Sensing and Actuating Tasks as Services and Its Quality of Services in Large Clustered Environments

Abstract

With the proliferation of sensors and actuators in today's world, we envisage the world as an inter-connected network of millions of Sensing and Actuating resources performing multiple tasks in everyday life. These distributed resources are capable of performing tasks that monitor and/or affect the parameters of the physical and environmental entities. To perform a task, the user might require a single or group of sensors and/or actuators, which are offered by multiple Service Providers in the market. The ability to trigger these tasks without the user having to determine the owner of the service, schedule tasks by searching and determining the availability of resources, in a location-independent manner, is provided by enabling the Sensor and Actuator resources as services. We propose an architecture called SATS (Sensing and Actuating Tasks as Service) that provides the ability to trigger sensing and actuating tasks over the Internet by selecting the best combination from the available resources, including the resources owned by other Service Providers. Selection of the best possible resources amongst the available resources is a challenge as many problems related to QoS have to be addressed. We propose a solution based on the Map-Reduce framework and develop the RS (Resource Selection) algorithm to address the problem of resource selection, in a network of service providers provide sensing and actuator services that are composed of large numbers of sensors and actuators.