Architekturen für Multihop-Netzwerke für die drahtlose Nachrichtenübertragung wurden vor über zehn Jahren vorgeschlagen. Derzeit werden verschiedene Systemaspekte, mögliche Anwendungen und die Standardisierung in der wissenschaftlichen Gemeinschaft diskutiert. Leitungsgebundene und drahtlose Verbindungen in der HAEC Box bilden ein Multihop-Netzwerk mit speziellen Rand­bedingungen, Eigenschaften und Möglichkeiten zur kooperativen Kommunikation. Das Projekt A03 möchte zur Erkenntnis der grundlegenden Funktionsweise von solchen Multihop- und Mehrwege-Netzwerken einen wissenschaftlichen Beitrag leisten. Dabei stehen insbesondere die Verhältnisse zwischen der Spektral- und der Energieeffizienz sowie die der Latenz und Komplexität im Mittelpunkt des Interesses.

Rolle innerhalb des SFB 912

© SFB 912

The project addresses one of the key parts of the high-speed computing platform and is therefore well connected to several other projects: The specification of the RF link is necessary to model the wireless links and their weights in the network graph appropriately. Therefore, discussions on channel models and development of statistical models for the wireless link are needed as input from project A05 (Plettemeier). The models for the fixed onboard links need to be developed and discussed together with project A07 (Ellinger/Wolter). The signal processing capabilities and beamsteering properties are developed jointly with project A02 (Fettweis/Fischer). The stochastic traffic model is needed as input from project A04 (Nagel/Müller). The output of this project to the projects A04 (Nagel/Müller) and B04 (Härtig) are descriptions of delay and success probabilities of the onboard and backplane wired and wireless communication including models of the energy consumption. The security aspect of network coding is going to be discussed in project A08 (Jorswieck/Franz). The developed network code design should be implemented on a simplified demonstrator component developed by A02 (Fettweis/Fischer) and A05 (Plettemeier).
In particular, in Phases II and III the input from projects A02 (Fettweis/Fischer), A04 (Nagel/Müller), A05 (Plettemeier) and A08 (Jorswieck/Franz) is of importance. The more abstract graph representation of the NoC will then be replaced by representatives of typical network layouts in the HAEC Box. The assumptions on the amount and type of traffic (sources, destinations, multicast, unicast) will be substantiated. We expect that the second development cycle will lead to fundamentally new problem statements and novel constraints for our network code design and analysis. As pointed out above, the vision is to have a real cross-layer design of network coding, beamforming, routing and parallel processing algorithms. This is an ambitious goal to be achieved after three phases.

Mitarbeiter 

Principal Investigator

• Prof. Dr.-Ing. Eduard Jorswieck

Postdoc

• Dr.-Ing. Christian Scheunert

PhD Student

• Dipl.-Ing. Bho Matthiesen

ehemalige Mitarbeitende

• M. Sc. Andrew Lonnstrom