The number of social network users in the last few years has increased at an unexpected rate. For example, the Facebook web site, which was first established in 2004, recorded 1.09 billion simultaneous active users in March 2016. This number represents every seventh inhabitant of the planet as an active user of online services of one of the providers of social networks. But considering that there are many online social network service providers, and that their users are located in different time zones, the estimated number is much bigger.
In the future, we can expect the creation of social networks of connected devices. The number of connected devices will grow at the growth rate of users of social networks, or probably much faster. Devices such as sensors, smart sensors and other unimaginable devices already do, and will in the future primarily communicate with each other (Machine-to-machine communication - M2M). Later, these devices will be connected to Internet through gateways to communicate with management server infrastructure, which is a term used today known as Internet of Things (IoT).
The estimated number of connected devices will be several times greater than today's present number of smart devices such as smart phones, tablets and computers on network. It is expected that the connected devices will be cheap and limited in computing power. And that they will not have as powerful hardware as today's smart devices. This means that they will not have the capacity of traditional Internet Protocol connectivity.
Internet Protocol assumes that all devices are alike on the network and as such represent a network of equal partners. Approach that will be discussed in this thesis will be different. The traditional network protocols such as TCP / IP are robustly designed with reliability, for use in many repetitive cycles and in large data streams. These features are also desirable in IoT networks, but as previously mentioned standard Internet network protocols are not the most appropriate as they require a lot of overhead data to function.
The ratio of the amount of overhead data needed in standardized Internet protocol compared to small sets of data sent by a typical device in the IoT is too high. Therefore, it makes sense to introduce new solutions that enable co-existence with devices that do not support IP or with those that cannot. The idea behind introducing a new solution lies in the fact that the burden on otherwise simple devices such as sensors, switches, etc., with a full network Internet protocol stack also significantly increases the final cost and complexity of a number of simple devices.
Traditional network protocols and applications are mainly designed in such a way that on one end is human and communications are consequently designed to optimize the user experience. This traditional approach requires a lot of overhead data, which is unnecessary in the case of M2M communication. The development of new architectural solutions for the forthcoming connection devices in the IoT should lead to the minimization of necessary overhead data.
The future of IoT is in connecting devices to the network, which have never communicated with each other before or were never connected to the Internet. Such systems or devices are not "smart" in the modern sense of the word, but they are very suitable for carrying out their primary role of specialized tasks and they will represent the end or edge devices in the IoT. The end devices may have an address that is unique in the local community, but is not necessarily globally unique. This issue will solve intelligent network edge devices with globally unique IP address. Intelligent network edge devices will constitute a communication path to the Internet for local connected community of devices. Such an approach would allow for rapid growth in the number of connected devices and greatly reduce the cost of introducing functionality of connections taking into account the anticipated number of connected devices.
M2M is the major part of communication in the IoT. As a result, we could quickly select the most appropriate mode of communication type as peer-to-peer. However, many end devices in the IoT will not be communicating with each other, since there is no need to exchange information between them.
As already mentioned, to burden of all end devices with Internet Protocol stack, adding computing power and memory to provide a complete communication peer is unnecessary. This would create more risk for errors, the possibility of errors in the configuration and management, and vulnerability to network attacks and infections with malware.
Adding end devices to the IoT is carried out in such a way that they would send or receive information only relevant to them. The universal approach for building the architecture of IoT increases the level of intelligence and capability of networking of devices in the direction which the end device is sending information. This concept allows for the transport of data through various points in the network without burdening each device included in IoT, with the same capacity of connectivity. In situations where endpoint devices are not capable of intelligent communication functionality, other devices must be set in the network to ensure the efficient transfer of data. The most realistic concept of the emerging architecture of the IoT is in the division of the network into three functional sections/units, which enables the employment of a variety of complex network functions in a particular part of the network, according to needs. Three functional units are:
- the end devices,
- the propagator nodes providing transport and gateways to the traditional Internet,
- the integrator functions offering analysis, control, and human interfaces to the IoT.
In this master thesis I intend to present and implement a practical example of the concept of the architecture for the forthcoming IoT with the construction of a typical representative of each functional unit.
The end device, which features easy to use, low power consumption and limited communication with the neighboring devices consists of a radio interface manufacturer Nordic Seminconductor model nRF24l01+ connected to a microcontroller made by manufacturer Arduino.
The propagator device with the functionality of networking and propagation of messages consists of a radio interface manufacturer Nordic Seminconductor model nRF24l01+ and Internet network interface connected to an Arduino microcontroller. By using the network interface capable of providing connectivity from source to sink at the level of Internet Protocol IP, it will send messages to the integral unit using a telemetry protocol MQTT.
A minimalist computer, Raspberry Pi, is used for the implementation of integrator functions for analysis, control and user interface for IoT. Messages transmitted from the propagator device are received and sent using a transport protocol MQTT. For integrator functions is used open source software OpenHAB.
|
