Nandan Parameswaran

Ontology mapping for the interoperability problem in network management

Interoperability between different network management domains, heterogeneous devices, and various management systems is one of the main requirements for managing complex enterprise services. While substantial advances have been made in low-level device and data interoperability using common data formats and specifications such as simple network management protocols (SNMPs) SMI and TMFs SID, various interoperability issues including semantic interoperability offer interesting research challenges. While semantic interoperability is a difficult problem in its own right, the semantic web that incorporates intelligent agents necessitates an interoperability solution requiring agents to communicate unambiguously and reason intelligently to perform cooperative management tasks. Agents need a formal representation of knowledge; an ontology is capable of modeling the rich semantics of the managed environment (and especially, relationships between managed entities) so that agents can act on them. This paper presents an ontology-driven approach for solving the semantic interoperability problem in the management of enterprise services, illustrated here with a router configuration management application.

Analyzing Tennis Tactics from Broadcasting Tennis Video Clips

This paper attempts to classify tennis games into 58 winning patterns for training purpose. It bases on tracking ball movement from broadcasting tennis video. Trajectory and landing position are used as the basic features for classification. We use an improved Bayesian networks to classify the landing position of different patterns. Intelligent agents are used to combine trajectories and landing positions as two features are in different dimensions. Semantic labels are granted after classification. The aim of the analysis is to provide a browsing tool for coachers or other personnel to retrieve tennis video clips.

Fault Repair Framework for Mobile Sensor Networks

In this paper, we propose a framework for fault repair in mobile sensor networks. A hierarchical structure which consists of replacement module, management policy module, knowledge module, decision making module, and evaluation module is adopted. We also propose a solution for faulty sensor replacement problem. Through the numerical results, we show that our algorithm is more efficient and achieves higher energy savings than the greedy approach to sensor replacement. We believe that the problem of faulty sensor nodes can be solved efficiently through the cooperation and communication across different modules, such as evaluation decision making, knowledge management, and replacement

Rules and Ontology in Compliance Management

Compliance management (CM) is the management process that an organization implements to ensure organizational compliance with relevant requirements and expectations. It is a continual, manual and labor intensive process that is proved to be of great challenge for many organizations. CM affects almost every aspect of an organization and is in nature a complex problem due to voluminous knowledge and data involved. In our attempts to automate and simplify compliance, we propose and examine a semantic rule-based approach for modeling compliance knowledge with the use of semantic Web rules (SWRL) and Web ontology language (OWL). We study the use of exception handling approach to create a more robust rule base to deal with data incompleteness in the semantic Web.

Awareness modelling in collaborative mobile e-health

E-health based on mobile wireless networks is called mobile e-health. Mobile e-health can facilitate computer-supported cooperative work, which encompasses tasks from email and instant messaging to wireless information sharing through broadband and telecommunication networks. In the development of cooperative management systems, the concept of ‘awareness level’ has been introduced. This is based on factors such as the location and actions of the user involved. Mobile e-health can help to achieve cooperation by providing the right awareness levels at the right time. The improved awareness levels allow health professionals to provide patient care with better quality and efficiency. Mobile e-health systems have the potential to take over the mundane tasks of the doctor so that better quality health services can be provided.

Detecting tactics patterns for archiving tennis video clips

Video processing has found many applications in sports such as slow motion replay, pattern analysis, statistics collection, video archiving, etc. This paper presents an automatic archiving system for tennis games. It extracts ball trajectory using calibrated cameras. The trajectory is then used to classify tennis video clips into 58 tactic patterns. Semantic annotation is then attached to each clip so future query can be made easily. The annotated video clips can be used for training, sports analysis, broadcasting, etc. It can also be used for online browsing and streaming.

An Awareness Framework for Agent-Based Mobile Health Monitoring

One of significant contributions in medical information technology is the design and implementation of e-health solutions, which provide patients with mobile services to support and optimize their treatment based on the monitoring of certain physiological parameters. This category of e-health is called mobile health monitoring. Although the mobile wireless technologies are now commercially available, there is a need for a computational framework to integrate the technology with the needs of mobile health monitoring. In this paper, we propose a Computer Supported Cooperative Work (CSCW) based technique called awareness modeling to realize the awareness of human roles involved in mobile health monitoring activities using software agents.

Flexible power consumption management in smart homes

The complexity of the power system keeps growing due to the penetration of alternative sources of electric energy such as solar plant and wind farm to meet the growing demand for electricity. In order to overcome the challenges due to complexity the power grid needs to be intelligent in all aspects. As the grid gets smarter, considerable efforts are being undertaken to make the houses and businesses smarter in consuming the electrical energy to minimize and level the electricity demand. It also necessitates that the conventional way to do modeling, control and energy management in all sectors needs to be improved or replaced by intelligent information processing techniques. We propose a policy based framework which allows intelligent and flexible energy management of home appliances in a smart home which is complex and dynamic in ways that save energy automatically. In this paper, we consider the challenges in formalizing the behavior of the appliances using states, and managing the energy consumption using policies. Our simulation results show that the proposed system can efficiently manage energy in a smart home in a user friendly way.

Peak reduction algorithms for a smart community

As the energy consumption in residential sector keeps growing, it is essential to find a systematic way of managing the energy consumption in a smart community consisting of a number of smart houses so that the available resource is utilized properly without causing peak demand. In this paper, we propose peak demand management techniques for a smart community using different types of coordination mechanisms for coordination of multiple house agents working in the same environment. In this we propose few peak reduction algorithms which use centralized model, decentralized model and Pareto resource allocation model for resource allocation. We describe the smart energy community architecture and propose user comfort model for the appliances based on user preferences and power reduction capability that comes from the load profile of appliances. We used performance indices such as community comfort, community involvement to analyze and compare our algorithms and finally find the best performing algorithm among them.

Towards Robust and Adaptive Semantic-Based Compliance Auditing

Compliance management (CM) is the management process that an organization implements to ensure organizational compliance with relevant requirements and expectations. Compliance auditing (CA) is a child-process of CM where compliance rules and policies are individually checked against the organization to determine the level of compliance achieved by the organization. In this paper, we arrange organizational knowledge and facts within OWL ontologies and model compliance rules as adaptive semantic-based rules for compliance audit automation. We study the issues of uncertainty and inconsistency in compliance and propose an adaptive human-like strategy for mimicking conventional compliance auditing.