SVM‐Based Speaker Verification System for Match‐on‐Card and Its Hardware Implementation

Using biometrics to verify a person's identity has several advantages over the present practice of personal identification numbers (PINs) and passwords. To gain maximum security in a verification system using biometrics, the computation of the verification as well as the storing of the biometric pattern has to take place in a smart card. However, there is an open issue of integrating biometrics into a smart card because of its limited resources (processing power and memory space). In this paper, we propose a speaker verification algorithm using a support vector machine (SVM) with a very few features, and implemented it on a 32‐bit smart card. The proposed algorithm can reduce the required memory space by a factor of more than 100 and can be executed in real‐time. Also, we propose a hardware design for the algorithm on a field‐programmable gate array (FPGA)‐based platform. Based on the experimental results, our SVM solution can provide superior performance over typical speaker verification solutions. Furthermore, our FPGA‐based solution can achieve a speed‐up of 50 times over a software‐based solution.     

Efficient Power‐Saving 10‐Gb/s ONU Using Uplink Usage‐Dependent Sleep Mode Control Algorithm in WDM‐PON

We propose and demonstrate an efficient power‐saving optical network unit (ONU) based on upstream traffic monitoring for 10‐Gb/s wavelength division multiplexed passive optical networks (WDM‐PONs). The power‐saving mode controller uses a μ‐processor and traffic monitoring modules followed by the proposed power‐saving processes to operate the sleep mode ONU. The power consumption of the ONU is effectively reduced from 19.3 W to 6.4 W when no traffic from the users is detected. In addition, we design a power‐saving mechanism based on a cyclic sleep mode operation to allow a connectivity check between the optical line terminal and ONU. Our calculation results show that the WDM‐PON ONU reduces the power consumption by around 60% using the proposed mechanism.     

Cost‐Effective Transition to 40 Gb/s Line Rate Using the Existing 10 Gb/s‐Based DWDM Infrastructure

In this paper, we propose and demonstrate a cost‐effective technique to upgrade the capacity of dense wavelength division multiplexing (DWDM) networks to a 40 Gb/s line rate using the existing 10 Gb/s‐based infrastructure. To accommodate 40 Gb/s over the link optimized for 10 Gb/s, we propose applying a combination of super‐FEC, carrier‐suppressed return‐to‐zero, and pre‐emphasis to the 40 Gb/s transponder. The transmission of 40 Gb/s DWDM channels over existing 10 Gb/s line‐rate long‐haul DWDM links, including 40×40 Gb/s transmission over KT's standard single‐mode fiber optimized for 10 Gb/s achieves successful results. The proposed upgrading technique allows the Q‐value margin for a 40 Gb/s line rate to be compatible with that of 10 Gb/s.     

Multi‐granular,multi‐purpose and multi‐Gb/s monitoring on off‐the‐shelf systems

As an attempt to make network managers’ life easier, we present M³Omon , a system architecture that helps to develop monitoring applications and perform network diagnosis. M³Omon behaves as an intermediate layer between the traffic and monitoring applications that provides advanced features, high performance and low cost. Such advanced features leverage a multi‐granular and multi‐purpose approach to the monitoring problem. Multi‐granular monitoring provides answers to tasks that use traffic aggregates to identify an event, and requires either flow records or packet data or even both to understand it and, eventually, take convenient countermeasures. M³Omon provides a simple API to access traffic simultaneously at several different granularities, i.e. packet‐level, flow‐level and aggregate statistics. The multi‐purposed design of M³Omon allows not only performing tasks in parallel that are specifically targeted to different traffic‐related purposes (e.g. traffic classification and intrusion detection) but also sharing granularities between applications, e.g. several concurrent applications fed from flow records that are provided by M³Omon . Finally, the low‐cost characteristic is brought by off‐the‐shelf systems (the combination of open‐source software and commodity hardware) and the high performance is achieved thanks to modifications in the standard NIC driver, low‐level hardware interaction, efficient memory management and programming optimization. Copyright © 2014 John Wiley & Sons, Ltd.     

Fine‐grained traffic classification based on functional separation

Current efforts to classify Internet traffic highlight accuracy. Previous studies have focused on the detection of major applications such as P2P and streaming applications. However, these applications can generate various types of traffic which are often considered as minor and ignorant traffic portions. As network applications become more complex, the price paid for not concentrating on minor traffic classes is in reduction of accuracy and completeness. In this context, we propose a fine‐grained traffic classification scheme and its detailed method, called functional separation. Our proposal can detect, according to functionalities, different types of traffic generated by a single application and should increase completeness by reducing the amount of undetected traffic. We verify our method with real‐world traffic. Our performance comparison against existing DPI‐based classification frameworks shows that the fine‐grained classification scheme achieves consistently higher accuracyand completeness. Copyright © 2013 John Wiley & Sons, Ltd.     

OW‐SVM: Ontology and whale optimization‐based support vector machine for privacy‐preserved medical data classification in cloud

Cloud is a multitenant architecture that allows the cloud users to share the resources via servers and is used in various applications, including data classification. Data classification is a widely used data mining technique for big data analysis. It helps the learners to discover hidden data patterns by training massive data collected from the real world. Because this trained model is the private asset of an entity, it should be protected from all other noncollaborative entities. Therefore, it is essential to take effective measures to preserve the confidential data. The objective of this paper is to preserve the privacy of the confidential data in the cloud environment by introducing the medical data classification method. In view of that, this paper presents a method for medical data classification using a novel ontology and whale optimization‐based support vector machine (OW‐SVM) approach. Initially, privacy‐preserved data are developed adopting Kronecker product bat approach, and then, ontology is built for the feature selection process. Ontology and whale optimization‐based support vector machine is then proposed by integrating ontology and whale optimization algorithm into SVM, in which ontology and whale optimization algorithm is used for the feasible selection of kernel parameters. The experiment is done using 3 heart disease datasets, such as Cleveland, Switzerland, and Hungarian. In a comparative analysis, the performance of the OW‐SVM approach is compared with that of K‐nearest neighbor, Naive Bayes, decision tree, SVM, and OW‐SVM, using accuracy, sensitivity, specificity, and fitness, as the evaluation metrics. The OW‐SVM approach could achieve maximum performance with accuracy of 83.21%, the sensitivity of 91.49%, specificity of 73%, and fitness of 81.955, outperforming existing comparative techniques.     

Design and Implementation of Open‐Loop Clock Recovery Circuit for 39.8 Gb/s and 42.8 Gb/s Dual‐Mode Operation

This paper proposes an open‐loop clock recovery circuit (CRC) using two high‐Q dielectric resonator (DR) filters for 39.8 Gb/s and 42.8 Gb/s dual‐mode operation. The DR filters are fabricated to obtain high Q‐values of approximately 950 at the 40 GHz band and to suppress spurious resonant modes up to 45 GHz. The CRC is implemented in a compact module by integrating the DR filters with other circuits in the CRC. The peak‐to‐peak and RMS jitter values of the clock signals recovered from 39.8 Gb/s and 42.8 Gb/s pseudo‐random binary sequence (PRBS) data with a word length of 2³¹?1 are less than 2.0 ps and 0.3 ps, respectively. The peak‐to‐peak amplitudes of the recovered clocks are quite stable and within the range of 2.5 V to 2.7 V, even when the input data signals vary from 150 mV to 500 mV. Error‐free operation of the 40 Gb/s‐class optical receiver with the dual‐mode CRC is confirmed at both 39.8 Gb/s and 42.8 Gb/s data rates.     

Low‐cost and high‐performance: VoIP monitoring and full‐data retention at multi‐Gb/s rates using commodity hardware

Voice over IP (VoIP) is increasingly replacing the old public switched telephone network (PSTN) technology. In this new scenario, there are several challenges for VoIP providers. First, VoIP requires a detailed monitoring of both users' quality of service (QoS) and experience (QoE) to a greater extent than in traditional PSTNs. Second, such a monitoring process must be able to track VoIP traffic in high‐speed networks, nowadays typically of multi‐Gb/s rates. Third, recent government directives require that providers retain information from their users' calls. Similarly, the convergence of data and voice services allows operators to provide new services such as full‐data retention, in which users' calls can be recorded for either quality assessment (call centers, QoE) or security purposes (lawful interception). This implies a significant investment in infrastructure, especially in large‐scale networks which require multiple points of measurement and redundancy. This paper proposes a novel methodology, architecture and system to fulfill such challenges, called VoIPCallMon, as well as the data structures and necessary hardware‐tuning knowledge for its development. As distinguishing features, VoIPCallMon provides very high performance, being able to process VoIP traffic on‐the‐fly at high bitrates, novel services and significant cost reduction by using commodity hardware with minimal interference with operational VoIP networks. The performance evaluation shows that the system copes with the VoIP load of real‐world operators. We further evaluated the system performance at a fully saturated 10 Gb/s link and no packet loss was reported, therefore demonstrating the potential of commodity hardware solutions. Copyright © 2014 John Wiley & Sons, Ltd.     

Quality‐of‐Service Mechanisms for Flow‐Based Routers

In this paper, we propose quality of service mechanisms for flow‐based routers which have to handle several million flows at wire speed in high‐speed networks. Traffic management mechanisms are proposed for guaranteed traffic and non‐guaranteed traffic separately, and then the effective harmonization of the two mechanisms is introduced for real networks in which both traffic types are mixed together. A simple non‐work‐conserving fair queuing algorithm is proposed for guaranteed traffic, and an adaptive flow‐based random early drop algorithm is proposed for non‐guaranteed traffic. Based on that basic architecture, we propose a dynamic traffic identification method to dynamically prioritize traffic according to the traffic characteristics of applications. In a high‐speed router system, the dynamic traffic identification method could be a good alternative to deep packet inspection, which requires handling of the IP packet header and payload. Through numerical analysis, simulation, and a real system experiment, we demonstrate the performance of the proposed mechanisms.     

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.     

Support vector machine‐based equalisation for direct‐sequence ultra wideband systems

We proposed the support vector machine (SVM)‐based equalisation schemes for direct‐sequence ultra wideband (UWB) systems. The severe intersymbol interference caused by the UWB channel was formulated as a pattern classification problem in the SVM‐based equaliser, which operates in two main modes: training and detection. We also applied the least squares support vector classifiers (LS‐SVCs) to reduce the training complexity and sparse LS‐SVCs to reduce the detection complexity, with little performance loss compared to SVCs. Simulation results confirm the outperformance of the proposed equalisers over the conventional rake receiver with the same order of complexity for detection, especially when no channel information is known at the receiver. Also, the SVM‐based equalisers in the line‐of‐sight scenario provide a performance close to the case with additive white Gaussian noise only. Copyright © 2011 John Wiley & Sons, Ltd.     

Exploiting packet‐sampling measurements for traffic characterization and classification

The use of packet sampling for traffic measurement has become mandatory for network operators to cope with the huge amount of data transmitted in today's networks, powered by increasingly faster transmission technologies. Therefore, many networking tasks must already deal with such reduced data, more available but less rich in information. In this work we assess the impact of packet sampling on various network monitoring‐activities, with a particular focus on traffic characterization and classification. We process an extremely heterogeneous dataset composed of four packet‐level traces (representative of different access technologies and operational environments) with a traffic monitor able to apply different sampling policies and rates to the traffic and extract several features both in aggregated and per‐flow fashion, providing empirical evidences of the impact of packet sampling on both traffic measurement and traffic classification. First, we analyze feature distortion, quantified by means of two statistical metrics: most features appear already deteriorated under low sampling step, no matter the sampling policy, while only a few remain consistent under harsh sampling conditions, which may even cause some artifacts, undermining the correctness of measurements. Second, we evaluate the performance of traffic classification under sampling. The information content of features, even though deteriorated, still allows a good classification accuracy, provided that the classifier is trained with data obtained at the same sampling rate of the target data. The accuracy is also due to a thoughtful choice of a smart sampling policy which biases the sampling towards packets carrying the most useful information. Copyright © 2012 John Wiley & Sons, Ltd.     

A PLC‐Based Optical Sub‐assembly of Triplexer Using TFF‐Attached WDM and PD Carriers

We have fabricated a planar lightwave circuit (PLC) hybrid‐integrated optical sub‐assembly of a triplexer using a thin film filter (TFF)‐attached wavelength division multiplexer (WDM) and photodiode (PD) carriers. Two types of TFFs were attached to a diced side of a silica‐terraced PLC platform, and the PD carriers with a 45° mirror on which pin‐PDs were bonded were assembled with the platform. A clear transmitter eye‐pattern and minimum receiver sensitivity of ‐24.5 dBm were obtained under 1.25 Gb/s operation for digital applications, and a second‐order inter‐modulation distortion (IMD2) of‐70 dBc was achieved for an analog receiver.     

An accurate traffic classification model based on support vector machines

Network traffic classification is a fundamental research topic on high‐performance network protocol design and network operation management. Compared with other state‐of‐the‐art studies done on the network traffic classification, machine learning (ML) methods are more flexible and intelligent, which can automatically search for and describe useful structural patterns in a supplied traffic dataset. As a typical ML method, support vector machines (SVMs) based on statistical theory has high classification accuracy and stability. However, the performance of SVM classifier can be severely affected by the data scale, feature dimension, and parameters of the classifier. In this paper, a real‐time accurate SVM training model named SPP‐SVM is proposed. An SPP‐SVM is deducted from the scaling dataset and employs principal component analysis (PCA) to extract data features and verify its relevant traffic features obtained from PCA. By employing PCA algorithm to do the dimension extraction, SPP‐SVM confirms the critical component features, reduces the redundancy among them, and lowers the original feature dimension so as to reduce the over fitting and increase its generalization effectively. The optimal working parameters of kernel function used in SPP‐SVM are derived automatically from improved particle swarm optimization algorithm, which will optimize the global solution and make its inertia weight coefficient adaptive without searching for the parameters in a wide range, traversing all the parameter points in the grid and adjusting steps gradually. The performance of its two‐ and multi‐class classifiers is proved over 2 sets of traffic traces, coming from different topological points on the Internet. Experiments show that the SPP‐SVM's two‐ and multi‐class classifiers are superior to the typical supervised ML algorithms and performs significantly better than traditional SVM in classification accuracy, dimension, and elapsed time.     

Fabrication of Transimpedance Amplifier Module and Post‐Amplifier Module for 40 Gb/s Optical Communication Systems

The design and performance of an InGaAs/InP transimpedance amplifier and post amplifier for 40 Gb/s receiver applications are presented. We fabricated the 40 Gb/s transimpedance amplifier and post amplifier using InGaAs/InP heterojunction bipolar transistor (HBT) technology. The developed InGaAs/InP HBTs show a cut‐off frequency (f_T) of 129 GHz and a maximum oscillation frequency (f_max) of 175 GHz. The developed transimpedance amplifier provides a bandwidth of 33.5 GHz and a gain of 40.1 dBΩ. A 40 Gb/s data clean eye with 146 mV amplitude of the transimpedance amplifier module is achieved. The fabricated post amplifier demonstrates a very wide bandwidth of 36 GHz and a gain of 20.2 dB. The post‐amplifier module was fabricated using a Teflon PCB substrate and shows a good eye opening and an output voltage swing above 520 mV.     

DNS‐Class: immediate classification of IP flows using DNS

Nowadays we see a tremendous growth of the Internet, especially in terms of the amont of data being transmitted and new network protocols being introduced. This poses a challenge for network administrators, who need adequate tools for network management. Recent findings show that DNS can contribute valuable information on IP flows and improve traffic visibility in a computer network. In this paper, we apply these findings on DNS to propose a novel traffic classification algorithm with interesting features. We experimentally show that the information carried in domain names and port numbers is sufficient for immediate classification of a highly significant portion of the traffic. We present DNS‐Class: an innovative, fast and reliable flow‐based traffic classification algorithm, which on average yields 99.8% of true positives and < 0.1% of false positives on real traffic traces. The algorithm can work as a major element of a modular system in a cascade architecture. Additionally, we provide an analysis on how various network protocols depend on DNS in terms of flows, packets and bytes. We release the complete source code implementing the presented system as open source. Copyright © 2014 John Wiley & Sons, Ltd.     

Real‐time performance evaluation of asynchronous time division traffic‐aware and delay‐tolerant scheme in ad hoc sensor networks

Wireless infrastructureless networks demand high resource availability with respect to the progressively decreasing energy consumption. A variety of new applications with different service requirements demand fairness to the service provision and classification, and reliability in an end‐to‐end manner. High‐priority packets are delivered within a hard time delay bound whereas improper power management in wireless networks can substantially degrade the throughput and increase the overall energy consumed. In this work a new scheme is being proposed and evaluated in real time using a state‐based layered oriented architecture for energy conservation (EC). The proposed scheme uses the node's self‐tuning scheme, where each node is assigned with a dissimilar sleep and wake time, based on traffic that is destined for each node. This approach is based on stream's characteristics with respect to different caching behavioral and storage‐capacity characteristics, and considers a model concerning the layered connectivity characteristics for enabling the EC mechanism. EC characteristics are modeled and through the designed tiered architecture the estimated metrics of the scheme can be bounded and tuned into certain regulated values. The real‐time evaluation results were extracted by using dynamically moving and statically located sensor nodes. A performance comparison is done with respect to different data traffic priority classifications following a real‐time asymmetrical transmission channel. Results have shown the scheme's efficiency in conserving energy while the topology configuration changes with time. Copyright © 2009 John Wiley & Sons, Ltd.     

基于M值概率分布的网络视频流分类

为了改善网络视频流的细粒度分类效果,该文分析视频流传输过程中的特征变化与流分类之间的关系。根据不同类型的视频流具有不同的下行传输速率变化模式,提出一种新的基于下行速率传输的视频流分类特征--M值概率分布,并使用支持向量机(SVM)实现网络视频流的分类。实验结果表明,M值概率分布相比较于已有的常见流特征,可以更好地实现6种典型的网络视频流分类。     

Pseudo Optical PAM‐N Signal Using Externally Modulated Lasers

We propose a pseudo optical N‐level pulse‐amplitude modulation (PO PAM‐N) signal using a few externally‐modulated lasers (EMLs) operating at different wavelengths, which is suitable for upgrading the transmission speed over an optical link of ＜ 10 km single‐mode fiber with low‐cost components. To compare a PO PAM‐N signal with that of a standard optical PAM‐N signal, we perform experiments for evaluating the performance of a 51.56‐Gb/s PO PAM‐4 signal and standard 51.56‐Gb/s optical PAM‐4 signal. The receiver sensitivity (at ) of the PO PAM‐4 signal is 1.5 dB better than the receiver sensitivity of a standard optical PAM‐4 signal. We also investigate the feasibility of PO PAM‐N signals operating at 103.12 Gb/s, considering relative intensity noise, timing jitter, extinction ratio (ER) of EMLs, and dispersion. From the results, a PO PAM‐8 signal performs better than PO PAM‐4 and PO PAM‐16 signals at 103.12 Gb/s. Finally, we suggest a timing control method to suppress the effect of dispersion in a PO PAM‐N signal. We show that the tolerance to dispersion of a 103.12‐Gb/s PO PAM‐8 signal can be improved to ±40 ps/nm by applying a proposed scheme.     

Efficient implementation of a frame aggregation unit for optical frame-based switching

Aggregating Ethernet frame or IP packet in large fixed-size frames allows for building scalable core network architectures. Classifying the arriving traffic based on destination core node information and quality of service parameters alleviates the need of performing table look-ups on packet basis. These advantages come at the cost of extra logic at the network egress, as regards implementation, and additional jitter due to the frame assembly process. This paper describes the efficient implementation of a frame aggregation unit that gathers Ethernet packets in G.709 containers, handles 10 Gb/s links, performs classification based on 24-byte headers, and includes a highly pipelined Queue Manager to cope with the considered rates while a specific scheduler controls the quality of service per core network flow. Based on the developed demonstrator, we provide results both as regards area and performance for an FPGA (field programmable gate array) Virtex-4 implementation as well as regarding the introduced jitter.