By Leslie Mulder, President of Exegin Technologies
Network reformation is crucial for modern wireless networks, which require near-continuous connectivity and typically contain a multitude of different nodes. It’s essential that when an entire network suffers a power loss to the extent that all nodes must restart, they are returned to operation as quickly and efficiently as possible.
Power loss to the entire network may occur for a variety of reasons. These include a fault or grid overload, an extreme weather event such as a hurricane, natural events such as nighttime at a solar farm or a windless period at a wind farm, or intentional energy conservation or maintenance such as streetlight networks during the day.
This concept is different from the network formation, as it is assumed that each of the nodes involved have already been authenticated onto the network and retained their security material over the duration of the power failure.
The ability for a network to recover quickly and effectively is crucial in environments where power disruptions may occur, and will be a determining factor for utilities and municipalities in choosing their network technology. To this end, a recent demonstration by Wi-SUN Alliance member Exegin Technologies used network simulation software in order to provide a realistic real-time demonstration of Wi-SUN FAN network reformation.
An insightful simulation
Exegin chose a set of 530 Wi-SUN FAN nodes (23 by 23 nodes, plus a border router) that were positioned on a perturbed regular grid with an average inter-node spacing of 1000 meters to represent a typical network topology. The set-up enables other configurations to be simulated, including stars, string of pearls, dense, sparse and combinations thereof, to stringently test Wi-SUN FAN’s network reformation capabilities.
Each node was set to have a transmit power of 0 dBm and a receive sensitivity of -100 dBm, with the system set to use the North American sub-GHz band of 902-928 MHz, and a symbol rate of 300 kbps.
The underlying simulation software functioned at the physical layer of the network to provide accurate modelling of the interaction between radio frequency nodes in a correct deployment. The use of a free space path-loss model – a mathematical concept used to describe how radio signals travel in free space with no obstructions affecting them – was used.
To bring the simulation closer to emulating real environments, random interference in the form of Gaussian white noise, with a worst case signal to noise ratio of 75 dB, was introduced to ensure uniquely different behavior was exhibited each time it was run. Additionally, the simulation provided results that are within 10% of the behavior of the actual, physically comparable deployments. This made it a suitable digital twin.
Explaining the methodology
A transmit power of 0 dBm (decibel-milliwatts) represents a moderate power level that balances signal strength with energy efficiency. This allows nodes to communicate effectively within their intended range without excessive power consumption.
Additionally, the receive sensitivity of -100 dBm indicates the minimum signal strength that a node can detect and process. This high sensitivity enables nodes to pick up even weak signals from distant or obstructed nodes, enhancing the network’s overall connectivity and resilience. Together, these specifications contribute to Wi-SUN FAN’s ability to create robust, far-reaching mesh networks that can reform quickly after power disruptions.
Demonstrable results
The following set of images – taken at various stages of the reformation process – capture the transition from a completely unjoined state to a fully joined one. A full video of the simulation can be found here.
The simulation was allowed to run many times to the point where all nodes in the network had rejoined. The statistics from those multiple simulations yielded an average reformation time (for 529 nodes to join the network) of 6 minutes and 45 seconds with half of the nodes being re-joining in less than 2 minutes and 20 seconds.
Of course, the behavior of a mesh is dependent on the density of nodes within the network. During the simulation, Exegin noted the average density seen by each node in the network was approximately 16.6 nodes/node. This means that there are on average 16.6 nodes in each node’s neighbor table.
Setting the industry standard
These results demonstrate the capabilities of Wi-SUN FAN implementations for optimal network reformation in free space environments. The underlying nature of radio communication will always mean there is a degree of randomness in network behavior, but the simulation demonstrated near-consistent results for the duration of testing.
In terms of what this means for the industry, it is clear Wi-SUN FAN can deliver and maintain reliable, seamless connectivity for large-scale wireless mesh networks. With the standard set to be expanded through the evolution to FAN 1.1, utilities and municipalities deploying intelligent grid and smart city solutions can rest assured they can rely on Wi-SUN technology when it comes to reformation and general performance.
To see the full results of the simulation, please visit Exegin’s whitepaper “Wi-SUN FAN Network Reformation Characterization”.