive dynamic bandwidth allocation algorithm with sorting report messages (Sort-DBA) and promotes utilization by reduction of unused slot remainders (USR) and distribution of excess bandwidth. For performance evaluation in terms of average delay, average queue size, loss rate and utilization, five DBA schemes are simulated, namely AP-Sort DBA, interleaved polling with adaptive cycle time (IPACT), dynamic bandwidth allocation with a modified grant table generation algorithm and fair-excess allocation (DBA2-FE), Sort-DBA and double phase polling algorithm (DPA). Self-similar traffic and OC-3 packet size distribution are used for frame generation. The results show that AP-Sort DBA can achieve up to 99% utilization, which is the highest among the tested alternatives, with up to 60% improved delay compared to the IPACT. Technical analysis is presented for further proof of the effectiveness of the proposed methodology.

Effect of Bainitic Packet Size Distribution on Impact Toughness and its Scattering in the Ductile-Brittle Transition Temperature Region of Q&T Mn-Ni-Mo Bainitic Steels

The impact toughness of Q&T Mn-Ni-Mo bainitic steel in the ductile–brittle transition temperature (DBTT) region was measured. The fracture micromechanism associated with the microstructural features was studied, by statistically analyzing the grain sizes of various structures. The results reveal that the cleavage fracture behavior is controlled by bainitic packets. There exists a critical bainitic packet size. When a microcrack formed in the bainitic packet with a size exceeding the critical size propagates into adjacent packets, brittle cleavage fracture occurs. The probability of finding the bainitic packets larger than the critical size dominates the impact toughness and its scattering. It is suggested that refining the bainitic packets and homogenizing their sizes is an effective method of improving the impact toughness and reducing its scattering in the DBTT region.

Robust application identification methods for P2P and VoIP traffic classification in backbone networks

Application identification plays an essential role in network management such as intrusion detection and security monitoring. But the continuous growth of bandwidth and massive amount of packets pose serious challenges for efficacious and accurate application identification. In this paper, we develop a new method to reduce the number of packets being processed while achieving the goal of accurate P2P and VoIP application identification. Firstly, we employ the Bi-flow model to aggregate traffic packets into Bi-flow, which can capture the exchange behavior characteristics between different terminals. Then we employ the signature of Packet Size Distribution (PSD) to capture flow dynamics, which is defined as the payload length distribution probability of the packets in one Bi-flow. Secondly, we collect PSD of several different P2P and VoIP applications and the analysis results show that PSD of different applications are different with each other, which can be used as features to perform traffic identification. We also find the PSD characteristics of one Bi-flow can be captured by its first few packets, which demonstrate our methods can identify the Bi-flow quickly after its establishment. We employ the Renyi cross entropy to perform identification by calculating the similarity between PSD of the Bi-flow being identified and that of specific application. If the similarity is higher than a selected threshold, the Bi-flow being identified is classified to the specific application. Finally, as the PSD is a type of probability feature which is not sensitive to packet lose, we integrate the Poisson sampling method into our framework to process the massive data in backbone networks. Experimental results using the artificial and actual traces collected from monitoring platform in the Northwest Center of CERNET (China Education and Research Network) verify the accuracy and robustness of our method.

Modeling and Analysis of IEEE 802.11 DCF MAC

Abstract In this paper, an analytic model is present for evaluating the MAC layer Collision Probability and mean delay at wireless nodes using the Distributed Coordination Function of IEEE 802.11 MAC Model. This model is valid for finite loads and can account for arbitrary arrival patterns, packet size distributions and number of nodes. Each node is modeled as a discrete time M/G/1 queue and impact of packet collisions, the resulting back-offs as well as the packet size distribution. We analyze this Model for different number of nodes in any network. The impact of arrival pattern on collision probability and delay is also analyzed.

Integrated service resource reservation using queueing networks theory

This study analyses multi-server multi-service queueing networks with service protection. To guarantee each service a certain quality-of-service and at the same time ensure high utilisation of servers, a minimum capacity is reserved each service. In addition, all services share the remaining non-reserved capacity and all buffers in such a way that the system becomes reversible. This implies the model has product form between nodes and is insensitive to the packet size distribution. An analytical model and algorithm for performance evaluation are presented.

Enabling free internet access at the edges of broadband connections

B roadband connection sharing constitutes a network resource pooling technique that could evolve toward supporting the establishment of the “Free Internet for all” notion. This technique could be implemented through the User-Provided Network model in which end-users could become suppliers of Internet connectivity. In this paper, the behavior of such system is investigated from a queuemanagement perspective. A numerical analysis that takes into consideration both the high variability of broadband link speeds and different packet-size distributions is presented. Based on these results, we propose a hybrid packet scheduling scheme that enables the application of Lessthan-Best-Effort services at the connection edges and allows for the development of different packet scheduling policies corresponding to different requirements. Some significant open issues and research directions are also highlighted.

Online Identification of Tor Anonymous Communication Traffic

PDF HTML阅读 XML下载 导出引用 引用提醒 Tor匿名通信流量在线识别方法 DOI: 10.3724/SP.J.1001.2013.04253 作者: 作者单位: 作者简介: 通讯作者: 中图分类号: 基金项目: 国家重点基础研究发展计划(973)(2010CB328104); 国家自然科学基金(61272054, 61202449, 61070161, 61003257,60903162); 国家科技支撑计划(2010BAI88B03, 2011BAK21B02); 国家核高基科技重大专项(软件类)项目(2010ZX01044-001-001);高等学校博士点专项科研基金(20110092130002); 江苏省自然科学基金(BK2008030); 江苏省网络与信息安全重点实验室资助项目(BM2003201); 教育部计算机网络与信息集成重点实验室(东南大学)资助项目(93K-9) Online Identification of Tor Anonymous Communication Traffic Author: Affiliation: Fund Project: 摘要 | 图/表 | 访问统计 | 参考文献 | 相似文献 | 引证文献 | 资源附件 | 文章评论 摘要:匿名通信技术的滥用给网络监管带来了新的挑战.有效识别出匿名通信流量,是阻止该类技术滥用的前提,具有重要的研究意义和应用价值.现有研究工作侧重于匿名通信关系的确认,无法用于匿名通信流量的识别和阻塞.针对这个问题,围绕广泛使用的Tor匿名通信系统,深入分析运行机制,归纳总结其流量特征.在此基础上,分别提出基于TLS指纹和基于报文长度分布的Tor匿名通信流量识别方法.对两种识别方法的优缺点和适用性进行了详细分析和讨论,并通过CAIDA数据集和在线部署对识别方法进行了验证.实验结果表明,基于TLS指纹和基于报文长度分布的识别方法均能有效识别出Tor匿名通信流量. Abstract:Abuse of anonymous communication systems has introduced new challenges into network administration. The effective identification of anonymous communication traffic is a prerequisite to prevent such abuse; thus, this is fundamentally important for both theoretical researches and practical applications. Existing researches mainly focus on the confirmation of anonymous communication relationship and cannot be used to identify and block anonymous communication traffic. To solve this problem, the operation mechanism is deeply analyzed and traffic characteristics are summarized for the widely used Tor anonymous communication system. On this basis, a TLS fingerprint-based and packet-size distributions based methods are proposed to identify Tor anonymous communication traffic, respectively. The advantages, disadvantages and applicability of these two methods are analyzed and discussed in detail, and are validated by CAIDA dataset and online deployment. Experimental results prove that both methods are effective in identifying Tor anonymous communication traffic. 参考文献 相似文献 引证文献

A Closed-Form Model for the IEEE 802.3az Network and Power Performance

We propose an analytical model able to accurately estimate both power consumption and network performance indexes of Energy Efficient Ethernet (EEE) links working at the three available speeds under various traffic load patterns and packet size distributions. The model is sufficiently flexible and accurate to consider different traffic parameters; among others, the packet size distribution, the average burst inter-arrival rate, the burst size distribution, etc. With relatively low complexity, since it addresses stationary queue behavior, the analysis allows obtaining the average energy consumption of the link and, unlike previous works, the mean latency time experienced by incoming packets in closed form, without any upper or lower bound approximations. This aspect makes the model suitable to be adopted in optimization frameworks for network design and control purposes. The numerical results of the model are validated against measurements on a real test bench.

Novel Packet Size-Based Covert Channel Attacks against Anonymizer

In this paper, we present a study on the anonymity of Anonymizer, a well-known commercial anonymous communication system. We discovered the architecture of Anonymizer and found that the size of web packets in the Anonymizer network can be very dynamic at the client. Motivated by this finding, we investigated a class of novel packet size-based covert channel attacks against Anonymizer. The attacker between a website and the Anonymizer server can manipulate the web packet size and embed secret signal symbols into the target traffic. An accomplice at the user side can sniff the traffic and recognize the secret signal. In this way, the anonymity provided by Anonymizer is compromised. We developed intelligent and robust algorithms to cope with the packet size distortion incurred by Anonymizer and Internet. We developed techniques to make the attack harder to detect: 1) We pick up right packets of web objects to manipulate to preserve the regularity of the TCP packet size dynamics, which can be measured by the Hurst parameter; 2) We adopt the Monte Carlo sampling technique to preserve the distribution of the web packet size despite manipulation. We have implemented the attack over Anonymizer and conducted extensive analytical and experimental evaluations. It is observed that the attack is highly efficient and requires only tens of packets to compromise the anonymous web surfing via Anonymizer. The experimental results are consistent with our theoretical analysis.

Cataloging Of Sessions in Genuine Traffic by Packet Size Distribution and Session Grouping

Cataloging Of Sessions in Genuine Traffic by Packet Size Distribution and Session Grouping

Packet traffic features of IPv6 and IPv4 protocol traffic

Nowadays, the IPv6 protocol is in a transition phase in operational networks. The ratio of its traffic volume is increasing day by day. The many provided facilities for IPv6 connection increasethe total IPv6 traffic load. IPv6-over-IPv4 tunnels, pilot programsto provide IPv6 connections, IPv6/IPv4 dual stack operating systems,and free IPv6 tunnel brokers cause the IPv6 protocol to expand quickly. For efficient resource utilization, the characteristics of network traffic should be determined accurately. Many traffic characterization studies regarding IPv4 have demonstrated that most of the network traffic is self-similar. Self-similarity causes significant impacts on network performance. With the increasing volume of IPv6 traffic, the characteristics of IPv6 traffic and differences between IPv4 traffic in terms of characterization should be explicitly revealed. In this study, we investigate the characteristics of IPv6 packet traffic and the differences between IPv6 and IPv4 packet traffic in terms of spectral density, autocorrelation, distribution, and self-similarity of packet interarrival time and packet size. The results obviously show that IPv6 traffic exhibits totally different properties in comparison to that of IPv4. Distribution fittings prove that packet interarrival time and packet size have different distributions in the 2 traffic types. While beta distribution models the empirical cumulative distribution of IPv4 packet size, log-logistic distribution gives more efficient results for IPv6 packet size. Furthermore, a significant difference is observed in self-similarity degrees. IPv6 protocol traffic gives greater self-similarity than that of IPv4. Results show that IPv6 traffic would cause greater performance degradations in computer networks in comparison to IPv4 due to high self-similarity.

Dynamic packet fragmentation based on particle swarm optimised prediction

Particle Swarm Optimisation (PSO) has been successfully applied in function optimisation, neural network design, classification, pattern recognition, signal processing, robot technology and so on. The new concept of analysis and prediction of network performance using PSO is proposed, and the main challenges come from network traffic for different applications. We collected and analysed traffic of BigBlueButton application, using Hurst calculation approach to judge the performance of network, and offered the necessary prediction. According to Hurst classification, using PSO algorithm to find the right packet size distribution under the network performance constraints. The paper proposes a new dynamic perspective about packet fragmentation and the related performance of the real-time applications. It can reduce network congestion and improve the end user experience. The paper offers a preliminary analysis and simulation with the biological algorithm.