Takashi Oshiba

Quick end-to-end available bandwidth estimation for QoS of real-time multimedia communication

We propose PathQuick, a quick end-to-end available bandwidth estimation method. When PathQuick is used in real-time multimedia communication, such as real-time photo sharing and video conferencing, just before the transmission of media data such as photo and video, it can quickly complete the estimation of the latest available bandwidth. Consequently, the bandwidth consumption of media data can be controlled to within the available bandwidth estimated at the beginning of media data transmission, and this will prevent delay and packet loss. Thus, QoS at the beginning of real-time photo sharing and video conferencing can be ensured. Conventional methods have a critical restriction that they require a long estimation time; e.g., several seconds or several ten seconds. Using these methods just before the media data transmission causes an unacceptably long waiting time until media data transmission begins, thus degrading real-time responsiveness. PathQuick achieves quick estimation by using a probing packet train (i.e., a set of multiple probing packets) with distinctive features: each packet within the packet train is placed at an equal time interval, and each packet size linearly increases as the packet sequence proceeds. Our evaluation of PathQuick has shown that its estimation duration is several hundred milliseconds, which is more than four times as fast as a conventional method. We also found that its probable range is more than a hundred times as wide as that of the conventional method when a limited waiting time is set to avoid degrading real-time responsiveness.

Personalized advertisement-duration control for streaming delivery

This paper describes the development of a streaming advertisement delivery system that controls the insertion of streaming advertisements into streaming content.Conventional personalization techniques lack a time-control function for advertisement insertion, so the advertisement exposure for each user access can become excessive, much to the annoyance of viewers. This could devalue streaming content by making it less attractive.In our technique, advertisement insertion control is based on the history of each viewer. This personalization method makes it possible to maintain a balanced ratio of the advertisement length to the content length. As a result, our technique should encourage the growth of Internet streaming services and enable more effective and less intrusive advertising.

Quick and simultaneous estimation of available bandwidth and effective UDP throughput for real-time communication

We propose a method, called PathQuick2, for quickly and simultaneously estimating both of the available bandwidth that is unused capacity of an end-to-end path and the effective UDP throughput that is the receiving rate of an UDP flow which can pass through the path, with a single measurement. In PathQuick2, a sender transmits a probing packet train (i.e., a set of multiple probing packets) that each packet is placed at an equal time interval, and each packet size increases as the packet sequence proceeds. A receiver produces the estimated available bandwidth and the estimated effective UDP throughput at the same time from the single packet train. To this end, the receiver detects a packet at which the observed time intervals begin increasing, and the per-packet receiving rate becomes the estimated available bandwidth. Then, the receiver detects a packet at which the observed per-packet receiving rates stop increasing even if the packet sequence proceeds, and the per-packet receiving rate becomes the estimated effective UDP throughput. Our evaluation of PathQuick2 has shown that its estimation duration is only 182 ms, and its probing load is about 90 kB which is more than 160 times as light as a direct measurement of effective UDP throughput. PathQuick2 can provide two useful metrics for real-time communication applications such as video chat and video conferencing; the available bandwidth indicates an upper limit of video sending rate which enables stable video communication without packet loss, and the effective UDP throughput indicates an upper limit of video receiving rate that can pass through an end-to-end path, with a given video sending rate.

Ad-Hoc Endpoint Pairing Based on Simultaneous Clicking for Ubiquitous Communications

The spread of ubiquitous communications is dramatically changing the way of daily work. In this paper, we present an ad-hoc endpoint pairing method that enables a user to easily establish a pair of endpoints, such as an IP-hard phone (fixed or mobile IP-phone) and a PC-based soft phone, simply by simultaneously clicking a GUI button on the display screen of each endpoint. Using the endpoint pairing information, the user can easily launch a PC-to-PC Web/data conferencing application during a phone-to-phone audio conversation. It can thus promote deeper mutual understanding between users.

Robust available bandwidth estimation against dynamic behavior of packet scheduler in operational LTE networks

We shed new light on the mechanism behind how the dynamic behavior of a packet scheduler at the link layer in mobile networks degrades the accuracy of conventional available bandwidth (i.e., unused capacity of an end-to-end path) estimation methods that use a probing packet train (i.e., a set of multiple probing packets). Most of the conventional methods, which were originally designed for wired networks, estimate available bandwidth at the receiver by detecting changes of the observed queuing delays of probing packets. They utilize a microscopic approach in which they check the difference of the queuing delay of each packet on a packet-by-packet basis in order to detect the queuing delay changes. We found that the dynamic behavior of a packet scheduler at the link layer dramatically disturbs the queuing delays observed at the receiver. The disturbed queuing delays make it tremendously difficult for the conventional microscopic approach to detect changes of the delays, resulting in degraded estimation accuracy.

Multi-step-ahead prediction of IP packet delay variation based on a GARCH model

We propose JitSeer, a long horizon prediction method for the delay variation of IP packets; i.e., jitter. In designing JitSeer, we directed our attention to similarities between delay variation, which is the variation of one-way delay, and volatility in finance, which is the variation of financial asset return. From a time series view, the similarities are that both vary over time, and during some periods they swing quickly while in other periods they stay calm. Based on these similarities, the delay variation is modelled with a GARCH model, which is a time series model originally developed in econometrics to capture the dynamics of volatility. Then the model predicts the future multi-step-ahead delay variation. Experimental results obtained from operational mobile networks show that the prediction accuracy of JitSeer is better than that of conventional methods; its prediction accuracy for the next 10 seconds is 74% over 3G and 88% over LTE.

Device pairing based on adaptive channel fluctuation control for large-scale organizations

We are developing device pairing methods that enable a user to establish two devices, such as a mobile phone and a PC as a device pair. In our previous developed method, a user can create a device pair by simultaneously clicking the pairing buttons shown on their screens. However, when the system handles many users, clicks from different users could overlap, resulting in request collisions.

Experimental comparison of machine learning-based available bandwidth estimation methods over operational LTE networks

We propose PathML, an available bandwidth (i.e., unused capacity of an end-to-end path) estimation method based on a data-driven paradigm that uses machine learning with a large amount of data. An experiment over an operational LTE network was performed to compare our method with prior work.

Empirical Evaluation and Analysis of Application-Layer Delay Reduction Methods over Wireless Access Networks

In this paper, we discuss the experimental results of application-layer delay reduction methods, which can reduce the delay during TCP applications. Due to the popularization of smart phones, Wireless Local Area Network (WLAN) access such as WLAN hot spot or WLAN tethering is increasing. This increase leads to a WLAN channel being shared by multiple data flows, resulting in the degradation of the communication performance. Specifically, the delay of the real time applications that communicate over TCP can be made longer. To reduce the delay, we developed three methods: reducing the amount of data to send, triggering TCP’s fast retransmission earlier, and returning TCP’s acknowledgment immediately. In our experiments, we measured the delay during the real time TCP applications connected with each other over WLAN and commercial fixed network and analyzed the effectiveness of our methods in the case of WLAN performance being degraded by the file transfer flow. Results indicate that our methods can reduce the 95th percentile of the delay by as much as 84%.