Positioning in Wi-Fi networks with high accuracy
One of the most popular categories of mobile applications today is the one that provides location-based services. Many people use positioning systems on their navigation devices, smartphones and tablets. The weak side of these solutions is that the use of the Global Positioning System (GPS) is not available indoors because of the strong redemption of signals by walls and floors of buildings. This opens a niche for reliable solutions with positioning in rooms.
Today, there are a number of approaches and technologies to solve this problem. For several years, Cisco has been working on positioning using Wi-Fi technology, given the prevalence of networks (in virtually every room) and devices (in almost every person).
The first developments began in 2007, when Cognio was acquired, whose spectral analysis engine was built into Cisco Aironet Wi-Fi access points. The ability to analyze the broadcast on the presence of interference and determine their impact on the performance of the Wi-Fi network has opened up new opportunities to ensure the reliability and performance of wireless networks. There were also new tasks - to understand where the sources of interference are, because to find them is not always easy, especially if the interference is intentionally created by the attackers.
Since then, the algorithm and solution portfolio have been improved many times and now, in 2015, Cisco offers the latest positioning solution with high accuracy, allowing to determine the Wi-Fi coordinate of the device with an accuracy of 1 m. In the coming days, Cisco will open this solution to the order in Russia.
Positioning in wireless networks can be implemented in several ways:
')
1. Pattern recognition method.
This method assumes that at each point the device sees a unique radio picture. The device scans the radio situation — the access points and the level of their signals, checks the received radio signal scheme with the list of templates and finds the device coordinate. To set up the entire network, it is necessary to carry out a long process of scanning the air of the entire premises, in practice several times, and also to conduct regular data calibration, since in reality, the room is constantly changing.
The method has the advantage of low equipment costs, but at the same time the cost of owning such a solution will be high and the positioning accuracy low.
2. On the access point to which the client is connected.
This method has the advantage of simplicity, but suffers exactly. Indeed, the coverage area of ​​a wireless network can be quite large, the diameter of the spot of illumination can be 50m or more. So this method is better than nothing, but it is unlikely to allow us to put together analytics that are valuable for understanding our clients.
The method allows you to determine the presence of the client rather than position it.
3. Triangulation.
This method has been used by Cisco for several years now and consists in determining the signal strength from the client at 3–4 Wi-Fi access points and in the intersection zone of the possible location of the client relative to each point to position the device. This method is quite informative. If the access points are correctly weighed, it makes it possible to determine the client’s position with an accuracy of 5–7 m with high probability. A good scenario is the access points around the perimeter of the room and in the center so that each point in space is “heard” by 3 to 4 Wi-Fi access points. Obstacles in the path of the radio signal will interfere with the accuracy of determining the coordinates. Static obstacles need to be modeled, and moving inevitably will have a negative impact on accuracy.
To improve positioning accuracy in a Wi-Fi network, access points should be installed more frequently, since Signal extinction and distance from the access point have an exponential dependence. Near the access point, when moving away from it, the signal level drop is significant, in the distance — when removed, the signal level decreases one unit distance less and it becomes more difficult to calculate the coordinate.
A Wi-Fi network with positioning capabilities in addition to access points and a wireless network controller receives another element that will perform coordinate calculations and accumulate data for analytics - the Mobility Services Engine (MSE). Today, MSE capacity allows you to accumulate data from 100,000 client devices for 2-8 years (depending on the amount of data generated, this usually corresponds to the frequency of movement of devices, and the type of MSE). However, in view of the emerging challenges associated with positioning in Wi-Fi networks, MSE in future versions will acquire a distributed architecture that allows you to collect data from a million devices. In this case, the collected information can be submitted to the external analytical engine via API in almost real time.
The implementation of positioning on a Wi-Fi network leads to additional costs for infrastructure deployment, with low operating costs for the network.
4. Angulation or positioning with the definition of the angle of the incoming signal.
The method is a revolutionary development of Cisco, which allows to achieve meter accuracy of Wi-Fi client positioning. The external precise positioning module connected to the Cisco Aironet modular access point with a special antenna allows you to further determine the angle at which the signal came and narrow the segment of the possible location of the Wi-Fi client to the beam. By applying the triangulation method to such information from 3–4 access points, we obtain a coordinate that gives a high probability of accuracy up to 1 m.
Physically, the device is a Cisco Aironet 3600 or 3700 access point with the precise positioning module turned on and a special antenna. An antenna is an array of 32 antennas, each of which receives a signal differently than the next. The algorithm allows us to calculate the angle at which the signal came from the collected data.
The use of radio beacons using Bluetooth Low Energy (BLE), an energy-efficient device capable of periodically signaling like a fire sensor, is becoming popular. Radio beacons are able to determine the approach of devices with BLE enabled to them and inform the server of the mobile application, which uses this information as a device coordinate, about it.
Radio bells are attractive for their low price, however, the cost of operating such a solution will be substantial, since it is necessary to track the real location and battery charge of beacons, while doing so by the physical presence of a specialist at the sites.
However, when integrating the two approaches, an interesting solution may arise. The Wi-Fi infrastructure provides general positioning, and the refinement of the coordinates of a particular object, for which it is crucial to have a device from it or to a museum exhibit, a cash register, and an entrance — it happens with a radio beacon. The integration of Wi-Fi and BLE networks at the same time allows you to reduce the cost of operating a hybrid network. On the interface to MSE, today you can track the actual location of radio beacons relative to the planned, “alien” radio beacons, and the exact positioning module has an integrated radio beacon in which the battery will not end.
In terms of developing a precise positioning solution, Cisco is also monitoring battery power and using the combined information from a Wi-Fi network and BLE devices for positioning and analytics.
For an innovative precision positioning module, Cisco received a number of awards - Best of INTEROP 2015 in Las Vegas (April 2015), in Tokyo (June 2015) and the Wireless Broadband Alliance in the Best WiFi Service Solution for Consumers or Enterprises nomination in San Jose (October 2015). ). The advantages of the method are high positioning accuracy, low operating costs, integration with BLE.
Positioning a device on a Wi-Fi network can be used for a variety of applications:
A) Detection of assets labeled with Wi-Fi tags.
B) connect to a Wi-Fi network based on the location of the client
C) navigation around the room
D) sending high-performance offers based on the location of the client
D) collection of analytics customer behavior
For applications C) -D), in addition to positioning accuracy, the frequency with which this coordinate is determined is important.
While the client device is not connected to the Wi-Fi network, the coordinate can be determined by probe (probe requests), which the device periodically sends. The device is not connected to a Wi-Fi network, but we are already receiving information about its movement. Samples are sent by Broadcast, i.e. on all frequency channels - all access points can hear them and triangulation by samples works great.
A minor problem is that the samples are sent by the device every 15-30-60 seconds, depending on the algorithm specified by the device manufacturer. In addition, mobile device manufacturers are striving to reduce the number of samples sent, because This increases the energy efficiency of the device. By collecting data in this way, we know that the client was at point A and after, for example, 30 seconds at point B, but we have no information about what he did between A and B. To navigate the room, as well as collect analytics about the behavior customers such intervals are also too big.
To increase the frequency of determining the coordinates of the Cisco developers have implemented a method of positioning the device for data traffic, which increased the frequency of data collection up to 10 times per minute - FastLocate.
FastLocate can be implemented on a separate module for the Cisco Aironet modular access points of the 3600 and 3700 series. The WSM module will scan the air and collect information to calculate the coordinates of the devices approximately every 8 seconds.
The second use case of FastLocate is suitable for any Cisco Aironet access points and does not require an additional module. The access point for a short period of time leaves the customer service mode and switches to the on-air scanning mode (Enhanced Local Mode - ELM). In ELM, the point, as in the previous case, collects information for calculating the coordinates of devices and switches back to customer service mode. This method has a price in the form of a Wi-Fi network performance decrease of approximately 15%, since access points cannot serve clients 100% of the time.
The client device must be connected to a Wi-Fi network, i.e. it is in the interest of the owner of the premises to offer Wi-Fi connectivity to customers, motivating customers with the information content of the application and special offers.
If, as a result, a large proportion of customers connect, the site owner will be able to extract a lot of information from such analytics:
Today, more than 2,000 organizations around the world use Cisco Wi-Fi positioning solutions and achieve a new level of customer satisfaction, increasing the efficiency of their business and marketing programs.
Additional materials
Cisco's corporate blog on high-precision positioning solutions.
Cisco's corporate blog on high-precision positioning solutions.
Cisco CMX 10.0 Feature overview and demo video
Cisco official web site for high-precision positioning solutions
Description of the high-precision positioning solution
An article on how to use positioning capabilities in Wi-Fi networks
Today, there are a number of approaches and technologies to solve this problem. For several years, Cisco has been working on positioning using Wi-Fi technology, given the prevalence of networks (in virtually every room) and devices (in almost every person).
The first developments began in 2007, when Cognio was acquired, whose spectral analysis engine was built into Cisco Aironet Wi-Fi access points. The ability to analyze the broadcast on the presence of interference and determine their impact on the performance of the Wi-Fi network has opened up new opportunities to ensure the reliability and performance of wireless networks. There were also new tasks - to understand where the sources of interference are, because to find them is not always easy, especially if the interference is intentionally created by the attackers.
Since then, the algorithm and solution portfolio have been improved many times and now, in 2015, Cisco offers the latest positioning solution with high accuracy, allowing to determine the Wi-Fi coordinate of the device with an accuracy of 1 m. In the coming days, Cisco will open this solution to the order in Russia.
Positioning methods
Positioning in wireless networks can be implemented in several ways:
')
1. Pattern recognition method.
This method assumes that at each point the device sees a unique radio picture. The device scans the radio situation — the access points and the level of their signals, checks the received radio signal scheme with the list of templates and finds the device coordinate. To set up the entire network, it is necessary to carry out a long process of scanning the air of the entire premises, in practice several times, and also to conduct regular data calibration, since in reality, the room is constantly changing.
The method has the advantage of low equipment costs, but at the same time the cost of owning such a solution will be high and the positioning accuracy low.
2. On the access point to which the client is connected.
This method has the advantage of simplicity, but suffers exactly. Indeed, the coverage area of ​​a wireless network can be quite large, the diameter of the spot of illumination can be 50m or more. So this method is better than nothing, but it is unlikely to allow us to put together analytics that are valuable for understanding our clients.
The method allows you to determine the presence of the client rather than position it.
3. Triangulation.
This method has been used by Cisco for several years now and consists in determining the signal strength from the client at 3–4 Wi-Fi access points and in the intersection zone of the possible location of the client relative to each point to position the device. This method is quite informative. If the access points are correctly weighed, it makes it possible to determine the client’s position with an accuracy of 5–7 m with high probability. A good scenario is the access points around the perimeter of the room and in the center so that each point in space is “heard” by 3 to 4 Wi-Fi access points. Obstacles in the path of the radio signal will interfere with the accuracy of determining the coordinates. Static obstacles need to be modeled, and moving inevitably will have a negative impact on accuracy.
To improve positioning accuracy in a Wi-Fi network, access points should be installed more frequently, since Signal extinction and distance from the access point have an exponential dependence. Near the access point, when moving away from it, the signal level drop is significant, in the distance — when removed, the signal level decreases one unit distance less and it becomes more difficult to calculate the coordinate.
A Wi-Fi network with positioning capabilities in addition to access points and a wireless network controller receives another element that will perform coordinate calculations and accumulate data for analytics - the Mobility Services Engine (MSE). Today, MSE capacity allows you to accumulate data from 100,000 client devices for 2-8 years (depending on the amount of data generated, this usually corresponds to the frequency of movement of devices, and the type of MSE). However, in view of the emerging challenges associated with positioning in Wi-Fi networks, MSE in future versions will acquire a distributed architecture that allows you to collect data from a million devices. In this case, the collected information can be submitted to the external analytical engine via API in almost real time.
The implementation of positioning on a Wi-Fi network leads to additional costs for infrastructure deployment, with low operating costs for the network.
4. Angulation or positioning with the definition of the angle of the incoming signal.
The method is a revolutionary development of Cisco, which allows to achieve meter accuracy of Wi-Fi client positioning. The external precise positioning module connected to the Cisco Aironet modular access point with a special antenna allows you to further determine the angle at which the signal came and narrow the segment of the possible location of the Wi-Fi client to the beam. By applying the triangulation method to such information from 3–4 access points, we obtain a coordinate that gives a high probability of accuracy up to 1 m.
Physically, the device is a Cisco Aironet 3600 or 3700 access point with the precise positioning module turned on and a special antenna. An antenna is an array of 32 antennas, each of which receives a signal differently than the next. The algorithm allows us to calculate the angle at which the signal came from the collected data.
The use of radio beacons using Bluetooth Low Energy (BLE), an energy-efficient device capable of periodically signaling like a fire sensor, is becoming popular. Radio beacons are able to determine the approach of devices with BLE enabled to them and inform the server of the mobile application, which uses this information as a device coordinate, about it.
Radio bells are attractive for their low price, however, the cost of operating such a solution will be substantial, since it is necessary to track the real location and battery charge of beacons, while doing so by the physical presence of a specialist at the sites.
However, when integrating the two approaches, an interesting solution may arise. The Wi-Fi infrastructure provides general positioning, and the refinement of the coordinates of a particular object, for which it is crucial to have a device from it or to a museum exhibit, a cash register, and an entrance — it happens with a radio beacon. The integration of Wi-Fi and BLE networks at the same time allows you to reduce the cost of operating a hybrid network. On the interface to MSE, today you can track the actual location of radio beacons relative to the planned, “alien” radio beacons, and the exact positioning module has an integrated radio beacon in which the battery will not end.
In terms of developing a precise positioning solution, Cisco is also monitoring battery power and using the combined information from a Wi-Fi network and BLE devices for positioning and analytics.For an innovative precision positioning module, Cisco received a number of awards - Best of INTEROP 2015 in Las Vegas (April 2015), in Tokyo (June 2015) and the Wireless Broadband Alliance in the Best WiFi Service Solution for Consumers or Enterprises nomination in San Jose (October 2015). ). The advantages of the method are high positioning accuracy, low operating costs, integration with BLE.
Coordinate update rate
Positioning a device on a Wi-Fi network can be used for a variety of applications:
A) Detection of assets labeled with Wi-Fi tags.
B) connect to a Wi-Fi network based on the location of the client
C) navigation around the room
D) sending high-performance offers based on the location of the client
D) collection of analytics customer behavior
For applications C) -D), in addition to positioning accuracy, the frequency with which this coordinate is determined is important.
While the client device is not connected to the Wi-Fi network, the coordinate can be determined by probe (probe requests), which the device periodically sends. The device is not connected to a Wi-Fi network, but we are already receiving information about its movement. Samples are sent by Broadcast, i.e. on all frequency channels - all access points can hear them and triangulation by samples works great.
A minor problem is that the samples are sent by the device every 15-30-60 seconds, depending on the algorithm specified by the device manufacturer. In addition, mobile device manufacturers are striving to reduce the number of samples sent, because This increases the energy efficiency of the device. By collecting data in this way, we know that the client was at point A and after, for example, 30 seconds at point B, but we have no information about what he did between A and B. To navigate the room, as well as collect analytics about the behavior customers such intervals are also too big.
To increase the frequency of determining the coordinates of the Cisco developers have implemented a method of positioning the device for data traffic, which increased the frequency of data collection up to 10 times per minute - FastLocate.
FastLocate can be implemented on a separate module for the Cisco Aironet modular access points of the 3600 and 3700 series. The WSM module will scan the air and collect information to calculate the coordinates of the devices approximately every 8 seconds.
The second use case of FastLocate is suitable for any Cisco Aironet access points and does not require an additional module. The access point for a short period of time leaves the customer service mode and switches to the on-air scanning mode (Enhanced Local Mode - ELM). In ELM, the point, as in the previous case, collects information for calculating the coordinates of devices and switches back to customer service mode. This method has a price in the form of a Wi-Fi network performance decrease of approximately 15%, since access points cannot serve clients 100% of the time.
The client device must be connected to a Wi-Fi network, i.e. it is in the interest of the owner of the premises to offer Wi-Fi connectivity to customers, motivating customers with the information content of the application and special offers.
If, as a result, a large proportion of customers connect, the site owner will be able to extract a lot of information from such analytics:
Today, more than 2,000 organizations around the world use Cisco Wi-Fi positioning solutions and achieve a new level of customer satisfaction, increasing the efficiency of their business and marketing programs.
Additional materials
Cisco's corporate blog on high-precision positioning solutions.
Cisco's corporate blog on high-precision positioning solutions.
Cisco CMX 10.0 Feature overview and demo video
Cisco official web site for high-precision positioning solutions
Description of the high-precision positioning solution
An article on how to use positioning capabilities in Wi-Fi networks
Source: https://habr.com/ru/post/270779/
All Articles