This paper proposes a novel method of UDP multi-pathization with SDN for NFS. Although current UDP protocol, which is one of the transport layer protocol of NFS, cannot handle multiple paths simultaneously, the proposed method can multi-pathize UDP by utilizing SDN without any changes of current TCP/IP protocols. The method distributes packets over multiple paths according to the congestion of the paths and the length of IP packets. Since the control is performed on SDN, no changes of existing protocols, softwares, and devices are required. The authors implemented their method by using OpenFlow and IoT devices, Raspberry Pi 3 as SDN protocols and IoT devices, respectively; they then confirmed the effectiveness of the proposed method by actual experiments.

This paper discusses the most appropriate transmission rate of IEEE 802.11 wireless LAN based on QoE for Web services. Since some QoS parameters have trade-off relationship between each other for the transmission rate, it is difficult to decide the appropriate value of the rate. Thus the authors clarify the relationship between QoE and QoS for Web services by the linear multiple regression analysis. They then adapt measured values of the packet loss rate and those of RTT as QoS parameters for the relationship, and estimate the transmission rate which maximizes the QoE. The authors targeted two actual Web services; shop service and map service; they performed experiments. Their experimental results show that the highest satisfaction can be given when the transmission rate are 9.125 Mb/s and 24.472 Mb/s for the shopping service and the map service, respectively.

In this paper, the authors develop a novel wireless LAN base station with DPI(Deep Packet Inspection) based QoS control, which can maintain QoE during congestion. In the QoS control, the base station always monitors the file type which is found in HTTP messages encapsulated in an IP packet and prioritizes the packet. Since the control is light weight, it can be easily implemented a LAN base station. The authors developed a prototype of the wireless LAN base station with WMM; and they evaluate it by QoE assessment with subjects for actual two Web services. From the experimental results, the authors show the effectiveness of the developed wireless LAN station.

This paper studies evaluation of influence of IP performance degradation on QoE of a Web service by means of the severity rating method. The authors consider Web usability as the QoE and tackle to evaluate it by using Nielsen's severity rating. The severity on the Web usability is one of measures that represent how a problem degrades the usability of a Web service. This paper defines the concrete scales about the severity and evaluates them for the two existing online shopping services by experiments, which consist of the two stages. One is done to clarify problems caused by the IP performance degradation, and the other is for assessment of Web usability with the severity rating. From the results, this study clarifies the four problems and measures the severity for each problem; it then shows how the problems aggravate the Web usability of the services. Consequently, the authors confirm that the severity rating method is effective in evaluation of the influence of the IP quality deterioration on QoE of a Web service.

This paper studies a multipoint-to-multipoint TV conferencing application over IP networks and assesses its user-level QoS with two types of QoS mapping. In utilizing the application, the user perceives quality of the communication with every other conferee; we refer to the quality as individual user-level QoS to a conferee. According to the individual user-level QoS, he/she totally judges the quality of the application, which is refer to as overall user-level QoS for the user. The overall user-level QoS of the application can be affected by the individual one to each conferee; therefore, it is difficult to clarify QoS parameters which affect the overall user-level QoS. This paper tackles the problem by utilizing two types of QoS mapping: mapping between the two kinds of user-level QoS and that between user-level QoS and application-level QoS. In this paper, an experiment with a simple task by three conferees was carried out. The user-level QoS is assessed by one of the psychometric methods. As a result of the two types of QoS mapping, we find two interesting results. First, when a user communicates with the other two conferees, the lower individual user-level QoS has more effect on the overall user-level QoS than the higher one. Second, the individual user-level QoS can depend on not only its application-level QoS but also that of the other conferees.

This paper proposes a session control method which provides a mechanism to achieve desirable user-level QoS (i.e., perceptual QoS) of audio-video transmission over IP networks. The proposed method, which is referred to as GPSQ (Guarantee of Psychologically Scaled Quality), utilizes a user-level QoS parameter of the interval scale in the psychometric analysis. GPSQ involves four kinds of components: media terminals, a SIP server, a QoS manager and bandwidth-controllable routers. From the media terminals, the SIP server obtains the information for this control, which is further delivered to the QoS manager. The QoS manager keeps a database of representative regression lines which express user-level QoS as a function of guaranteed bandwidth of audio and that of video. Using the regression line, the QoS manager calculates necessary guaranteed bandwidth according to user-level QoS specified by the user. Then, the QoS manager sets up routers to guarantee the calculated bandwidth. We implemented GPSQ in a simple experimental network and confirmed its effectiveness.

This paper studies a method of utilizing the cross- modal influences of audio and video, which is referred to as the mutually compensatory property here, in order to enhance user- level QoS (i.e., perceptual QoS) in IP networks. For that purpose, we clarify how the mutually compensatory property depends on the content type. We conducted an experiment on audio-video transmission over a simple IP network with ten contents which are classified into five types. In the experiment, we transmitted CBR audio-video streams with various values of guaranteed bandwidth for audio and that for video under the condition that the total guaranteed bandwidth of the two streams is constant. Then, we assess user-level QoS and perform QoS mapping between user-level and application-level by multiple regression analysis. Using the obtained regression lines, we examine the dependence of the property on the content type. As a result, we confirm that the mutual compensation relationship between audio and video can be classified into three kinds. Furthermore, we investigate how we should allocate the bandwidth between audio and video according to the content type in order to maximize the user-level QoS.

This paper proposes a method for testing feasibility of QoS control based on QoS mapping. We focus on the issue in wireless IP networks with IEEE 802.11 MAC. In wireless multimedia applications, it is important to investigate how we enhance user-level QoS of audio-video transmission. As one of the methods, we can control QoS at lower levels so that user-level QoS keeps high by means of QoS mapping. However, there is no guarantee that a given scheme of QoS control at lower-levels can give desirable user-level QoS. Therefore, we must examine the feasibility of the QoS control. In order to test it, we introduce concepts of the feasibility space and the decision boundary, which are obtained by the QoS mapping with multiple regression analysis. To show the effectiveness of our method, we utilize the simulation results in a previous work by the authors' research group. The simulation treats application-level QoS control and MAC-level one; we adopt EDCA of IEEE 802.11e as MAC-level control.

This paper proposes a real-time estimation method of perceptual QoS (i.e., user-level QoS) of audio-video transmission over IP networks. The estimated quality reflects cross-modal influences of audio and video in the framework of ITU-T Recommendation J. 148. As objective measures of audio and video for the use of the perceptual QoS estimation, we select application-level QoS parameters for media synchronization quality. The perceptual quality is expressed in terms of the user-level QoS parameter of the interval scale, which is referred to as the psychological scale. As the multimedia quality integration function, we propose QoS parameter mapping between the application-level and the user-level. We made subjective quality assessment of simulated audio-video transmission over an IP network by the rating-scale method. We then calculate the psychological scale, utilizing the law of categorical judgment. Furthermore, we obtain the QoS mapping relations by the principal component analysis and multiple regression analysis. The regression formula thus derived can estimate the psychological scale with high accuracy unless the video frame rate is low. Finally we present an extension of the proposed method to continuous time estimation of user-level QoS.

This paper investigates user-level (or perceptual) QoS of audio-video transmission over IP networks with terminals of different size monitors. As the first step of our research, we treat large-sized monitors. By experiment, we assess the user-level QoS for three monitor sizes (80, 60 and 40 inches) and perform QoS mapping. In our experiment, we changed the monitor size by using a projector. The user-level QoS is assessed by the method of successive categories, which is one of the psychometric methods. Moreover, we performed QoS mapping from application-level to user-level with multiple regression analysis. The experimental result indicates that when the degradation of the quality of the audio-video transmission is not noticeable, the user-level QoS of the large monitor size (i.e., 80 inch) becomes the highest. We also observe that when we can notice a low level of degradation of the transmission, the user-level QoS parameter of the large monitor size becomes the smallest. When the degradation is too much, we hardly notice any difference in the user-level QoS between the monitor sizes. We thus confirmed that the monitor size is one of important application-level QoS parameters we must consider

This paper proposes a scheme for quantitative assessment of user-level (or perceptual) quality of service (QoS) for audio-video transmission by means of two psychometric methods: the method of paired comparisons and the law of comparative judgment. Moreover, we discuss QoS mapping from application-level QoS to user-level QoS by principal component analysis and multiple regression analysis. In the assessment, we simulate the transmission of an audio-video stream over a loaded network. In order to investigate the effect of the contents on QoS mapping, we treat two types of audio-video streams. By experiment, we demonstrate that our scheme can construct an interval scale as the user-level QoS parameter for each stream and represent it as a function of two application-level QoS parameters with high accuracy. We notice that the multiple regression line depends on the contents. We also propose the concept of control gain by media synchronization, which indicates how much media synchronization control subjectively lightens the average network load.

This paper investigates the effect of the mean and jitter of end-to-end delay on use-level QoS of live audio-video transmission. The temporal structure of continuous media can be disturbed by the delay jitter of packets. The temporal disturbance causes a subjective degradation of the media. On the other hand, long end-to-end delay also degrades the quality of live media subjectively. Therefore, it is necessary to study the effect of the mean delay and the delay jitter on user-level QoS degradation. In this paper, we assess the effect by experiment and a psychometric method. In our experiment, we actually delay the packets of live audio and video according to the Pareto-normal distribution with a network emulator and assess user-level QoS of the live transmission with the method of successive categories, which is a psychometric method. As a result, we show that the standard deviation of' delay is more dominant than the mean delay in the effect on the user-level QoS. We also perform QoS mapping from application-level to user-level with multiple regression analysis. From the QoS mapping, we clarify the effect of degradation of the application-level QoS caused by the mean and standard deviation of delay on the user-level QoS.

This paper quantitatively evaluates the mutually compensatory property in audio-video transmission over the Internet by the psychometric method. In our previous study, we used MOS (mean opinion score) as the user-level QoS parameter; however, MOS is an ordinal scale, and therefore, its value does not necessarily represent exact difference between the observers sensory magnitude. In order to solve this problem, we construct an interval scale from the MOS values, utilizing the law of categorical judgment. We then examine how closely MOS approximates an interval scale. Furthermore, we obtain QoS mapping relations between application-level and user-level by the principal component analysis and multiple regression analysis. The regression formula thus derived can demonstrate how much one medium compensates for the other in a quantitative way with the interval scale.

Recently, many kinds of real-time applications have become available over IP networks. It is important to measure network performance for such applications before making use of real applications. The authors developed a general purpose traffic measurement tool for IP networks. This system can generate any kind of traffic flexibly and calculate network performance such as throughput, delay and loss rate according to the packers to be observed. In this paper, the concept and implementation of this tool are described in detail. As an example of network measurements, we also conducted subjective assessment tests for Internet telephony and demonstrated the validity of this tool.