Michael A. Jaeger

Self-organizing broker topologies for publish/subscribe systems

Distributed publish/subscribe systems are usually deployed on top of an overlay network that enables complex routing strategies implemented in the application layer. Up to now, only little effort has been spent on the design of the broker overlay network assuming that it is either static or manually administered. As publish/subscribe systems are increasingly targeted at dynamic environments where client behavior and network characteristics vary over time, static overlay networks lead to suboptimal performance. In this paper, we present a self-organizing broker overlay infrastructure that adapts dynamically to achieve a better efficiency on both, the application and the network layer. This is obtained by taking network metrics as well as notification traffic into account.

Self-stabilizing publish/subscribe systems: algorithms and evaluation

Most research in the area of publish/subscribe systems has not considered fault-tolerance as a central design issues. However, faults do obviously occur and masking all faults is at least expensive if not impossible. A potential alternative (or sensible supplementation) to fault masking is self-stabilization which allows a system to recover from arbitrary transient faults such as memory perturbations, communication errors, and process crashes with subsequent recoveries. In this paper we discuss how publish/subscribe systems can be made self-stabilizing by using self-stabilizing content-based routing. When the time between consecutive faults is long enough, corrupted parts of the routing tables are removed, while correct parts are refreshed in time, and missing parts are inserted. To judge the efficiency of self-stabilizing content-based routing, we compare it to flooding, which is the naive implementation of a self-stabilizing publish/subscribe system. We show that our approach is superior to flooding for a large range of practical settings.

Reconfiguring Publish/Subscribe Overlay Topologies

Distributed content-based publish/subscribe systems are usually implemented by a set of brokers forming an overlay network. Most existing publish/subscribe middleware assumes that the overlay topology is static. Those that consider reconfiguration, exchange a single link that is torn down by another link that comes up. However, they do not guarantee that no notifications are lost or duplicated nor do they ensure any message ordering policy. In this paper, we discuss the dynamic reconfiguration of publish/subscribe systems which are built on content-based routing. We present algorithms that allow for reconfigurations without notification loss or duplication that can also ensure FIFO-publisher and causal ordering. Moreover, the efficiency of reconfigurations is increased by limiting their effects to those parts of the network which are directly affected by the reconfiguration.

MESH Mdl event spaces - A coordination middleware for self-organizing applications in ad hoc networks

Mobile ad hoc networks (MANETs) are gaining importance as a promising technology for flexible, proximity-based, mobile communication. However, the inherent dynamics of MANETs imposes strong limitations on the design of distributed applications. They need to be able to adapt to changing conditions quickly and organize themselves in terms of component placement and communication habits. In this paper, we present MESHMdl, a middleware that provides a high level of awareness and decoupling for application components to make them more flexible and adaptable. We focus on the Event Space as the central communication medium of MESHMdl. The Event Space offers a simple, unified communication interface for inter-agent communication as well as for communication with the middleware and resource access. Furthermore, it serves as a means for flexibly extending a MESHMdl daemon. We investigate the performance of the Event Space on different mobile devices and show that it is superior to comparable systems.

Reconfiguring self-stabilizing publish/subscribe systems

Recent work on self-stabilizing routing in publish/subscribe systems showed that it is feasible to automate reconfigurations in case of faults by enabling the system to recover from arbitrary transient faults. In this paper, we discuss how to incorporate planned reconfigurations of the broker topology into self-stabilizing publish/subscribe systems without service interruption. We present an algorithm that uses a coloring mechanism to enable the system to be automatically switched from one system configuration to another. The colors thereby synchronize the broker overlay and the publish/subscribe routing layer.

Self-organizing publish/subscribe

The increasing availability of broadband Internet access coming along with cheap flatrate tariffs changed the way the Internet is used over the last years. As a result, we observe more users that become producers and collaborate instead of solely consuming information as they did before. Besides that, the Internet has become a tool for daily use and information dissemination. As push-based information services are heavily needed for collaboration and information dissemination, it seems like the breakthrough of distributed push publish/subscribe on the open Internet is just imminent today, although research has been around already for many years now. One reason why most publish/subscribe applications are still based on a rather basic centralized mechanism and not on a distributed scalable notification service, as it is commonly proposed in state-of-the-art publish/subscribe systems in research, is the unsolved management issue.This extended research abstracts is intended to give an overview of my Ph.D. project on self-organizing publish/subscribe, where I want to tackle management issues for better applicability. After giving an introduction to the pub/sub communication paradigm and motivating the necessity to make such systems self-organizing, I will point out the contributions of my work. They are centered around making publish/subscribe systems self-stabilizing and adaptive by means of self-organizing mechanisms. While my work on self-stabilization is nearly finished, introducing adaptivity by self-organization is still in an early stage.

Self-organizing and self-stabilizing role assignment in sensor/actuator networks

Developing distributed applications for sensor/actuator networks is challenging, particularly, with regard to unreliable nodes and communication links Splitting applications into roles eases the development significantly but presumes a reliable role management that autonomously assigns roles to devices depending on their capabilities In this paper, we present a self-organizing and self-stabilizing role assignment mechanism as an integral part of a light-weight, flexible middleware The deployed algorithms tolerate the addition and removal of devices at runtime and are also able to recover from any transient fault Considering that resources are restricted on many devices, we analyze the proposed algorithms with respect to communication overhead, memory usage, and stabilization time.

Decoupling Components of an Attack Prevention System Using Publish/Subscribe

Distributed and coordinated attacks can disrupt electronic commerce applications and cause large revenue losses. The prevention of these attacks is not possible by just considering information from isolated sources of the network. A global view of the whole system is necessary to react against the different actions of such an attack. We are currently working on a decentralized attack prevention framework that is targeted at detecting as well as reacting to these attacks. The cooperation between the different entities of this system has been efficiently solved through the use of a publish/subscribe model. In this paper we first present the advantages and convenience in using this communication paradigm for a general decentralized attack prevention framework. Then, we present the design for our specific approach. Finally, we shortly discuss our implementation based on a freely available publish/subscribe message oriented middleware.

PayFLUX: Secure electronic payment in mobile ad hoc networks

Electronic payment is a key building block of distributed business applications in mobile ad hoc networks (MANETs). However, existing payment systems do not fulfill the requirements imposed by the highly dynamic and decentralized nature of MANETs. Either they rely on digital coins that suffer from usability problems, or they build on cellular phone technology which is bound to the availability of a fixed infrastructure. Therefore, we propose PayFlux, as a new system for electronic payment in MANETs. It is based on the light-weight Simple Public Key Infrastructure (SPKI) that allows for the decentralized creation and delegation of authorizations. Adopting the well-known abstraction of direct debits and enhancing it with new useful features, it offers good usability and can be easily integrated into the existing banking system.

An alert communication infrastructure for a decentralized attack prevention framework

The cooperation between the different entities of a decentralized prevention system can be solved efficiently using the publish/subscribe communication model. Here, clients can share and correlate alert information about the systems they monitor. In this paper, we present the advantages and convenience in using this communication model for a general decentralized prevention framework. Additionally, we outline the design for a specific architecture, and evaluate our design using a freely available publish/subscribe message oriented middleware