NEXT: New enhanced available bandwidth measurement technique, algorithm and evaluation

Paul, Anup Kumar; Tachibana, Akira; Hasegawa, Teruyuki

doi:10.1109/pimrc.2014.7136206

Cited by 8 publications

(10 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The NEXT model [1] is well suited to describing and explaining how end-to-end ABW measurement methods work. Therefore, we used our extended model to study the impact of wireless bottleneck links on ABW estimates.…”

Section: A Brief Overview Of Next and The Curve-fitting Techniquementioning

confidence: 99%

“…NEXT has a high-density region of the packet rate between the P 1 and the P 2 region by using two different spread factors 2 γ 1 and γ 2 and adaptively fits that high-density region to the possible ABW for higher accuracy. Details of the bandwidth estimation logic, rate adjustment algorithm, and structure of the probing streams can be found in [1]. Here, we will describe the curvefitting technique for accurately determining the ABW, and, to measure the time required for one estimate, we will formulate the total chirp duration of NEXT-FIT.…”

Section: A Brief Overview Of Next and The Curve-fitting Techniquementioning

confidence: 99%

“…Wireless networks are often used for nomadic access to the Internet from smartphones or laptops. Therefore, it is important to study the performance behavior of end-to-end active ABW estimation tools [1], [2], [3], [4] when wireless links are part of the end-to-end path. If end-to-end active ABW estimation tools are to be widely accepted as performance measurement tools, it is necessary to ensure that they report correct values in wireless networks as well.…”

Section: Introductionmentioning

confidence: 99%

“…The existing methods differ in how probe-packets are sent and the logic used to infer the ABW. NEXT [1] used Pathchirp's excursion detection algorithm (EDA) [2] to find the ABW. However, the algorithm's accuracy in a real network heavily depends on the proper tuning of different parameters and hence it results in inaccurate ABW estimation when there are inadequate thresholds of different parameters.…”

Section: Introductionmentioning

confidence: 99%

“…Our simulation scenario as well as real test scenario consists of one access link to an access point (or base station), since this is the most common structure of wireless networks. To produce measurement results, we used the newly proposed extended version (NEXT-FIT) of our previously proposed method NEXT [1], which measures the ABW quite accurately and in a few RTT. For comparison, we used the tool Pathchirp [2].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

NEXT-FIT: Available Bandwidth Measurement over 4G/LTE Networks -- A Curve-Fitting Approach

Paul

Tachibana

Hasegawa

2016

2016 IEEE 30th International Conference on Advanced Information Networking and Applications (AINA)

Self Cite

View full text Add to dashboard Cite

Available bandwidth (ABW) estimation in wireless networks is a critical issue for quality-of-service (QoS) provisioning. Nowadays, a network path often contains at least one wireless link e.g., mobile users having wireless connection to the Internet through laptop PCs or smartphones. Most of the existing tools for measuring ABW have been developed and evaluated in wired networks. Since the characteristics of wired links and wireless links differ in many respects, such as fluctuations in capacity and stability due to the shared and unreliable nature of the wireless links, ABW estimation tools also need to be evaluated for network paths containing at least one wireless link. In this paper, we have extended our previous work, New Enhanced Available Bandwidth Estimation Technique (NEXT) by introducing a parameter-independent curve-fitting technique to detect the ABW from a one-way queuing delay signature and conducted a real test over a radio interface in a 4G/LTE mobile communication network. For the experiments, a commercial 4G/LTE mobile network of a Japanese mobile operator was used. Extensive simulations over an IEEE 802.11 network were also performed. Simulation results as well as real-world experimental results demonstrate that it is feasible to achieve reliable estimates under certain circumstances. Our extended approach, which we call NEXT-FIT, has very good ABW estimation results under conditions of different packet sizes, dynamic wireless link rates, and channel noises.

show abstract

Section: A Brief Overview Of Next and The Curve-fitting Techniquementioning

confidence: 99%

Section: A Brief Overview Of Next and The Curve-fitting Techniquementioning

confidence: 99%