The Internet of Things consists of thousand of devices, which we use in everyday life – at homes, offices, enterprises, or even wear on ourselves. They create a firm ecosystem of connected online gadgets and move the world towards digitalization.
Every device produces and transfers a high quantity of data, which should be sent to the analysis and further manipulations through the internet. Sometimes they are located far from the area with good coverage, as well as it may be costly to provide each device with an internet access point. To overcome those difficulties a new approach to maintaining the work of IoT networks was introduced, which is a mesh network.
The non-mesh network enables the devices to communicate directly with a base station, while a mesh network provides a device-to-device connection. Each device communicates through nodes, which branch off to other devices or nodes creating a wireless stream. A Mesh network can be composed of thousands of nodes, thus covering a wide area.
The mesh network is composed of wirelessly connected nodes. Nodes exchange the data between one another: they transfer and receive the information. With the help of software, nodes understand what to do with information and how to interact with it. The network also includes getaways connected to the Internet. It may also include endpoints, which are the devices that only transfer their own data, never passing the data of other nodes.
The connected nodes exchange information through numerous paths. Thus the resilience of the network increases even if a node or connection fails. If one of the nodes or paths goes out of order, the network reconfigures itself with the help of self-healing algorithms. Every other node extends the radio signal, such as Wi-Fi or cellular, further than the last one. Thus we exclude the possibility of spots without coverage.
Decentralization of a mesh network creates different routes for information to run among the nodes and reach the getaway. There are two main ways of data transmitting, such as flooding and routing.
The flooding technique corresponds to broadcasting data among each node in the network. As all nodes may not be available at the same time, data is delivered by a subset of them. A portion of the data is stored in each node. To enhance throughput, a protocol selects the senders for each data transmission. The information passes through very fast, though this technique requires quite a lot of power.
Another way of data transfer is called routing. It is based on the principle of self-healing algorithms, such as Shortest Path Bridging. Only one node receives the data. If the chosen node is out of reach, the data is sent via another path.
A variety of embedded devices compose an IoT network. Temperature sensors, lighting controls, automatic sunshades, and others are among them. All of them need to be connected to the internet as it’s only one way to send data to be checked, analyzed, and updated if needed. A Wi-Fi chip can be used for this reason. But this solution may not meet the need of specific projects. For example, it’s not applicable for those IoT devices, which work on batteries, as Wi-Fi consumes a significant quantity of energy. It is also costly, so it won’t work for devices for mass production. Also if you need to transfer a small amount of data once in a while you don’t need a constant connection to the internet. If the devices are located in areas with poor cellular or internet services you won’t be able to use Wi-Fi for its signal limitations.
Another way out is to connect all the devices to internet-connected IoT gateways, as in the star network topology. But it’s good until the number of linked devices exceeds the capabilities of a certain IoT gateway. Communication collisions can occur when too many devices share the same frequency.
For the star network, it’s also crucial that every node is available. The sensors can be connected to one another, but only the authorized IoT gateway can transfer data to the internet. If one sensor fails to access an IoT gateway, it loses the connection and can’t send data anymore. There’s where a wireless mesh network can help you out.
First of all, the IoT mesh network is self-healing, due to specific algorithms, for example, Shortest Path Bridging. If one of the nodes fails to connect, this algorithm passes the information to an available one through the shortest route. So that a single failure doesn’t affect the whole network. It’s also beneficial for security reasons, as if one node is attacked, you can easily replace it. Though it’s hard and time-consuming to find a failed node.
The configuration and administration of the mesh network are also easy. The new nodes immediately calibrate and join to the network after they have been added to the network.
The mesh network is scalable: the number of nodes doesn’t affect the work of the network. Vice versa, the more nodes you add, the more paths appear for data transmitting.
Last but not least, it can reduce the cost of the project, as you can use a number of mesh tools for free. Also, sensors which use such tech as LoRa and Bluetooth are less expensive than those using 3G or Wi-Fi for work.
A large amount of transmitted data through a mesh network decreases the network performance, so load balancing is of vital importance for sustainable work.
As long as nodes join and disconnect from the mesh anytime, pay attention to updating routes of data transmission. Also, you should manage routes from time to time, for if data sent through the nodes didn’t get to the IoT gateway, it should be sent again. If the IoT gateway is inaccessible the network should break current routes. If you fail to do it, the linked nodes will still send data and consume battery power for no reason.
You will also need an effective but still power-constrained CPU to manage the software implementing both the mesh network nodes and channel access method logic.
A mesh network has the advantage over other network topologies in that if a node is too far away from the hub, it may still interact with a neighboring node and thus the data reaches a router. It may be used for home and medical monitoring, security systems, industrial monitoring, and control.
It is specifically good for the devices, which need to communicate constantly and are located in areas with poor wireless signal or whenever a large coverage is required. For example in-door environments or city-scale systems, including street lighting and sensor-based measurement.
For big indoor locations, IoT systems must cover a vast region. It may be warehouses, stores, and office spaces. Wireless communication systems that rely on cellular connectivity for in-door conditions might be problematic since cellular communication can be inconsistent. This is especially true in basements and deep structures. And as we already know, a
Several IoT systems are now in operation at the municipal scale, such as lighting for streets, monitoring of air pollution, or trash can usage. As a mesh network can cover wide regions and extend with additional devices, it is a viable choice for city-scale systems. Also, weather conditions can affect the signal, and devices within wireless mesh networks pick the most appropriate network structure.
To sum up, the IoT mesh network is great for sending small amounts of data and scaling by connecting more nodes while maintaining good performance. The ability to self-healing reduces the risks of losing data and security breaches, as if one device is out of order it doesn’t break the whole system and the algorithm finds another path to transfer the information. You may also choose between full mesh topology, connecting all nodes, or partial, which connects directly specific nodes. This depends on the network's general traffic pattern and which nodes or connections can fail to work.