Mar 13, 2023
LoRaWAN network topology – Understanding Network and Security for Far-Edge Computing
Notice that, unlike Wi-Fi, LoRaWAN inserts gateways as intermediaries between the devices and the network. While some large enterprise Wi-Fi networks have a similar topology, it is something manufacturers bolted on later for scale-related reasons and is not part of the original 802.11x specification.
With Wi-Fi, a device is only ever associated with one access point at a time, and when they move around, their session is cut over between them. LoRaWAN, on the other hand, sends its traffic to all of the gateways it can see simultaneously. If the server needs to send a message back to the device, it will choose the best gateway to use for that purpose:
Figure 3.36 – LoRaWAN network topology
This architecture is known as star-on-star. It yields advantages that are relevant to typical LoRaWAN use cases:
Redundancy: If a gateway fails or needs to be taken offline for maintenance, devices in the network are not affected.
Affordability: Because this form of redundancy is a fundamental part of the LoRaWAN specification, it is cheaper to implement both in terms of hardware and deployment effort.
Scalability: The number of gateways a network server can manage is limited only by the processing power of that server. When that is exhausted, additional servers can be added to scale the system horizontally. There are LoRaWAN networks with 40,000 gateways that support many millions of devices 15.
15 https://www.thethingsnetwork.org/map
Direct communication between devices
The LoRaWAN protocol does not support direct communication between end nodes. This can be confusing because LoRaWAN-capable devices exist that communicate without involving the gateways. However, this is done using a different protocol such as RadioHead16 or something proprietary to that manufacturer.
16 https://www.airspayce.com/mikem/arduino/RadioHead/
Geolocation
All battery-powered LoRaWAN devices such as tags or sensors can move while they are communicating without increasing the power budget. Additionally, it is not always practical to track physical coordinates when deploying stationary devices.
Fortunately, the LoRaWAN protocol provides two inbuilt methods for determining the position of devices. Nothing needs to be added to existing LoRaWAN-capable endpoints for these to work, making it a lower-cost alternative to adding GNSS to all devices. In some cases, even when the devices do have GNSS positioning, LoRaWAN geolocation is used as a check on that position:
Received Signal Strength Indication (RSSI): This measures the received signal power in milliwatts, and is measured in dBm. This method works for coarse positioning in the 1,000-to-2,000-meter range.
Time Difference of Arrival (TDOA): Each gateway must have a tightly synchronized time source for this method. Usually, this is obtained from a GNSS network such as GPS. The network server converts the timestamp of when messages were received by each gateway into a distance. It then plots those distances and estimates the devices’ location at the intersection. If a device can reach three or more gateways, its position can be calculated to be between 20 and 200 meters.
Regardless of which method is used, a good rule of thumb is that rural deployments will see accuracies toward the lower end of the range while accuracy in urban environments will be toward the higher end. Both methods will benefit from higher gateway density.
More Details