MOBILE-BASED POSITIONING USING RECEIVED SIGNAL POWER AND TIMING MEASUREMENTS - Patent application

Abstract

A hybrid method of estimating a mobile device's location utilizing both received signal power and timing measurements. Received signal power of signals received by the mobile device from multiple cells is measured and corresponding received signal power measurements are stored. The method further includes measuring, at the mobile device, arrival times of the signals received from the multiple cells. Multiple time difference of arrival (TDOA) measurements are determined from the arrival times. A power-time hybrid Gaussian maximum likelihood estimator and positioning assistance data for the multiple cells are used to generate a maximum likelihood estimate of the mobile device's location by evaluating the joint conditional probability of the received signal power measurements and the multiple TDOA measurements. Gaussian random variables may be used to represent the received signal power measurements and the TDOA measurements.

Description

[Technical field]

[0001] SUMMARY The present disclosure relates to systems and methods for determining the position of a user terminal or other communication device based on time-of-arrival measurements in a wireless environment. [Background technology]

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to and the benefit of U.S. Provisional Patent Application No. 63 / 173,239, filed April 9, 2021, entitled MOBILE-BASED POSITIONING USING MEASUREMENTS OF RECEIVED SIGNAL POWER AND TIMING, the disclosure of which is incorporated by reference in its entirety into this application for all purposes.

[0003] Measurements of the time of arrival (TOA) for signals from a set of wireless base stations can help determine a user's position or location. For example, existing standards allow wireless devices to determine their own location or positioning based on signals from base stations in a manner that can replace or complement GPS or WiFi-aided positioning strategies (see, for example, U.S. Pat. Nos. 5,993,137, 5,971,142, 5,971,151, and 5,971,161). See, for example, 3GPP TS 36.133, "Requirements for support of radio resource management", V15.5.0.

[0004] The prior art in downlink cellular positioning of wireless devices (also referred to as "user equipment" or "UE") is UE-assisted (UE-A). In the UE-A approach, a location server provides assistance data to nearby cells (also known as base stations, eNBs for 4G LTE, gNBs for 5G NR), which communicate this to the UE. The UE makes measurements (i.e., power, timing measurements) on the current downlink radio conditions and transmits these measurements in the uplink back to the location server. The location server then uses the measurements to estimate the location of the UE.

[0005] One example positioning method is OTDOA (Observed Time Difference of Arrival), where the UE makes TDOA (also known as RSTD in 3GPP) measurements and the location server makes hyperbolic TDOA position calculations. Another positioning method is Enhanced Cell ID (E-CID), where the UE makes power measurements of surrounding cells and round-trip timing estimation of the UE serving cell. Release 9 3GPP E-CID further incorporates Angle of Arrival (AoA) measurements at the eNB. The most basic cellular positioning method is Basic Cell ID, where the location is estimated to be at the centroid of the serving cell or the transmission point of the serving cell.

[0006] Together with the UE measurements, the location server has the additional information needed to estimate the UE location, including cell locations, cell transmit powers, relative cell timing offsets, antenna orientations and aperture details.

[0007] For example, the LTE Positioning Protocol (LPP) described in 3GPP TS 36.355 version 13.3.0 referenced in Release 13 of the 3GPP LTE specification provides UE-A positioning. For example, according to the LTE positioning protocol, positioning reference signal (PRS) subframes are embedded in designated orthogonal frequency division multiplexing (OFDM) symbols over specific time intervals, sometimes called positioning occasions. The UE can measure the time of arrival (TOA) of the PRS subframes from each accessible base station (which the 3GPP LTE specification calls eNodeB). The user equipment preferably measures at least one reference signal time difference (RSTD) between two different eNodeBs (one called reference and the other called neighbor). The reference signal time difference is related to the established strategy for observed time difference of arrival (OTDOA) described in the LTE positioning protocol.

[0008] Determining the location of the user equipment proceeds by measuring the TOA of a first reference signal path from each eNodeB of interest, followed by determining a reference signal time difference (RSTD) between a pair of designated eNodeB base stations using the respective measured arrival times at the user equipment. Depending on the particular configuration specified in the standard, the TOA and RSTD measurements may be made to a particular number of base stations and over different corresponding combinations of RSTD measurements between base stations among the set of base stations.

[0009] The fundamental modulation scheme in LTE for transmitting bits over the air uses OFDM. That is, bits are generated by applying quadrature amplitude modulation (QAM) to each active subcarrier that constitutes an OFDM symbol. In practice, an LTE OFDM symbol can have 2048 time samples representing 600 active subcarriers out of a maximum of 1200 subcarriers. Each subcarrier may be assigned a function at the receiver, such as transmitting a bit that is known in advance to the receiver, thereby enabling different calculations. These calculations can include channel impulse response (CIR) estimation, and positioning-related measurements.

[0010] FIG. 1 provides a functional block diagram of an apparatus for determining position using observed time difference of arrival (OTDOA) based on reference signal time difference (RSTD) measurements defined in LTE. The illustrated user equipment receiver 110 receives multiple OFDM symbols from two base stations 101, 103. The receiver 110 receives the symbols using one or more antennas. FIG. 1 illustrates the position determination function using signals received from two base stations 101, 103 as an example with the receiver 110 using a single antenna, which is the minimum configuration for RSTD measurements. This configuration can be extended to a larger number of base stations and a larger number of user equipment antennas.

[0011] Since the user equipment receiver 110 complies with the LTE standard, the receiver can process the received OFDM symbols to provide a best estimate of the transmitted bits. Such a receiver 110 can identify the first path using one or more first path identification (FP-ID) modules 130, 140 corresponding to subcarriers assigned to calculate positioning information. Each first path identification module 130, 140 corresponds to information 132, 142 provided by the user equipment receiver 110 about the subcarriers used for positioning measurements. For example, the information may be stored in a table in a non-volatile memory.

[0012] The first path identification units 130, 140 identify the respective first paths for OFDM symbols received from known eNodeBs. Reference signal time difference (RSTD) measurements are typically based on a predefined duration of OFDM symbols to achieve a desired accuracy. In LTE, this may span at least one subframe of OFDM symbols, defined to be 14 OFDM symbols.

[0013] The output from each first path identification module 130, 140 is the time of arrival (TOA) at the user equipment of the signal from the corresponding base station. In general, in LTE, the RSTD between base stations indexed as k and j is k,j teeth, RSTD k,j =TOA k -TOA j EQ.1 It is decided that:

[0014] 1 shows that module 150 of receiver 110 provides the Equation 1 reference signal time difference calculation as its output 152. This output RSTD 0,1 152 is the result of the calculation of the output 134 of the first path identification module 130 minus the output 144 of the first path identification module 140 .

[0015] TOA k and TOA j Given a reliable estimate of , and knowing the structure of the signal received at the first path identification FP-ID module 130, 140, RSTD k,j The calculation of is simple. Standards such as the LTE standard prescribe a symbol structure that can be generalized as shown in Figure 2. Modern wireless systems that transmit from one source, such as a base station, to multiple users within a coverage area require a further division of the transmission into "subchannels". This is not very different in concept from FM or AM radio transmissions, but wireless transmissions have very high bits / second / hertz targets for a given spectrum. In the case of modern wireless technologies, channel identification can be achieved using orthogonal schemes, including OFDM and Code Division Multiple Access (CDMA). In the near future, wireless standards can increase capacity by using quasi-orthogonal channels, which can be achieved in a myriad of spatial and temporal strategies.

[0016] Figure 2 simplifies the description of the signals involved in observed time difference of arrival (OTDOA) measurements by showing how orthogonal channelization is assumed, i.e., orthogonality is preserved and crosstalk between channels is kept to a low level.

[0017] The horizontal axis 201 of FIG. 2 represents time, qualitatively representing the time occupied by the received symbols. The vertical axis of FIG. 2 represents a second channel dimension so that FIG. 2 qualitatively illustrates the channels without overlap. The channel separation on the vertical axis can represent a division of frequency as in the case of OFDM, or an indexing of different codes in CDMA. For example, in the LTE standard, the division on the frequency axis can represent 15 kHz of bandwidth for subcarriers, and an OFDM symbol may possibly consist of up to 600 active subcarriers out of a total of 1024 subcarriers in one symbol. This is merely an example, other allocations are known. Thus, for example, the extent of each square in FIG. 2 can represent 15 kHz (y-axis) by 71.4 μs (x-axis) in a frequency by time grid. The value 71.4 μs is determined by dividing the duration of an LTE subframe of 1000 μs by the number of OFDM symbols, 14, defined to constitute an LTE subframe. In LTE technology, each 15 kHz (y-axis) by 71.4 μs (x-axis) block in the grid is called a Resource Element (RE). A time-frequency allocation of 7 symbols (500 μs) by 12 subcarriers (180 kHz) is called a Resource Block (RB) in 3GPP LTE.

[0018] While the following discussion of Figure 2 focuses on OFDM transmission, it should be understood that Figure 2 may equally illustrate other transmission systems. For example, Figure 2 may illustrate transmission strategies such as those for use in Next Generation Wireless (5G), as well as other orthogonal schemes such as CDMA transmission. Among other transmission strategies, orthogonal or quasi-orthogonal transmission strategies may be used for subchannels or for signaling related to observed time difference of arrival (OTDOA) measurements.

[0019] To enable a user equipment terminal to determine location by calculating OTDOA, certain wireless standards allocate subcarriers within the grid to be used to determine location or to implement OTDOA functionality. To simplify this discussion, exemplary OTDOA subcarriers are designated as "Location Pilots" (LP) 212, 214, 216 in FIG. 2. The term pilot is used to denote subcarriers that have a known transmit modulation at the receiver. These pilot subcarriers are in contrast to data subcarriers that have unknown modulation characteristics because they are encoded with unknown information bits. This pilot scheme allows compatible terminals to implement a variety of measurement techniques. User equipment terminals generally need to estimate the channel impulse response (CIR) and other parameters for successful reception and modulation of OFDM symbols. As a result, the grid shown in FIG. 2 will likely include other subcarriers designated as pilots. These persistent pilots are designated as Estimated Pilots (EP) and are shown in FIG. 2 as 221, 223, 225. In LTE, these locations, and the bits in the modulation of the EP, are known to the receiver as they are defined by the LTE specification. In a UE-assisted system, the user equipment terminal uses these pilots to make TDOA measurements and sends the measurements to a location server configured to make a hyperbolic TDOA position calculation.

[0020] 3, a simplified representation of a conventional UE-aided positioning system 300 is provided. As shown, the system 300 includes a network infrastructure 310 including a base station almanac (BSA) 320, a location assistance data calculator 324, and a location estimator 328 located, for example, in a location server. The location server in a conventional UE-aided positioning system is composed of the location assistance data calculator 324 and the location estimator 328. A mobile network operator (MNO) maintains the BSA 320 and provides the BSA 320 as an input to the location server. For control plane solutions the location server is referred to as an E-SMLC and for user plane solutions the location server is known as an SLP. The location estimator 328 is configured to provide a location estimate based on measurements 330 received from the UE 350, in a known manner. A power and timing measurement module 354 generates the measurements 330 based on the assistance data 358 received from the location assistance data calculator 324. In operation, the location server can provide information from the BSA 320 to nearby cells to facilitate positioning calculations. Unfortunately, assistance data 358 generated from information in the BSA 320 must be downloaded to the UE 350 and measurements 330 must be uploaded to the position estimator 328 every time the position is updated. This creates network congestion and compromises the battery life of the UE 350.

[0021] In the current state of the art, the assistance data 358 consists of several dozen cells used for positioning. For example, in 3GPP TS 36.355, "LTE Positioning Protocol (LPP)," the number of OTDOA neighbor cells for a given frequency layer is 24. The location server derives a "best set" of the 24 cells using the ECGI of the serving cell, which is a globally unique identifier of the cell. However, when the location server derives the 24 cells, it only has a very rough estimate of the location of the UE 350. This rough estimate may be, for example, the centroid or transmission point of the serving cell, i.e., the basic cell ID location. As shown in FIG. 4, the initial position estimate used to derive the assistance data 358 is defined as a seed estimate 410. However, a seed estimate 410 based on basic cell ID positioning may result in a poor set of cells selected for the assistance data 358.

[0022] UE-based ("UE-B") cellular positioning is similar to UE-assisted where a location server provides assistance data (AD) to the UE. In a UE-B system, the assistance data includes additional information that the UE needs to locally estimate its location. UE-B has technical advantages over UE-A. In UE-B, the UE does not need to transmit measurements to a location server. Uplink transmissions are expensive in terms of battery consumption, which puts the UE-A approach at a relative disadvantage to the UE-B approach in terms of battery life. UE-B also relieves network congestion since measurements are not transmitted in the uplink. For 5G large-scale Internet of Things (IoT) scenarios with thousands or even millions of connected devices per cell, excessive uplink transmissions for UE-A positioning can exhaust spectrum and time resources. On the other hand, UE-B reduces uplink traffic and facilitates large-scale service deployment.

[0023] Cellular positioning of both UE-A and UE-B offers technical advantages over traditional GNSS positioning in terms of indoor coverage and lower power consumption. These cellular positioning methods also offer advantages over WiFi location and Bluetooth location, which relies on intrinsic lookup tables and prohibitively expensive beacon deployment. [Prior art documents] [Patent documents]

[0024] [Patent Document 1] US Patent Application Publication No. 2022 / 0334215 [Patent Document 2] China Patent Application Publication No. 105188082 [Patent Document 3] European Patent Application Publication No. 1552321 [Patent Document 4] European Patent Application Publication No. 2893367 Summary of the Invention [Means for solving the problem]

[0025] A hybrid method of estimating a mobile device's location utilizing both received signal power and timing measurements is disclosed herein. The method includes measuring, at the mobile device, received signal power of signals received from a plurality of cells and storing corresponding received signal power measurements. The method further includes measuring, at the mobile device, a plurality of times of arrival of the plurality of signals received from the plurality of cells. A plurality of time difference of arrival (TDOA) measurements are determined from the plurality of times of arrival. A power-time hybrid Gaussian maximum likelihood estimator and positioning assistance data for the plurality of cells are then used to generate a maximum likelihood estimate of the mobile device's location by evaluating the joint conditional probability of the received signal power measurements and the plurality of TDOA measurements. The received signal power measurements are represented by a first Gaussian random variable and the plurality of TDOA measurements are represented by a second Gaussian random variable.

[0026] The present disclosure also relates to a mobile device including a processor and a memory in communication with the processor. The memory stores program instructions that, when executed by the processor, cause the processor to measure received signal power of signals received from a first plurality of cells and store corresponding received signal power measurements in the memory. The instructions further cause the processor to measure first plurality of arrival times of the first plurality of signals received from the first plurality of cells and determine first plurality of time difference of arrival (TDOA) measurements from the first plurality of arrival times. The processor is further configured to generate a maximum likelihood estimate of the location of the mobile device by evaluating a joint conditional probability of the received signal power measurements and the first plurality of TDOA measurements using a power-time hybrid Gaussian maximum likelihood estimator and first positioning assistance data for the first plurality of cells. The received signal power measurements are represented by a first Gaussian random variable and the first plurality of TDOA measurements are represented by a second Gaussian random variable.

[0027] The program instructions may include additional instructions that, when executed by the processor, cause the processor to measure a second plurality of times of arrival of a second plurality of signals received from a second plurality of cells and determine a second plurality of time difference of arrival (TDOA) measurements from the second plurality of times of arrival. The processor may be further configured by the instructions to generate an updated estimate of a position of the mobile device using the second positioning assistance data for the second plurality of cells and the second plurality of TDOA measurements.

[0028] The processor may be further configured, by the instructions, to store cell parameters for the first plurality of cells and the second plurality of cells in a micro-BSA established in the memory, and to derive the first positioning assistance data and the second positioning assistance data from the cell parameters.

[0029] Additionally, disclosed herein is a mobile-based positioning system and method that uses assistance data provided by a repository of Base Station Almanac (BSA) information downloaded from a server and stored on-board a mobile device (or user equipment, UE) as a micro-BSA. The disclosed system and method improve upon the state of the art for downlink cellular positioning by providing, for example, improvements in network efficiency, positioning accuracy, cost, and device battery life. These improvements are achieved at least in part by leveraging the UE's location awareness resulting from the disclosed UE-based positioning methods.

[0030] As discussed herein, the micro-BSA stored on the UE is used for improved assistance data generation, which may be refined based on measurements and location estimates made on the device. This higher quality assistance data allows a more accurate location estimate to be generated by the UE relative to estimates generated using state-of-the-art TDOA techniques. Furthermore, on-board generation of assistance data allows the UE to roam and generate highly accurate updated location estimates without the need for network interaction. This therefore reduces network congestion and improves device battery life.

[0031] In one aspect, the disclosure relates to a method performed in a mobile device for estimating a position of the mobile device. The method includes receiving, from a network server, observed time difference of arrival (OTDOA) assistance data for a first plurality of cells from a base station almanac (BSA) accessible from the network server. The OTDOA assistance data is stored in a memory of the mobile device as a first micro-BSA. An initial position estimate for the mobile device is determined based on time difference of arrival (TDOA) measurements associated with an initial subset of the first plurality of cells and the initial OTDOA assistance data corresponding to the initial subset of the first plurality of cells. The initial OTDOA assistance data is generated by the micro-BSA based on an initial seed estimate. An improved position estimate for the mobile device is determined based on TDOA measurements associated with an additional subset of the first plurality of cells and the improved OTDOA assistance data corresponding to the additional subset of the first plurality of cells. The improved OTDOA assistance data is generated by the micro-BSA based on the initial position estimate.

[0032] The method may further include storing, in a memory of the mobile device, additional OTDOA assistance data for a second plurality of cells from the BSA as a second micro-BSA, the second plurality of cells including at least one cell not included in the first plurality of cells.

[0033] According to another aspect of the disclosure, determining the improved location estimate further includes calculating a first location estimate for the mobile device based on TDOA measurements associated with a first configuration of cells in the first plurality of cells and first improved OTDOA assistance data corresponding to the first configuration of cells, the first configuration of cells being characterized by a first geometric dilution of precision (GDOP). A second location estimate for the mobile device is determined based on TDOA measurements associated with a second configuration of cells in the first plurality of cells and second improved OTDOA assistance data corresponding to the second configuration of cells, the second configuration of cells being characterized by a second GDOP. A third location estimate for the mobile device may further be calculated using an alternating positioning method unrelated to the TDOA measurements. Then, determining the improved location estimate may further include selecting from among the first location estimate, the second location estimate, and the third location estimate.

[0034] In a further aspect of the disclosure, the mobile device transitions to a low-power sleep mode after determining the initial position estimate. In this case, the method for estimating a position of the mobile device further includes, when the mobile device temporarily transitions from the low-power sleep mode, determining an updated position estimate for the mobile device based on additional time difference of arrival (TDOA) measurements associated with the initial subset of the first plurality of cells and additional OTDOA assistance data corresponding to the initial subset of the first plurality of cells, the additional OTDOA assistance data being generated by micro-BSA based on the initial position estimate. The mobile device is configured to transition to the low-power sleep mode after determining the updated position estimate.

[0035] The present disclosure further relates to a mobile device including a processor, a receiver in communication with the processor, and a memory including program code executable by the processor to estimate a position of the mobile device. The program code includes code for receiving Observed Time Difference of Arrival (OTDOA) assistance data for a first plurality of cells from a network server, the OTDOA assistance data being included in a base station almanac (BSA) accessible from the network server. The program code further includes code for storing the OTDOA assistance data in the memory as a first micro-BSA. The program code further includes code for determining an initial position estimate for the mobile device based on time difference of arrival (TDOA) measurements associated with an initial subset of the first plurality of cells and the initial OTDOA assistance data corresponding to the initial subset of the first plurality of cells. The initial OTDOA assistance data is generated by the micro-BSA based on an initial seed estimate. The program code further includes code for determining an improved position estimate for the mobile device based on TDOA measurements associated with an additional subset of the first plurality of cells and the improved OTDOA assistance data corresponding to the additional subset of the first plurality of cells. Improved OTDOA assistance data is generated by micro-BSA based on the initial position estimate.

[0036] The present disclosure further relates to a method executed in a mobile device for estimating a location of the mobile device. The method includes receiving from a network server first Observed Time Difference of Arrival (OTDOA) assistance data for a first plurality of cells and second OTDOA assistance data for a second plurality of cells from a base station almanac (BSA) accessible from the network server, the first plurality of cells corresponding to a first geographic area and the second plurality of cells corresponding to a second geographic area different from the first geographic area. The second plurality of cells includes at least one cell not included in the first plurality of cells. The method further includes storing in a memory of the mobile device the first OTDOA assistance data as a first micro-BSA and the second OTDOA assistance data as a second micro-BSA. A first location estimate for the mobile device is then determined based on a first Time Difference of Arrival (TDOA) measurement associated with a subset of the first plurality of cells and the first OTDOA assistance data corresponding to the subset of the first plurality of cells without receiving additional OTDOA assistance data from the network server. The first OTDOA assistance is generated by the first micro-BSA based on the previous location estimate. The method further includes determining a second location estimate for the mobile device based on a second Time Difference of Arrival (TDOA) measurement associated with a subset of the second plurality of cells and the second OTDOA assistance data corresponding to the subset of the second plurality of cells without receiving additional OTDOA assistance data from the network server. The second OTDOA assistance is generated by the second micro-BSA based on the first location estimate.

[0037] In yet another aspect, the present disclosure relates to a method performed in a mobile device for estimating a location of the mobile device. The method includes receiving, from a network server, Observed Time Difference of Arrival (OTDOA) assistance data for a first plurality of cells from a Base Station Almanac (BSA) accessible from the network server. The OTDOA assistance data is stored as a first micro-BSA in a memory of the mobile device. The method further includes determining a first location estimate for the mobile device based on Time Difference of Arrival (TDOA) measurements associated with a subset of the first plurality of cells and initial OTDOA assistance data corresponding to the subset of the first plurality of cells. An artificial intelligence (AI) management module predicts a route along which the mobile device is expected to travel. Observed Time Difference of Arrival (OTDOA) assistance data for a second plurality of cells from the BSA is received from the network server. The second plurality of cells are determined by the AI ​​management module to be proximate to a route along which the mobile device is expected to travel, and the second plurality of cells includes at least one cell not included in the first plurality of cells. A second position estimate is then determined based at least in part on second OTDOA assistance data corresponding to a subset of the second plurality of cells.

[0038] The present disclosure further relates to a method performed in a mobile device for estimating a position of the mobile device. The method includes determining a first position estimate for the mobile device based on time difference of arrival (TDOA) measurements associated with at least a subset of a first plurality of cells and observed time difference of arrival (OTDOA) assistance data corresponding to the subset of the first plurality of cells. The TDOA measurements are associated with a first TDOA measurement vector, and the OTDOA assistance data is stored in a micro-BSA in a memory of the mobile device. The method further includes determining a quality of the first position estimate by at least calculating a first TDOA residual vector using the first position estimate, and detecting a bad cell included in the subset of the first plurality of cells based on the first TDOA residual vector. A second TDOA measurement vector is then constructed by removing the TDOA measurements associated with the bad cell from the first TDOA measurement vector, and a quality of a second position estimate for the mobile device is determined based on the second TDOA measurement vector.

[0039] In an additional aspect, the present disclosure relates to a method performed in a mobile device for estimating a location of the mobile device. The method includes determining a time of arrival (TOA) estimate of a signal received from a plurality of cells in a vicinity of the mobile device and determining a value of a quality metric associated with the TOA estimate. One of the plurality of cells is selected as a reference cell based on the value of the quality metric associated with the TOA estimate. The method further includes determining a time difference of arrival (TDOA) measurement between the reference cell and a neighboring cell included in the plurality of cells. An initial estimate of the location of the mobile device is determined using the plurality of TDOA measurements and associated observed time difference of arrival (OTDOA) measurements stored in a micro-BSA in a memory of the mobile device. The plurality of TDOA measurements are associated with a subset of the TOA estimates having a value of a quality metric relative to a value of a quality metric corresponding to a TOA estimate associated with another TDOA measurement of the TDOA measurements. The method further includes determining successive additional location estimates by updating the initial estimate of the location of the mobile device using additional TDOA measurements of the TDOA measurements and associated OTDOA assistance data until a stopping criterion is met. Determining the successive additional position estimates may include judging a quality of the successive additional position estimates by at least one of: (i) evaluating a contour in a position estimate cost function; and (ii) calculating a time difference of arrival (TDOA) residual vector.

[0040] The present disclosure further relates to a method including sending, by a network server to a mobile device, Observed Time Difference of Arrival (OTDOA) assistance data for a first plurality of cells from a base station almanac (BSA) accessible from the network server. The OTDOA assistance data is stored as a first micro-BSA in a memory of the mobile device. The mobile device is configured to determine a first position estimate for the mobile device based on Time Difference of Arrival (TDOA) measurements associated with a subset of the first plurality of cells and initial OTDOA assistance data corresponding to the subset of the first plurality of cells. A route along which the mobile device is expected to travel is predicted by an artificial intelligence (AI) management module. The network server then sends, to the mobile device, Observed Time Difference of Arrival (OTDOA) assistance data for a second plurality of cells from the BSA. The second plurality of cells are determined by the AI ​​management module to be proximate to a route along which the mobile device is expected to travel, and the second plurality of cells includes at least one cell not included in the first plurality of cells. The mobile device is configured to determine a second position estimate based at least in part on second OTDOA assistance data corresponding to a subset of the second plurality of cells.

[0041] The present disclosure further relates to a method that includes determining a time of arrival (TOA) estimate for signals received from a plurality of cells in a vicinity of a mobile device, the method including determining a time difference of arrival (TDOA) measurement between a reference cell and a neighboring cell included within the plurality of cells, and determining an estimated location of the mobile device using the plurality of TDOA measurements and an associated observed time difference of arrival (OTDOA). The method further includes determining an uncertainty in the estimated location by evaluating a contour of a cost function formed from the plurality of TDOA measurements, the uncertainty corresponding to an area associated with one of the contours.

[0042] The present invention will become more fully understood in connection with the following detailed description taken in conjunction with the accompanying drawings, in which: [Brief description of the drawings]

[0043] [Figure 1] 1 is a functional block diagram of an apparatus for determining position using observed time difference of arrival (OTDOA) based on reference signal time difference (RSTD) measurements defined in the LTE standard. [Diagram 2] FIG. 1 is a generalized diagram of a symbol structure of the type used in the LTE standard. [Diagram 3] FIG. 1 provides a simplified representation of a conventional UE-assisted positioning system. [Figure 4] FIG. 2 illustrates the use of an initial position estimate in the form of a seed estimate to derive positioning assistance data. [Diagram 5] FIG. 1 is a functional diagram of a positioning system, according to one embodiment. [Figure 6] 4 is a flowchart of a sequence of operations performed by a UE according to one embodiment. [Figure 7] FIG. 6 illustrates a geographic footprint associated with multiple micro-BSAs stored on a UE included in the system of FIG. 5. [Figure 8] 6 illustrates a bad cell detection process implemented by a bad cell detection unit for improving the quality of location estimates produced by a location estimation unit included in the UE of FIG. 5. [Figure 9] FIG. 1 illustrates a method for generating circular sector assistance data to aid in reducing GDOP in accordance with the present disclosure. [Figure 10] FIG. 13 is a screenshot capture of a map of a portion of New York City showing an arrangement of cells that lacks geometric diversity. [Figure 11] FIG. 13 is a screenshot capture of a map showing the placement of cells selected in accordance with the circular sector assistance data generation method to have improved geometric diversity. [Figure 12] FIG. 2 includes a block diagram representation of a particular implementation of a UE configured in accordance with one embodiment. [Figure 13]A diagram showing example and potential strategies for forming cell pairs consisting of a reference cell and a neighboring cell. [Figure 14] A diagram showing example and potential strategies for forming cell pairs consisting of a reference cell and a neighboring cell. [Figure 15] A diagram showing example and potential strategies for forming cell pairs consisting of a reference cell and a neighboring cell. [Figure 16] FIG. 1 illustrates an estimated confidence ellipse formed by evaluating multiple observations. [Figure 17] FIG. 13 illustrates an example error surface contour generated to establish a confidence region. [Figure 18] FIG. 1 illustrates attenuation profiles associated with commonly used attenuation models. [Figure 19A] FIG. 1 illustrates the results of a simulation performed using a hybrid power-time positioning method, according to one embodiment. [Figure 19B] FIG. 1 illustrates the results of a simulation performed using a hybrid power-time positioning method, according to one embodiment. [Figure 19C] FIG. 1 illustrates the results of a simulation performed using a hybrid power-time positioning method, according to one embodiment. [Figure 20] FIG. 1 provides a block diagram of a first embodiment of a power-time hybrid positioning system in accordance with the present disclosure. [Figure 21] FIG. 2 provides a block diagram of a second embodiment of a hybrid power-time positioning system in accordance with the present disclosure. [Figure 22] FIG. 10 provides a block diagram of a third embodiment of a hybrid power-time positioning system in accordance with the present disclosure. [Diagram 23] FIG. 13 provides a block diagram of a fourth embodiment of a power-time hybrid positioning system in accordance with the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0044] Turning attention to Figure 5, Figure 5 illustrates a functional diagram of a positioning system 500 according to one embodiment. As discussed below, the system 500 includes a UE 510 configured to generate positioning assistance data 526 based on a relatively small subset of a base station almanac (BSA) 512 accessible from a BSA server 511 in a network 514. The BSA 512 may be on a server in communication with the BSA server 511 or may be included on the BSA server 511. In one embodiment, information from the BSA 512 is provided to the UE 510 by a micro-BSA cloud assistance server 516 in the network 514.

[0045] The functional elements of the UE 510 include one or more micro-BSAs 520 configured to store information corresponding to a subset of the BSA 512 received by the UE 510 from the micro-BSA cloud assistance server 516. The micro-BSAs 520 may be calculated using information of the UE 510 serving cell ECGI to provide a rough estimate of the UE location. A positioning assistance data calculation unit 524 is configured to receive cell parameters from the micro-BSA 520 for use in generating assistance data 526. The positioning assistance data calculation unit 524 may use the serving cell ECGI to derive a rough estimate of the UE location for calculating the assistance data 526. As shown, the assistance data 526 is provided to a power and timing measurements module 530 and a position estimator 540. As discussed below, the calculation unit 524 is further configured to respond to measurement feedback 532 and position estimation feedback 534 in providing cell selection feedback 536 to aid in intelligently updating the content of the micro-BSA 520. The position estimator 540 is configured to provide a position estimate and position estimate feedback 534 based on the assistance data 526 and measurements 542 received from the power and timing measurement module 530. The power and timing measurement module 530 generates the measurements 542 and measurement feedback 532 based on the assistance data 526 and feedback 546 received from the position estimator 540.

[0046] In one embodiment, the location estimator 540 performs OTDOA calculations based on the assistance data 526 and measurements by the UE 510 of the time of arrival (TOA) of positioning reference signals (PRS) received from base stations (e.g., eNodeBs) of multiple cells associated with the assistance data 526. The location estimator 540 subtracts the TOA of a reference cell (which may be selected by the UE 510 using known techniques) from the measured TOAs corresponding to such multiple cells to form reference signal time difference (RSTD), or time difference of arrival (TDOA) measurements. These TDOA measurements can be used together with the assistance data 526 to constrain the location of the UE 510 to a set of hyperbolic curves. If the TOA measurements made by the UE 510 were completely free of noise and interference, these hyperbolic curves would intersect at a single point corresponding to the location of the UE 510. In practice, however, such noise and interference limits the accuracy with which the location of the UE 510 can be estimated. As discussed herein, the use of the micro-BSA 520 in accordance with the present disclosure improves the accuracy and efficiency of estimating the location of the UE 510.

[0047] In terms of scale, in an exemplary embodiment, the BSA 512 may contain information corresponding to hundreds of thousands of cells, the micro-BSA 520 may contain information for hundreds of cells, and the assistance data 526 may relate to tens of cells.

[0048] This BSA 512 is typically managed by a mobile network operator (MNO) and includes a database containing cell parameters that define the network layout. Each cell in the BSA 512 database is typically characterized by a unique cell identifier (ECGI), latitude and longitude of the cell transmission point, physical cell index (PCI), antenna aperture and orientation details, transmit power, and various other parameters. A cloud assistance server 516 interacts with the BSA 512 to provide the UE 510 with a small subset of the contents of the BSA 512. As mentioned above, the resulting micro-BSA 520 from which the assistance data 526 is derived may consist of hundreds of cells in the vicinity of the UE 510's serving cell. The storage and download requirements associated with a further 1000-cell micro-BSA 520 are modest. For example, if roughly 120 bits are needed to represent the cell parameters for a given cell, then only approximately 15 kB would be required for a 1000-cell micro-BSA520 (ie, 1000 cells x 120 bits / cell x 1 kB / 8000 bits = 15 kB).

[0049] The information including this 15 kB, 1000-cell micro-BSA can be transferred to the UE 510 over a radio link in a few seconds while the UE is in LTE connected mode. A smaller micro-BSA may be required for shorter download times and less storage, and a larger micro-BSA may be required for greater coverage and less overall interaction with the cloud-assisted server 516 or otherwise with the network 514. As a reference, for a typical cell density of 1 cell / km2, the 1000-cell micro-BSA 520 provides coverage for an area of ​​1000 km2. Many positionings can be obtained using a single micro-BSA 520. Thus, once the information for the micro-BSA 520 is downloaded, the UE 510 requires minimal further interaction with elements of the network 514.

[0050] Thus, during operation of the UE 510, the location estimator 540 may generate many location estimates even though the UE 510 is operating solely based on measurements 542 and assistance data 526 derived from information in the micro-BSA 520. This advantageously improves battery life of the UE 510 and reduces network congestion because no almanac information is provided by the BSA 512 and no assistance information is otherwise provided to the UE 510 by the network associated with each location estimate generated by the location estimator. Moreover, the UE 510 may, in some embodiments, utilize filtering and other techniques to average or otherwise smooth the location estimates generated locally by the location estimator 540, thereby enhancing positioning accuracy relative to the case where such almanac information, and / or assistance data is provided to the UE 510 to facilitate each position measurement. The current state of the art UE assisted methods are considered as "single shot" estimation where assistance data is provided to the UE from a location server, the UE then reports measurements, and the location server uses a single set of measurements to estimate the location. In this approach, the estimation algorithm does not really benefit from filtering, since continuous measurement reporting is not feasible from both a UE battery drain and network congestion perspective. Measurements 542 are more efficiently fed to the location estimator 540 to enhance the estimation process.

[0051] Since the UE 510 is aware of its current location, it may be configured to sense when new BSA information is needed. For example, when the location estimator 540 is deriving a high quality location estimate, new BSA information from the micro-BSA cloud assistance server 516 is not needed. The location estimator 540 can determine if the estimate is of high quality by examining the contour of the likelihood or recursive function surface, or by calculating a TDOA residual vector. The TDOA residual vector is denoted by e, where

number

number

number

number

[0052] In one embodiment, the position estimator 540 uses a TDOA hyperbolic location signal model to generate the position estimate.

[0053] TDOA Hyperbolic Location Signal Model In an embodiment in which the UE 510 performs downlink time difference of arrival (TDOA) hyperbolic location estimation, the UE 510 performs time of arrival (TOA) estimation for the surrounding cells. In the following, the term "cell" is used interchangeably with "base station" or "transmission point". In addition, the term "RSTD" (i.e., "reference signal time difference" as defined in the 3GPP standard) is used interchangeably with TDOA. The surrounding cells are assumed to be time synchronized and to have transmitted positioning reference signals ("pilots") at some time close to τ seconds. More specifically, the kth cell is assumed to be located at time τ k = τ + α k However, α k is a relatively small transmission synchronization period. The TOA of the kth cell is

number

number

[0054] Assume that the UE 510 is synchronized to a serving cell, which may or may not be the closest cell to the UE 510. Without loss of generality, this serving cell is indexed by k=0, and each of the other cells is indexed by k=1, 2,...,K. To synchronize, the UE 510 determines the TOA of the serving cell.

number

number

number

number

[0055] The UE 510 then performs an estimate of the relative TOA.

number

[0056] Here, the unknown location of UEx is the relative TOA estimate

number

number

number

number

[0057] It is convenient to represent TDOA measurements in vector form.

number

number

[0058] The i(m)th cell is considered as the RSTD neighbor cell for the mth measurement, and the j(m)th cell is considered as the RSTD reference cell for the mth measurement. In the 3GPP specification, a common RSTD reference cell is used, and there are k ref For ∈{0,1,...,K}, j(m)=k refThe remaining cells are RSTD neighbor candidates. There may be other strategies for pairing RSTD neighbor candidates to the reference cell. For example, one such potential pairing strategy is characterized as all "N choose two", cell pairs, where N=K+1. See, for example, F. Gustafsson and F. Gunnarsson, "Positioning using time-difference of arrival measurements," 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP'03), Hong Kong, China, 2003. Another possibility is a neighbor index approach, where i(m)=j(m)+1 and j(m)=m-1, for m=1, 2,..., M=K. These enumerated strategies do not exclude other possibilities, and the described strategies have their benefits and costs. For example, a single reference cell is advantageous when a low-error reference cell is selected, but may not perform as well as a nearest neighbor approach when a high-error reference cell is selected. An exhaustive "N choose two" can provide richer information to the location estimation algorithm at the cost of higher computational complexity.

[0059]

[0071] Turning attention now to Figures 13-15, Figures 13-15 show exemplary potential strategies for forming cell pairs composed of reference and neighbor cells. Specifically, Figure 13 shows a single-cell reference cell pairing strategy, Figure 14 shows a neighbor index cell pairing strategy, and Figure 15 shows an N-choose-two cell pairing strategy.

[0060] TDOA position estimation part The TDOA position estimator 540 functions to determine a good estimate for x given the measurements at r. The position estimator 540 may utilize a variety of different methods in making this determination, including, for example, least squares, weighted least squares, and Gaussian maximum likelihood. The methods described below do not preclude the use of other possibilities.

[0061] The location estimation unit 540 detects the location of the UE 510.

number

number

number

number

number

number

number

[0062] Assume that some measurements at r are of higher quality than others. Therefore, it may be beneficial for the location estimator 540 to give more weight to estimates of higher quality and less weight to estimates of lower quality. A weighted least squares (WLS) estimator can achieve this.

number

[0063] Now assume that statistical information is available about the measurement vector r. Let p(r|x) be the probability of the measurement vector r being at the location of the UE 510 at x. This is known as the likelihood function and is

number

[0064] A special case of the ML estimator is when the likelihood function is

number

number

[0065] The generalized weighted least squares estimator is

number

number

[0066] The implementation of the location estimator 540 as a GML estimator described above assumes that the TOA is obtained from a Gaussian distribution. A known generalization of this framework is to assume that the TOA is derived from a Gaussian mixture model (GMM). See, for example, F. Perez-Cruz, C. Lin, and H. Huang, "BLADE: A Universal, Blind Learning Algorithm for ToA Localization in NLOS Channels," 2016 IEEE Globecom Workshops (GC Wkshps), Washington, DC, USA, 2016. The GMM framework better accounts for the multipath nature of cellular radio frequency (RF) environments. Furthermore, the ML estimator implemented by the location estimator 540 can be generalized to a maximum a posteriori (MAP) estimator using prior statistical information about the location of the UE 510.

[0067] The weighted least squares estimator derived above is a quadratic cost function Q(x)=(rh(x)) T W(rh(x)) Minimize. The minimization can be done by numerical sampling of a rectangular or hexagonal grid, or by statistical sampling methods such as Markov Chain Monte Carlo (MCMC), of which the Metropolis Hastings algorithm is an example. Alternatively, the minimization can be solved analytically through a Taylor series expansion, see for example Torrieri, DJ, "Statistical Theory of Passive Location Systems", IEEE Trans. on Aerospace and Electronic Systems AES-20, 2 (Mar. 1984). In the case of Gaussian maximum likelihood, this cost function is the logarithm of the likelihood function, known as the log-likelihood. Similarly, for maximum a posteriori (MAP) estimation, a similar formulation is derived by incorporating the so-called a priori probability.

[0068] Adaptive Generation of Assistance Data from Micro-BSA Information The BSA coherence time can be defined as the time duration during which the BSA information remains relatively fixed and useful for positioning. The BSA coherence time is large relative to the position measurement update rate. For example, the position measurement may be updated once per hour, whereas the BSA coherence time may be on the order of days or months. This allows the same BSA information in the micro-BSA 520 to be used across multiple position measurement events.

[0069] Consider the case where geofencing is used. The owner of a valuable asset attaches it to a geofencing tracker device (which may be a simplified implementation of UE 510). The owner wants to be notified if the asset moves beyond a certain area. The asset may remain in the specific area for days or months. Over this duration of time, the contents of micro-BSA 520 also remain practically fixed, and location estimator 540 produces high-quality location estimates. Under these conditions, the implementation of UE 510 as a tracker device will not require any further BSA information and therefore will not need to interact with BSA server 511 or with other cloud servers hosting BSA 512.

[0070] Notably, in the current cellular positioning state-of-the-art, the UE is not location aware. Therefore, this network mitigation feature cannot be realized. The current state-of-the-art is inefficient in that assistance data must be downloaded to the UE and measurements must be uploaded to the location server for every location update. This creates network congestion issues and impairs the device's battery life. In contrast, the UE 510 is "location aware," which enables applications such as geofencing better and more efficiently. This location awareness also allows for faster positioning, reduced latency, and improved time to first fix (TTFF). These features enable battery-efficient breadcrumb applications when the UE is primarily in a low-power sleep mode. It wakes up instantly, updates its location estimate, and returns to a lower power state. The faster the location update, the more efficient the resolution.

[0071] As mentioned in the Background, in the current state of the art, the assistance data consists of tens of cells used for positioning, and conventional location servers use globally unique identifiers of the cells to derive the "best set" of cells. However, this only provides the location server with a very rough seed estimate of the UE location when deriving the set of cells used in generating the assistance data.

[0072] Turning now to FIG. 6, FIG. 6 is a flow chart of a sequence of operations 600 performed by the UE 510 highlighting one way in which intelligent and adaptive generation of assistance data 526 in the UE 510 may be utilized to address this shortcoming in the current state of the art. Additionally, in one embodiment, the micro-BSA 520 includes information on more cells than those represented in the assistance data 526. This larger collection of cells in the micro-BSA 520 provides improved flexibility in generating a good set of cells for the assistance data 526. Referring to FIG. 6, an initial set of assistance data cells is generated using the basic cell ID 604 for a seed estimate (stage 608). An initial position estimate is then derived using a position estimate that is better than the basic cell ID estimate (stage 612). This new estimate is used to re-derive assistance data, thereby generating improved assistance data (stage 616). This is done efficiently on the UE 510 using the local micro-BSA 520 without requiring interaction with the micro-BSA cloud assistance server 516 or otherwise with the network 514. An improved position estimate may then be generated (stage 620) using the improved assistance data. This method 600 of improved seed estimation to generate better assistance data may of course be repeated over time.

[0073] Device initialization and Micro-BSA population The BSA 512 is controlled by the network operator and has parameters for all cells in the network (e.g., .about.700,000 cells for the largest operator). The micro-BSA 520 typically contains a very small subset of information in the BSA 512 for cells in the vicinity of the UE 510. For example, the micro-BSA 520 may contain parameters for 1,000 cells that make up a metropolitan area including a downtown and surrounding areas of the city. For UE-based OTDOA, the micro-BSA 520 has the parameters necessary to execute the OTDOA algorithm. The OTDOA algorithm consists, for example, of (i) generating assistance data (AD) from the micro-BSA 520 and (ii) using the assistance data (AD) to make TOA / TDOA measurements and estimating the location of the UE using the TOA / TDOA measurements plus the assistance data (e.g., cell latitude / longitude).

[0074] The AD is a subset of the micro-BSA 520. For example, the AD may consist of parameters of 50 cells. BSA->micro-BSA->AD 700,000 cells -> 1,000 cells -> 50 cells

[0075] The UE 510 may have one or more micro-BSA 520. For example, the UE 510 may have two or more micro-BSA 520 to serve a few different areas around a town that the UE 510 frequents. To initialize the device, the UE 510 communicates with the BSA server 511 via the micro-BSA cloud assistance server 516. The UE 510 informs the BSA server 511 of the ECGI of the serving cell, and possibly the ECGI or PCI of neighboring cells. For example, the UE 510 may inform the BSA server 511 via the micro-BSA cloud assistance server 516 that the serving cell ECGI is "xyz" and that the UE 510 would like a 200-cell micro-BSA 520. The parameters that make up a cell are roughly 120 bits, so a 200-cell micro-BSA 520 would be 200*120 / 8 / 1000=3 kilobytes. These 3 kB of BSA information are retrieved from the BSA 512 by the BSA server 511 and then communicated to the UE 510 by the micro-BSA cloud assistance server 516 in the downlink channel for storage.

[0076] With the micro-BSA 520 instantiated on the UE 510, the UE 510 effectively has access to a "mini-map" of those 200 cells. Since the UE 510 is aware of its location, it knows if it remains within the coverage area of ​​these 200 cells. If the UE 510 roams outside of these cells, it may want to request a new micro-BSA 520. If the UE 510 is stationary and has not detected as many cells covering the same number as the expected predetermined contents of the current micro-BSA 520, it may be the case that the topology of cells has changed and it may be a good time to have the micro-BSA 520 refreshed.

[0077] The ECGI of the serving cell can provide a seed estimate for the UE 510 from which the set of micro-BSA cells is derived. The seed estimate can simply be the transmission point of the serving cell. With further neighborhood information, the server can derive a better seed estimate, such as the centroid of the serving cell and surrounding cells.

[0078] To determine a good set of AD cells, the seed estimate can be similar to the seed estimate used to obtain the micro-BSA 520 from the BSA server 511, i.e., similar to a cell ID. Alternatively, assume that the UE 510 is roaming. The serving cell may change from one ECGI to another ECGI due to handover. At this time, the timing reference at the UE 510 will likely change as the UE 510 synchronizes with some new serving cell. This time change can be logged so that the current set of timing measurements is adjusted to the current set of cells being monitored. The positioning assistance data calculation 524 on the UE 510 will likely want to obtain a new set of AD cells from the micro-BSA 520. The seed estimate for the new AD can be the latest UE location estimate (e.g., using OTDOA). If the new serving cell is within the micro-BSA 520, no further action is required. Otherwise, the UE 510 will need to retrieve the new micro-BSA 520 from the BSA server 511 that hosts or has access to the BSA 512. If the UE 510 is at the edge of the serving area of ​​the micro-BSA 520 used for AD, the positioning assistance data calculation unit 524 and / or the location estimation unit 540 can cause the UE 510 to retrieve the BSA information corresponding to the new micro-BSA 520 from the BSA server 511.

[0079] It may be desirable to have the BSA server 511 constantly monitor the cells whose information is stored in the micro-BSA 520 of the UE 510. That way, the BSA server 511 can provide the UE 510 with information about new cells that are not duplicates of the current micro-BSA. When populating a new micro-BSA, it may be advantageous to send only the differences from the previous micro-BSA.

[0080] 5 and 6, improved assistance data 616 is made possible by improved seed estimates at 612. Feedback 532 also allows for improved assistance data. Consider, for example, the case where the serving cell is much farther away from the UE 510 than other surrounding cells. For example, this can occur when the serving cell is transmitting at high power at the top of a hill that is in line of sight with the receiver of the UE 510. The receiver of the UE 510 can sense that this hilltop cell has the highest signal-to-noise-plus-interference ratio (SINR) of all its surrounding cells and can use this as its serving cell. In this example, there may be other closer cells that possibly transmit at even lower power or are not in line of sight. Setting the initial seed estimate to the serving cell location 604 may be suboptimal in this case since the serving cell will be relatively far from the target receiver of the UE 510. This condition can eliminate cells that are actually much closer than the serving cell.

[0081] The current state of the art suffers from this scenario because the Assistance Data is derived at the location server, rather than adaptively derived on the device, such as the UE 510, configured with a superset of the Assistance Data, micro-BSA 520. In an embodiment of the system, measurements 530 are provided as feedback 532 to the Positioning Assistance Data Calculator 524, resulting in efficient adaptation and improved location accuracy.

[0082] In this described scenario of a far away serving cell, the timing measurement at 530 may detect a closer cell (with a negative delay relative to the timing of the serving cell). For example, a high quality (PAPR, or SINR, low variance, etc.) negative TOA of 1000 meters may be present in the list of detected cells. This implies that the negative TOA cell is 1000 meters closer to the UE than the serving cell. It may be beneficial to re-seed the assistance data calculation by incorporating this information. For example, the latitude / longitude coordinates of the closer cell may be used as a new assistance data seed estimate 612.

[0083] Similarly, the timing advance (TA) at the receiver can be used to detect a serving cell that is far away. In a cellular system, the TA is used to signal a far away receiver to transmit earlier, so that the uplink transmissions of the far away and closer devices arrive at the base station receiver at approximately the same time. Thus, the assistance data calculation unit can improve the assistance data generation strategy using the TA information available at the host modem 1224 in FIG. 12. For example, if the TA is high and implies a serving cell that is far away, the number of cells in the assistance data for TOA measurements may extend to a greater distance from the serving cell. Thus, when a closer TOA is detected, the assistance data can be recalculated as described above.

[0084] Multiple on-board Micro-BSA and tracking use cases The present system advantageously allows devices such as UE 510 to specialize location services for a wider range of use cases. The current state of the art only provides assistance data for a few dozen cells intended for a single use case, mainly emergency services (e911). This current state of the art is not well suited for roaming use cases, for example. The present system solves this problem by the use of a micro-BSA 520 stored on the UE 510. For UEs implemented as dedicated location devices, more memory can be allocated on the UE 510 to store the micro-BSA 520. This enables roaming use cases and minimizes interactions with the micro-BSA cloud assistance server 516 or with other elements of the network 514. For example, 15 bytes per cell is sufficient to represent cell parameters in the assistance data. This representation includes parameters such as, for example, the physical cell ID (PCI), latitude / longitude coordinates of the cell, etc. Instead of storing 24 cells, as is done in the current state of the art, the UE 510 can store 1000 cells using 15 kilobytes of memory. Assuming a cell density of one cell per square kilometer, the micro-BSA 520 can have a location coverage area of ​​1000 square kilometers, allowing the UE 510 to roam. An example use case here is tracking rental scooters as the devices roam around town. Drivers of these scooters may want to track their location both indoors and outdoors, and with the present system this feature can be delivered at low cost.

[0085] Turning now to FIG. 7, an illustration of a geographic footprint associated with multiple micro-BSAs 520 stored on a UE 510 is provided. In some use cases, it may be desirable for the UE 510 to store multiple micro-BSAs across a larger geographic region. As shown, information related cells located within a first geographic footprint 710 corresponding to an area of ​​high cell density are stored within a first micro-BSA 5201. Similarly, information related cells located within a second geographic footprint 720 corresponding to an area of ​​medium cell density are stored within a second micro-BSA 5202, and information related cells located within a third geographic footprint 730 corresponding to an area of ​​low cell density are stored within a third micro-BSA 5203.

[0086] In this use case illustrated in Figure 7, the UE 510 is known to typically travel across a first geographic footprint 710, a second geographic footprint 720, and a third geographic footprint 730. By storing the micro-BSAs 5201, 5202, 5203 for each geographic footprint 710, 720, 730, the UE 510 has all the needed cell information to perform location functions without interacting with the micro-BSA cloud assistance server 516 or with other elements of the network 514. Yet, this can be done using less storage than storing the superset, represented by the circle 750, of the geographic footprints 710, 720, 730. As the UE 510 roams throughout the geographic footprints 710, 720, 730, the UE 510 intelligently derives its assistance data by making cell selection across multiple locally stored micro-BSAs 5201, 5202, 5203.

[0087] A particular use case illustrated in Figure 7 involves a company tracking tools used at different work sites. For example, a construction company may have expensive generators or reciprocating saws that are moved from work site to work site. Suppose the company has three work sites (located in geographic footprints 710, 720, 730, respectively) and a manager misplaces a tool. In this case, the manager can use a small tracker (an implementation of UE 510) attached to the tool to determine its location.

[0088] Artificial intelligence (AI)-assisted micro-BSA management In use cases where a device such as UE 510 is moving over large distances (e.g., while attached to or associated with a container in a truck traversing an interstate highway in the United States), the UE 510 can sense the high degree of mobility (e.g., using Doppler estimation) and the cell parameter information downloaded to the device's micro-BSA 520 can be adapted accordingly. For example, in this case, it may be advantageous to provide parameter information in the micro-BSA 520 to cells that are on the expected route of the truck.

[0089] In such use cases, the micro-BSA "artificial intelligence" (AI) management module 550 can help manage the information contained in the micro-BSA 520. For example, the micro-BSA AI management module 550 can implement pattern recognition algorithms that can identify with a high likelihood that the UE 510 will best benefit from a certain set of micro-BSA cells when it is located on an interstate highway and traveling at a certain speed. Similarly, when the UE 510 is determined to be stationary in a town center, the UE 510 is more likely to benefit from a different strategy. In this latter case, the UE 510 may be attached to a smart meter, a traffic sign, or an automated teller machine (ATM) cash machine that is not intended to move for the life of the UE 510. For these applications, the micro-BSA download management performed by the AI ​​management module 550 will be different than for high speed interstate traveling use cases.

[0090] This management of micro-BSA information using AI can benefit underground worker use cases. For example, if the device serving cells are underground in an urban subway system, the AI ​​management module 550 may only consider serving underground cells in the micro-BSA 520. Considering another use case, the AI ​​management module 550 can learn from patterns in commuting. For example, a commuter line will have a finite number of travel routes. The download of information to the micro-BSA 520 of the UE 510 being transported by the line can benefit from including cells in the most common travel routes, which may depend on the time of day / week. As another example, the AI ​​management module 550 can "learn" that a device traveling in a hurry on a particular highway during a particular time (e.g., 20 miles east of El Centro, Tuesday, 9:00 a.m., on Interstate 8) will arrive in Glendale, Arizona with a 90% likelihood. This knowledge may then be used to download to the micro-BSA 520 information regarding cells that are more likely to be used by the UE 510 when traveling along such highways at a particular time.

[0091] In other embodiments, the AI ​​management module 550 in the BSA server 511 may be supplemented by an optional AI management module 552 disposed in the UE 510. The optional AI management module may be configured to perform at least some of the processing otherwise performed by the AI ​​management module 550.

[0092] Defective cell detection section FIG. 8 shows a bad cell detection process 800 implemented by the bad cell detection unit 544 (FIG. 5) to improve the quality of the position estimate produced by the position estimation unit 540. As mentioned above, the TDOA residual vector provides insights into the quality of the position estimate, which can be exploited in the process 800. First, a position estimate is calculated using the TDOA measurements in a first TDOA measurement vector r1 (stage 810). The resulting position estimate is used to build a TDOA residual vector e1 (stage 820). Cell 5 is detected and labeled as a "bad cell" due to its relatively high error value (stage 830). Cell 5 is then removed to build a new TDOA measurement vector r2 (stage 840). This measurement vector is used to improve the resulting position estimate (stage 850). This bad cell detection process 800 can be combined with other criteria to determine the quality of the position estimate. For example, a minimum number of cells may be required. For 2D hyperbolic TDOA positioning, measurements from at least three distinct cell sites are required. It may be advantageous for more than three distinct cell sites to be required for extra redundancy and added robustness in the position calculation. Furthermore, the GDOP between the estimated UE location and the cell used for positioning can be calculated. If the GDOP is raised above a certain threshold, it can be determined that the environment is not well suited to hyperbolic TDOA. In this case, the position estimation may return no results, return an erroneous result, or fall back to another positioning method such as E-CID. The threshold settings for the residual, the minimum number of cells, and the minimum GDOP can be dynamically determined, or multiple static configurations can operate independently in parallel. For example, config A can be a strict configuration, config B can be a less strict configuration, and config C can be E-CID. When reporting, the position estimate of config A is used if available, otherwise config B is used if available, otherwise config C is used as a fallback.Alternatively, the thresholds can be dynamically set. In a dense cellular environment where many cells are measured, for example, the minimum threshold can be raised. In a coastal environment where the cell locations are unidirectional (many on land, some at sea, and a few at sea), the GDOP thresholds may start low but gradually increase to relax the requirements for this given environment. The E-CID can incorporate power measurements at every v-shift to obtain RSRP for multiple surrounding cells, and cell TOA detection capability rates can be incorporated to obtain a determination of the UE angle relative to the serving cell.

[0093] The UE 510 may be configured to utilize feedback from the bad cell detector 544 to improve the assistance data 526. In the particular case of FIG. 8, cell 5 being deemed poor by the bad cell detector 544 is useful information that may be included in the feedback 534 to the positioning assistance data calculator 524. In this manner, information about cell 5 may be excluded from inclusion in future assistance data 526.

[0094] Considering again the example of FIG. 8, "cell 5" may be considered to be of low quality due to difficult multipath channel conditions that impair TOA estimation. Alternatively, cell 5 may be relatively unsynchronized with respect to other cells in the group. In the latter case, the bad cell detector 544 provides an algorithmic means to deal with poorly synchronized networks. In particular, improved performance is achieved by removing a small number of cells in the list that are relatively unsynchronized with other cells. Moreover, the level of asynchrony may be a relatively fixed amount that is estimated and compensated for in the location estimator 540.

[0095] Another feature of the system is to filter out cells in the Assistance Data 526 that are rarely or never detected. Trying to detect cells that cannot be detected wastes computing resources of the UE 510. Thus, there is an efficiency gain to be made by ignoring cells that are difficult to detect. The estimation algorithm performed by the Position Estimator 540 can monitor which cells are detected and which are not. This information can be included in the feedback 534 provided to the Positioning Assistance Data Computer 524 to allow for incremental and efficient improvements.

[0096] Good cell selection section In an alternative embodiment, a location estimation method can be performed by the good cell selector (GCS) 560 of the location estimator 540 instead of the method performed by the bad cell detector 544. First, the good cell selector 560 ranks the estimated TOAs of the surrounding cells in terms of quality. The quality metric can be based on the signal-to-noise-plus-interference ratio (SINR) or peak-to-average power ratio (PAPR) of the correlator output. See, for example, Thompson et al., "Communication System Determining Time of Arrival Using Matching Pursuit," U.S. Pat. No. 10,749,778. Alternatively, the estimated TOAs of the surrounding cells can be ranked by the peak-to-average power ratio of the pseudospectrum in the multiple signal classification (MUSIC) super-resolution algorithm. See, for example, X. Li and K. Pahlavan, "Super-Resolution TOA Estimation With Diversity for Indoor Geolocation," IEEE Transactions on Wireless Communications, vol. 3, no. 1, Jan. 2004.

[0097] It is advantageous to use the best quality TOA as the RSTD reference cell and the remaining TOAs as the RSTD neighbor cells. Thus, index i(m) is set to the best quality TOA and RSTD neighbor cell index j(m) is set to the remaining cells. The objective of the good cell selector (GCS) is to select a subset of good measurements P≦M out of the M available TDOA measurements. This assumes that some of the measurements are poor, and this may be due to various error sources including timing errors, non-line-of-sight (NLOS) bias, and TOA estimation errors. One procedure for identifying good cells is to start with an initial few best quality measurements to form an initial location estimate by minimizing a cost function Q(x). The 2D hyperbolic estimate requires at least two TDOA measurements from three geographically distinct cells. Thus, the initial estimate requires at least P=2 high quality TDOA measurements corresponding to three geographically distinct cells.

[0098] Once the initial estimate is established, the (P+1)th TDOA measurement is included in r and a new Q(x) is formed and minimized. With the inclusion of a new cell, the updated location estimate is examined to determine if it would be beneficial to include the new cell. This can be done in a variety of ways. For example, the quantity

number

[0099] When the good cell selector 560 tests new candidate cells (i.e., TDOA measurements), it is advantageous to use cells in their ranked order of quality. It may also be beneficial to select a next trial cell that improves the estimated geometry (i.e., reduces the geometric dilution of precision (GDOP)). Given the estimated location of the current UE, the surrounding cells can be categorized in terms of circular sectors. Cells in underrepresented sectors can be prioritized for improved geometry. This circular sector method is similar to the selection of assisting data cells using circular sector assistance data (CSAD) described with reference to Figures 9-11. The estimated UE location is placed at the origin of the circular sector, and for example, six sectors are established around this origin. Assume that there are r cells in sectors 0 to 3, but there are no r cells in sectors 4 and 5. In this case, it may be beneficial to introduce the next cell candidates from the underrepresented sectors 4 and 5.

[0100] The good cell selector 560 continues to run until the desired number of cells are included in the position calculation. This stopping criterion may be established by a threshold on P, or once a desired GDOP level is reached, or

number

[0101] Improvement of Micro-BSA in response to measurements In one embodiment, the on-device micro-BSA 520 can be refined and pruned using feedback 532 from radio condition measurements and location estimation performed by module 530. For example, if a particular cell cannot be detected, it may be beneficial to remove the cell from the micro-BSA 520 to free up memory. Similarly, a cell may be further removed if it is consistently deemed "bad" in the bad cell detector 544.

[0102] BSA request optimization If the position estimator 540 is producing a reliable estimate, then as described above, no further BSA information is needed. Other methods for determining whether new BSA information is needed include:

[0103] Cell statistical data tracking (received power, timing measurements, etc.). If these have not changed significantly over time, the UE 510 can assume that the network configuration has not changed and therefore no BSA updates are required.

[0104] Cell Scanning. The UE 510 may scan for all cells (PCI, PRS ID). If a cell is detected that is not in the micro-BSA 520, this may trigger an update of the micro-BSA 520 where further information is requested from the BSA 512.

[0105] BSA updates in response to UE mobility. When the UE 510 detects high mobility, for example using a Doppler estimator, or using a high serving cell change rate, efficiency gains can be obtained by pausing BSA updates until a slower rate is achieved. This depends on the number of stored cells in the micro-BSA 520. For example, if 1000 cells are stored and the cell density is 1 cell per square kilometer, high mobility can be maintained in a serving area of ​​1000 square kilometers. However, if 100 cells are stored and the cell density is 10 cells per square kilometer, high mobility can result in outdated information in the micro-BSA 520. If a location update is required by an application, the micro-BSA 520 can be updated to provide positioning during times of high mobility. However, if an application is not requesting a location update, it may be advantageous to pause updates of the contents of the micro-BSA 520 until the UE 510 returns to a stationary state.

[0106] Reduced Latency For example, in existing positioning applications such as e911, a request is made regarding the location of the UE 510. Such a request traditionally triggers the complete process described in the Background section, i.e., assistance data delivery from the network to the UE, measurements on the UE, communication of measurements from the UE to the network, and finally, location estimation and delivery to the application. The delays involved in such traditional approaches can be tens of seconds.

[0107] The system reduces the time delay between application position request and position delivery. Since the UE 510 is location aware, delivery can be "instantaneous" on the order of tens of milliseconds if the application is hosted by the micro-BSA cloud assistance server 516 or otherwise cloud-based in the network 514. If the application is executed by the UE 410, the time required to transmit the latitude and longitude coordinates of the UE location in the uplink can be on the order of tens of nanoseconds. The location awareness of the UE 510 is possible with the system thanks to the UE-based positioning method. The intelligent handling of the micro-BSA 520 can eliminate the need for the UE 510 to interact with the micro-BSA cloud assistance server 516 or with other elements of the network 514 in the event of a position request. Thus, the UE 510 can periodically update its position estimate between position requests without interacting with the micro-BSA cloud assistance server 516 or with other elements of the network 514. This type of location awareness on the part of a mobile device between positioning requests to the network cannot be achieved by existing techniques, for example, the location of a fast moving conventional UE can change substantially between the times when location requests are made to the network.

[0108] When either a cloud-based application or an application running on the UE 510 requests a location, an "instant" location estimate is immediately made available by the system and delivered to the application. A timestamp may also be provided along with the estimate to signal the application when this latest location update occurred. For example, the location estimate may be updated hourly in the background. Consider the case where an administrator at a tool company wants to find the location of the company's generators. In this case, the breadcrumb application may inform the administrator that the device was at the work site 10 minutes ago. This may be a sufficient amount of information in this use case, and the device does not require further interaction with the network. Alternatively, the administrator may want to know the location of the device at the current time, and if so, the tracking device (e.g., a simplified implementation of the UE 510) may update the location estimate accordingly.

[0109] Geometry optimization and support data selection for micro-BSA Attention is now directed to FIGS. 9-11, to which reference is made for a description of geometry optimization and assistance data selection in micro-BSA according to the present disclosure. In one embodiment, the positioning assistance data calculation unit utilizes intelligence in selecting the micro-BSA information and assistance data 526. Cellular position estimates are subject to geometric dilution of precision (GDOP). Thus, in one implementation, the assistance data 526 is generated using a seed estimate of the location of the UE 510, as described above. However, simply selecting the N closest cells to the UE seed estimate may be a suboptimal strategy in terms of geometry. It may be advantageous to select cells that improve the GDOP and the overall stability of the position calculation.

[0110] FIG. 9 illustrates a method 900 of circular sector assistance data generation that aids in lowering GDOP according to the present disclosure. Using the assistance data seed as the center, six circular sectors 904 are formed with equal central angles of 60°. The assistance data 526 begins to be assembled in sectors 0 and (9040), rotating counterclockwise and successively adding one cell from each sector. In the example of FIG. 9, the first cell 9101 included in the assistance data is in sector 0 (9040). The second cell 9102 is in sector 2 (9042), even though a closer cell (9103) is in sector 0. The cell 9102 in sector 2 is preferred since it provides a more geometrically diverse set. After the cells of sector 2 are added, the third cell 9103 is in sector 0.

[0111] Turning now to FIG. 10, a screenshot capture of a map 1000 of a portion of New York City is shown. The map 1000 shows a UE's location 1002, which is a difficult location from a positioning perspective since there are no adjacent cell sites to the west and south of the UE. In this case, a suboptimal approach would simply add all information associated with the closest cells to the location 1002 to the assistance data seed. However, since these closest cells are only to the north and east of the location 1002, this approach does not provide geometric diversity. As a result, the GDOP associated with this approach is poor, and the resulting position estimation error 1004 is approximately 280 meters.

[0112] Figure 11 shows a screenshot capture of a map 1100 that corresponds to the same screenshot capture shown in Figure 10. In this case, however, the circular sector assistance data generation method of the present disclosure is used to select cells whose assistance data will be used in determining the UE's location 1102. The number of cells in the assistance data is the same as in the example of Figure 10, but the set of cells is more geometrically diverse and includes cells 1110 that are relatively far from the UE's location 1102, resulting in a lower GDOP. The resulting position estimation error 1114 is reduced from 280 meters to approximately 3 meters.

[0113] Further mitigating network congestion Current state-of-the-art LPP / SURL assistance data methods can add unnecessary congestion to the network. For example, in 3GPP Rel-14, the number of muting bits can be as high as 1024 per cell. The exchange of 1024 bits per cell may not be necessary since the number of unique bit sequences will be less than 2^1024. Unique sequences can be stored on the device and fewer bits can be transmitted over the network. In another Rel-13 example using 16-bit muting sequences, there can be only 70 unique sequences. These 70 sequences can be stored in a 70*16=1120-bit lookup table on the device, and 2^(ceil(log2(70)))=7 bits per cell would need to be accessed from the cloud server. This results in a savings of 9 bits (56%) in muting bit sequence download.

[0114] Another example is the 3GPP ECGI described in the LPP, which consists of MCC (mobile country code), MNC (mobile network code), and a 28-bit cell identity. MCC and MNC require 24 bits, but they can be common with the network operator. Thus, instead of using all 52 bits for the 3GPP ECGI, 28 bits for the cell identity can be used for a particular deployment, resulting in a reduction of 46%.

[0115] Data compression algorithms such as Lempel-Ziv can also be used to reduce network congestion. Complete micro-BSA bit sequences are concatenated so that the entropy of the sequence is likely to be less than 1, so the sequences can be concatenated and a compression algorithm used to compress the bit pattern.

[0116] Exemplary UE Implementation Attention is now directed to Figure 12, which includes a block diagram representation of a particular implementation of a UE 1200, in this case a mobile or cellular phone, configured in accordance with the present disclosure. It will be apparent that, although certain details and functions of the UE 1200 have been omitted for clarity, in various implementations, various additional features of a mobile device will be included as known. Furthermore, those skilled in the art will appreciate that the UE 1200 need not be implemented as a personal communications device, such as a mobile or cellular phone, and that in other implementations may include a tracking device or the like that lacks certain features and characteristics of the implementation of Figure 12.

[0117] 12, the UE 1200 includes a processor 1220 operably coupled to a touch-sensitive display 1204 configured to present a user interface 1208. In other embodiments, the user interface 1208 may include a physical keypad or keyboard, a voice input device, and / or any other device capable of receiving user input or instructions. The UE 1200 includes a memory 1240, which may be comprised of, for example, one or more of random access memory (RAM), read-only memory (ROM), flash memory, and / or any other medium that enables the processor 1220 to store and retrieve data. As shown, the memory 1240 stores the micro-BSA 520 and programs including programs or instructions executable by the processor 1220. These modules include a positioning assistance data calculation unit 524, a power and timing measurement module 530, a position estimator 540, and a micro-BSA AI management module 552.

[0118] The UE 1200 includes a wireless transceiver and modem 1224 for communication with a network, such as the network 514, which may include, for example, the Internet, and / or a wireless network, such as a cellular network, and / or other wired or wireless networks. The UE 1200 may further include a camera 1228 and other auxiliary modules.

[0119] Uncertainty Calculation for UE-Based Positioning A common approach to estimating the quality of a location estimate is to examine many position estimates and form a confidence ellipse. See, for example, Chew, V., "Confidence, Prediction, and Tolerance Regions for the Multivariate Normal Distribution," Journal of the American Statistical Association. 61, 315 (Sept. 1966) 605-617, and Owens, T. and McConville, D., "Geospatial Applications: Estimating the Spatial Accuracy of Coodinates Collected Using the Global Positioning System," Tech. rep., National Biological Service, Environmental Management Technical Center, Onalaska, Wisconsin (Apr. 1996). For example, FIG. 16 shows an example in which multiple (i.e., 50) position estimates (observations 1610) are used to form a confidence ellipse 1602.

[0120] 17, an exemplary error surface contour 1700 is shown that is generated to establish a confidence region, according to an embodiment.

number

[0121] The on-board micro-BSA positioning method described herein allows the confidence region (location estimation uncertainty) to be calculated on the UE 510, which has technical advantages over the state of the art UE-assisted location estimation. For example, performing multiple estimations to form a confidence ellipse can be done using the currently disclosed method by interacting with the UE 510 in receive-only mode without interacting with the network 514. This method can be done when the UE 510 is technically in RRC idle, eDRX, or PSM state from a data communication perspective. Similarly, the error surface contour method can be performed on the UE 510 using a single set of TDOA measurements, which advantageously does not require multiple observations as opposed to the method of forming a confidence ellipse.

[0122] Hybrid Power-Timing Based Positioning In the TDOA positioning method discussed above, timing measurements are used to estimate the location of the UE. As described herein, TDOA is a hyperbolic location method where each cell pair results in a hyperbola on a two-dimensional map. The intersection between multiple hyperbolae is the estimated location of the device. When the number of detected cells is low, the limited number of hyperbolae can increase the uncertainty in the estimation. When only two cells are detected, only a single hyperbola is drawn, greatly increasing the uncertainty in the estimated location of the UE.

[0123] In order to reduce the uncertainty of location estimates obtained using only timing measurements, a hybrid power-time positioning method is described in this section. It can be appreciated that the relationship between power measurements versus distance is usually less clear than that between time measurements versus distance. As a result, power-only positioning methods usually underperform timing-only measurement positioning methods as long as there are a sufficient number of detected cells. However, when only two cells are detected, the power-only method can outperform the time-only method, and the hybrid power-time method performs best.

[0124] UE, M p Assume that power measurements are made on surrounding transmitters. The received signal power measurement for the mth cell can be modeled as follows: r p,m =P TX,m +A m (θ m )-PL avg,m (||xx m ||)+X m

[0125] dBm, where m=1,2,...,M p , P TX,m is the mth transmitter power in dBm, and A m (θ m ) is the mth transmitter attenuation factor in dB, and PL avg,m (||xx m ||) is ||xx m is the average path loss in dB experienced by the mth transmitted signal traveling || meters, x is the unknown UE location, and x m is the known location of the mth transmitter, and X m is the standard deviation σ p,m is a zero-mean Gaussian random variable with

[0126] The usual path loss model is the so-called log-distance model, where

number

[0127] See, e.g., Rappaport, TS Wireless Communications: Principles and Practices. Prentice Hall, 2002.

[0128] PL avg,close(d0) is the near-field path loss in dB at some near field distance d0 meters, and n is the path loss exponent. m ~N(0,σ p,m ) characterizes the large-scale shadow fading between the transmitter and the receiver. The parameter set {PL avg,close ,d0,n,σ p,m} fully characterizes the model and can be fitted empirically to the acquired dataset.

[0129] Figure 18 shows the attenuation profiles associated with a commonly used attenuation model, see 3GPP TR 36.814: Table A.2.1.1-2, "Further advancements for E-UTRA physical layer aspects". This model

number

number

[0130] Next, the subscript “t” is introduced to incorporate the timing measurements for the power-timing hybrid method, where the mth TDOA measurement is r t,m =h t,m (x)+n t,m However, h t,m (x) is the ground truth TDOA component, n t,m is the TDOA noise component. Its vector form is r t =h t (x)+n t It is.

[0131] Since the power and TDOA measurements have different units (dBm for the power measurements and seconds for the timing measurements), it is unclear how exactly to combine the measurements, and has not been suggested by previous positioning techniques. To circumvent this problem, maximum likelihood methods such as the Gaussian maximum likelihood method are put forward.

[0132] In particular, the shadow fading term X m of

number

number

number

[0133] Similarly, n t,m of

number

number

number

[0134] Assuming statistical independence across power and time measurements, the joint conditional probability of all measurements is

number

number

number

number

[0135] While the use of a Gaussian Maximum Likelihood (GML) estimator to combine the power and TDOA measurements within the location estimator may be generally preferred, other implementations of the location estimator may be suitable for use in alternative embodiments. For example, since the GML estimator discussed above assumes that the TOA is derived from a Gaussian distribution, a generalization of this approach would be to assume that the TOA is derived from a Gaussian Mixture Model (GMM). As noted above, the GMM framework better accounts for the multipath nature of cellular radio frequency (RF) environments. Alternatively, the location estimator may be implemented as an ML estimator, such as a Maximum A Posteriori (MAP) estimator. In yet other embodiments, the power and TDOA measurements may be modeled as non-Gaussian impulsive noise.

[0136] Turning attention to Figures 19A-19C, Figures 19A-19C show simulation results illustrating that positioning using the power-time hybrid approach disclosed herein can be advantageous in an environment with a limited number of cells. Figure 19A shows the results of estimating a UE's position in an environment where the UE receives signals from two cells using only power. Figure 19B shows the results of estimating a UE's position in the same two cell environment as Figure 19A using only time. Figure 19C shows the results of estimating a UE's position in the same two cell environment using the power-time hybrid approach. As can be seen by comparing the errors associated with the estimated positions in Figures 19A-19C, the power-time hybrid approach illustrates the relative benefits to the power-only and time-only methods in this example having two cells.

[0137] 20, a block diagram diagram of a first embodiment of a power-time hybrid positioning system 2000 in accordance with the present disclosure is provided. The system 2000 includes a UE 2010 configured to generate positioning assistance data based on a relatively small subset of a base station almanac (BSA) 2012 accessible from a BSA server 2011 in a network 2014. The BSA 2012 may be on a server in communication with the BSA server 2011 or may be included on the BSA server 2011. In one embodiment, information from the BSA 2012 is provided to the UE 2010 by a micro-BSA cloud assistance server 2016 in the network 2014.

[0138] As shown, the UE 2010 includes a memory 2040 and a processor 2020 operatively coupled to a wireless transceiver and modem 2024. The memory 2040 may be comprised of, for example, one or more of a random access memory (RAM), a read only memory (ROM), a flash memory, and / or any other medium that enables the processor 2020 to store or retrieve data. As shown, the memory 2040 stores a micro-BSA 2022 and programs including programs or instructions executable by the processor 2020. These modules include a positioning assistance data calculation unit 2026, a power and timing measurement module 2030, a power-time hybrid position estimator 2042, and a micro-BSA AI management module 2052. The power-time hybrid position estimator 2042 may be configured to implement the power-time hybrid positioning techniques described herein.

[0139] It will be apparent that, although certain details and functions of the UE 2010 have been omitted for clarity, in various implementations various additional features of a mobile device (e.g., a display, user interface elements, etc.) will be included as known. Further, while the UE 2010 may be implemented as a mobile or personal communication device such as a cellular phone, in other implementations the UE 2010 may include a tracking device or the like that lacks certain features and characteristics of such personal communication devices.

[0140] The one or more micro-BSAs 2022 are configured to store information corresponding to a subset of the BSAs 2012 received by the UE 2010 from the micro-BSA cloud assistance server 2016. The micro-BSAs 2022 may be calculated using information of the UE 2010 serving cell ECGI to provide a rough estimate of the UE location. The positioning assistance data calculation unit 2026 is configured to receive cell parameters from the micro-BSAs 2022 for use in generating positioning assistance data. The positioning assistance data calculation unit 2026 may use the serving cell ECGI to obtain a rough estimate of the UE location from which to calculate the positioning assistance data, which is provided to the power and timing measurements module 2030 and the power-time hybrid position estimator 2042.

[0141] As shown, the power-time hybrid position estimator 2042 may include a bad cell detector 2044 configured to improve the quality of the position estimate produced by the estimator 2042 by identifying cells that cause relatively high errors in the residual vector. The power-time hybrid position estimator 2042 may alternatively include a good cell selector (GCS) 2060 instead of the bad cell detector 2044. As discussed above, the good cell selector 2060 may first rank the estimated TOAs of the surrounding cells in terms of quality and select the TOA with the highest quality as the RSTD reference cell. The bad cell detector 2044 or the good cell selector 2060 may be useful in situations where the UE 2010 is within range of a significant number of different cells. In other cases (e.g., when the UE 2010 is within only two or three cells), the bad cell detector 2044 or the good cell selector 2060 may not be utilized.

[0142] FIG. 21 provides a block diagram diagram of a second embodiment of a power-time hybrid positioning system 2100 according to the present disclosure. As can be seen by comparing FIGS. 20 and 21, the power-time hybrid positioning system 2100 is substantially similar to the system 2000, except that the memory 2140 of the UE 2110 in the system 2100 includes a TDOA position estimator 2142. In one embodiment, the TDOA position estimator 2142 is implemented substantially similarly or identically to the position estimator 540. As previously mentioned, the embodiment of the power-time hybrid positioning system is particularly useful when the UE is within range of a relatively small number of cells. To the extent that the UE 2110 moves to a location within range of a relatively larger number of cells, it may be true that the positioning uncertainty is minimized by utilizing the TDOA position estimator 2142 rather than the power-time hybrid position estimator 2042 to generate a position estimate.

[0143] 22, which illustrates a third embodiment of a power-time hybrid positioning system 2200 in accordance with the present disclosure. The system 2200 includes a UE 2210 in communication with a server arrangement. The server arrangement includes a BSA server 2211 in a network 2214. The BSA 2212 may be on a server in communication with the BSA server 2011 or may be included on the BSA server 2011. In one embodiment, information from the BSA 2212 is provided to the UE 2210 by a location assistance server 2250 in the network 2214.

[0144] As can be seen by reference to Figures 20 and 22, the UE 2210 is similar to the UE 2010 of Figure 20, but lacks the micro-BSA and micro-BSA AI management modules. The memory 2240 of the UE 2210 includes a positioning assistance data calculation unit 2026, a power and timing measurement module 2030, and a power-time hybrid position estimator 2042. In the embodiment of Figure 22, cell parameters from the BSA 2212 are provided to the UE 2210 by a location server 2250 on an as-needed or scheduled basis. The positioning assistance data calculation unit 2026 derives positioning assistance data from the received cell parameters to facilitate timing measurements made by the power and timing measurement module 2030 and position estimates generated by the power-time hybrid position estimator 2042. The absence of a micro-BSA in the UE 2210 may tend to increase network traffic and power consumption as a result of the continuous need to obtain cell parameters from the location assistance server 2250, but the server 2250 may not be configured to populate the micro-BSA or the UE may not be configured with such a micro-BSA. However, in these situations, the UE 2210 may nevertheless be configured with a power-time hybrid position estimator 2042 that can operate in the absence of a micro-BSA or other locally stored repository of cell parameters.

[0145] FIG. 23 illustrates a fourth embodiment of a power-time hybrid positioning system 2300 according to the present disclosure. As can be understood by comparing FIG. 22 and FIG. 23, the systems 2200 and 2300 are substantially similar, except that the memory 2340 of the UE 2310 in the system 2300 further includes a TDOA position estimator 2340. As mentioned above, the power-time hybrid position estimator 2042 may be particularly useful in an environment where the UE 2310 is within range of a relatively small number of cells. To the extent that the UE 2310 moves to a location within range of a relatively larger number of cells, it may be true that positioning uncertainty is minimized by utilizing the TDOA position estimator 2342 rather than the power-time hybrid position estimator 2042 to generate a position estimate.

[0146] Where the above method indicates that certain events are performed in a certain order, the ordering of certain events may be modified. Furthermore, some of the events may be performed simultaneously in a parallel process where possible, as well as sequentially as described above. Although various modules in different devices are shown to be located within a processor of a device, they may also be located / stored within the memory of the device (e.g., software modules) and accessible and executable by the processor. Accordingly, this specification is intended to cover all such various modifications and variations of the disclosed embodiments that are within the spirit and scope of the appended claims.

[0147] The above description, for purposes of explanation, uses specific terminology to achieve a thorough understanding of the claimed systems and methods. However, it will be apparent to one skilled in the art that specific details are not required to practice the systems and methods described herein. Thus, the above descriptions of specific embodiments of the described systems and methods have been presented for purposes of illustration and description. They are not intended to be exhaustive, nor are they intended to limit the scope of the claims to the precise forms disclosed. Obviously, many modifications and variations are possible in light of the above teachings. The embodiments were chosen and described in order to best explain the principles of the described systems and methods and their practical applications, so that others skilled in the art can best utilize the described systems and methods and various embodiments with various modifications as appropriate for the particular use contemplated. The following claims and their equivalents are intended to define the scope of the systems and methods described herein.

[0148] The various methods or processes outlined herein may be coded as software executable on one or more processors employing any one of a variety of operating systems or platforms. Moreover, such software may be written using any suitable programming language and / or programming or scripting tools, and may be compiled as executable machine language code or intermediate code that runs on a framework or virtual machine.

[0149] Examples of computer code include, but are not limited to, microcode or microinstructions, machine instructions such as produced by a compiler, code used to create a web service, and files containing high-level instructions executed by a computer using an interpreter. For example, embodiments may be implemented using a formal programming language (e.g., C, Fortran, etc.), a functional programming language (Haskell, Erlang, etc.), a logical programming language (e.g., Prolog), an object-oriented programming language (e.g., Java, C++, etc.), or other suitable programming language and / or development tools. Further examples of computer code include, but are not limited to, control signals, encryption code, and compression code.

[0150] In this regard, the various inventive concepts may be embodied as a computer-readable storage medium (or multiple computer-readable storage media) (e.g., a computer memory, one or more floppy disks, compact disks, optical disks, magnetic tapes, flash memories, circuitry in field programmable gate arrays or other semiconductor devices, or other non-transitory or tangible computer storage media) encoded with one or more programs that, when executed on one or more computers or other processors, perform methods for implementing the various embodiments of the invention as discussed above. The computer-readable medium may be portable such that the programs stored thereon can be loaded into one or more computers or other processors to implement various aspects of the invention as discussed above.

[0151] The terms "program" or "software" are used herein in a general sense to refer to any type of computer code or set of computer-executable instructions that can be employed to program a computer or other processor to perform various aspects of the embodiments discussed above. Furthermore, it should be understood that, according to one aspect, one or more computer programs that, when executed, perform the methods of the present invention need not reside on a single computer or processor, but may be distributed in a modular manner among several different computers or processors to perform various aspects of the present invention.

[0152] Computer-executable instructions may be in many forms, such as program modules, executed by one or more computers or other devices. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The exemplary functionality of the program modules may be combined or distributed as desired in various embodiments.

[0153] Further, the data structures may be stored in the computer-readable medium in any suitable form. For ease of explanation, the data structures may be depicted as having fields that are related through locations within the data structure. Such relationships may also be achieved by assigning storage for the fields locations within the computer-readable medium that convey the relationships between the fields. However, any suitable mechanism may be used to establish relationships between information within fields of the data structures, including through the use of pointers, tags, or other mechanisms that establish relationships between data elements.

[0154] Moreover, various inventive concepts may be embodied as one or more of the example methods provided. The acts performed as part of a method may be ordered in any suitable manner. Thus, embodiments may be constructed in which acts are performed in an order different from that illustrated, and may include performing some acts simultaneously even though in the illustrated embodiment they are shown as sequential acts.

[0155] All definitions defined and used herein should be understood to control over dictionary definitions, definitions in documents incorporated by reference, and / or ordinary meanings of the defined terms.

[0156] The indefinite articles "a" and "an," as used in this specification and the claims, unless clearly indicated to the contrary, should be understood to mean "at least one."

[0157] The phrase "and / or" as used herein and in the claims should be understood to mean "either" or "both" of the elements so conjoined, i.e., elements that are conjunctive in some cases and disjunctive in other cases. Multiple elements listed with "and / or" should be interpreted in the same manner, i.e., to mean "on or more" of the elements so conjoined. Other elements (whether related to the elements explicitly identified) not the elements explicitly identified by the "and / or" clause may optionally be present. Thus, as a non-limiting example, when used in conjunction with open-ended phraseology such as "comprising," a reference to "A and / or B" can refer in one embodiment to only A (optionally including elements other than B), in another embodiment to only B (optionally including elements other than A), in yet another embodiment to both A and B (optionally including other elements), etc.

[0158] As used herein and in the claims, "or" should be understood to have the same meaning as "and / or" as defined above. For example, when items are separated in a list, "or" or "and / or" shall be interpreted as inclusive, i.e., including not only the inclusion of at least one of an element or list of elements, and optionally additional unlisted items, but two or more of them. Only terms clearly indicated to the contrary, such as "only one of" or "exactly one of" or "consisting of" when used in the claims, shall refer to the inclusion of exactly one element of an element or list of elements. In general, the term "or" as used herein shall only be interpreted to indicate exclusive alternatives (i.e., "one or the other but not both") when preceded by terms of exclusivity, such as "either", "one of", "only one of", "exactly one of", etc. As used in the claims, "consisting essentially of" shall have its ordinary meaning as used in the field of patent law.

[0159] As used herein and in the claims, the phrase "at least one" in connection with a list of one or more elements should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each and every element specifically listed in the list of elements, nor excluding any combination of elements in the list of elements. This definition further allows for the optional presence of elements (whether related to an explicitly identified element or not) that are not explicitly identified elements in the list of elements to which the phrase "at least one" refers. Thus, as a non-limiting example, "at least one of A and B" (or, in other words, "at least one of A or B" or, in other words, "at least one of A and / or B") can refer to, in one embodiment, at least one A that includes two or more As, optionally with no B (optionally including elements that are not B); in another embodiment, at least one B that includes two or more Bs, optionally with no A (optionally including elements that are not A); and, in yet another embodiment, at least one A that optionally includes two or more As and at least one B that optionally includes two or more Bs (and optionally includes other elements), etc.

[0160] As herein above, as in the claims, all transitional phrases such as "comprising," "including," "carrying," "having," "containing," "involving," "holding," and "composed of" are to be understood to be open-ended, i.e., to mean without limitation what is inclusive. As set forth in the United States Patent Office Manual of Patent Examining Procedure, Section 2111.03, only the transitional phrases "consisting of" and "consisting essentially of" shall be closed or semi-closed transitional phrases, respectively.

Claims

1. 1. A method for estimating a location of a mobile device, the method comprising: measuring received signal power of signals received from a plurality of cells; measuring a plurality of times of arrival of a plurality of signals received from the plurality of cells; determining a plurality of time difference of arrival (TDOA) measurements based on the plurality of times of arrival; generating a maximum likelihood estimate of the location of the mobile device by evaluating a joint conditional probability of the received signal power measurements and the plurality of TDOA measurements using a hybrid power-time Gaussian maximum likelihood estimator and positioning assistance data for the plurality of cells; Including, A method comprising:

2. A method as described in claim 1, characterized in that the received signal power measurement value is represented by a first Gaussian random variable and the multiple TDOA measurement values ​​are represented by a second Gaussian random variable.

3. 2. The method of claim 1, wherein the received signal power measurements are calculated using a vector r ｐ , i.e. r ｐ =h ｐ (x)+n ｐ is included in, except for h ｐ，ｍ (x) is the average received signal power from the mth cell of the plurality of cells, and n ｐ，ｍ =X ｍ is a measured power error term for the mth cell of the plurality of cells, and the method further comprises: ｐ，ｍ But, dispersion [0010] , average h ｐ，ｍ (x), and the conditional probability density function [0025] m=1, 2, . .. .. , Mp be the first Gaussian random variable with ｍ of [0030] The method of claim 1, further comprising modeling the

4. 4. The method of claim 3, wherein the plurality of TDOA measurements are represented by a vector r ｔ , i.e. r ｔ =h ｔ (x)+n ｔ is included in, except for h ｔ (x) is the TDOA ground truth vector, ｔ is the TDOA noise component, and the method comprises: ｔ，ｍ But, dispersion [0045] , average h ｔ，ｍ (x), and the conditional probability density function [0050] m=1, 2, . .. .. , Mt. to be the second Gaussian random variable with ｔ，ｍ of [006] The method of claim 1, further comprising modeling the

5. 2. The method of claim 1 , receiving first cell parameters for a first plurality of cells from a Base Station Almanac (BSA) accessible from a network server, the first plurality of cells forming a subset of the first plurality of cells; storing the first cell parameters as a first micro-BSA in a memory of the mobile device; deriving said positioning assistance data from said subset of said first cell parameters; The method further comprising:

6. 6. The method of claim 5, deriving initial positioning assistance data from an initial subset of first cell parameters associated with an initial subset of the first plurality of cells, wherein one or more cells included in the subset of the first plurality of cells are not included in the initial subset of the first plurality of cells; measuring received signal power of an initial plurality of signals received from an initial subset of the first plurality of cells; measuring initial times of arrival of signals received from the initial subset of the first plurality of cells; determining an initial plurality of TDOA measurements from the initial plurality of times of arrival; generating an initial maximum likelihood estimate of the location of the mobile device by evaluating a joint conditional probability of the initial plurality of received signal power measurements and the initial plurality of TDOA measurements using a power-time hybrid Gaussian maximum likelihood estimator and the initial positioning assistance data; Further comprising: A method comprising:

7. 7. The method of claim 6, wherein the initial positioning assistance data is generated by the first micro-BSA based at least in part on an initial seed estimate.

8. 6. The method of claim 5, predicting a route along which the mobile device is expected to travel; receiving second cell parameters for a second plurality of cells determined to be proximate to the expected route from the BSA accessible to the network server, the second plurality of cells including at least one cell not included in the first plurality of cells; storing the second cell parameters as a second micro-BSA in the memory of the mobile device; The method further comprising:

9. 9. The method of claim 8, measuring received signal power of signals received from a subset of the second plurality of cells; measuring second times of arrival of signals received from the subset of the second plurality of cells; determining a second plurality of TDOA measurements based on the second times of arrival; generating a maximum likelihood estimate of a second location of the mobile device by evaluating a joint conditional probability of the second received signal power measurement and the plurality of second TDOA measurements using the power-time hybrid Gaussian maximum likelihood estimator and second positioning assistance data for the subset of the second plurality of cells derived from the second cell parameters; The method further comprising:

10. 10. The method of claim 8, wherein the predicting comprises detecting a Doppler shift in frequency of a signal received by the mobile device.

11. 10. The method of claim 8, wherein the predicting comprises determining that the mobile device is traveling on a road and extracting route information from a map that includes the road.

12. 10. The method of claim 8, further comprising selecting the second plurality of cells based on pattern recognition information regarding cells proximate to the expected route previously used by the mobile device for position estimation.

13. 2. The method of claim 1, wherein the plurality of TDOA measurements are contained within a first TDOA measurement vector, the method comprising: determining a quality of the maximum likelihood estimate of the location of the mobile device by at least calculating a first TDOA residual vector using the maximum likelihood estimate of the location of the mobile device; detecting a faulty cell included in the plurality of cells based on the first TDOA residual vector; constructing a second TDOA measurement vector by removing the TDOA measurements associated with the bad cells from the first TDOA measurement vector; generating an updated maximum likelihood estimate of the location of the mobile device using the hybrid power-time Gaussian maximum likelihood estimator and the second TDOA measurement vector; determining a quality of the updated maximum likelihood estimate of the location of the mobile device based on the second TDOA measurement vector; The method further comprising:

14. 14. The method of claim 13, wherein determining the quality of the updated maximum likelihood estimate includes calculating a second TDOA residual vector using the updated maximum likelihood estimate of the location of the mobile device, the method further comprising detecting additional bad cells by evaluating the second TDOA residual vector.

15. 14. The method of claim 13, wherein determining the quality of the maximum likelihood estimate of the location of the mobile device comprises determining whether at least a minimum number of cells are included in the first TDOA measurement vector.

16. 14. The method of claim 13, wherein determining the quality of the maximum likelihood estimate of the location of the mobile device includes determining a first geometric dilution of precision (GDOP) between the maximum likelihood estimate of the location of the mobile device and a location of the cell in the first TDOA measurement vector.

17. 17. The method of claim 16, further comprising determining that the first GDOP exceeds a threshold.

18. The method of claim 17, further comprising indicating an error condition, determining the location of the mobile device using an alternate positioning method, or any combination thereof.

19. 6. The method of claim 5, further comprising storing second cell parameters for a second plurality of cells as a second micro-BSA in the memory of the mobile device, the second plurality of cells including at least one cell not included in the first plurality of cells.

20. 6. The method of claim 5, further comprising determining an improved estimate of the location of the mobile device based at least in part on TDOA measurements associated with an additional subset of the first plurality of cells and improved positioning assistance data corresponding to the additional subset of the first plurality of cells.

21. 21. The method of claim 20, further comprising determining the additional position estimate based on further refined positioning assistance data in response to a previous one of the additional position estimates for the mobile device.

22. 22. The method of claim 21, further comprising determining a quality of the additional location estimate by evaluating a contour line in a location estimation cost function.

23. 23. The method of claim 22, further comprising requesting updated positioning assistance data from at least one of the micro-BSA or the network server based on the quality falling below a threshold.

24. 1. A mobile device comprising: one or more processors; and a memory in communication with the one or more processors, the memory storing instructions that, when executed by the one or more processors, cause the mobile device to: measuring received signal power of signals received from a first plurality of cells; measuring a first plurality of times of arrival of a first plurality of signals received from the first plurality of cells; determining a first plurality of time difference of arrival (TDOA) measurements based on the first plurality of times of arrival; generating a maximum likelihood estimate of a location of the mobile device by evaluating a joint conditional probability of the received signal power measurements and the first plurality of TDOA measurements using a power-time hybrid Gaussian maximum likelihood estimator and first positioning assistance data for the first plurality of cells; To carry out A mobile device comprising:

25. The mobile device of claim 24, wherein the received signal power measurement is represented by a first Gaussian random variable and the plurality of TDOA measurements are represented by a second Gaussian random variable.

26. 25. The mobile device of claim 24, wherein execution of the instructions causes the mobile device to: measuring a second plurality of times of arrival of a second plurality of signals received from a second plurality of cells; determining a second plurality of TDOA measurements based on the second plurality of times of arrival; generating an updated estimate of the location of the mobile device using second positioning assistance data for the second plurality of cells and the second plurality of TDOA measurements; The mobile device further comprises:

27. 25. The mobile device of claim 24, wherein execution of the instructions causes the mobile device to: storing cell parameters for the first plurality of cells and the second plurality of cells in a micro-BSA in the memory of the mobile device; deriving the first positioning assistance data and the second positioning assistance data based on the stored cell parameters; The mobile device further comprises:

28. 1. A method for estimating a location of a mobile device, the method comprising: measuring received signal power of signals received from a plurality of cells; measuring a plurality of times of arrival of a plurality of signals received from the plurality of cells; determining a plurality of time difference of arrival (TDOA) measurements based on the plurality of times of arrival; generating an estimate of the location of the mobile device based at least in part on a maximum likelihood estimator using the plurality of TDOA measurements, the received signal power measurements, and positioning assistance data for the plurality of cells; The method according to claim 1, further comprising:

29. The method of claim 28, wherein the maximum likelihood estimator is based on at least the plurality of TDOA measurements being derived from a Gaussian mixture model.

30. A method for estimating a location of a mobile device, the method being performed by the mobile device, comprising: measuring received signal power of signals received from a plurality of cells; measuring a plurality of times of arrival of a plurality of signals received from the plurality of cells; determining a plurality of time difference of arrival (TDOA) measurements based on the plurality of times of arrival; using the plurality of TDOA measurements, the received signal power measurements, and positioning assistance data for the plurality of cells to generate an estimate of the location of the mobile device based at least in part on modeling the plurality of TDOA measurements and the received signal power measurements as non-Gaussian impulsive noise; The method according to claim 1, further comprising:

Images