Accuware Wi-Fi Location Monitor leverages a set of nodes that need to be deployed in the area to monitor. The nodes are programmed to detect Wi-Fi signals transmitted by nearby Wi-Fi devices and report this information to the Accuware server. This means that a proper planning of the placement of the nodes is the key of a successful installation of Wi-Fi Location Monitor.

Factors that affect the accuracy: before reading the page below you should be already aware of all the factors that affect the accuracy provided by Accuware of Wi-Fi Location Monitor, described in this page.

1.1 – Use cases

LIST OF USE CASES

Monitor the locations of all the Wi-Fi devices in a small test area (lab/office/house)
Monitor the locations of all the Wi-Fi devices in production
Detect/count the number of Wi-Fi devices in proximity of/passing by specific spots
Count the number of Wi-Fi devices entering and exiting from an area

Monitor the locations of all the Wi-Fi devices in a small test area (lab/office/house)

Nodes – Number

Very often we are asked what is the minimum number of nodes for a test installation of Wi-Fi Location Monitor inside an office/lab/house. We recommend to carry out a test installation with at least 6 nodes. This is the right number of nodes to get meaningful results and understand how the system behave with the nodes arranged in different configurations. Keep in mind that 6 nodes can provide meaningful results in an area of about 450 sqm. If you want to test the system in a bigger area then you should refer to the use case below Monitor the locations of all the Wi-Fi devices in production to understand how to estimate the correct number and placement of nodes in a bigger area.

Nodes – Deployment model

As explained inside the Accuracy page, the most accurate results can be obtained when tracking devices inside the ideal perimeter that can be defined by connecting the most outer nodes. This is the reason why the first step during the deployment of the the nodes, is to place them along the perimeter of the area that will contain devices to be tracked. Once a first set of nodes has been placed along the perimeter then you can place the remaining nodes uniformly inside the area that you want to monitor.

Accurate estimate: feel free to use this form and provide us:

  • a floor plan with a unit of measurement and the highlighted areas that need to be monitored.
  • a floor plan with the locations of the electricity outlets (this is needed in order to minimize the number of extension cords).
  • 8-10 meaningful pictures of the environment (if available).
  • any information regarding constraints related to the placement of the nodes.

We will be pleased to provide an estimate regarding the number of nodes needed and the deployment model suggested.

Since this is an installation with 5-6 nodes, you will have to reach one of these configurations:

EXAMPLE 1 – Officeaccuware_wifitracker_office_esample1

EXAMPLE 2 – Officeaccuware_wifitracker_office_esample2

EXAMPLE 3 – Officeaccuware_wifitracker_office_esample3

Nodes – Physical locations

Please look at the paragraph 2.3 – Physical location of this support page.

Nodes – Number of nodes connected to Internet (gateways)

If you are installing a total of 5-6 nodes it’s enough to connect to Internet only one node (to act as a gateway). If possible, we recommend to place the gateway node at the center of the mesh network so that all the nodes are able to communicate directly with the gateway.

If the gateways are behind a firewall, then you have to check if ports 53/tcp (DNS), 80/tcp (HTTP) and 443/tcp (HTTPS) are open on your firewall and if the following domains are accessible:

  • its.accuware.com
  • s3.amazonaws.com
  • cloudtrax.com (and all subdomains)
  • downloads.openwrt.org

As a general rule, you have to check that the Ethernet cables used by the gateways provide a valid the Internet connection (you can test this easily by connecting a laptop and checking if you are able to surf the net and access the domains listed above).

Data collection

With this deployment you can collect the information regarding the locations of all the WiFi devices detected. The location of all the WiFi devices detected can be retrieved:
  • REAL-TIME – calling every 5 seconds this API and using at the LOC field of EXAMPLE 2 
  • HISTORICALLY – by downloading every day the CSV daily reports​​ and using the fourth and fifth fields of each line of the CSV files

Monitor the locations of all the Wi-Fi devices in production

Nodes – Number

There is not a unique and standard equation to compute the number of nodes and there is no limits regarding the maximum number that can be managed by our server (we are able to scale our resources dynamically and so there are virtually no limits regarding the computational power that we can provide as explained in this page).

Increasing the nodes density (and especially increasing the number of gateways) will always increase the system accuracy. If you want a rough estimate regarding the number of nodes needed to monitor all the Wi-Fi devices within a specific area (with a specific level of accuracy) you can use these formulas:

    \[WiFi Nodes = (\frac{\sqrt{Area}}{Node Spacing}+1)^{2}\]

Here is a rough estimate of how the level of accuracy varies with the NODES SPACING:

 HIGH accuracy (3 meters of avg. radius of accuracy) -> ~12 meters of NODES SPACING
LOW accuracy (5 meters of avg. radius of accuracy) -> ~18 meters of NODES SPACING

You can use this tool to estimate the number of nodes needed to cover a specific area.

Nodes – Deployment model

The most accurate results can be obtained when tracking devices inside the perimeter defined by the nodes. This is why the first step during the deployment of the the nodes, is to place them along the perimeter of the area that will contain devices to be tracked. Once a first set of nodes has been placed along the perimeter then you can place the remaining nodes uniformly inside the area that you want to monitor.

Accurate estimate: feel free to use this form and provide us:

  • a floor plan with a unit of measurement and the highlighted areas that need to be monitored.
  • a floor plan with the locations of the electricity outlets (this is needed in order to minimize the number of extension cords).
  • 8-10 meaningful pictures of the environment (if available).
  • any information regarding constraints related to the placement of the nodes.

We will be pleased to provide an estimate regarding the number of nodes needed and the deployment model suggested.

EXAMPLE 4 – Small store/retail

accuware_wifitracker_store_esample1

EXAMPLE 5 – Office space

accuware_wifitracker_office_esample4

Nodes – Physical locations

Please look at the paragraph 2.3 – Physical location of this support page.

Nodes – Number of nodes connected to Internet (gateway nodes)

The number of nodes that need to be connected to the Internet to act as gateways is function of different things:

  • Total number of nodes: if you have only one node, obviously that node must be connected to Internet and act as a gateway. If you have more than one node, as a rule of thumb, we suggest to connect to Internet at least 1 node every 5 nodes (e.g. if you have a total of 30 nodes you should connect to Internet 6 nodes).
  • Deployment model: if the nodes are close enough to create a reliable mesh network and transmit the collected information to the gateway(s), then 1 gateway every 5 nodes is enough. If the nodes aren’t close enough to each other to create a reliable mesh network, then you will have to connect to Internet also the nodes with a poor connection. The reliability of the mesh network can be checked easily inside the Cloudtrax dashboard.
  • Internet connectivity: if you want to provide the Internet connectivity to people then we recommend to connect to Internet as many nodes as possible. As a rule of thumb, we suggest to connect to Internet at least 1 node every 3 nodes.  At this link you can find some of the many success stories of (big) companies that are using the Open Mesh nodes to provide internet connectivity.

If the gateways are behind a firewall, then you have to check if ports 53/tcp (DNS), 80/tcp (HTTP) and 443/tcp (HTTPS) are open on your firewall and if the following domains are accessible:

  • its.accuware.com
  • s3.amazonaws.com
  • cloudtrax.com (and all subdomains)
  • downloads.openwrt.org

As a general rule, you have to check that the Ethernet cables used by the gateways provide a valid the Internet connection (you can test this easily by connecting a laptop and checking if you are able to surf the net and access the domains listed above).

Data collection

With this deployment you can collect the information regarding the locations of all the WiFi devices detected. The location of all the WiFi devices detected can be retrieved:
  • REAL-TIME – calling every 5 seconds this API and using at the LOC field of EXAMPLE 2 
  • HISTORICALLY – by downloading every day the CSV daily reports​​ and using the fourth and fifth fields of each line of the CSV files

Detect/count the number of Wi-Fi devices in proximity of/passing by specific spots

WARNING: before planning an installation of Accuware Wi-Fi Location Monitor with nodes placed ONLY in the spots of interest, please be aware that with this type of installation Accuware Wi-Fi Location Monitor will not be able to triangulate the positions of the Wi-Fi devices detected. This means that the locations of each one of the Wi-Fi devices detected will be equal to the location of the node that is detecting the highest RSS (when looking at the Accuware dashboard and when calling the Accuware Wi-Fi Location Monitor “stations” API).

Despite this, with this type of installation, you will be able to make assertions regarding the proximity of the Wi-Fi devices detected with respect to the nodes by analyzing the RSS values returned by the Accuware Wi-Fi Location Monitor API. Please look at this article.

Nodes – Number

It is necessary to install one node for each one of the spots that need to be monitored. If the area to be monitored is complex or big then it is necessary to sub-divide the area in sub-squares and install 1 node in the center of each square.

IMPORTANT INFO: you can find all the information about how to properly handle multiple Accuware sites/Levels/Cloudtrax networks inside this page.

Nodes – Deployment model

We suggest to install each one of the nodes at the center of each specific spot that needs to be monitored. If the area to be monitored is complex or big then it is necessary to sub-divide the area in sub-squares and install 1 node in the center of each square.

Accurate estimate: feel free to use this form and provide us:

  • a floor plan with a unit of measurement and the marked spots that need to be monitored.
  • (OPTIONALLY) a floor plan with the locations of the electricity outlets (this is needed in order to minimize the number of extension cords).
  • (OPTIONALLY) 8-10 meaningful pictures of the environment (if available).
  • (OPTIONALLY) any information regarding constraints related to the placement of the nodes.

We will be pleased to provide an estimate regarding the number of nodes needed and the deployment model suggested.

EXAMPLE 1 – Office

Accuware WiFi Tracker - Spots

The deployment of the example in the image above allows to monitor 3 spots. All the Wi-Fi devices detected by specific nodes with a RSS value equal or greater a chosen threshold will be considered as inside a specific area (inside a specific radius). Example of rules and thresholds:

  • If the RSS detected by node A is greater or equal to -50 dBm then the Wi-Fi device is inside the VIOLET area.
  • If the RSS detected by node B is greater or equal to -35 dBm then the Wi-Fi device is inside the YELLOW area.
  • If the RSS detected by node C is greater or equal to -70 dBm then the Wi-Fi device is inside the GREEN area.

RSS vs distance (radius): even if it is true that the RSS value decreases as the distance from the node increases, it is also true that the RSS value does not decrease linearly and it is also affected by different materials in the environment. For these reasons we suggest to perform some empirical tests to determine how the RSS varies with the distance from the nodes (the spots) in your specific environment. These kind of empirical tests can be performed following the instructions in this support page.

Nodes – Physical locations

Please look at the paragraph 2.3 – Physical location of this support page.

Nodes – Number of nodes connected to Internet (gateway nodes)

The number of nodes that need to be connected to the Internet to act as gateways is function of different things:

  • Total number of nodes:
    • If you have only one node, obviously that node must be connected to Internet and act as a gateway.
    • If you have more than one node, as a rule of thumb, we suggest to connect to Internet at least 1 node every 5 nodes (e.g. if you have a total of 30 nodes you should connect to Internet 6 nodes).
  • Deployment model:
    • If the nodes are close enough to create a reliable mesh network and transmit the collected information to the gateway(s), then 1 gateway every 5 nodes is enough.
    • If the nodes aren’t close enough to each other to create a reliable mesh network, then you will have to connect to Internet also the nodes with a poor connection (the reliability of the mesh network can be checked easily inside the Cloudtrax dashboard the menu Manage > Access points).
  • Internet connectivity: if you want to provide the Internet connectivity to people then we recommend to connect to the Internet as many nodes as possible. As a rule of thumb, we suggest to connect to the Internet at least 1 node every 3 nodes.  At this link you can find some of the many success stories of known brands/companies that are using the Open Mesh nodes to provide internet connectivity.

If the gateways are behind a firewall, then you have to check if ports 53/tcp (DNS), 80/tcp (HTTP) and 443/tcp (HTTPS) are open on your firewall and if the following domains are accessible:

  • its.accuware.com
  • s3.amazonaws.com
  • cloudtrax.com (and all subdomains)
  • downloads.openwrt.org

As a general rule, you have to check that the Ethernet cables used by the gateways provide a valid the Internet connection (you can test this easily by connecting a laptop and checking if you are able to surf the net and access the domains listed above).

Data collection

Raw data

With this deployment you can collect the information regarding the proximity/distance of all the Wi-Fi devices with respect to the nodes by analyzing the RSS values detected by Wi-Fi Location Monitor. The raw data regarding the proximity/distance of the WiFi device with respect to the nodes can be retrieved:

  • in REAL-TIME – calling every 5 seconds this API and looking at the RSS field of the JSON returned.
  • HISTORICALLY – by downloading every day the CSV daily reports​​ and looking the sixth and seventh fields of each line of the CSV files.
Aggregated data

If you are not interested in the raw data and you are looking for aggregated metrics (the number of Wi-Fi devices detected within a certain distance) then you can retrieve these information using Accuware Analytics:

Count the number of Wi-Fi devices entering and exiting from an area

Small area/store

Nodes – Number

It is necessary to install one node for each area/store that needs be monitored.

Nodes – Deployment model

We suggest to install each one of the nodes at the center of each area/store that needs to be monitored.

EXAMPLE – Small store

Accuware_Count_People_Small_Store_Enter_Exit

The deployment of the example in the image above allows to monitor a store. All the Wi-Fi devices detected by the node (at the center of the store) with a RSS value equal or greater a chosen threshold (e.g. -60 dBm) will be considered as inside the store (inside a specific radius).

RSS vs distance (radius): even if it is true that the RSS value decreases as the distance from the node increases, it is also true that the RSS value does not decrease linearly and it is also affected by different materials in the environment. For these reasons we suggest to perform some empirical tests to determine how the RSS varies with the distance from the nodes (the spots) in your specific environment. These kind of empirical tests can be performed following the instructions in this support page.

Nodes – Physical locations

Please look at the paragraph 2.3 – Physical location of this support page.

Nodes – Number of nodes connected to Internet (gateway nodes)

Since only one node is used, the node must be connected to Internet and act as a gateway. If the node is behind a firewall, then you have to check if ports 53/tcp (DNS), 80/tcp (HTTP) and 443/tcp (HTTPS) are open on your firewall and if the following domains are accessible:

  • its.accuware.com
  • s3.amazonaws.com
  • cloudtrax.com (and all subdomains)
  • downloads.openwrt.org

As a general rule, you have to check that the Ethernet cable used by the node provides a valid the Internet connection (you can test this easily by connecting a laptop and checking if you are able to surf the net and access the domains listed above).

Data collection

Raw data

With this deployment you can count the number of Wi-Fi devices entering and exiting from an area by analyzing the time-stamps and the RSS values of the Wi-Fi devices detected by Wi-Fi Location Monitor. The raw data regarding the proximity/distance of the WiFi device with respect to the node can be retrieved:

  • in REAL-TIME – calling every 5 seconds this API and looking at the RSS field of the JSON returned.
  • HISTORICALLY – by downloading every day the CSV daily report​​ and looking the sixth and seventh fields of each line of the CSV files.

1.2 – Multi-floor buildings

Accuware Wi-Fi Location Monitor can work on multi-floor buildings, allowing to distinguish Wi-Fi devices located on multiple floors/levels. Anyway, the attenuation introduced by the floors is not enough to ensure that the Wi-Fi devices detected are on the same level in which the nodes are installed. This because the propagation of the Wi-Fi signal is spherical and the nodes are able to detect the Wi-Fi signal of devices within a spherical radius of 100-150 feet. This means that Accuware Wi-Fi Location Monitor triangulates the locations of Wi-Fi devices located not only on the floor in which the nodes are installed, but also on the floors above and below.

In order to create a distinction between Wi-Fi devices located on multiple floors, there are 2 options:

  1. the first (more reliable) solution is to install a comparable number of nodes on the floors above and below the floor of interest. The nodes must be placed inside the Accuware dashboard on different logical Levels according to the physical floors in which they are installed (this can be done by adding new levels using the Levels section of the dashboard). In addition by keeping the same vertical alignment of the Wi-Fi access points and/or iBeacons of different levels the results will be better (reducing the possibility of jumps between the levels)
  2. the second solution is to analyze the RSS value provided by the Accuware Wi-Fi Location Monitor “station” API (or provided inside the CSV files) to create a filter that considers on a different floor, all the Wi-Fi devices with a very low RSS value (i.e. minor or equal to -80 dBm) and thus physically on the floor above or below. Please take a look at the RSS fields returned in the EXAMPLE 2 of this page.

Even if it is true that the RSS value, for a Wi-Fi device, decreases as the distance from the nodes increases. It is also true that the RSS value does not decrease linearly and it is also affected by different materials in the environment. For these reasons we suggest to perform some empirical tests to determine how the RSS varies with the distance from the nodes in your specific environment. These kind of empirical tests can be performed following the instructions in this support page.

PLEASE READ:

  • The field “Altitude (meters)” that can be added for each “Level” is just for your reference but it is not taken into account by the Accuware Wi-Fi Location Monitor algorithm at this time.
  • For each Site you will find the default level with LevelID 0. This level can not be eliminated.
  • Using the Accuware dashboard it is possible to create up to 30 Levels (in addition to the Level 0). Additional levels can be created using this PUT API call.