A compressive sensing‐based network tomography approach to estimating origin–destination flow traffic in large‐scale backbone networks

A traffic matrix can exhibit the volume of network traffic from origin nodes to destination nodes. It is a critical input parameter to network management and traffic engineering, and thus it is necessary to obtain accurate traffic matrix estimates. Network tomography method is widely used to reconstruct end‐to‐end network traffic from link loads and routing matrix in a large‐scale Internet protocol backbone networks. However, it is a significant challenge because solving network tomography model is an ill‐posed and under‐constrained inverse problem. Compressive sensing reconstruction algorithms have been well known as efficient and precise approaches to deal with the under‐constrained inference problem. Hence, in this paper, we propose a compressive sensing‐based network traffic reconstruction algorithm. Taking into account the constraints in compressive sensing theory, we propose an approach for constructing a novel network tomography model that obeys the constraints of compressive sensing. In the proposed network tomography model, a framework of measurement matrix according to routing matrix is proposed. To obtain optimal traffic matrix estimates, we propose an iteration algorithm to solve the proposed model. Numerical results demonstrate that our method is able to pursuit the trace of each origin–destination flow faithfully. Copyright © 2014 John Wiley & Sons, Ltd.     

Time‐varying network link loss rate tracking from unicast end‐to‐end measurements

Time‐varying network link loss rate is a useful information for network managers to discover and locate the network link failures. This paper proposes a method to track time‐varying network link loss rates from unicast end‐to‐end measurements. The method first trains a state transition matrix to capture the spatio‐temporal characters of packet link transmission probabilities by sending end‐to‐end probe packets and then estimates the time‐varying link loss rates using the state transition matrix and the end‐to‐end measurements obtained from background traffic (the existed packets in network). We also introduce a validation step to check and retrain the state transition matrix in order to ensure the accuracy of the state transition matrix. Our method is capable of tracking the variation of link loss rates without incessantly sending probe packets, which is very feasible for many practical applications. The ns‐2 simulation results show the good performance of our method.     

基于离散小波变换的网络流量多重分形模型

网络流量过程中所蕴含的分形尺度特性对网络性能有显著的影响。因此研究能全面准确地刻画网络流量过程在小时间／空间尺度上的复杂奇异性特征和大时间／空间尺度上的长程依赖性特征的流量模型对Internet网络工程有重要的意义。本文对实测的流量数据(从著名的校园网和国内著名的ISP)进行了分析，利用小波技术构建了一个新的网络流量的多重分形模型，通过模拟验证，发现该新模型能以较简洁的形式捕捉实际网络流量特性，并具有刻画真实流量数据中的多重分形特征的能力。     

Multiplicative multifractal modelling of long‐range‐dependent network traffic

We present a multiplicative multifractal process to model traffic which exhibits long‐range dependence. Using traffic trace data captured by Bellcore from operations across local and wide area networks, we examine the interarrival time series and the packet length sequences. We also model the frame size sequences of VBR video traffic process. We prove a number of properties of multiplicative multifractal processes that are most relevant to their use as traffic models. In particular, we show these processes to characterize effectively the long‐range dependence properties of the measured processes. Furthermore, we consider a single server queueing system which is loaded, on one hand, by the measured processes, and, on the other hand, by our multifractal processes (the latter forming a MF_e/MF_g/1 queueing system model). In comparing the performance of both systems, we demonstrate our models to effectively track the behaviour exhibited by the system driven by the actual traffic processes. We show the multiplicative multifractal process to be easy to construct. Through parametric dependence on one or two parameters, this model can be calibrated to fit the measured data. We also show that in simulating the packet loss probability, our multifractal traffic model provides a better fit than that obtained by using a fractional Brownian motion model. Copyright © 2001 John Wiley & Sons, Ltd.     

MPEG-4视频业务的多重分形分析与建模

基于对象编码的MPEG-4视频业务将是网络业务中的主流业务，因而研究MPEG-4视频业务的特性对网络设计，容量规划，性能评估，接入控制与性能分析有重要的意义。通过应用多重分形分析方法对MPEG-4视频业务的分形行为进行分析，结果表明MPEG-4视频业务多重分形特性的存在，因而采用多重分形小波模型对MPEG-4视频业务进行建模分析，通过仿真试验结果的比较和分析，表明该模型能够真实反映MPEG-4视频业务的突发特性。     

Identifying LDoS attack traffic based on wavelet energy spectrum and combined neural network

As a special type of denial of service (DoS) attacks, the TCP‐targeted low‐rate denial of service (LDoS) attacks have the characteristics of low average rate and strong concealment, so it is difficult to identify such attack traffic. As multifractal characteristics exist in network traffic, a new identification approach based on wavelet transform and combined neural network is proposed to classify normal network traffic and LDoS attack traffic. Wavelet energy spectrum coefficients extracted from the sampled traffic are used for multifractal analysis of traffic over different time scale. The combined neural network is designed to classify these multiscale spectrum coefficients that show different multifractal characteristics belonging to normal network traffic and LDoS attack traffic. Test results of test‐bed experiments indicate that the proposed approach can identify LDoS attack traffic accurately.     

QoS‐based radio network dimensioning for LTE networks with heavy real‐time traffic

Applications of video streaming and real‐time gaming, which generate large amounts of real‐time traffic in the network, are expected to gain considerable popularity in Long Term Evolution networks. Maintaining the QoS such as packet delay, packet loss ratio, median, and cell border throughput requirements in networks dominated by real time traffic, is critical. The existing dimensioning methodology does not consider QoS parameters of real‐time traffic in network dimensioning. Moreover, exhaustive and time‐consuming simulations are normally required to evaluate the performance and QoS of real‐time services. To overcome this problem, we propose an improved radio network dimensioning framework that considers the QoS of real‐time traffic in network dimensioning. In this framework, an analytical model is proposed to evaluate the capacity and performance of real‐time traffic dominant Long Term Evolution networks. The proposed framework provides a fast and accurate means of finding the trade‐off between system load, packet delay, packet loss ratio, required median, and cell border throughput. It also provides network operators with an analytical means for obtaining the minimum number of sites required by jointly considering coverage, capacity and QoS requirements. The accuracy of the proposed model is validated through simulations. Copyright © 2012 John Wiley & Sons, Ltd.     

Bounded‐delay and loss‐free guarantees for real‐time video in integrated services packet networks

In this paper, we analyse upper bounds on the end‐to‐end delay and the required buffer size at the leaky bucket and packet switches within the network in the context of the deterministic bandwidth allocation method in integrated services packet networks. Based on that formulation, we then propose a CAC method suitable to ISPN to guarantee the bounded end‐to‐end delay and loss‐free packet transmissions. As an example application, the GOP–CBR MPEG‐2 is considered. In that case, we also show tighter bounds by slightly modifying the coding method of GOP–CBR MPEG‐2. Using the actual traced data of GOP–CBR MPEG‐2, we discuss the applicabilities of our analytical results and proposed CAC by comparing with simulation. Numerical results show that the loose upper bounds can also achieve more utilization even in the context of deterministic bandwidth allocation compared with the peak bandwidth allocation strategy. Copyright © 2000 John Wiley & Sons, Ltd.     

A delay‐centric parallel multi‐path routing protocol for cognitive radio ad hoc networks

In this paper, we develop a delay‐centric parallel multi‐path routing protocol for multi‐hop cognitive radio ad hoc networks. First, we analyze the end‐to‐end delay of multi‐path routing based on queueing theory and present a new dynamic traffic assignment scheme for multi‐path routing with the objective of minimizing end‐to‐end delay, considering both spectrum availability and link data rate. The problem is formulated as a convex problem and solved by a gradient‐based search method to obtain optimal traffic assignments. Furthermore, a heuristic decentralized traffic assignment scheme for multi‐path routing is presented. Then, based on the delay analysis and the 3D conflict graph that captures spectrum opportunity and interference among paths, we present a route discovery and selection scheme. Via extensive NS2‐based simulation, we show that the proposed protocol outperforms the benchmark protocols significantly and achieves the shortest end‐to‐end delay. Copyright © 2015 John Wiley & Sons, Ltd.     

A time‐efficient shrinkage algorithm for the Fourier‐based prediction enabling proactive optimisation in software‐defined networks

This paper focuses on the problem of time‐efficient traffic prediction. The prediction enables the proactive and globally scoped optimisation in software‐defined networks (SDNs). We propose the shrinkage and selection heuristic method for the trigonometric Fourier‐based traffic models in SDNs. The proposed solution allows us to optimise the network for an upcoming time window by installing flow entries in SDN nodes before the first packet of a new flow arrives. As the mechanism is designed to be a part of a sophisticated routing‐support system, several critical constraints are considered and taken into account. Specifically, the system is traffic‐ and topology‐agnostic, thus the prediction mechanism must be applicable to the networks with highly variable traffic loads (e.g., observed inside intra‐DCNs: datacentre networks). Furthermore, the system must effectively optimise routing in large‐scale SDNs comprised of numerous nodes and handling millions of flows of a dynamic nature. Therefore, the prediction must be simultaneously accurate as well as being time efficient and scalable. These requirements are met by our Fourier‐based solution, which subtracts consecutive harmonics from the original signal and compares the result with an adaptive threshold adjusted to the signal's standard deviation. The evaluation is performed by comparing the proposed heuristic with the well‐known Lasso method of proven accuracy. The results show that our solution is able to retain prediction accuracy at a comparable level. Moreover, in accordance with our main aim, we operate in a manner which is always significantly faster. In some cases, computation times are reduced by as much as 50 times.     

An adaptive end‐to‐end RTO calculation framework for DTNs with scheduled connectivity

In this work, we propose an end‐to‐end retransmission framework for dynamically calculating efficient retransmission time‐out intervals in delay‐tolerant networks (DTNs) with scheduled connectivity. The proposed framework combines deterministic and statistical information about the network state to calculate worst‐case estimates about the expected round trip times. Such information includes connectivity schedules, convergence layer protocols specifics, communication link characteristics, and network statistics about the maximum expected packet error rates and storage congestion. We detail the implementation of the proposed framework within the end‐to‐end application data conditioning layer proposed for the DTN architecture, realized by the Delay‐Tolerant Payload Conditioning protocol, as part of the Interplanetary Overlay Network–DTN reference implementation, and evaluate its performance in a complex deep‐space emulation scenario in our DTN testbed. Our results show that our approach achieves great accuracy in round‐trip time estimations and, therefore, faster retransmissions of lost data, in comparison to the statically configured retransmission mechanism of the original Delay‐Tolerant Payload Conditioning protocol. As a result, in‐order data reception rate and storage requirements on the receiver side are significantly improved, at minimum or even zero extra cost in transmission overhead due to duplicate transmissions.     

网络业务量的一种稳定多分形模型及其检验

目前网络业务量研究中有自相似和多分形两种主要的先验模型。自相似模型简洁,运用单一的H 指数来描述业务量特性,但它并不能完全表达业务量的信息;多分形模型更准确,但它运用函数形式的多分形谱来刻画信号,不便于实际分析。本文在无穷可分层叠的理论框架下提出了网络业务量的一种稳定多分形模型。运用小波变换方法,通过多个业务量记录的检验,证实此模型从尺度行为和边缘分布两方面都能准确刻画业务量。作为一般多分形过程的特例,该模型把多分形指数的表达式确定下来,并归结为4个参数的估计。     

Goodput and throughput comparison of single‐hop and multi‐hop routing for IEEE 802.11 DCF‐based wireless networks under hidden terminal existence

We investigate how multi‐hop routing affects the goodput and throughput performances of IEEE 802.11 distributed coordination function‐based wireless networks compared with direct transmission (single hopping), when medium access control dynamics such as carrier sensing, collisions, retransmissions, and exponential backoff are taken into account under hidden terminal presence. We propose a semi‐Markov chain‐based goodput and throughput model for IEEE 802.11‐based wireless networks, which works accurately with both multi‐hopping and single hopping for different network topologies and over a large range of traffic loads. Results show that, under light traffic, there is little benefit of parallel transmissions and both single‐hop and multi‐hop routing achieve the same end‐to‐end goodput. Under moderate traffic, concurrent transmissions are favorable as multi‐hopping improves the goodput up to 730% with respect to single hopping for dense networks. At heavy traffic, multi‐hopping becomes unstable because of increased packet collisions and network congestion, and single‐hopping achieves higher network layer goodput compared with multi‐hop routing. As for the link layer throughput is concerned, multi‐hopping increases throughput 75 times for large networks, whereas single hopping may become advantageous for small networks. The results point out that the end‐to‐end goodput can be improved by adaptively switching between single hopping and multi‐hopping according to the traffic load and topology. Copyright © 2015 John Wiley & Sons, Ltd.     

Application‐Level Traffic Monitoring and an Analysis on IP Networks

Traditional traffic identification methods based on well‐known port numbers are not appropriate for the identification of new types of Internet applications. This paper proposes a new method to identify current Internet traffic, which is a preliminary but essential step toward traffic characterization. We categorized most current network‐based applications into several classes according to their traffic patterns. Then, using this categorization, we developed a flow grouping method that determines the application name of traffic flows. We have incorporated our method into NG‐MON, a traffic analysis system, to analyze Internet traffic between our enterprise network and the Internet, and characterized all the traffic according to their application types.     

Performance optimization of duty‐cycled MAC in delay‐energy constrained sensor network under uniform and nonuniform traffic generation

Duty‐cycle at the media access control (MAC) layer plays a key role in energy savings and network lifetime extension. It consists in putting a node's radio in the sleep state as soon as it has no communication activity. Traditional wireless sensor network MAC protocols are designed with short duty‐cycles at the cost of long delays. Careful design is required for joint energy‐delay constrained applications, where the optimal parameters should be thoroughly derived. The present paper deals with this issue and mathematically derives optimal values of key MAC parameters under low data rate applications for 3 well‐known duty‐cycled MAC protocols, WiseMAC, SCP‐MAC, and LMAC as representatives of 3 MAC protocol categories, respectively, preamble‐sampling, slotted contention‐based, and frame‐based. The analysis provides also the optimum traffic sampling rate that guarantees the minimum energy consumption. It shows the role of these parameters in achieving the targeted end‐to‐end delay constraints under network models with uniform traffic generation, for ring and grid topologies. As a second contribution, the model is extended to nonuniform traffic scenarios, where a certain percentage of deployed nodes are relays whose role is to balance traffic forwarding and save the overall network energy. The results reveal that different optimal internal MAC parameters and traffic generation rates can be found for different configurations of relay nodes deployment, which achieve minimal network energy consumption while satisfying the application required end‐to‐end delay threshold.     

Improvement in DRX Power Saving for Non‐real‐time Traffic in LTE

A discontinuous reception (DRX) operation is included in the Long Term Evolution (LTE) system to achieve power saving and prolonged battery life of the user equipment. An improvement in DRX power saving usually leads to a potential increase in the packet delay. An optimum DRX configuration depends on the current traffic, which is not easy to estimate accurately, particularly for non‐real‐time applications. In this paper, we propose a novel way to vary the DRX cycle length, avoiding a continuous estimation of the data traffic when only non‐real‐time applications are running with no active real‐time applications. Because a small delay in non‐real‐time traffic does not essentially impact the user's experience adversely, we deliberately allow a limited amount of delay in our proposal to attain a significant improvement in power saving. Our proposal also improves the delay in service resumption after a long period of inactivity. We use a stochastic analysis assuming an M/G/1 queue to validate this improvement.     

Network traffic prediction method based on wavelet transform and multiple models fusion

Accurate prediction of network traffic is an important premise in network management and congestion control. In order to improve the prediction accuracy of network traffic, a prediction method based on wavelet transform and multiple models fusion is presented. Mallat wavelet transform algorithm is used to decompose and reconstruct the network traffic time series. The approximate and detailed components of the original network traffic can be obtained. The characteristics of approximate components and detail components are analyzed by Hurst exponent. Then, according to the different characteristics of the components, autoregressive integrated moving average model (ARIMA) is chosen as the prediction model for the approximate component. Least squares support vector machine (LSSVM) is used to predict detail component. Meanwhile, an improved particle swarm optimization (PSO) algorithm is proposed to optimize the parameters of the LSSVM model. Gauss‐Markov estimation algorithm is adapted to fuse the predicted values of multiple prediction models. The variance of fusion prediction error is smaller than that of single prediction model, and the prediction accuracy is improved. Two actual datasets of network traffic are studied. Compared with other state‐of‐the‐art models, the case study results indicate that the proposed prediction method has a better prediction effect.     

An end‐to‐end alternative to TCP PEPs: Initial Spreading,a TCP fast start‐up mechanism

While Transmission Control Protocol (TCP) Performance Enhancing Proxy (PEP) solutions have long been undisputed to solve the inherent satellite problems, the improvement of the regular end‐to‐end TCP congestion avoidance algorithms and the recent emphasis on the PEPs drawbacks have opened the question of the PEPs sustainability. Nevertheless, with a vast majority of Internet connections shorter than ten segments, TCP PEPs continue to be required to counter the poor efficiency of the end‐to‐end TCP start‐up mechanisms. To reduce the PEPs dependency, designing a new fast start‐up TCP mechanism is therefore a major concern. But, while enlarging the Initial Window (IW) up to ten segments is, without any doubt, the fastest solution to deal with a short‐lived connection in an uncongested network, numerous researchers are concerned about the impact of the large initial burst on an already congested network. Based on traffic observations and real experiments, Initial Spreading has been designed to remove those concerns whatever the load and type of networks. It offers performance similar to a large IW in uncongested network and outperforms existing end‐to‐end solutions in congested networks. In this paper, we show that Initial Spreading, taking care of the satellite specificities, is an efficient end‐to‐end alternative to the TCP PEPs. Copyright © 2015 John Wiley & Sons, Ltd.     

Investigation of maximum lifetime and minimum delay trade‐off in underwater sensor networks

Underwater acoustic sensor networks (UASNs) are subjected to harsh characteristics of underwater acoustic channel such as severe path losses, noise, and high propagation delays. Among these constraints, propagation delay (more generally, end‐to‐end delay) is the most dominating limitation especially for time‐critical UASN applications. Although the minimization of end‐to‐end delay can be achieved by using the minimum hop routing, this solution cannot lead prolonged lifetimes since nodes consume excessive energy for transmission over long links. On the other hand, the maximization of network lifetime is possible by using energy efficient paths, which consist of relatively short links but high number of hops. However, this solution results in long end‐to‐end delays. Hence, there is a trade‐off between maximizing the network lifetime and minimizing the end‐to‐end delay in UASNs. In this work, we develop a novel multi‐objective–optimization (MOO) model that jointly maximizes the network lifetime while minimizing the end‐to‐end delay. We systematically analyze the effects of limiting the end‐to‐end delay on UASN lifetime. Our results reveal that the minimum end‐to‐end delay routing solution results in at most 72.93% reduction in maximum network lifetimes obtained without any restrictions on the end‐to‐end delay. Nevertheless, relaxing the minimum end‐to‐end delay constraint at least by 30.91% yields negligible reductions in maximum network lifetimes.     

How to reconstruct end-to-end traffic based on time-frequency analysis and artificial neural network

End-to-end traffic, which describes the inherent characteristics and end-to-end behaviors of communication networks, is the crucial input parameter of network management and network traffic engineering. This paper proposes a new reconstruction algorithm to develop the research on reconstruction of end-to-end traffic in large-scale communication networks. We firstly conduct the time-frequency analysis on end-to-end traffic, and then localize its features to gain its time-frequency properties before decomposing it into the low-frequency and high-frequency components. We find that if decomposing appropriately, the low-frequency component of end-to-end traffic can accurately reflect its change trend, while its high-frequency component can well show the burst and fluctuation nature. This motivates us to find a reasonable time-frequency decomposition strategy to extract the low-frequency and high-frequency components of end-to-end traffic. Moreover, this further inspires us to use the regressive model to model the low-frequency part, exploit artificial neural network to characterize the high-frequency component, and then combine these two parts according to the regressive model and artificial neural network to precisely reconstruct end-to-end traffic. Simulation results show that in contrast to previous methods our algorithm is much more effective and promising.