Digitale Kommunikationssysteme


Approximation To Capacity

The characterization of the capacity, i.e., the highest rate for reliable communication, for multi-user systems is an long-standing open problem. Given the ever-increasing demand for higher data-rates anytime and anywhere it is of paramount importance to understand those fundemental limits to obtain

- design guidelines

- important insights

for developing transceivers operating close to optimum, as state-of-the-art strategies are not able to fullfil this demand.

The approach taken in this reseach thrust is to take an engineering approach by slightly changing the objective: Instead of characterizing the capacity exactly, an approxative characterization is pursued, in wich upper and lower bounds on the capacity are derived which differ only by a contant gap. A finer characterization is then obtained by minimizing this gap.  The beauty of this approach is that it provides useful insights for the missing puzzle-pieces for a good unterstanding of the system behavior.

Information Theoretic Security

The central question to which this topic is addressed is whether it is possible to transmit a message to a legitimate receiver in a way that makes it impossible for an unintended receiver or a wire-tapper to decode. Due to the broadcast nature of the wireless  medium, the message transmitted is overheard by all elements in the network for free, so the goal is to construct efficient encoding schemes such that the intended receiver is able to decode the message error-free while the  eavesdropper is unable to get any information from the transmitted signal.

Interference alignment and interference networks

Interference Alignment is a linear pre-coding technique that attempts to align interfering signals in time, frequency or space. In MIMO (Multiple-Input Multiple-Output) networks, where there are multiple transmitter-receiver pairs operating simultaneously in the same wireless channel, one particular transmitter would ideally transmit a signal to one particular receiver and this receiver would solely receive signals from this transmitter. However, in reality, the receivers observe noisy linear combinations of signals from each transmitter present in the same space. The big question is how to effectively ensure that each receiver is able to identify and separate the signal transmitted by its corresponding transmitter from the undesired interference signals. An answer to this lies in interference alignment. Each receiver has its own signal space and therefore, using this technique, all the interference is aligned into one half of this space at each receiver, thereby leaving the other half of the signal space for the desired signal. The alignment is optimized by maximizing the overlap between the signal spaces of all interferences at each receiver. However, the extent to which interference can be aligned over a finite number of signaling dimensions remains unknown. Coordination between transmitter-receiver pairs makes it possible to design the transmit strategies such that the interference aligns at each receiver. This is of special importance in cellular and ad hoc networks because it shows that coordination can help overcome the effects of interference generated by the simultaneous transmission of multiple transmitters. Interference alignment still faces some challenges such as the fact that for it to be successful, global channel knowledge is required.

In multiuser communication, due to the broadcast nature of the wireless communication channel, each receiver observes a noisy version of the sum of the signal transmitted by the corresponding transmitters and the unwanted signals transmitted by closely-located transmitters. Such a communication network is referred to as an interference network. Interference is one of the defining features of a wireless network. How to optimally deal with interference is one of the most critical and least understood aspects of multiuser communication where multiple transmitters wish to communicate with corresponding receivers. An interference network is one of the centerpieces in improving our understanding of wireless systems. It is therefore not surprising that considerable energy has been invested in the research of interference networks by the information theory research community and other related research communities over the past few decades. Even though this research has lead to remarkable advances over the years, many questions remain unanswered.

MIMO und Codierung

MIMO and Space-Time Coding

 

The ground-breaking result by A. Paulraj (Stanford) and E. Telatar (EPFL) showed that by using multiple antennas at both the transmitters and receivers, thus multiple-input multiple-output (MIMO), can significantly improve the data rate and reliability for communication systems.  In order to achieve those promised gains, new tranmission schemes have been developed. Among those, the most famous ones are

- the Alamouti scheme for systems with two transmit antennas: The elegancy of the Alamouit scheme is that it reduces the  optimal ML detector to a linear scheme due to the inherent orthogonality in the transmitted codewords. In addition, if the receiver is equipped with one antenna only, the Alamouti scheme is capacity achieving.

- the stacked Alamouti scheme for systems with 2n transmit anntenas: By backing off slightly from the linear detection, this scheme is achieved the capacity with one receive antenna and approaches the capacity asymtotically if the number of receive antennnas is less then or equal to n antennas.

- Quasi-orthogonal schemes: By backing of slightly from ortogonality,  the important property of achieving unity rate as the Alamouti scheme can be preserved for more than two transmit antennas by achieving full diversity

- BLAST: An architecture which aims at achieving the high rates promised by the theory on MIMO systems

 

MIMO now is considered a standard in today communication systems like in WiFi, MIMO Radar, Cellular systems etc.

 

 

Coding and its applications

Transmitted messages, like data from a satellite, are always subject to noise. It is important, to be able to detect the errors if happened; or encode a message in such a way that after noise scrambles it, it can be decoded to its original form.

Essensially, there are two aspects to coding theory:

  1. Data compression (or, source coding), such as Zip, JPEG and MP3 etc;

  2. Error detection/correction (or, channel coding), such as ISBN numbers and wide applications in storage (e.g.: CDs/DVDs) and communication (e.g.: NASA Mariner deep-space probes) systems.

 

Prototypenentwurf

Das Konzept des Software-Defined Radio (SDR) bezeichnet das Bestreben, möglichst viel der Funktionalität eines Übertragungssystems in Software ausführen zu lassen. Dabei übernehmen digitale Signalprozessoren (DSPs) oder FPGAs die entscheidenden Aufgaben wie Modulation/Demodulation, Signalerzeugung und Codierung. Die Verwendung solcher rekonfigurierbarer Hardware erlaubt eine hohe Flexibiltät. Es können eine Vielzahl von Übertragungsstandards mit der gleichen Hardware in unterschiedlichen Frequenzbändern realisiert werden. Im Idealfall können die Signale von der Antenne direkt über einen A/D-Wandler digitalisiert werden. Da jedoch für hohe Frequenzen häufig keine kostengünstigen A/D-Wandler zur Verfügung stehen, enthalten viele praktische Aufbauten von SDRs eine analoge Hochfrequenzschaltung als erste Mischstufe. Am Lehrstuhl stehen für Lehre und Forschung USRP-Geräte (Universal Software Radio Peripheral) der Firma Ettus Research zur Verfügung, die das SDR-Konzept realisieren. Sie erlauben drahtlosen, gleichzeitigen Sende- und Empfangsbetrieb (Voll-Duplex) für einen Frequenzbereich von 50 MHz bis 2,2 GHz. Die Geräte verfügen über eine Gigabit-Ethernet-Schnittstelle, wodurch eine Ansteuerung von einem PC aus über das IP-Netzwerk möglich ist. Die digitale Signalverarbeitung kann wahlweise auf dem geräteinternen FPGA oder auf einem PC mit Netzwerkzugang durchgeführt werden.

Relay, Multi-Hop, Multi-Way

Relay

Cooperative communication and relaying is one of the important research topics in wireless network information theory. The basic model to study this problem is the 3-node relay channel in which there is a signal transmitter, a relay over which the signal is transmitted and a signal receiver. The purpose of the relay is to support the exchange of information between the transmitter and its intended receiver. The relay is of particular interest in a multi-user setting where there are multiple pairs of corresponding transmitters and receivers. Taking interference, noise and decoding of the signal into account particularly makes relaying a topic of interest as an optimal relaying strategy still remains unknown.

 

Multi-Hop

In cellular networks, communication occurs between a base station and nodes within a cell. A challenge for system designers is to achieve Quality of Services (Q.O.S) for all users, especially those near the cell-edge in urban areas. Promising solutions to this challenge are multi-hop strategies. In multi-hop wireless networks, one or more intermediate nodes are placed along the path between the base station and the receiver. These intermediate nodes serve the purpose of wirelessly receiving and forwarding the packets of information between the base station and the receiver – exactly the same principle as that of an athletic relay race where the baton that is passed from one athlete to the next can be imagined as the packet of information and the athletes as the intermediate nodes. The advantages of multi-hop networks as opposed to networks with single wireless links are that multi-hop networks can extend the coverage of a network over a large area, can improve network connectivity and can be deployed in a cost-effective way. However, multi-hop network methods are still an open research problem, therefore research is still being carried out to tackle questions such as how to optimally forward the information.

Robust Design

  • Treating interference as noise: Optimality
  • Robust Beamforming
  • Imperfect Channel Knowledge
  • Topological Interference Alignment

Systems Characterization, Diagnosis and Prediction

  • Material characterization
  • Self-interference channel estimation and removal
  • Machine diagnosis
  • Predictive Maintenance

Vollduplex-Kommunikation

Die Kommunikation im Vollduplexbetrieb (IBFD, in-band full-duplex) ermöglicht es, gleichzeitig auf dem gleichen Frequenzband zu senden und zu empfangen. Mit dieser IBFD-Technologie kann prinzipiell die spektrale Effizienz (Bits pro Sek. und Hz) verdoppelt werden, ohne dass ein zusätzlicher Bedarf an Bandbreite entsteht. In zukünftigen Kommunikationsstandards wie 5G werden die Anforderungen an die Datenrate erheblich ansteigen. Daher wird eine Neuentwickelung von Kommunikationssystemen benötigt, wobei die Vollduplex-Kommunikation als ein aussichtsreicher Kandidat gilt. IBFD kann auch Verbesserungen in höheren Protokollschichten nach sich ziehen. Das Auftreten von Kollisionen kann besser vermieden oder auch zusätzliche Optionen wie ein sofortiger Rückkanal ermöglicht werden. Es wird erwartet, dass zukünftige Standards auch höhere Anforderungen an die erlaubten Latenzzeiten stellen. Auch hier bietet IBFD das Potenzial zu Verbesserungen.