Locations are normally estimated with a certain level of uncertainty, referred to as accuracy. This is true regardless of the technique used to determine that location. Higher accuracy means lower uncertainty.
In this page you can find information regarding all the factors that influence the Wi-Fi signals detected by the nodes and thus the level of accuracy provided by Wi-Fi Location Monitor. Being aware about these factors will allow you to know what to expect from each installation and how to bring out the most from Wi-Fi Location Monitor to get the highest positioning accuracy.
1 – Environment
1.1 – Number of obstacles
The best results with Accuware Wi-Fi Location Monitor are obtained when there is an high attenuation of the Wi-Fi signals (RSSI) detected by the nodes. Accuware Wi-Fi Location Monitor determines the position of Wi-Fi devices by measuring the attenuation of the Wi-Fi signal they transmit. This means that, as strange as it might sound, it works better inside building with walls because, since the attenuation of the radio signals is stronger, it is easier to triangulate the position of Wi-Fi devices. In open spaces the attenuation of the Wi-Fi is weaker and this leads to a lower accuracy.
- You can expect an average radius of accuracy of 3 meters when tracking Wi-Fi devices inside a shopping mall, hospital, school campus, airport, hotels, train station, subway, museum, corporate building (with cubicles or offices), jail, retail store and in general inside buildings with walls.
- You can expect an average radius of accuracy of 5 meters when tracking Wi-Fi devices inside a large warehouse, trade show, industrial plant, stadium etc.. and in general inside buildings with only few walls.
1.2 – Types of obstacles
Although Wi-Fi Location Monitor works well in environments with walls, it must be said that different type of obstacles made of different materials affect the accuracy of Wi-Fi Location Monitor:
- Thick walls – even if Wi-Fi Location Monitor works better inside building with walls, if walls are too thick (and the Wi-Fi signal absorption of the walls is too high) there can be some communication problems between the nodes. This means that the repeater nodes will not be able to send their reports to the gateway nodes and so the Accuware serve will have less data for the computation of the Wi-Fi devices locations. Anyway this is quite rare because the average spacing between the nodes is usually small enough to avoid these kind of communication problems.
- People – one source of interference is the human body that, with its high percentage of water, attenuates the RF signal. This means that if you make your tests while keeping your Wi-Fi device (Wi-Fi tag, smartphone..) in your pockets or your hands, you will see variations of the RSS depending on how you hold the device and the orientation of your body with respect to the nodes. These RSS variations lead to small variations of the locations of the Wi-Fi devices. It should be kept in mind that this effect is more evident in open spaces where there is no attenuation introduced by the walls and the attenuation introduced by the human body can be higher than the attenuation introduced by the real distance between the Wi-Fi device and the nodes
- Metal objects and mirrors – big metal objects and/or mirrors affect negatively the level of accuracy because of the reflection of these kind of objects on the Wi-Fi signal propagation.
Anyway, even if you have to be aware about the important notions just reported, you do not have to worry too much because the Accuware Wi-Fi Location Monitor algorithm is able to adapt itself to the environment in which the nodes are placed.
2 – Nodes
2.1 – Amount
Increasing the nodes density (and especially increasing the number of gateways) will always increase the system accuracy. Anyway be aware that the number of nodes is a function of the objectives of your project as explained inside the next Support page: Plan your installation.
2.2 – Deployment model
The best results are obtained when the nodes are installed uniformly (ideally in a in a grid pattern) and diminishing the spacing between the nodes will always increase the system accuracy. Even if the accuracy is affected by multiple factors (and not just by the spacing), here is a rough estimate of how the level of accuracy varies with the spacing between the nodes:
Here is a rough estimate of how the level of accuracy varies with the NODES SPACING:
~12 meters of NODES SPACING -> average accuracy of 3 meters
~18 meters of NODES SPACING -> average accuracy of 5 meters
Please keep in mind also these notions:
- placing the nodes too close to each other in the same room, does not improve the accuracy because all the nodes will receive the same signal strengths from the Wi-Fi devices regardless the location of the Wi-Fi devices
- if you do not need the same level of accuracy everywhere, you can use a different spacing for different areas
Anyway be aware that the spacing between the nodes is in function of the objectives of your project as explained inside the next Support page: Plan your installation.
2.3 – Physical location
The physical location of the nodes affects the system accuracy. For this reason we suggest to install the nodes in these physical locations (in order of preference):
- on the ground.
- on the lower part of a wall/column.
- on an item that is very close to the ground. Please make sure the signal source is attached to an object that won’t be moved.
- on the ceiling.
- inside a false-ceiling.
- on the upper part of a wall/column.
- an an item that is very close to the ceiling. Please make sure the signal source is attached to an object that won’t be moved.
Do not place the nodes halfway between ceiling and floor (e.g.on desks) and completely avoid mixed configurations (e.g. half of the nodes on the ground and half of the nodes on the ceiling) because in both cases the performances will be negatively affected due to a different propagation of the Wi-Fi signal.
There is not a maximum height for the ceiling installation but it should be noted that the accuracy drops as the height increases. From our experience, we do not recommend to install the nodes at a height over 4 meters. The fact that the height of the ceiling is not uniform is not a problem as long as the difference between the min and max height is not higher than 1 meter.
3 – Usage conditions
3.1 – Amount of traffic generated
The level of accuracy provided by Accuware Wi-Fi Location Monitor, for each Wi-Fi enabled device detected, is in function of the amount of traffic generated by each device. As explained in the previous Support page (Devices detected), it could be that a device generates Wi-Fi traffic every 5 seconds (and so is seen by the nodes every 5 seconds) but it could be also that a Wi-Fi device does not transmit constantly some Wi-Fi traffic every 5 seconds (and so is seen by the nodes less often). The higher the amount of traffic generated (and detected by the nodes) and the higher is the accuracy provided by the system. For this reason we invite you to:
- read again the page Devices detected
- determine your current scenario (among those illustrated inside the paragraph Detection frequency of Wi-Fi devices)
- determine if there are actions that can be taken in order to move into a better scenario.
3.2 – Position with respect to the nodes
Another factor that affects the accuracy is the relative location of the Wi-Fi devices compared to the area covered by the nodes. The nodes are able to detect WiFi devices within the following radius:
- OM2P (with external antennas): up to 150 feet indoor (45 meters), up to 600 feet outdoor (182 meters)
- OM2P-HS: up to 150 feet indoor (45 meters), up to 600 feet outdoor (182 meters)
Despite the fact that the detection radius is pretty large, you need to be aware that the only accurate location computed by our algorithm are those related to Wi-Fi devices that are physically located inside the ideal perimeter that can be defined by connecting the most outer nodes. The system is not accurate when the Wi-Fi devices are physically outside the nodes perimeter. In addition to this, the accuracy outside the perimeter decreases very quickly. The technical explanation of this kind of behavior is the Geometrical Dilution of Precision (GDOP). The Geometrical Dilution of Precision (GDOP) measures how much the positioning errors are amplified as a consequence of the position of the nodes relative to the WiFi device to be tracked. When you step outside the nodes perimeter, the GDOP value rapidly increases.
In addition to what has been just explained you need to be aware that Wi-Fi devices, outside the nodes perimeter, far away from the nodes and detected by just two out of all the nodes, will be placed along straight lines like those inside the red ovals of the image below. Obviously those positions are not accurate and should not be taken into account.
In order to clean the data (provided by our API/CSV files) from the wrong locations you have to develop a geo-fences mechanism in order to exclude all the Wi-Fi devices with a latitude and logitude that is outside the perimeter defined by the most outer nodes. To develop such a mechanism we strongly recommend to use the Google API. Google provides a function that allows you to know if a set of latitude and longitude is inside or outside a geo-fence (as polygon) previously defined. Please check this link for more details.
Clusters of WiFi devices outside the perimeter of the WiFi nodes could be related to one or a mix of the following reasons:
- WiFi devices that are physically on the floors above or below the floor with the WiFi nodes.
- a greater concentration of WiFi devices that is physically inside the area with the cluster.
- some obstacles that are blocking the reception of the part of the WiFi nodes.
- a poor reception of part of the WiFi nodes (e.g. the antennas are not all tightly screwed or are not all placed in vertical).
The “cluster” effect is empowered when:
- the WiFi nodes are installed in open spaces (e.g. trade-shows).
- the WiFi nodes are not placed on the ground.