Sensor Networks’ Integration - Application and Multidisciplinary Aspects of Wireless Sensor Networks - page 108

Resource
User

RLI

ELI SQI

Entity Lookup Mapping Table

REI

Query Analysis

SCI Enitity

Publication

Abstract Plan
Data Base

Ontology Task Planning

RLI

ELI

Entity Directory

Semantic Query Resolver

EPI

EPI

R

RCI

RAI

RAI

Request
Management

RDI

Dynamic Resource
Creation

Component
Data Base

REP

Resource

SCI R

Session
Monitoring

Resource Host

Resource Deployment

RPI

R

Execution Manager

RPI

Dynamic Resource Creator

RLI

Resource Directory

Resource
Publication

Resource
Data Base

Resource Lookup

Fig. 5.3 SENSEI resource layer architecture

SENSEI Resource layer architecture is given in refrence to Fig. 5.3. The following
components comprise this layer:

• (Real-world) resource provider consists of one or more real-world resources.

It provides access to these resources via Resource Access interface (RAI) and
is responsible for interaction with the rendezvous component to publish infor-
mation about its resources.
Rendezvous makes the glue between resource clients and resources. Its purpose •

is twofold: to provide mechanisms for resources to publish their capabilities and
functionalities (resource publishing interface - RPI), and to provide mechanisms for
the resource clients to lookup particular resources (resource lookup interface - RLI).
It also stores resources descriptions of all available resources at any given time
in the resource directory.
Semantic Query Resolver (SQR) is an advanced component, responsible for •

analysis of complex queries and their decomposition to simple queries. These
simple queries are then used to search for adequate resources in the resource
directory. In case no adequate resources exist, the SQR can trigger Dynamic
resource creation component to create a new resource based on the available
resources.
Dynamic resource creation is a component capable of dynamically combining •

several resources into one when required to meet specific functional
requirements.
Execution management manages long-lasting interactions and handles changes •

in resource availability at runtime.