Nazario Cipriani

Making the World Wide Space happen: New challenges for the Nexus context platform

Context-aware applications rely on models of the physical world. Within the Nexus project, we envision a World Wide Space which provides the conceptual and technological framework for integrating and sharing such context models in an open, global platform of context providers. In our ongoing research we tackle important challenges in such a platform including distributed processing of streamed context data, situation recognition by distributed reasoning, efficient management of context data histories, and quality of context information. In this paper we discuss our approach to cope with these challenges and present an extended Nexus architecture.

NexusDS: a flexible and extensible middleware for distributed stream processing

Techniques for efficient and distributed processing of huge, unbound data streams have made some impact in the database community. Sensors and data sources, such as position data of moving objects, continuously produce data that is consumed, e.g., by location-aware applications. Depending on the domain of interest, e.g. visualization, the processing of such data often depends on domain-specific functionality. This functionality is specified in terms of dedicated operators that may require specialized hardware, e.g. GPUs. This creates a strong dependency which a data stream processing system must consider when deploying such operators. Many data stream processing systems have been presented so far. However, these systems assume homogeneous computing nodes, do not consider operator deployment constraints, and are not designed to address domain-specific needs. In this paper, we identify necessary features that a flexible and extensible middleware for distributed stream processing of context data must satisfy. We present NexusDS, our approach to achieve these requirements. In NexusDS, data processing is specified by orchestrating data flow graphs, which are modeled as processing pipelines of predefined and general operators as well as custom-built and domain-specific ones. We focus on easy extensibility and support for domain-specific operators and services that may even utilize specific hardware available on dedicated computing nodes.

ProRea: live database migration for multi-tenant RDBMS with snapshot isolation

The consolidation of multiple tenants onto a single RDBMS instance turned out to be benefical with respect to resource utilization and scalability. The consolidation implies that multiple tenants share the physical resources available for the RDBMS instance. If the available resources tend to get insufficient to meet the SLAs agreed with the tenants, migration of a tenants database from one RDBMS instance to another is compelling. Highly available services demand for live migration techniques that come with minimal service interruption and low performance impact. This paper meets the demand for live migration techniques by contributing ProRea. ProRea is a live database migration approach designed for multi-tenant RDBMS that run OLTP workloads under snapshot isolation. ProRea extends concepts of existing live database migration approaches to accomplish minimal service interruption, high efficiency and very low migration overhead. Measurements of a prototypical ProRea implementation underpin its good performance.

Tool support for the design and management of context models

A central task in the development of context-aware applications is the modeling and management of complex context information. In this paper, we present the NexusEditor, which can ease this task by providing a graphical user interface to design schemas for spatial and technical context models, interactively create queries, send them to a server and visualize the results. One main contribution is to show how schema awareness can improve such a tool: The NexusEditor dynamically parses the underlying data model and provides additional syntactic checks, semantic checks, and short-cuts based on the schema information. Furthermore, the tool helps to design new schema definitions based on the existing ones, which is crucial for an iterative and user-centric development of context-aware applications. Finally, it provides interfaces to existing information spaces and visualization tools for spatial data like GoogleEarth.

A System for Distributed Context Reasoning

Context aware systems use context information toadapt their behaviour accordingly. In order to derive highlevel context information from low level context, such as sensorvalues, context reasoning methods that correlate observablecontext information, are necessary. Several context reasoningmechanisms have been proposed in the literature. Usually thesemechanisms are centralized, leading to suboptimal utilization ofnetwork resources and poor system performance in case of largescalescenarios. Therefore to increase the scalability of contextreasoning systems the development of methods that distribute thereasoning process is necessary. Existing distributed approachesare method specific and do not provide a generic formalizationfor distributed reasoning. In this paper we introduce a novelsystem which enables distributed context reasoning in a genericway that is independent of the reasoning algorithm.

DiSCO: A Distributed Semantic Cache Overlay for Location-Based Services

Location-based services (LBS) have gained tremendous popularity with millions of simultaneous users daily. LBS handle very large data volumes and face enormous query loads. Both the data and the queries possess high locality: spatial data is distributed very unevenly around the globe, query load is different throughout the day, and users often search for similar things in the same places. This causes high load peaks at the data tier of LBS, which may seriously degrade performance. To cope with these load peaks, we present DiSCO, a distributed semantic cache overlay for LBS. DiSCO exploits the spatial, temporal and semantic locality in the queries of LBS and distributes frequently accessed data over many nodes. Based on the Content-Addressable Network (CAN) peer-to-peer approach, DiSCO achieves high scalability by partitioning data using spatial proximity. Our evaluation shows that DiSCO significantly reduces queries to the underlying data tier.

Distributed context-aware visualization

We present a visualization framework integrated in a context-aware system that uses a common underlying stream processing middleware for tight integration of data accessing, processing, and visualization. Context-aware systems are often realized on mobile devices that do not have the computational power to perform complex tasks. Therefore, a dedicated hardware infrastructure might be required for data processing. In our case, stream processing is used, supporting parallelism on distributed and shared memory multiprocessors. We present the integration of visualization modules into a Java-based stream processing framework for context-aware systems, with focus on efficient communication and parallelization. Our approach is demonstrated for the example of a flow visualization scenario.

Exploiting constraints to build a flexible and extensible data stream processing middleware

A wide range of real-time applications process stream-based data. To process this stream-based data in an application-independent manner, many stream processing systems have been built. However, none of them reached a huge domain of applications, such as databases did. This is due to the fact that they do not consider the specific needs of real-time applications. For instance, an application which visualizes stream-based data has stringent timing constraints, or may even need a specific hardware environment to smoothly process the data. Furthermore, users may even add additional constraints. E.g., for security reasons they may want to restrict the set of nodes that participates in processing. Thus, constraints naturally arise on different levels of query processing. In this work we classify constraints that occur on different levels of query processing. Furthermore we propose a scheme to classify the constraints and show how these can be integrated into the query processing of the distributed data stream middleware NexusDS.

M-TOP: multi-target operator placement of query graphs for data streams

Nowadays, many applications processes stream-based data, such as financial market analysis, network intrusion detection, or visualization applications. To process stream-based data in an application-independent manner, distributed stream processing systems emerged. They typically translate a query to an operator graph, place the operators to stream processing nodes, and execute them to process the streamed data. The operator placement is crucial in such systems, as it deeply influences query execution. Often, different stream-based applications require dedicated placement of query graphs according to their specific objectives, e.g. bandwidth not less than 500 MBit/s and costs not more that 1 cost unit. This fact constraints operator placement. Existing approaches do not take into account application-specific objectives, thus not reflecting application-specific placement decisions. As objectives might conflict among each other, operator placement is subject to delicate trade-offs, such as bandwidth maximization is more important than cost reduction. Thus, the challenge is to find a solution which considers the application-specific objectives and their trade-offs. We present M-TOP, an QoS-aware multi-target operator placement framework for data stream systems. Particularly, we propose an operator placement strategy considering application-specific targets consisting of objectives, their respective trade-offs specifications, bottleneck conditions, and ranking schemes to compute a suitable placement. We integrated M-TOP into NexusDS, our distributed data stream processing middleware, and provide an experimental evaluation to show the effectiveness of M-TOP.

NexusVIS: A distributed visualization toolkit for mobile applications

Many mobile pervasive applications need to visualize information about the users geographic surroundings combined with data from sensors, which determine the users context. We demonstrate NexusVIS, a distributed visualization toolkit for mobile applications. By building upon an existing data stream processing system we enable applications to define distributed visualization processes as continuous queries. This allows applications to define visualization semantics descriptively. Moreover, NexusVIS is capable of adapting the visual query at runtime, and thus allows to navigate in the visualized scene both automatically and manually through user control.