InitRech 2015/2016, sujet 23 : Différence entre versions
m (→Summary) |
m (→Summary) |
||
Ligne 20 : | Ligne 20 : | ||
Like the IP/TCP, uses to communicate in a network, is running on this way, and is a problem for our tiny device with few Mhz frequency and kilo byte of memory. | Like the IP/TCP, uses to communicate in a network, is running on this way, and is a problem for our tiny device with few Mhz frequency and kilo byte of memory. | ||
This paper claim a Personal Area Network (PAN) wherein tiny devices can easily discuss. A user (standard web browser) with more power than our peripheral can be running and do an more complicated execution. | This paper claim a Personal Area Network (PAN) wherein tiny devices can easily discuss. A user (standard web browser) with more power than our peripheral can be running and do an more complicated execution. | ||
+ | This is the main goal of the Internet of thing. The capabilty to tiny devices to discuss, interact without user activity. | ||
This state of art is the beginning for our contributors and expose the problem by replying it. | This state of art is the beginning for our contributors and expose the problem by replying it. |
Version du 15 juin 2016 à 18:55
Summary
This paper discuss the technical feasibility of embedded Web servers and propose a solution for tiny devices such as Arduino or Funcard.
The number of devices running in our environment increase and the desire of connect all of them and do it with the minimum of program and memory is even bigger than 2009 (Date of the article). The well-known standard for this kind of system is the UPnP, a network protocol which have the ability to making discuss peripherals. But this technology have several drawbacks like security or the unavailability to create hierarchical network. Because of the hardware constrained devices, it's important to keep in mind that the connexion are available in a little perimeter, the physical layer of this communication is therefore limited to (USB, Bluetooth, ZigBee...). Another technology which uses a non standard protocol is described by the authors and could be use on tiny hardware constraint devices. The JXTA is mentionned for its capability to be run on this kind of hardware, instead of the the successor of the UpnP : DPWS. This last one have bigger memory and cpu usage need but include a secure authentication system. To understand the dilemma, the authors described the actual design of the Web of Things, by explaining the AJAX model (Asynchronous JavaScript and XML)
This web design is done in two phase :
- The reception of static files (HTML, CSS, Javascript).
- The execution of the files get in the first phase and the asynchronous exchange between server and user.
This methodology reduces web traffic because "dynamic content are semantic information and the formation rules are loaded only once in the initialization phase." But if we dig even deeper the protocol uses to discuss are running in a different way and allow multiple request in short time. Like the IP/TCP, uses to communicate in a network, is running on this way, and is a problem for our tiny device with few Mhz frequency and kilo byte of memory. This paper claim a Personal Area Network (PAN) wherein tiny devices can easily discuss. A user (standard web browser) with more power than our peripheral can be running and do an more complicated execution. This is the main goal of the Internet of thing. The capabilty to tiny devices to discuss, interact without user activity.
This state of art is the beginning for our contributors and expose the problem by replying it.
Main Contribution
la mise en relief de la contribution principale de l'article en une dizaine de lignes
- Solution for IoT design
Applications
applications découlant de l'article en une vingtaine de lignes.