Research of the QoC-Aware Service Selection for Middleware Based Pervasive Applications

Di Zheng, Hang Yan, Jun WangInternational Journal of Engineering and Manufacturing 2011 · DOI

Abstract

With the rapid development of information technology, distributed mobile computing is evolving toward pervasive computing, where information space and physical space fuse seamlessly. As pervasive applications increasingly rely on Service-Oriented Architecture (SOA), selecting suitable services from a vast pool of candidates becomes critical. Unlike traditional service selection, pervasive environments require careful consideration of context and its quality. However, many existing approaches ignore Quality of Context (QoC), leading to unreliable selections. This paper proposes a middleware-based service selection scheme designed to support efficient, QoC-aware service selection.

Introduction

In the evolving landscape of pervasive computing, applications must adapt to the user's physical environment. This adaptation relies on context-aware services that respond to data from various sensors. Traditionally, service selection has focused on Quality of Service (QoS) metrics, such as network latency and availability. However, in pervasive environments, the reliability of the underlying context data is just as important. If a system makes a decision based on stale or imprecise sensor data, the resulting service selection may be incorrect or even hazardous. This paper introduces a framework that prioritizes Quality of Context (QoC) to ensure robust decision-making.

The Proposed Middleware Architecture

The authors propose a specialized middleware architecture that acts as a quality-oriented filter between raw environmental data and high-level applications. The architecture is designed to manage the complexity of heterogeneous sensors and provide a unified interface for service selection.

The Context Manager

The core of this framework is the Context Manager. Rather than simply passing raw data through, the Context Manager interprets and reasons about the environment to ensure data integrity. It utilizes threshold controls to discard low-certainty data and resolution logic to resolve inconsistencies between overlapping sensor readings.

The internal architecture of the Context Manager, showing the flow from sensing to reasoning and storage.

As illustrated in the architecture above, the process begins with the Context Sensor, which feeds data into the Context Interpreter. The Context Reasoner then evaluates the interpreted data against stored historical context and repository rules before making it available to the application layer via the Context Access interface.

QoC Modeling and Parameters

To standardize how quality is measured across different domains, the system employs an ontology-based approach. This allows the middleware to categorize context into four primary domains: computation, location, user, and activity. Within these domains, the system evaluates six specific QoC dimensions:

  1. Security: The probability that the context data remained secure during transmission and storage.
  2. Precision: The level of granularity or detail provided by the data.
  3. Resolution: The spatial or temporal granularity (e.g., distinguishing between a specific room versus a whole building).
  4. Freshness: The time elapsed since the data was captured, ensuring the system doesn't act on outdated information.
  5. Certainty: The statistical probability that the data accurately reflects the physical state.
  6. Completeness: The ratio of available context information to the total information required for a decision.
The Quality Ontology and a practical example of location-based QoC parameters.

By linking these non-functional constraints directly to service metadata, a service can define minimum quality requirements. For instance, a high-security access service might refuse to bind if the "Certainty" of the user's location data falls below a specific threshold.

Experimental Evaluation

The framework was tested in a real-world university building environment, managing context-based access control and climate systems. Sensing data was captured at 5-second intervals over a 24-hour period and processed on a standard server (Intel Core Duo 2.8GHz, 4GB RAM).

Key Results

  • Performance Overhead: The introduction of QoC filtering adds computational cost. Processing 1,000 context items took approximately 3.5 seconds with QoC-aware logic, compared to 2 seconds for basic processing. The authors argue this 1.5-second trade-off is acceptable given the significant increase in reliability.
  • Selection Accuracy: The evaluation demonstrated that QoC-aware algorithms (filtering by certainty or freshness) significantly outperformed random selection methods. By prioritizing high-quality data, the system achieved a much higher "true probability" of correct environmental understanding.

Conclusion and Future Work

This research demonstrates that integrating QoC into middleware is essential for reliable pervasive computing. While the current framework effectively filters noisy data, future work will address the computational overhead to make the system suitable for high-speed or safety-critical applications. Additionally, the researchers plan to investigate the interdependencies between different QoC indicators—such as how a decrease in freshness might automatically impact certainty—and develop more sophisticated "repair" methods for dynamic service rebinding.

Summary updated

Abstract

With the rapid development of information technology, it is inevitable that the distributed mobile computing will evolve to the pervasive computing gradually whose final goal is fusing the information space composed of computers with the physical space in which the people are working and living in. Furthermore, with the development of SOA, more and more pervasive applications have been composed of different kinds of services. So how to choose a suitable service from all the useable services is the most important step for pervasive computing. Compare to the traditional service selection we must take more care of the context as well as the quality of them in pervasive environment. However, most of existing researches pay no attention to QoC (Quality of Service) which may lead to unreliable selections. Therefore we proposed a middleware based service selection scheme to support QoC-aware service selection efficiently.
BibTeX
@article{zheng2011research,
  author = {Di Zheng and Hang Yan and Jun Wang},
  title = {{Research of the QoC-Aware Service Selection for Middleware Based Pervasive Applications}},
  journal = {International Journal of Engineering and Manufacturing},
  year = {2011},
  doi = {10.5815/ijem.2011.03.09},
}
Figure