Sandford Bessler

Introducing Aspect-oriented Space Containers for efficient publish/subscribe scenarios in Intelligent Transportation Systems

The publish/subscribe paradigm is a common concept for delivering events from information producers to consumers in a decoupled manner. Some approaches allow durable subscriptions or the transportation of events even to mobile subscribers in a dynamic network infrastructure. However, in the safety-critical telematics durable delivery of events is not sufficient enough. Short network connectivity time and small bandwidth limit the number and size of events to be transmitted hence relevant information needed for safety-critical decision making may not be timely delivered.

Using tuple-spaces to manage the storage and dissemination of spatial-temporal content

Structured, spatial-temporal content arises in application areas such as mobile computing, intelligent transportation, urban mobility, and ubiquitous sensing. For the distributed storage and dissemination of such content, peer-to-peer solutions appear to be the natural choice. However, a closer analysis shows that distributed hash tables (DHT) alone are not enough: firstly, they do not maintain the original data structure needed to efficiently access the comprising attributes, and secondly, they lead to high signaling traffic when the data is short lived, such as in high mobility scenarios. In order to address these two problems we propose a novel content dissemination architecture based on an overlay of space-based containers. Furthermore, we apply the proposed concept to realize a concrete application in the field of intelligent transportation, and present the results of the performance evaluation conducted with the system prototype.

A P2P Network of Space Containers for Efficient Management of Spatial-Temporal Data in Intelligent Transportation Scenarios

The effectiveness of Intelligent Transportation Systems (ITS) depends on their ability to collect contextual data from various sources and appropriately generate and transport comprehensible, reliable and timely content to users. In such applications, the exchanged content is structured in space and time. Peer-to-peer(P2P) networks are the natural choice these applications due to their fault-tolerance, self-organization and scalability properties. However, a closer analysis of the available Distributed Hash Tables (DHT) protocols shows that the structure of the datagets lost and its short liveness leads to high signalling traffic. In this work we propose a novel overlay network of so called Space Containers for storing, accessing, manipulating and structuring dynamic geo-located content. The benefits of combining Space Containers and DHT are: clean application programming logic and efficient content retrieval while preserving the properties of DHTs. We describe the system architecture applied to a transportation scenario and show preliminary evaluation results.

Controlled Probing - A system for targeted floating car data collection

Since vehicle to infrastructure communication will become state of the art in a few years, data from mobility and environment sensors in the vehicle could produce valuable information for traffic management and vehicle safety applications. The massive deployment of such data sources could however lead to high data traffic that would unnecessarily strain both the communication network, processing and storage resources. In this paper we present Controlled Probing, a system solution for floating car data collection which allows to select exactly the location, the time period and the type of sensor data to be collected. We address the system performance with respect to the accuracy and timeliness of the received probes.

Integration of Shareable Containers with Distributed Hash Tables for Storage of Structured and Dynamic Data

Structured, spatial-temporal data arises in many applications areas such as transportation, sensor networks or mobile services. Peer to peer networks are the natural choice for the distributed architecture required by these applications. However, a closer analysis of the available distributed hash table (DHT) based approaches shows their inefficiency since the data structure gets lost and the short liveness of the data leads to a high signalling traffic. In this work we propose a novel storage network based on an overlay of Space-based Computing Containers for storing, accessing and manipulating dynamic data.

A generic load balancing framework for cooperative ITS applications

The deployment of cooperative ITS applications is due to start as soon as 2015. Large investments in the roadside unit (RSU) infrastructure will be necessary to create a dense network and accommodate an increasing number of services, leading to a discussion about the trade-off between distributed processing and storage solutions on the RSU nodes and the centralized alternative. A strictly central solution might not be scalable, whereas the decentralized approach faces the problem that load in the form of CPU and memory usage may be unequally distributed among the nodes, causing performance bottlenecks on some of the RSUs. This work presents a solution for this problem, in form of a generic framework that balances the load between the nodes and reduces in this way the RSU costs. The interactions are based on a flexible coordination pattern for load balancing that is realized using customizable containers provided by a distributed systems middleware. This mechanism is applied to a probe data collection scenario in which individual messages are aggregated by RSU nodes, causing both CPU and memory load. Simulation results illustrate the operation in dynamic load situations.