DCNET 2019 Abstracts


Full Papers
Paper Nr: 4
Title:

Estimating TCP Congestion Control Algorithms from Passively Collected Packet Traces using Recurrent Neural Network

Authors:

Naoki Ohzeki, Ryo Yamamoto, Satoshi Ohzahata and Toshihiko Kato

Abstract: Recently, as various types of networks are introduced, a number of TCP congestion control algorithms have been adopted. Since the TCP congestion control algorithms affect traffic characteristics in the Internet, it is important for network operators to analyse which algorithms are used widely in their backbone networks. In such an analysis, a lot of TCP flows need to be handled and so the automatically processing is indispensable. Thin paper proposes a machine learning based method for estimating TCP congestion control algorithms. The proposed method uses a passively collected packet traces including both data and ACK segments, and calculates a time sequence of congestion window size for individual TCP flows contained in the trances. We use s recurrent neural network based classifier in the congestion control algorithm estimation. As the results of applying the proposed classifier to ten congestion control algorithms, the major three algorithms were clearly classified from the packet traces, and ten algorithms could be categorized into several groups which have similar characteristics.

Paper Nr: 8
Title:

Using Network Traces to Generate Models for Automatic Network Application Protocols Diagnostics

Authors:

Martin Holkovič, Ondřej Ryšavý and Libor Polčák

Abstract: Network diagnostics is a time-consuming activity that requires an administrator with good knowledge of network principles and technologies. Even if some network errors have been resolved in the past, the administrator must spend considerable time removing these errors when they reoccur. This article presents an automated tool to learn the expected behavior of network protocols and possible variations. The created model can be used to automate the diagnostic process. The model presents a finite automaton containing protocol behavior for different situations. Diagnostics of unknown communication is performed by checking the created model and searching for error states and their descriptions. We have also created a proof-of-concept tool that demonstrates the practical potential of this approach.

Short Papers
Paper Nr: 2
Title:

Evaluation of the Impact of e-Services on Enterprise Broadband Demand in the German Market

Authors:

Erik Massarczyk and Peter Winzer

Abstract: The worldwide broadband penetration and internet usage are increasing. However, often the research regarding the demand for a better broadband availability and higher broadband connection speeds focuses on private households, whereas the need for higher bandwidths of enterprises is mostly unconsidered. Although some market overviews also consider the broadband requirements of enterprises, the research lacks a study with a clear focus on the broadband market of enterprises in Germany. To increase the research-based knowledge about the needs of enterprises regarding the usage of broadband connections, a survey of enterprises in Germany (with focus on the Rhine-Main area) has been performed. To strengthen the insight about the impact factors for the use of higher broadband connection speeds, elements like e.g. the expected performance or price-performance ratio will be analyzed by means of the "Unified Theory of Acceptance and Use of Technology 2". The first results of the survey indicate that the main drivers for internet usage of enterprises are the availability of higher connection speeds in combination with better price-performance-ratios.

Paper Nr: 5
Title:

Power-aware Algorithms for Energy-efficient Elastic Optical Backbone and Metro Networks

Authors:

Georgia A. Beletsioti, Stathis Mavridopoulos, Georgios A. Tziroglou, Constantine A. Kyriakopoulos, Georgios I. Papadimitriou, Petros Nicopolitidis and Emmanouel Varvarigos

Abstract: Research in Optical Networking has recently focused on Elastic Optical Network architectures, that support elastic band connections to increase spectrum availability, support high transmission rates and reduce network costs. Elastic optical networks offer flexibility in the way capacity is assigned to connections and are considered the most prevalent solution for the next generation metro/backbone networks. Reduction in energy consumption is an important issue in such networks. In this work, a new power aware algorithm is introduced, which selectively switches off network links under low utilization scenarios supporting energy efficiency. A new power-aware scheme is proposed, which reduces the total energy consumption, while maintaining a low blocking probability under dynamic traffic. Extensive simulation results are presented, which indicate that the proposed heuristic algorithm achieves a power saving of up to 9%, compared to a simple energy unaware dynamic RSA algorithm.

Paper Nr: 6
Title:

A Novel Hop-distance Sensitive Approach to Elastic Optical Networks RSA Algorithms

Authors:

Stathis B. Mavridopoulos, Georgia A. Beletsioti, Georgios A. Tziroglou, Constantine A. Kyriakopoulos, Petros Nicopolitidis, Georgios I. Papadimitriou and Emmanouel Varvarigos

Abstract: Elastic optical networks (EON) allow great flexibility through finer spectrum allocation granularity when compared to traditional WDM solutions. Their improved spectrum efficiency makes them a promising solution for next-generation backbone and metropolitan networks. Distant connections in elastic optical networks that are routed through multiple hops suffer from increased bandwidth blocking probability (BBP), in contrast to easier formulation of more direct connections. Traditional BBP as a metric fails to capture this phenomenon. In this work, a normalization of BP to the connection’s hop distance is proposed and a novel low complexity algorithm is presented that takes this new metric into consideration. Simulation results show that the proposed scheme improves network performance and fairness with no deterioration of BBP, when compared to the FirstFit RSA algorithm.