The widespread use and increasing popularity of broadband service has prompted a focus on the measurement and analysis of its empirical performance in recent studies. The worldwide view of broadband performance has been examined over the short term with Speedtest.net, but research in this area has not yet provided a long-term evolutionary insight on how DSL, Cable, and Fiber access technologies have influenced on user experience. In this study, we present 6 years of measurement results, from 2006 to 2011, of broadband performance with fast developing broadband networks in Korea. With 57% Fiber penetration in 2011, our data consist of a total of 29M test records and 10M subscribers. Over the 6 years, we have observed a 2.9-fold improvement in download speed (57Mbps), 2.8-fold increase in upload speed (38Mbps), and 0.7-fold decrease in latency due to the high penetration rate of Fiber broadband service and the advanced Cable modem technology. In addition, we carried out longitudinal analysis of various aspects of services, providers, regions, and cost-performance. We believe that the evolutionary Korean broadband measurement results can shed light on how high-speed access technologies are substantially enhancing on end-to-end performance.

Round-trip time (RTT) is an important performance metric. Traditional RTT estimation methods usually depend on the cooperation of other networks and particular active or passive measurement platforms, whose global deployments are costly and difficult. Thus a new RTT estimation algorithm, ME algorithm, is introduced. It can estimate the RTT of two hosts communicating through border routers by using TCP CUBIC bulk flow data from those routhers without the use of extra facilities, which makes the RTT estimation in large-scale high-speed networks more effective. In addition, a simpler and more accurate algorithm — AE algorithm — is presented and used when the link has large bandwidth and low packet loss rate. The two proposed algorithms suit sampled flow data because only duration and total packet number of a TCP CUBIC bulk flow are inputs to their calculations. Experimental results show that both algorithms work excellently in real situations. Moreover, they have the potential to be adapted to other TCP versions with slight modification as their basic idea is independent of the TCP congestion control mechanism.

Available bandwidth, along with latency and packet loss rate, is an essential metric for the efficient operation of overlay network applications. However, the measurement of available bandwidth creates a larger traffic overhead than other metrics. Measurement conflicts on route-overlapping paths can also seriously degrade measurement accuracy and cause a non-negligible increase in the network load. In this paper, we propose a distributed method for measuring the available bandwidth in overlay networks that can reduce measurement conflicts while maintaining high measurement accuracy at low cost. Our main idea is that neighboring overlay nodes exchange route information to detect overlapping paths and share the measurement results of overlapping paths to configure parameter settings for available bandwidth measurements. Our simulation results show that the relative errors in the measurement results of our method are approximately only 65% of those of the existing method. The measurement accuracy of our method remains better than that of the existing method when the total measurement traffic loads of both methods are equal.

BitTorrent is one of the most popular P2P file sharing applications worldwide. Each BitTorrent network is called a swarm, and millions of peers may join multiple swarms. However, there are many unreachable peers (NATed (network address translated), firewalled, or inactive at the time of measurement) in each swarm; hence, existing techniques can only measure a part of all the peers in a swarm. In this paper, we propose an improved measurement method for BitTorrent swarms that include many unreachable peers. In essence, NATed peers and those behind firewalls are found by allowing them to connect to our crawlers by actively advertising our crawlers' addresses. Evaluation results show that the proposed method increases the number of unique contacted peers by 112% compared to the conventional method. Moreover, the proposed method increases the total volume of downloaded pieces by 66%. We investigate the sampling bias among the proposed and conventional methods, and we find that different measurement methods yield significantly different results.

Measuring network resource information, including available bandwidth, propagation delay, and packet loss ratio, is an important task for efficient operation of overlay network services. Although measurement accuracy can be enhanced by frequent measurements, performing measurements with high frequency can cause measurement conflict problem that increases the network load and degrades measurement accuracy. In this paper, we propose a low-cost, distributed and conflict-aware measurement method that reduces measurement conflicts while maintaining high measurement accuracy. The main idea is that the overlay node exchanges the route information and the measurement results with its neighboring overlay nodes while decreasing the measurement frequency. This means our method trades the overhead of conducting measurements for the overhead of information exchange to enhance measurement accuracy. Simulation results show that the relative error in the measurement results of our method can be decreased by half compared with the existing method when the total measurement overheads of both methods are equal. We also confirm that exchanging measurement results contributes more to the enhancement of measurement accuracy than performing measurements.

Predicting network throughput is important for network-aware applications. Network throughput depends on a number of factors, and many throughput prediction methods have been proposed. However, many of these methods are suffering from the fact that a distribution of traffic fluctuation is unclear and the scale and the bandwidth of networks are rapidly increasing. Furthermore, virtual machines are used as platforms in many network research and services fields, and they can affect network measurement. A prediction method that uses pairs of differently sized connections has been proposed. This method, which we call connection pair, features a small probe transfer using the TCP that can be used to predict the throughput of a large data transfer. We focus on measurements, analyses, and modeling for precise prediction results. We first clarified that the actual throughput for the connection pair is non-linearly and monotonically changed with noise. Second, we built a previously proposed predictor using the same training data sets as for our proposed method, and it was unsuitable for considering the above characteristics. We propose a throughput prediction method based on the connection pair that uses ν-support vector regression and the polynomial kernel to deal with prediction models represented as a non-linear and continuous monotonic function. The prediction results of our method compared to those of the previous predictor are more accurate. Moreover, under an unstable network state, the drop in accuracy is also smaller than that of the previous predictor.

Network testbeds have been used for network measurement and experiments. In such testbeds, resources, such as CPU, memory, and I/O interfaces, are shared and virtualized to maximize node utility for many users. A few studies have investigated the impact of virtualization on precise network measurement and understood Internet traffic characteristics on virtualized testbeds. Although scheduling latency and heavy loads are reportedly affected in precise network measurement, no clear conditions or criteria have been established. Moreover, empirical-statistical criteria and methods that pick out anomalous cases for precise network experiments are required on userland because virtualization technology used in the provided testbeds is hardly replaceable. In this paper, we show that ‘oversize packet spacing’, which can be caused by CPU scheduling latency, is a major cause of throughput instability on a virtualized network testbed even when no significant changes occur in well-known network metrics. These are unusual anomalies on virtualized network environment. Empirical-statistical analysis results accord with results at previous work. If network throughput is decreased by the anomalies, we should carefully review measurement results. Our empirical approach enables anomalous cases to be identified. We present CPU availability as an important criterion for estimating the anomalies.

With the rapid increase of link speed in recent years, packet sampling has become a very attractive and scalable means in collecting flow statistics; however, it also makes inferring original flow characteristics much more difficult. In this paper, we develop techniques and schemes to identify flows with a very large number of packets (also known as heavy-hitter flows) from sampled flow statistics. Our approach follows a two-stage strategy: We first parametrically estimate the original flow length distribution from sampled flows. We then identify heavy-hitter flows with Bayes' theorem, where the flow length distribution estimated at the first stage is used as an a priori distribution. Our approach is validated and evaluated with publicly available packet traces. We show that our approach provides a very flexible framework in striking an appropriate balance between false positives and false negatives when sampling frequency is given.

In the present paper, ImTCP-bg, a new background TCP data transfer mechanism that uses an inline network measurement technique, is proposed. ImTCP-bg sets the upper limit of the congestion window size of the sender TCP based on the results of the inline network measurement, which measures the available bandwidth of the network path between the sender and receiver hosts. ImTCP-bg can provide background data transfer without affecting the foreground traffic, whereas previous methods cannot avoid network congestion. ImTCP-bg also employs an enhanced RTT-based mechanism so that ImTCP-bg can detect and resolve network congestion, even when reliable measurement results cannot be obtained. The performance of ImTCP-bg is investigated through simulations, and the effectiveness of ImTCP-bg in terms of the degree of interference with foreground traffic and the link bandwidth utilization is also investigated.

This paper proposes a new method of estimating real-time traffic matrices that only incurs small errors in estimation. A traffic matrix represents flows of traffic in a network. It is an essential tool for capacity planning and traffic engineering. However, the high costs involved in measurement make it difficult to assemble an accurate traffic matrix. It is therefore important to estimate a traffic matrix using limited information that only incurs small errors. Existing approaches have used IP-related information to reduce the estimation errors and computational complexity. In contrast, our method, called spike flow measurement (SFM) reduces errors and complexity by focusing on spikes. A spike is transient excessive usage of a communications link. Spikes are easily monitored through an SNMP framework. This reduces the measurement costs compared to that of other approaches. SFM identifies spike flows from traffic byte counts by detecting pairs of incoming and outgoing spikes in a network. A matrix is then constructed from collected spike flows as an approximation of the real traffic matrix. Our experimental evaluation reveals that the average error in estimation is 28%, which is sufficiently small for the method to be applied to a wide range of network nodes, including Ethernet switches and IP routers.

We have developed a measuring system for high-Tc superconducting single-flux quantum circuits and evaluated its performance in terms of bit error rate (BER) measurement for given signal voltage levels. The system includes magnetic shields and a high-frequency test fixture mounted on a closed-cycle cooler. The test fixture is made of non-magnetic material. The transmission characteristics of the measuring system were evaluated by using a vector network analyzer at frequencies ranging from 40 MHz to 20 GHz. The operating temperature of the measuring system ranges from 20 K to room temperature. We connected a 12-GHz wideband pulse amplifier to the system and evaluated its high-speed transmission characteristics. We used a standard 50-Ω microstrip line as an impedance-matched sample. The signal used in the experiment was a 215-1 pseudo random bit signal (PRBS) at 3 Gbps. As a result, the output voltage required for an output driver under the experimental condition was 18.8 mV in order to obtain a resolution of BER measurement of 10-12.