Method for calculating the position of WiFi-enabled devices

The present invention relates to location-based services, more specifically, methods of continuously optimizing or improving the quality of WiFi location data in the system.

In recent years, the number of portable arithmetic units has been increasing dramatically, creating a need for more advanced portable wireless devices. Mobile email, mobile phone services, multiplayer games, call follow, etc. are examples of how new applications are being created on mobile devices. In addition, users are beginning to request / request applications that not only use their current location but also share their location information with others. Parents want to keep track of their children, supervisors need to track the location of company delivery vehicles, and commercial travelers try to find a recent pharmacy to receive their prescriptions. In any of these examples, the individual needs to know his or her current location or the current location of others. To date, we all rely on asking directions, calling to ask about our location, and having our employees report their location accordingly.

<p> Location-based services are a new field of mobile applications that take advantage of the capabilities of new devices that calculate current geographic locations and report them to users or services. These services include, for example, local weather, traffic updates, vehicle directions, child tracking devices, friend hunting, and urban concierge services. These new position detectors all rely on a variety of technologies that use the same overall concept. Using radio signals coming from known reference points, these devices mathematically calculate the user's position with respect to these reference points. Each of these approaches has advantages and disadvantages based on the wireless technology and the positioning algorithm it employs.</p><p> The Global Positioning System (GPS), operated by the US government, uses dozens of orbiting satellites as reference points. These satellites broadcast radio signals received by GPS receivers. The receiver measures the time it takes for the signal to reach the receiver. After receiving signals from three or more GPS satellites, the receiver triangulates its position on Earth. For the system to work effectively, the radio signal must arrive with little or no interference. Since the receiver requires a clear view of up to three or more satellites, interference can occur from the weather, buildings or structures, and plants. Interference can also be caused by a phenomenon known as multipath. Due to the reflection of radio signals from satellites onto physical structures, multiple signals from the same satellite reach the receiver at different times. Multipath signals can confuse the receiver and lead to serious errors, as receiver calculations are based on the time it takes for the signal to reach the receiver.</p><p> Cell tower triangulation is another method used by mobile operators to determine the location of users and devices. The wireless network and the mobile device communicate with each other and share signal information that the network can use to determine the location of the mobile device. This approach was initially considered a better model than GPS, as these signals do not require a direct line of sight and drop buildings better. Unfortunately, these approaches have proven to be the next best approach due to the non-uniformity of the cell tower hardware, along with the lack of uniformity in the transmission of multipath signals and the positioning of the cell tower.</p><p> Assisted GPS is a newer model that combines GPS with mobile tower technology to produce more accurate and reliable location calculations for mobile device users. In this model, the wireless network attempts to help GPS improve signal reception by transmitting information about the clock offset of GPS satellites and the user's approximate position based on the location of the connected cell tower. These technologies address the weak signals experienced indoors by GPS receivers and help them gain a recent satellite "selection" that provides a faster "first read." These systems suffer from problems such as slow reaction times and inadequate accuracy over 100 meters in downtown areas.</p><p> Other models have been developed in recent years to address and address known issues related to GPS, A-GPS, and cell tower positioning. One of them, known as TV-GPS, utilizes signals from television towers (eg Muthukrishnan, Maria Lijding, Paul Havinga, "Towards a smart environment: enabling technology and technology for localization". , Lecture Notes in Computer Science, Vol. 3479, Jan 2Hazas, M., Scott, J., Krumm, J .: "The Arrival of the Position Recognition Era", IEEE Computer, 37 (2): 95-97, Feb See 2004005, Pa005, pages 350-362). The concept is based on the fact that most metropolitan areas have three or more TV towers. A dedicated hardware chip receives TV signals from these various towers and uses the known locations of these towers as reference points. The challenges faced by this model are the cost of new hardware receivers and the limitations of using such a small set of reference points. For example, if the user is outside the perimeter of the tower, the system is difficult to provide reasonable accuracy. A typical example is a user on the coastline. Since there are no TV towers in the ocean, there is no way to harmonize between reference points, resulting in a much more inland location than the user.</p><p> Microsoft Corporation and Intel Corporation (through a research group known as PlaceLab) is deploying a WiFi location system that uses access point locations obtained from amateur scanners (commonly known as "war drivers") who submit WiFi scan data to public community websites (through a research group known as PlaceLab). For example, LaMarca, A., et.al., PlaceLab: See "Devices that use radio beacons in the wilderness"). For example, WiGLE, Wi-FiMaps.com, Netstumbler.com, and NodeDB. Both Microsoftm and Intel are developing their own client software that uses this public wardrive data as a reference position. As individuals voluntarily provide data, the system has a number of performance and reliability issues. First, the data in the entire database is not contemporary. Some are new and some are 3-4 years old. The age of the access point location is important because the access point may move offline over time. Second, data is acquired using a variety of hardware and software structures. All 802.11 radios and antennas have different signal reception characteristics that affect the display of signal strength. Each scanning software implementation scans WiFi signals in different ways at different time intervals. Third, the data provided by the user suffers from deviations from the trunk line. Since the data is self-reported by individuals who do not follow the designed scan route, the data tends to collect in high-traffic areas. The trunk line deviation brings the resulting position closer to the main trunk line regardless of where the user is currently located, resulting in a large accuracy error. Fourth, these databases are 802. 11 Includes the calculated location of the scanned access point, not the raw scan data obtained by the hardware. Each of these databases uses a basic weighted average formula to calculate the access point location separately. As a result, many access points are shown as far away from their actual location, and some are inaccurately shown as if they were underwater.</p><p> Numerous commercially available location systems for indoor positioning (eg, Kavitha Muthukrishnan, Maria Lijding, Paul Havinga, "Towards a Smart Environment: Enable Technology and Technology for Localization", Lecture Notes in Computer Science, Vol. 3479, Jan 2Hazas, M., Scott, J., Krumm, J .: "The Arrival of the Position Recognition Era", IEEE Computer, 37 (2): 95-97, Feb 2004 See pages 005, Pa005, 350-362). These systems are designed to track assets and people within a controlled environment such as a corporate premises, hospital facility, or delivery yard. A typical example is to have a system that can monitor the exact location of an emergency cart in a hospital so that hospital staff do not waste time locating equipment in the event of cardiac arrest. .. The accuracy requirements for such use are very strict and require an accuracy of 1-3 meters. These systems use a variety of techniques to fine-tune the system to measure the propagation of radio signals, including conducting a detailed location survey on a square-by-site basis. These systems also require a certain network connection so that the access point and client radios can exchange synchronization information. While these systems are becoming more reliable for these indoor use cases, they are not effective for widespread deployment. It is not possible to perform detailed location surveys that span the entire city, and to the extent required by these systems, one cannot rely on constant communications channels with 802.11 access points that span the entire metropolitan area. Most importantly, these indoor positioning algorithms are of little use in a wide range of scenarios, as outdoor radio propagation is fundamentally different from indoor radio propagation.</p><p> There are many 802.11 location scan clients available that record the presence of 802.11 signals along with GPS location readings. These software applications are run manually and create reading log files. These applications are, for example, Netstumber, Kismet, and Wi-Fi FoFum. Some geeks use these applications to mark the location of 802.11 access point signals they detect and share with each other. All management of this data and sharing of information is done manually. These applications do not make calculations about the physical location of the access point, so they simply mark the location where the access point was detected.</p><p> The performance and reliability of the underlying positioning system are key factors for the successful deployment of any location-based service. Performance refers to the level of accuracy that a system achieves in a given use case. Confidence refers to the percentage of time that a required accuracy level is achieved.<img file="JP4980247B2_D0001.tif" /></p>

<p> The present invention provides methods and systems for continuously optimizing data in WiFi positioning systems. For example, data is monitored to infer whether a WiFi access point has been moved or new. In this way, the data is continuously optimized. Similarly, such a system can be used to avoid suspicious data when locating WiFi-enabled devices.</p><p> In one aspect of the invention, a location-based service system uses a WiFi-enabled device to monitor WiFi access points within the target area and indicate whether the WiFi access points have moved relative to a previously recorded location. The WiFi-enabled device communicates with the WiFi access point within the range of the WiFi-enabled device so that the observed WiFi access point identifies itself. A reference database is accessed to obtain location information recorded for each observed WiFi access point in the target area. The recorded location information is used for each observed WiFi access point in combination with a predetermined rule that estimates whether the observed WiFi access point has moved relative to the recorded location. The identified information of the observed WiFi access point, which is presumed to have moved, is transmitted to the reference database.</p><p> In another aspect of the invention, the location-based service system uses a WiFi-enabled device to monitor the WiFi access point in the target area, which indicates whether the WiFi access point has been newly observed. The WiFi-enabled device communicates with a WiFi access point within the range of the WiFi-enabled device so that the observed WiFi access point identifies itself. Access the reference database to obtain information that specifies the location recorded for each observed WiFi access point in the target area. Observed WiFi access points are identified where the reference database does not have the corresponding recorded location information. The recorded location information for each of the observed WiFi access points is used to calculate the location of the WiFi enabled device. The reference database is notified of the WiFi access point (no information in the database) and is provided with the calculated location along with it to serve as location information for the newly observed WiFi access point.</p><p> In another aspect of the invention, the location-based service system for WiFi-enabled devices calculates the location of WiFi-enabled devices. The WiFi-enabled device communicates with a WiFi access point within the range of the WiFi-enabled device so that the observed WiFi access point identifies itself. Access the reference database to obtain information that specifies the location recorded for each observed WiFi access point. The location information recorded for each observed WiFi access point is combined with certain rules to determine whether the observed WiFi access point should be included in or excluded from the set of WiFi access points. used. The recorded location information of only the WiFi access points included in the set is used, and the recorded location information of the excluded WiFi access points is excluded when calculating the geographic information of the WiFi enabled device.</p>

A preferred embodiment of the present invention provides a system and method for continuously maintaining and updating location data within a WiFi positioning system (WPS) using public and private 802.11 access points. Preferably, the client utilizing the location data collected by the system utilizes a technique of avoiding error-prone data when determining the WiFi location to improve the quality of the previously collected and determined location information. Use the newly discovered location information for. One embodiment communicates with a central location access point reference database to provide the location of newly discovered access points. Another embodiment informs the central location access point reference database of access points that are outside the bounds of the reading at which the reading should be predicted, based on the reading at the previous location. Access points whose readings are outside the bounds of the readings to be predicted are marked as suspicious and removed from the triangulation formula to avoid introducing erroneous data into the location calculation.

A preferred embodiment of the present invention is U.S. Pat. No. 11,261,988, filed October 28, 2005, "Position-based service that selects location algorithms based on access points detected within the user's equipment." Based on, but not limited to, the techniques, systems, and methods disclosed in previously filed applications. The full text of the content of the above patent is incorporated herein by reference. These applications are specific for collecting high quality location data for WiFi access points so that the data can be used in location-based services to determine the geographic location of WiFi-enabled devices that utilize these systems. I teach the method. In this case, for example, a new technique for continuously monitoring and improving the above data by a user who detects a new access point in the target area or estimates that the access point has moved is disclosed. However, the technology is not limited to the systems and methods disclosed in the incorporated patent application. Rather, those applications only disclose one framework or situation in which the technology is feasible. Therefore, while the above systems and application examples may be useful, they are not considered necessary to understand the present embodiment or the present invention.

In one embodiment of the invention, the WPS client device scans the access point to determine the physical location of the WPS client device and then compares the observed readings with the readings recorded in the database. By doing so, the quality of the current access point location in the access point reference database is calculated. The access point is marked as suspicious if the client determines that the observed reading is outside the bounds of the reading that should be expected based on the recorded reading. Suspicious readings are logged back into the feedback system to report back to the central location access point reference database.

In another embodiment of the invention, the WPS client device removes the identified suspicious access point from the triangulation point calculation of the WPS client device in real time so as not to introduce erroneous data into the location calculation.

In another embodiment of the invention, the WPS client device scans the access point to determine the physical location of the device and identifies the access point that does not exist in the current access point reference database. After the known access points have been used to calculate the current location of the device, those newly discovered access points will help determine the location along with the observed output readings. The location is used and reported back to the central location access point reference database.

In another embodiment of the invention, the device-centric WPS client device periodically connects to the central location access point reference database to download the latest access point data. The WPS client device also uploads all feedback data for newly observed and suspicious access points. This data is then sent to the central location access point reference database process to recalibrate the entire system.

In another embodiment of the invention, the network-centric WPS client device records feedback data about newly observed and suspected access points directly in the central location access point reference database in real time.

By requiring WPS client devices to continually update the access point reference database with information about new and suspicious access points, WiFi positioning systems provide higher quality data than systems scanned only by providers. .. Over time, WiFi access points will continue to be added and moved. Embodiments of the invention described above ensure that the access point reference database provides optimal positioning data that self-heales and self-expands to continuously reflect additions and changes to available access points. I will provide a. The more user-client devices are deployed, the more frequently the information in the database is updated, which improves the quality of the access point reference database.

FIG. 1 shows some of the preferred mobile phones for WiFi Positioning Systems (WPS). The positioning system includes positioning software (103) provided in the user arithmetic unit (101). There are fixed radio access points (102) that broadcast information using control / common channel broadcast signals over a particular range area. The client device monitors the broadcast signal or requests transmission through a probe request. Each access point contains a unique hardware identifier known as the MAC address. The client positioning software receives a signal beacon or probe response from an 802.11 access point within range and uses features from the received signal beacon or probe response to calculate the geographic location of the arithmetic unit.

The positioning software will be described in detail with reference to FIG. 2, which shows typical components of the positioning software 103. Typically, the user equipment embodiment shown in FIG. 1 includes an application or service (201) that utilizes position readings to provide numbers to end users (eg, travel directions). This location application makes a request to the positioning software regarding the location of the device at that particular moment. The location application is started continuously at regular intervals (eg, every second) or once at the request of another application or user.

In Figure 2, the location application queries all access points in range at a given point in time to determine which access point is suspicious because the observed data does not match the calculated location in the reference database. To request the positioning software. The information about the suspicious access point collected by the location application is used in real time or later to optimize the location information in the access point reference database.

In the embodiment shown in FIG. 2, a location application or service request initiates a scanner (202) that makes a "scan request" to the 802.11 radio (203) on the device. The 802.11 radio sends probe requests to all 802.11 access points (204) within range. According to the 802.11 protocol, upon receiving a probe request, those access points send a broadcast beacon containing information about the access point. The beacon contains information about the device's MAC address, network name, how to connect to the device, as well as the exact version and security configuration of the protocols supported by the beacon. The 802.11 radio calculates the received signal strength (RSS) of each observed access point and sends the identifier and RSS information back to the scanner.

The scanner passes an array of this access point to a locator (206) that checks the MAC address of each observed access point against the access point reference database (205). This database is located remotely on the device or over a network connection. The access point reference database is unprocessed 802. 11 Includes scan data and the calculated location for each access point known to the system. Figure 5 shows the access point evaluation process in more detail. A list of observed access points (501) is taken from the scanner and the locator (206) searches each access point in the access point reference database. The recorded location is searched for each access point found in the access point criteria database (502). The locator passes a set of known access point (502) location information, along with the signal characteristics returned from each access point to the quality filter (207). This filter determines if any of the access points have been moved since they were added to the access point reference database and operates continuously to improve the overall system. Quality filters mark access points that do not follow their quality algorithm as "suspicious" (504). After removing the erroneous data recording, the filter sends the remaining access points to the location calculator (208). Using a set of verified reference data from the access point reference database and signal strength readings from the scanner, the position calculator calculates the position of the device at that time. The location calculation unit calculates the location of the newly observed access point (503) that cannot be found in the access point reference database. The raw scan data and the location of the new access point are stored in the feedback file (212), as shown in FIG. This feedback is stored locally on the device for subsequent transmission to the server, or is transmitted to the server in real time. Positional data about known access points is processed by a smoothing engine (209) that averages a series of past position readings to eliminate erroneous readings from previous calculations before being sent back to the locator. To.

The calculated location readings generated by the locator are propagated to these location-based applications (201) via the application interface (210), which includes the application program interface (API), or virtual GPS performance (211). .. GPS receivers transmit position readings using their own messages or using position criteria such as those developed by the National Marine Electronics Association (NMEA). Search for messages by connecting to a device using a standard interface such as a COM port on the receiver. Specific embodiments of the present invention include virtual GPS performance that allows GPS compatible applications to communicate with this new positioning system without the need to change communication models or messages.

Positioning values are generated using a series of positioning algorithms aimed at transforming noisy data flows into reliable and stable position readings. The client software compares the list of observed access points with the calculated signal strength weighting the user's position to determine the exact position of the device user. Various methods are adopted, such as a simple signal strength weighted average model, a closest model combined with triangulation, and adaptive smoothing based on the speed of the instrument. Different algorithms work better under different scenarios and tend to be used together in complex deployments that produce the most accurate final readings. A preferred embodiment of the present invention can use a number of positioning algorithms. The decision to use which algorithm depends on the number of access points observed and the application of the user using it. The filtering model differs from traditional positioning systems because traditional positioning systems rely on known reference points that never move. In the model of the preferred embodiment, no guess is made about the fixed position of the access point. The access point is not owned by the positioning system and can be moved offline. Filtering methods presume that some access points are no longer co-located, leading to incorrect location calculations. Therefore, the filtering algorithm attempts to isolate the access point that has moved after the location was recorded. The filter is dynamic and changes based on the number of access points observed at that time. Smoothing algorithms include not only simple position averaging, but also advanced Bayesian theory involving particle filters. The speed algorithm calculates the speed of the device by estimating the Doppler effect from observing the signal strength of each access point. Optimizing the quality of the latest access point data

The quality filter (207) section compares data from known and observed access points in local or remote access point reference databases. For observed access points whose MAC addresses are located in the access point reference database, the quality filter unit compares the observed information with the location of the access points stored in the database.

An advanced function of the quality filter (207) is to improve the accuracy of position estimation as a result of eliminating suspicious access points from position calculation. The quality filter uses only access points that are located in the access point criteria database. In some cases, the quality filter does not have a history of current client device locations used to determine quality. The process of identifying suspicious access points for historical location estimation is stored in the database<u style="single">Cluster with the largest number of access points</u>Based on the position of. The location of all observed access points recorded in the access point reference database is considered and the average location of the access points in the largest cluster is used as the reference point. A cluster refers to distance-based clustering and is a group of access points having a distance of each access point from at least two or more access points in a subthreshold cluster. The clustering algorithm is shown as follows and is understood as "node n belongs to cluster K if there is at least one element in cluster K such as nu whose distance from n is less than the threshold".<img file="JP4980247B2_D0002.tif" />If the cluster cannot be found, the access point is mathematical<u style="single">Median</u>Serves as an optimal estimate of the distance average of the majority of access points.

If the distance of an individual access point to a reference point is calculated to exceed a given distance, it is defined as a suspicious access point and recorded in the feedback file for return to the access point reference database. The suspicious access points are then removed from the list of access points used to locate the user device.

To identify access points that are suspicious to the client device when there is a user's travel history, it is based on the client device's previous location. Figure 6 shows a typical execution of this determination. In an embodiment with a position history, the position of the client device is continuously calculated every time interval, usually every second. If the distance between an individual observed access point (602) and its past reference point (previous location calculation) exceeds a given distance (603), it is defined as a suspicious access point and added to the feedback file. And excluded from the calculation. The purpose of this filter is to try and use the access point closest to the user / device (601) in order to provide the highest possible accuracy. This filter is called an adaptive filter because the threshold distance for removing suspicious access points varies dynamically. The threshold distance used to identify suspicious access points is based on the number of access points considered to be of good quality for calculating the location of client equipment. Therefore, the adaptive filter includes two factors: 1) the minimum number of access points required to locate the user equipment, and 2) the minimum threshold distance to identify suspicious access points. The adaptive filter starts at the lowest threshold distance. If the number of access points within that distance exceeds the minimum number of access points required to calculate the client location, the location of the device is calculated. For example, if we find 5 access points within 20 meters of the previous reading, we exclude all observed access points over 20 meters. If the filter criteria are not met, the minimum number of access points is considered, or the adaptive filter threshold (603) of the distance is increased until the maximum acceptable distance is reached, after which the access points within the threshold distance are on the user's equipment. Used to locate. If the access point is not found within the maximum threshold distance (604) from the previous position, the position is not calculated.

Positioning software will continue to try to locate the device based on its previous position until a given maximum duration. If the position cannot be determined during this period, the maximum threshold distance is adjusted using the calculated speed of the device. It is known that the vehicle accelerates up to 6 m / s / s, and if it was previously calculated to travel at 20 mph, it could occur more than 42 meters from the last position after 2 seconds. If the earlier time adaptive filter did not work, this upper limit of the 42 meter distance is used to adjust the limit of the distance threshold. If it is very difficult to calculate the actual speed of the client device, the maximum speed threshold is used. If the access point is calculated to be more than the maximum threshold distance away from the reference point, the access point is marked "suspicious" and recorded in the feedback file. If the access point is not located within the maximum threshold distance within the period, the adaptive filter ignores the history, treats the next position determination example as a case without history, and sends it back to the clustering filter described above.

&lt;Real-time filtering of suspicious access points&gt; Suspicious access points are excluded from the input to the triangulation calculation, and valid access point locations are used to measure the device position trigonometrically (502). The input to the triangulation algorithm is the set of valid access points returned by the quality filter (207). The triangulator reads a list of observed valid access point locations along with their respective signal strengths and calculates latitude and longitude along with the horizontal position error (estimated accuracy error at that time). Since the triangulation process applies the smoothing process, it also takes into account the position before adding additional filters to the scan. By eliminating suspicious access points, we provide the triangulation algorithm with a more reliable set of reference points to calculate in the light. Since access points can be moved at any time, positioning software must take into account the dynamic nature of the reference point. Without filtering, the calculated locations can result in hundreds or thousands of miles apart.

Suspicious access points are not completely destroyed. By adding the newly observed location to the database via a feedback file (212) with different characteristics that indicate suspicious, the server either moves the official location of the access point or the new location. You can decide if you just want to keep it until it is confirmed. By retaining, this access point does not interfere with the location calculation of other users.

&lt;Add new access point data&gt; Observed access points found in the access point reference database of known access points are used to calculate the location of client devices after the elimination of suspicious access points. Observed access points whose MAC address is not found in the access point reference database represent new access points (302) (503) added after the database was created or updated. Those observed access points that are not found in the known access point criteria database are added to the feedback file as new access points. These newly discovered access points are marked with the position of the client device calculated by the positioning system, along with the observed signal strength. This situation occurs in many scenarios. In many cases, new access points are purchased and deployed in the vicinity since the last physical scan by the scanning vehicle. Due to the rapid expansion of W-Fi, this is a very common case. In another case, the access point is located deep inside the center of the building and the scanning vehicle cannot detect it from the street. In another example, the access point may be located on the upper floors of a skyscraper. These access points are difficult to detect from the street below where the scanning vehicle operates, but may be received by a client device passing near the building by a walking user or entering the building itself.

By "self-expanding" the system in this way, the coverage area of the system gradually expands deep into the building and into the upper floors of the skyscraper. The system also utilizes a large number of new access points deployed daily around the world.

&lt;Update central database server&gt; Referring to FIG. 3, in some embodiments, the access point reference database of known access points is located on a central network server away from the client equipment. The provision of this connection is possible via the available network connection and is managed by the data exchange unit (303). Once authenticated, the client device (103) identifies all suspicious and new access point data from the local storage feedback file (212) and uploads that data to the access point reference database (205).

In another embodiment, the client device is always connected to the access point reference database using a network connection. FIG. 4 shows how a network-centric embodiment works. Rather than storing reference data locally, the locator (201) uses a set of real-time network interfaces (401) to communicate with the access point reference database. The locator sends a list of observed access points to the network, and the network interface returns the list of observed access points and whether the database recorded the location or a new access point was discovered. The process continues the quality filter to mark suspicious access points as described above, but the list of suspicious access points is sent to the access point reference database in real time. After the calculator determines the location of the user device, the list of newly discovered access points is marked at the current location and sent back to the database in real time. This keeps the database up-to-date and eliminates the need for a data exchange.

After receiving the feedback data, in either the device-centric or network-centric model, the access point reference database decides whether to put the suspicious access point in a "pending state" so as not to interfere with another user's device location request. judge. Many techniques have been developed to optimize how this feedback data from suspicious access points can be used to improve the quality of the overall database. If more than one user places an access point in a new location, there may be a voting mechanism that decides which access point to move to the new location. If only one user marks the access point as suspicious, the access point is marked as a poor quality reading at the new location. Once the new location is authenticated by another user, the quality characteristics of the access point are enhanced, reflecting the higher level of reliability system being in the new location. The more people who support the new location of the access point, the higher the quality level. The system's client software then chooses an access point with a high quality rating over an access point with a low quality rating.

In a device-centric or network-centric model, the access point reference database collects information-specific access point, client device location, and access point signal strength information for newly discovered access points from client devices. Once an acceptable number of newly discovered access point readings have been collected by the access point reference database, the location for the new access point can be calculated based on the systems and methods described in the relevant application. .. The newly discovered access point can then be provided to the client device for use in location calculation.

The scope of the present invention is not limited to the above embodiments, but is defined by the appended claims, and it is understood that these claims include modifications and improvements of the contents described.

<figref num="1">A specific embodiment of a WiFi positioning system is shown.</figref><figref num="2">A typical architecture of positioning software according to a specific embodiment of the present invention is shown.</figref><figref num="3">The data transfer process in a specific client equipment-centric embodiment is shown.</figref><figref num="4">The data transfer process in a particular network-centric embodiment is shown.</figref><figref num="5">Shows the data flow of the quality filtering and feedback process.</figref><figref num="6">The behavior of the adaptive filter in a particular embodiment is shown.</figref>