Armin Schmidt

On the feasibility of UMTS-based Traffic Information Systems

Intelligent Transportation Systems (ITS) are a hot topic in the communications society. Currently, research is primarily focusing on setting up Vehicular Ad Hoc Networks (VANETs) based on WLAN technology. However, VANETs are heavily dependent on high market penetration or infrastructure support. Third-generation (3G) networks might complement these efforts. They are already widely deployed and can serve as the basis for Car-to-Infrastructure (C2I) applications. We developed a simulation framework for holistic analysis of complex UMTS-based ITS. This framework couples simulation models with corresponding protocols of the UMTS link level, of higher network layers, and of road traffic. Based on our simulation framework and real-world 3G network coverage data, we evaluated a UMTS-based Traffic Information Systems (TIS) in a typical highway scenario in which information about traffic jams needed to be communicated to other cars for optimized route planning. The evaluation clearly outlines the capabilities of the simulation framework and evaluation results are consistent with all expectations. For example, we show that the availability of UMTS multicast distribution services are demanded for an efficient operation of the TIS application.

Network coding assisted mobile-to-mobile file transfer

This paper is a sequel of previous work, in which we have studied the problem of congestions on the typically overloaded downlink channels in UMTS (Universal Mobile Telecommunication System) networks. Our concept is based on the uplink/downlink traffic imbalance in 3G wireless networks and is realized by organizing seamless cooperation between peer-to-peer and cellular networks using a unified radio interface for both systems. The primary goal is to improve the spectral efficiency of cellular networks by enabling direct mobile-to-mobile (m2m) communication on temporarily unused uplink channels for the distribution of a large popular content in a non-real time multicast manner. The aim of this work is to further improve the interaction between cellular and peer-to-peer networks by generalizing the traditional scheduling paradigm. A network coding technique is embedded as a solution to the scheduling problem in the distributed dynamic environment of wireless large-scale networks. We investigate the performance of the system in terms of dependability of information distribution among m2m users. Simulations demonstrate the enhanced performance of the file distribution in terms of file download time. Furthermore, the obtained results highlight that network coding based m2m data transfer allows distribution of popular files to a large number of users while placing minimal bandwidth requirements on the central server.

Trellis-based equalization schemes for physical layer network coding

In this paper, we consider two-way relaying in a physical layer network coded system in a scenario where transmissions are subject to frequency-selective fading. Physical Layer Network Coding (PNC) enabled single-carrier transmission systems operating in flat fading environments have been subjected to scrutiny extensively, yet approaches for tackling the problem of extending them to environments causing multi-path fading are quite limited. We present three equalization schemes that can be used at the relay nodes to retrieve the data to be relayed in the presence of intersymbol interference (ISI). The most basic scheme jointly estimates both data streams and combines them afterwards. Furthermore, a scheme that already combines state and transition probabilities inside the trellis diagram is devised. Finally, we present a complexity-reduced algorithm where state combinations that yield the same bit to be relayed can be combined into a single state. All of the three algorithms can be regarded as different types of a decode-and-forward (DF) scheme.

Maximum SNR transmit filtering for linear equalization in physical layer network coding

We consider a two-way relaying system employing physical layer network coding (PNC) in channels suffering from frequency-selective fading. In order to mitigate the distortions introduced by the channel, a linear equalizer is used at the relay node. We introduce transmit filters that generate identical overall channel impulse responses for both source nodes, while achieving the maximum signal-to-noise ratio (SNR) for linear zero-forcing (ZF) and minimum mean-squared error (MMSE) equalization at the relay node. The performance gain of our novel approach is compared against a benchmark filtering scheme which also creates identical channel impulse responses to enable linear equalization at the relay.

Optimal Tomlinson-Harashima precoding for physical layer network coding

We consider a two-way relaying system employing physical layer network coding in channels suffering from frequency-selective fading. In order to mitigate the distortions introduced by the channel, Tomlinson-Harashima precoding is applied at the source nodes. We introduce an optimal precoding scheme that provides intersymbol interference free source-to-relay transmission without requiring identical overall channel impulse responses, which is in contrast to previously proposed equalization approaches. The performance of our novel approach is compared against that of a benchmark precoding scheme, motivated by the optimal decision-feedback equalization for physical layer network coding, which creates identical overall channel impulse responses in order to be able to use identical feedback filters for precoding at both source nodes.

Nonlinear Equalization Approaches for Physical Layer Network Coding

We consider a two-way relaying system employing physical layer network coding in channels suffering from frequency-selective fading. We study decision-feedback equalization (DFE), a technique based on delayed decision-feedback sequence estimation (DDFSE), and Tomlinson-Harashima precoding (THP) approaches for mitigating the distortions introduced by the channel. For DFE, we introduce transmit filtering schemes that generate identical overall source-to-relay channel impulse responses for both source nodes, while achieving the maximum signal-to-noise ratio after equalization for the zero-forcing and the minimum mean-squared error criterion, respectively. The advocated DDFSE approach also relies on transmit filtering at the source nodes. For the THP scheme, we derive the optimal precoding filters that guarantee an intersymbol interference free source-to-relay transmission. In order to obtain a two-way relaying system, which uses the same equalization techniques in both transmission directions, we design a THP-based compromise precoding scheme for the relay-to-destination transmission. The performance of the proposed techniques is compared with that of benchmark schemes with simpler filtering/precoding, where identical overall source-to-relay channel impulse responses are enforced by straightforward but suboptimum choices for the transmit filters. Our results reveal that the developed schemes enable significant gains compared with the benchmark schemes.