Markus P. J. Fromherz

Localization from mere connectivity

It is often useful to know the geographic positions of nodes in a communications network, but adding GPS receivers or other sophisticated sensors to every node can be expensive. We present an algorithm that uses connectivity information who is within communications range of whom to derive the locations of the nodes in the network. The method can take advantage of additional information, such as estimated distances between neighbors or known positions for certain anchor nodes, if it is available. The algorithm is based on multidimensional scaling, a data analysis technique that takes O(n3) time for a network of n nodes. Through simulation studies, we demonstrate that the algorithm is more robust to measurement error than previous proposals, especially when nodes are positioned relatively uniformly throughout the plane. Furthermore, it can achieve comparable results using many fewer anchor nodes than previous methods, and even yields relative coordinates when no anchor nodes are available.

Localization from connectivity in sensor networks

We propose an approach that uses connectivity information - who is within communications range of whom - to derive the locations of nodes in a network. The approach can take advantage of additional information, such as estimated distances between neighbors or known positions for certain anchor nodes, if it is available. It is based on multidimensional scaling (MDS), an efficient data analysis technique that takes O(n/sup 3/) time for a network of n nodes. Unlike previous approaches, MDS takes full advantage of connectivity or distance information between nodes that have yet to be localized. Two methods are presented: a simple method that builds a global map using MDS and a more complicated one that builds small local maps and then patches them together to form a global map. Furthermore, least-squares optimization can be incorporated into the methods to further improve the solutions at the expense of additional computation. Through simulation studies on uniform as well as irregular networks, we show that the methods achieve more accurate solutions than previous methods, especially when there are few anchor nodes. They can even yield good relative maps when no anchor nodes are available.

Improvements on Ant Routing for Sensor Networks

Ad-hoc wireless sensor networks have been an active research topic for the last several years. Sensor networks are distinguished from traditional networks by characteristics such as deeply embedded routers, highly dynamic networks, resource-constrained nodes, and unreliable and asymmetric links. Ant routing has shown good performance for communication networks; in this paper, we show why the existing ant-routing algorithms do not work well for sensor networks. Three new ant-routing algorithms are proposed and performance evaluations for these algorithms on a real application are conducted on a routing simulator for sensor networks.

On-line planning and scheduling: an application to controlling modular printers

This paper summarizes recent work reported at ICAPS on applying artificial intelligence techniques to the control of production printing equipment. Like many other real-world applications, such as mobile robotics, this complex domain requires real-time autonomous decision-making and robust continual operation. To our knowledge, this work represents the first successful industrial application of embedded domain-independent temporal planning. At the heart of our system is an on-line algorithm that combines techniques from state-space planning and partial-order scheduling. For example, our planning-graph-based planning heuristic takes resource contention into account when estimating makespan remaining. We suggest that this general architecture may prove useful as more intelligent systems operate in continual, online settings. Our system has enabled a new product architecture for our industrial partner and has been used to drive several commercial prototypes. When compared with stateof-the-art off-line planners, our system is hundreds of times faster and often finds better plans.

Message-initiated constraint-based routing for wireless ad-hoc sensor networks

Most existing routing mechanisms differ in routing objectives and routing strategies; however, all the existing routing protocols have in common that the routing objective and destination specification are fixed, and that the routing objective is incorporated only implicitly. We propose a general message specification mechanism to encode explicitly the routing destinations, constraints and objectives in messages, so that general-purpose instead of objective or destination specific routing strategies can be applied. Using general-purpose routing strategies while specifying quality-of-service (QoS) properties explicitly at the application layer in messages, QoS-aware strategies for individual messages are obtained. We also propose two frameworks of general-purpose routing strategies for this type of message specification.

Model-based computing for design and control of reconfigurable systems

Complex electro-mechanical products, such as high-end printers and photocopiers, are designed as families, with reusable modules put together in different manufacturable configurations, and the ability to add new modules in the field. The modules are controlled locally by software that must take into account the entire configuration. This poses two problems for the manufacturer. The first is how to make the overall control architecture adapt to, and use productively, the inclusion of particular modules. The second is to decide, at design time, whether a proposed module is a worthwhile addition to the system: will the resulting system perform enough better to outweigh the costs of including the module? This article indicates how the use of qualitative, constraint-based models provides support for solving both of these problems. This has become an accepted part of the practice of Xerox, and the control software is deployed in high-end Xerox printers.

Constrained flooding: a robust and efficient routing framework for wireless sensor networks

Flooding protocols for wireless networks in general have been shown to be very inefficient and therefore are mainly used in network initialization or route discovery and maintenance. In this paper, we propose a framework of constrained flooding protocols. The framework incorporates a reinforcement learning kernel, a differential delay mechanism, and a constrained and probabilistic retransmission policy. This type of protocol takes the advantages of robustness from flooding, but maintains energy efficiency by constraining retransmissions. Without the use of any control packets, such a protocol adapts to the specific routing requirements of the task and the dynamic changes of the network. We analyze this framework in simulation using a real-world application in sensor networks.

Positioning using local maps

It is often useful to know the positions of nodes in a network. However, in a large network it is impractical to build a single global map. In this paper, we present a new approach for distributed localization called Positioning using Local Maps (PLM). Given a path between a starting node and a remote node we wish to localize, the nodes along the path each compute a map of their local neighborhood. Adjacent nodes then align their maps, and the relative position of the remote node can then be determined in the coordinate system of the starting node. Nodes with known positions can easily be incorporated to determine absolute coordinates. We instantiate the PLM framework using the previously proposed MDS-MAP(P) algorithm to generate the local maps. Through simulation experiments, we compare the resulting algorithm, which we call MDS-MAP(D), with existing distributed methods and show improved performance on both uniform and irregular topologies.

Model-Based Computing: Using Concurrent Constraint Programming for Modeling and Model Compilation

Writing software for simulating, controlling or diagnosing real-world physical devices is non-trivial, as it requires encoding the complex interactions of device components and processes. Model-based computing is an approach to developing such software that employs multi-use declarative machine descriptions to derive information from which machine software can be constructed automatically. Specifically, we have been using this approach to develop code for scheduling reprographic machines.

Smart Routing with Learning-Based QoS-Aware Meta-strategies

Conventional Quality of Service (QoS) routing cannot be applied easily to wireless ad-hoc sensor networks due to the unreliable and dynamic nature of such networks. For these networks, we have proposed a framework of Message-initiated Constraint-Based Routing (MCBR), which consists of a QoS specification and a set of QoS-aware meta-strategies. In contrast to most existing ad-hoc routing with no QoS support, MCBR is able to take QoS specifications into account. In this paper, we focus on learning-based meta-strategies. In contrast to most existing QoS routing approaches, learning-based meta-strategies do not create and maintain explicit routes; instead, packets discover and improve the routes during the search for the destination.