Detailseite
Projekt Druckansicht

Nichtkohärente Übertragungsverfahren für MIMO-Kanäle und deren Erweiterung auf drahtlose Netze mit Relaisstationen

Fachliche Zuordnung Elektronische Halbleiter, Bauelemente und Schaltungen, Integrierte Systeme, Sensorik, Theoretische Elektrotechnik
Förderung Förderung von 2011 bis 2015
Projektkennung Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 210322147
 
Erstellungsjahr 2016

Zusammenfassung der Projektergebnisse

The project investigates physical-layer aspects of wireless networks in the noncoherent setting, i.e. in the absence of a priori channel knowledge at the network nodes. The study is motivated by several practical scenarios where channel state information is either not available or comes at a large cost. This situation arises for example in the high mobility scenario where it is not desirable for the receiver to feed back channel state information (CSI) to the users because it may be outdated. It may also arise in the initialization phase of a communication link, e.g., when a mobile terminal joins a cellular network. This, for example, is the case with the emerging scenarios of Machine-Type Communication (MTC) where: i) the message of each user is very short, such that there is no dedicated metadata to detect the activity; ii) the users (devices) are simple and can neither use precoding nor invest resources to enable channel estimation at the receiver. The issues related to acquisition of channel/network knowledge and, to communication overhead in general, emerge as particularly relevant in light of the novel 5G architectures and technologies, such as cloud-radio access networks and large MIMO systems. In particular: i) the requirement on sufficiently accurate channel state information (CSI) to facilitate coherent processing at multiple antennas is crucial for the performance of large (massive) MIMO systems; ii) in a Cloud-Radio Access Network (C-RAN), where baseband processing functionality is implemented at a centralized cloud processor or Central Unit (CU) on behalf of multiple distributed Remote Radio Heads (RRHs), the capacity limitations of the fronthaul links (along with the associated latency) are understood to offer the most significant challenge to the implementation of C-RANs. Motivated by these observations, the project focuses on two main aspects. First, we derive results for the performance limits of some of the fundamental channels (information-theoretic models) which serve as building blocks of more complex wireless networks. In particular, we study the multiple access channel, both in the ergodic and in the transient regime (i.e. the random access scenario), the relay-aided interference channel and the two-way relay channel. In addition, significant attention is devoted to the (noncoherent) multiple access channel with random activation of users (i.e. the random access channel), with particular reference to massive access in MTC and Internet-of-Things (IoT) applications. Based on the insights from these analyses, in the second part of the project we address problems of relevance in C-RANs and large MIMO systems. In the context of C-RANs, we study the impact of the fronthaul limitations on the random access performance in a system with potentially massive number of users. In the context of large antenna systems, we assess the performance of simple noncoherent receiver architectures based on energy detection. Importantly, the project served as catalyst for intensive collaboration with colleagues from international institutions (NJIT, Chalmers University of Technology, Aalborg University and University Carlos III in Madrid). The collaboration resulted in joint publications and served as basis for project proposals and research visits.

Projektbezogene Publikationen (Auswahl)

 
 

Zusatzinformationen

Textvergrößerung und Kontrastanpassung