A Framework for Heterogeneous Resource Allocation in Sensor-Cloud Environment

With the advent of sensor technologies, cloud applications are able to acquire sensed data from remotely located sensors which are geographically distributed. Cloud computing, when integrated with sensor devices facilitates pervasive and shared access to the sensor data from geographically distributed sensors. There have been some research works targeting integration of cloud and sensors. However, so far integration of sensors and applications are done at the application layer. On the other hand, our objective is to propose a sensor-cloud architecture which creates virtual sensors, i.e. a layer of abstraction at the Infrastructure as a Service level. In this architecture, remotely located sensing resources are treated as citizens of generic resource family just like CPU, memory and storage devices. However, in order to realize this architecture, a resource model is required. The resource model should be sufficiently generic, so that it can incorporate computational, as well as sensing resources. In this paper one such generic resource model along with request, reservation and allocation facilities for both computational and sensing resources is presented.