Georg Carle | Technische Universität München (original) (raw)

Papers by Georg Carle

This contribution discusses the benefits of AAL-level FEC (Forward Error Correction) scheme and t... more This contribution discusses the benefits of AAL-level FEC (Forward Error Correction) scheme and the implementation complexity for both user's and network's point of views. Associated with all service categories (i.e., CBR, VBR, ABR, UBR), an AAL-level FEC scheme achieves some benefits, while improving the end-to-end service quality. Implementing an AAL-level FEC scheme will not require a large effort. A simple definition of additional SSCS code points would allow to indicate the use of an AAL-level FEC scheme for specific virtual circuits. For the end-systems, the support of AAL-level FEC is optional. Source:

The management of networks is automated by closed loops. Concurrent closed loops aiming for indiv... more The management of networks is automated by closed loops. Concurrent closed loops aiming for individual optimization cause conflicts which, left unresolved, leads to significant degradation in performance indicators, resulting in sub-optimal network performance. Centralized optimization avoids conflicts, but impractical in large-scale networks for time-critical applications. Distributed, pervasive intelligence is therefore envisaged in the evolution to B5G networks. In this letter, we propose a Q-Learning-based distributed architecture (QLC), addressing the conflict issue by encouraging cooperation among intelligent agents. We design a realistic B5G network slice auto-scaling model and validate the performance of QLC via simulations, justifying further research in this direction.

ArXiv, 2019

Cyber-Physical Systems (CPS) consist of distributed engineered environments where the monitoring ... more Cyber-Physical Systems (CPS) consist of distributed engineered environments where the monitoring and surveillance tasks are governed by tightly integrated computing, communication and control technologies. CPS are omnipresent in our everyday life. Hacking and failures of such systems have impact on critical services with potentially significant and lasting consequences. In this paper, we review which requirements a CPS must meet to address the challenges of tomorrow. Two key challenges are understanding and reinforcing the resilience of CPS.

ArXiv, 2021

Network automation is gaining significant attention in the development of B5G networks, primarily... more Network automation is gaining significant attention in the development of B5G networks, primarily for reducing operational complexity, expenditures and improving network efficiency. Concurrently operating closed loops aiming for individual optimization targets may cause conflicts which, left unresolved, would lead to significant degradation in network Key Performance Indicators (KPIs), thereby resulting in sub-optimal network performance. Centralized coordination, albeit optimal, is impractical in large scale networks and for time-critical applications. Decentralized approaches are therefore envisaged in the evolution to B5G and subsequently, 6G networks. This work explores pervasive intelligence for conflict resolution in network automation, as an alternative to centralized orchestration. A Q-Learning decentralized approach to network automation is proposed, and an application to network slice auto-scaling is designed and evaluated. Preliminary results highlight the potential of th...

Cognitive Autonomous Networks (CAN) deploys learning based Cognitive Functions (CF) instead of co... more Cognitive Autonomous Networks (CAN) deploys learning based Cognitive Functions (CF) instead of conventional rule-based SON Functions (SF) as Network Automation Functions (NAF) to increase the system autonomy. These CFs work in parallel sharing the same resources which give rise to conflicts among them which cannot be resolved using conventional rule based approach. Our main target is to design a Controller which can resolve any type of conflicts among the CFs in a dynamic way.

This contribution discusses the scope of SSCOP and the potential benefits of combining SSCOP with... more This contribution discusses the scope of SSCOP and the potential benefits of combining SSCOP with an AAL-level FEC scheme. SSCOP is a data link level protocol that provides reliable point-to-point services. It is designed as an SSCS on top of AAL5-CPCS. Its mechanisms for error and flow control are suitable for a wide range of connection parameters, such as error rate, bandwidth, and round trip delay. SSCOP employs frame-based selective retransmissions. Initially, SSCOP has been specified as an AAL protocol for signaling. It also can be applied for reliable communication between ATM end systems in general. In a pure ATM environment, SSCOP allows higher performance than TCP due to its superior error and flow control mechanisms. While SSCOP can achieve better performance than TCP, erroneous frames lead to retransmissions that can severely degrade service quality. Combining SSCOP with an AAL-level FEC scheme allows to provide a better service quality by reducing the probability of retr...

A number of protocol architectures for the provision of scalable reliable multicast services over... more A number of protocol architectures for the provision of scalable reliable multicast services over the Internet have been proposed. These protocol architectures could also be applied for ATM networks. However, since ATM networks are based on switches rather than on broadcast-based subnetworks (e.g., Ethernet and FDDI), some architectural modification is needed, while a modification the end-station's protocol is not forced. As shown in this memo, in order to provide efficient and scalable reliable multicast over an ATM cloud, we need an AAL-level FEC scheme to provide sufficiently small packet error ratio. Notice : This contribution has been prepared to assist the ATM Forum, and is offered by the above affiliations as a basis for discussion. The above listed affiliations reserves the right to add, amend or withdraw the statements contained herein.

2021 IFIP/IEEE International Symposium on Integrated Network Management (IM), 2021

Cognitive Autonomous Networks (CAN) promises to raise the level of operational autonomy in mobile... more Cognitive Autonomous Networks (CAN) promises to raise the level of operational autonomy in mobile networks through the introduction of Artificial Intelligence (AI) and Machine Learning (ML) in the network processes. In CAN, learning based functions, called Cognitive Functions (CF), adjust network control parameters to optimize their objectives which are different Key Performance Indicator (KPI). As the CFs work in parallel, there is often an overlap among their activities regarding control parameter adjustment, i.e., at one point of time, multiple CFs may want to change a single control parameter albeit by different degrees or to different values depending on their respective levels of interest in that parameter. To resolve this dispute, a coordination mechanism is required for sharing the parameter among the independent CFs according to their individual interest levels. In this paper we provide the design of such a Controller in CAN to determine the optimal control parameter value....

Intent Based Networks (IBNs) are mainly used to transform a user's intent into network config... more Intent Based Networks (IBNs) are mainly used to transform a user's intent into network configuration, operation, and maintenance strategies. IBN is a prominent feature for designing the AI-enabled next generation networks. In this paper we propose intent-driven orchestration of cognitive autonomous RAN for managing network control parameters.Our proposed design enables the Controller to detect and remove contradiction, which may arise from a single intent, in the runtime. We also provide a brief overview of existing standards on this aspect and standardization impact of our research to show that it conforms with the worldwide mobile network management standardization efforts.

As most of today's Internet traffic is provided by TCP, this also means that the overall performa... more As most of today's Internet traffic is provided by TCP, this also means that the overall performance is heavily dependent on the quality of these TCP flows. To make it possible to evaluate the state of connections or networks for further potential problem assessment, it is necessary to introduce metrics that act as key performance indicators. This paper introduces multiple of these performance metrics and classifies them into latency, packet loss, throughput and other indicators. As a next step, various methods to conduct measurements and data processing approaches to extract valuable information are presented and discussed. Finally, this allows to draw conclusions about a flow's or network's quality and state.

Vor Ihnen liegen die Proceedings der Seminare "Future Internet" (FI) und "Innovative Internettech... more Vor Ihnen liegen die Proceedings der Seminare "Future Internet" (FI) und "Innovative Internettechnologien und Mobilkommunikation" (IITM). Wir sind stolz, Ihnen Ausarbeitungen zu aktuellen Themen, die im Rahmen unserer Seminare im Sommersemester 2016 an der Fakultät für Informatik der Technischen Universität München verfasst wurden, präsentieren zu dürfen. Den Teilnehmerinnen und Teilnehmern stand es wie in der Vergangenheit frei, das Paper und den Vortrag in englischer oder in deutscher Sprache zu verfassen. Dementsprechend finden sich sowohl englische als auch deutsche Paper in diesen Proceedings. Unter allen Themen, die sich mit Aspekten der Computernetze von morgen befassen, verliehen wir in jedem der beiden Seminare einen Best Paper Award. Im IITM Seminar ging dieser an Herrn Frederic Naumann, der in seiner Ausarbeitung "Garbled Circuits" die Funktionsweise verschlüsselt ablaufender Berechnungen betrachtet. Im FI Seminar wurde dieser Herrn Dominik Scholz verliehen für seine Ausarbeitung "Diving into Snabb", in welcher er das Paketverarbeitungsframework Snabb analysierte. Einige der Vorträge wurden aufgezeichnet und sind auf unserem Medienportal unter https://media. net.in.tum.de abrufbar. III Wir hoffen, dass Sie den Beiträgen dieser Seminare wertvolle Anregungen entnehmen können. Falls Sie weiteres Interesse an unseren Arbeiten habe, so finden Sie weitere Informationen auf unserer Homepage https://www.net.in.tum.de.

Wir hoffen, dass Sie den Beiträgen dieser Seminare wertvolle Anregungen entnehmen können. Falls S... more Wir hoffen, dass Sie den Beiträgen dieser Seminare wertvolle Anregungen entnehmen können. Falls Sie weiteres Interesse an unseren Arbeiten habe, so finden Sie weitere Informationen auf unserer Homepage http://www.net.in.tum.de.

In the past few years network security threats have increased significantly. Methods for attacks ... more In the past few years network security threats have increased significantly. Methods for attacks have not only grown in diversity but also became more sophisticated. The increased need for security mechanisms and countermeasures requires a comprehensive understanding of those attacks and their characteristics. To organize the knowledge of attacks a large variety of classifications were proposed in form of taxonomies and ontologies. The development of these classifications has emerged as an effective means for developing awareness systems and creating common descriptive languages. However, due to the high diversity of attacks no standard classification of network attacks exists so far. In this paper, a survey of existing attack taxonomies and ontologies is presented to create an overview of conducted work in this field of research. Furthermore, issues and drawbacks are discussed in a comparative analysis. The conducted survey has shown the need of a flexible, standardized classification of attacks and vulnerabilities to enable efficient knowledge sharing among the scientific community.

Decentralized or P2P (peer-to-peer) VPNs (virtual private networks) are popular due to their adva... more Decentralized or P2P (peer-to-peer) VPNs (virtual private networks) are popular due to their advantage over centralized VPNs or the classic client-server model. This paper surveys the main difficulties of P2P VPNs in general, but most importantly of bootstrapping a new peer in such a network which is the mechanism of a new peer finding and joining it. Different solutions are indicated on existing implementations and evaluated separately. Especially the two contrasting models of supernodes versus the joining over an existing public overlay are examined for they are most commonly deployed in existing P2P VPNs. It becomes apparent that the presented solutions only partly cover the needed functionality, especially in terms of security and the traversal of NAT devices.

ArXiv, 2019

ArXiv, 2021

2021 IFIP/IEEE International Symposium on Integrated Network Management (IM), 2021