Rudolf Sollacher

Continuum percolation of wireless ad hoc communication networks

Wireless multi-hop ad hoc communication networks represent an infrastructure-less and self-organized generalization of todays wireless cellular networks. Connectivity within such a network is an important issue. Continuum percolation and technology-driven mutations thereof allow to address this issue in the static limit and to construct a simple distributed protocol, guaranteeing strong connectivity almost surely and independently of various typical uncorrelated and correlated random spatial patterns of participating ad hoc nodes.

Impact of interference on the wireless ad-hoc networks capacity and topology

It is wide spread belief that wireless mobile ad-hoc networks will be a further evolutionary step towards ubiquitous communication and computing. Due to the mobility of the network nodes, the strongly varying radio propagation conditions and the varying data traffic load these networks constitute a very dynamic environment. One essential step in evaluating the true benefit of this new technology consists of estimates and constraints concerning the scalability and performance of such networks. Using a simple model we discuss analytically the effect of interference on the link quality and connectivity of large networks. It turns out that the outage probability rapidly increases with increasing traffic load. Furthermore, we investigate the connectivity of the network under varying traffic load and find a percolation phase transition at a particular value of the traffic load. We discuss the dependence of these effects on parameters characterizing the receiver and the radio propagation conditions.

Impact of network structure on the capacity of wireless multihop ad hoc communication

As a representative of a complex technological system, the so-called wireless multihop ad hoc communication networks are discussed. They represent an infrastructure-less generalization of todays wireless cellular phone networks. Lacking a central control authority, the ad hoc nodes have to coordinate themselves such that the overall network performs in an optimal way. A performance indicator is the end-to-end throughput capacity. Various models, generating differing ad hoc network structure via differing transmission power assignments, are constructed and characterized. They serve as input for a generic data traffic simulation as well as some semi-analytic estimations. The latter reveal that due to the most-critical-node effect the end-to-end throughput capacity sensitively depends on the underlying network structure, resulting in differing scaling laws with respect to network size.

Nonlinear Speed-control for a Continuum Theory of Traffic Flow

Abstract A nonlinear controller creating a homogeneous flow is designed for a traffic model given by partial differential equations. To be able to apply nonlinear control techniques, a co-moving coordinate frame is set up leading to a reduced model of stop-and-go waves. The controller stabilizing this model is re-transformed and simplified. Properties of the controlled system indicating global stability are confirmed by simulations. Finally, a two-point controller is defined, bridging the gap to speed limits.

Distributive routing and congestion control in wireless multihop ad hoc communication networks

Due to their inherent complexity, engineered wireless multihop ad hoc communication networks represent a technological challenge. Having no mastering infrastructure the nodes have to selforganize themselves in such a way that for example network connectivity, good data traffic performance and robustness are guaranteed. In this contribution the focus is on routing and congestion control. First, random data traffic along shortest path routes is studied by simulations as well as theoretical modeling. Measures of congestion like end-to-end time delay and relaxation times are given. A scaling law of the average time delay with respect to network size is revealed and found to depend on the underlying network topology. In the second step, a distributive routing and congestion control is proposed. Each node locally propagates its routing cost estimates and information about its congestion state to its neighbors, which then update their respective cost estimates. This allows for a flexible adaptation of end-to-end routes to the overall congestion state of the network. Compared to shortest-path routing, the critical network load is significantly increased.

Quantum pattern recognition with liquid-state nuclear magnetic resonance

(Dated: April 21, 2009)A novel quantum pattern recognition scheme is presented, which combines the idea of a classicHopfield neural network with adiabatic quantum computation. Both the input and the memorizedpatterns are represented by means of the problem Hamiltonian. In contrast to classic neural net-works, the algorithm can return a quantum superposition of multiple recognized patterns. A proof ofprinciple for the algorithm for two qubits is provided using a liquid state NMR quantum computer.

Efficient online learning with Spiral Recurrent Neural Networks

Distributed intelligent systems like self-organizing wireless sensor and actuator networks are supposed to work mostly autonomous even under changing environmental conditions. This requires robust and efficient self-learning capabilities implementable on embedded systems with limited memory and computational power. We present a new solution called spiral recurrent neural networks with an online learning based on an extended Kalman filter and gradients as in real-time recurrent learning. We illustrate its performance using artificial and real-life time series and compare it to other approaches. Finally we describe a few potential applications.

Towards a Service Oriented Architecture for Wireless Sensor Networks in Industrial Applications

Abstract We discuss the introduction of service oriented architectures to wireless sensor networks (WSN) in industrial applications. We give an example for a WSN architecture as applied in the EU project SOCRADES in order to explain constraints preventing a full-fledged service oriented approach. Such an approach appears to be beneficial for applications like diagnostics or monitoring where service composition can provide new functionalities. However, the limited resources in WSN must be taken into account. For control applications additional constraints like determinism or latency bound severly limit a loose coupling of services. As a consequence we propose a support by appropriate design and engineering tools.

Self-⋆ topology control in wireless multihop ad hoc communication networks

Wireless multihop ad hoc communication networks represent an infrastructure-less generalization of todays cellular networks. Since a central control authority is missing, the complex network has to self-⋆ itself for various operating tasks. Key to a self-⋆ realization is the design of simple, yet robust distributive control rules, which allow the overall network to perform well. Two examples from topology control are given. The first one addresses the connectivity issue, where a self-⋆ rule is presented and shown to lead to strong network connectivity almost surely. A generic system analysis is used in the second example to first develop a phenomenological description of the networks end-to-end throughput capacity and then to sketch further steps towards a self-⋆ rule for obtaining a large throughput performance.

Smart wireless Sub-Metering

This paper presents a new approach for using wireless sensor networks for Sub-Metering applications. Parameters regarding reliability and energy consumption have been investigated.