Abstract:
Wireless Sensor Networks (WSNs) are increasingly attractive means to monitor
environmental conditions. A WSN consists of large numbers of cooperating small-scale
nodes capable of limited computation, wireless communication, and sensing. By
correlating sensor output of multiple nodes, the WSN as a whole can provide
functionality that an individual node cannot. The basic mode of operation of WSNs is
significantly different from traditional computer networks, due to their tight integration
with the physical world. Additionally, WSNs have some unique characteristics that make
the development of applications non-trivial.
The sensing tasks are usually rather high level, such as “report size, speed and direction
of vehicles over 40 tons moving through a certain area”. On the other hand individual
sensor nodes typically provide very simple functionality, such as determining movement
at certain place. In order to solve complex high-level sensing tasks, the limited sensor
nodes have to coordinate and split the task among themselves taking into account the
individual characteristics of nodes.
Middleware for WSNs should support the implementation and basic operation of a sensor
as outlined above. But currently we have no such standard middleware.
Our project is focused on research issues in designing middleware for WSNs. We studied
different middleware models proposed for WSNs and according to the unique
requirements posed by WSNs, we proposed our Middleware for WSN that successfully
v
addresses the issues like scalability, robustness and energy efficiency. And as a part of
our proposed Middleware, we simulated a data centric high level communication
paradigm for static as well as for Mobile Adhoc WSNs.
