Article Preview
TopIntroduction
The emerging field of IoT systems is new, and the underlying infrastructure is not yet mature to solve all the problems associated with the application of this technology. The IoT system is based on certain devices connected to the Internet using cloud computing, and the communication model used by the system is mainly a publication-subscription mechanism. IoT devices in general have limited memory and computing power, for example, the pic microcontroller from the microchip company does not allow recursive calls. The computation is therefore generally performed on the cloud. “maximum recursion depth exceeded” is a famous compiler message for developers on IoT devices. According to our programming experience, the esp8266 can reach a limit of 19 recursive calls. Moreover, a system with ten resources only needs at least 54 recursive calls using the Bron-Kerbosch algorithm Coen Bron and Joep Kerbosh (1973). This fact pushes us to think about developing a distributed algorithm of group classification without any recursion at all, where most of the distributed algorithms in the literature rely on recursion using a backtracking approach.
For that reason, efficiency is a pertinent concern for IoT systems. Many suggested propositions in the literature to fulfill this goal. One of these directions is using distributed ledger technology. We can cite, for example, Singh et al (2019) where the authors proposed an efficient consensus mechanism to reduces the overhead of broadcasting a new block by more than 50% coupled with saving CPU computation power along with network bandwidth. Due to the increasing number of IoT devices and the tremendous volume of data generated every day. We can mention another direction based on data compression and optimization Technique on the cloud Hossain et al (2019) . In Hossain et al (2019) the authors survey different techniques of data compressions and present the future of IoT data management in the context of cloud, fog, and mobile edge computing. The work presented in this paper solves one issue of the efficiency of IoT systems but at the level of IoT devices. The work of Singh et al (2019) and Hossain et al(2019) presented above can be combined with the approach presented here for a more advanced situation.
This fact of decoupling the computation from the user interface allows the system to evolve easily. in addition, the use of telecommunication technology of OTA (over the air update) with the IoT system makes it possible to easily fix any bug software or to extend any limitation, for example, the Chinese company Itead has released the SonOff IoT switch with a handicap of using the conventional switch in parallel with it; for example, children cannot turn on the light with the wall-mounted switch unless they ask their parents to do so for them. Itead later fixes this problem by updating the firmware of the IoT switch using OTA by allowing it to work in combination with the conventional wall-mounted switch. Although this update the switch remains limited in the absence of the Internet for example if the switch is fastened to the roof of the room, the switch mounted on the wall does not work hence we need a ladder to activate the mechanical button mounted on the housing of the SonOff switch to solve the problem which is not practical. Later Itead solves the problem by allowing IoT devices to operate in LAN mode. In addition, sharing resources and using IoT devices in combination for achieving certain functions caused the development of the notion of scenario on the cloud using a specific language for this purpose. For example, you can create a scenario to create the illusion that you are in your home. This fact prevents theft from your home. For this reason, you can schedule this in the cloud by turning on the toilet light for 5 minutes, then turning it off, then turning on the bathroom light, after that, you turn on the TV, and finally the air conditioner.