Research

Problem

The target environment I consider in my thesis is the Disconnected Mobile Ad hoc Network (D-MANETs). If you need any information about this special type of network, please refer to our team website.

The increasing popularity of D-MANET has turned it in to popular a distributed environment. However, designing and implementing applications for such a complex environment is a challenging task given the extra challenges posed by this type of MANET:

Devices’ limited resources
Frequent disconnections
Long delay paths
No APIs

Traditional Solution:

Developers need to REPROGRAM traditional applications in order to have mostly mobile versions of them

Solution

What do you know about Middleware?

If you know nothing, then you should know!

Click here to read what Wikipedia say about middleware.

Ok now what do you think of this?

As you see, middleware acts a critical level between D-MANETs and application levels. So all of services offered by D-MANETs will be moved from multiple instances into a centralized offering. This central provision of services makes the developers independent from D-MANETs challenges.

Pros:

It provides the developer with well known APIs which could be used by application’s developer in order to construct more complex and flexible applications.
traditional applications could be Deployed over D-MANETs with almost no change in the source code.

Cons:

Simply, there is no one ;-)

The worldwide developers can simply focus on writing standard applications, and the middleware will deploy them over any type of MANET.

Write Apps Once, Deploy Apps Everywhere

Take a look

Okay now, that is the moment of truth. Put on your glasses and get ready to give sneak peek at my thesis research.

I have just started my thesis by developing a communication middleware called JOMS. JOMS is a JMS provider specifically designed for D-MANETs. JOMS overview along with the latest version of source code can be found here.

However, the communication is usually used by the participants in order to coordinate their actions and achieve a mutual understanding between them. That's why I have also developed a coordination middleware for D-MANETs which I called JION. JION, as you will see, is a JavaSpaces implementation specifically designed for D-MANETs. JION overview along with the latest version of source code can be found here.

In the near future

Well I well try to enhance the performance of my two middlewares using standard programming abstractions such as "future object".