Angle of arrival positioning system for object tracking

Abstract

A system and method are described for determining a user equipment (UE) location within a wireless network using a reference signal of the wireless network. The disclosed system and method utilizes multiple in-phase and quadrature (I / Q) samples generated from signals provided by receive channels associated with two or more antennas of a wireless system. Based on received reference signal parameters, a reference signal within a range of signals from each receive channel among the receive channels is identified. Based on the identified reference signal from each receive channel, an angle of arrival between a baseline of the two or more antennas and incident energy from a UE onto the two or more antennas is determined. The angle of arrival is then used to calculate the location of the UE. The angle of arrival may be a horizontal angle of arrival and / or a vertical angle of arrival.

Description

[Technical field]

[0001] SUMMARY OF THE DISCLOSURE

[0002] Embodiments of the present invention relate to wireless communication and network systems and systems for radio frequency (RF) based identification, tracking and object location, including RTLS (Real Time Location Services).

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit under 35 U.S.C. §119(e) of U.S. Provisional Patent Application No. 62 / 239,195, filed October 8, 2015, entitled “ANGLE OF ARRIVAL (AOA) POSITIONING METHOD AND SYSTEM FOR POSITIONAL FINDING AND TRACKING OBJECTS USING REDUCED ATTENUATION RF TECHNOLOGY”, and U.S. Provisional Patent Application No. 62 / 290,087, filed February 2, 2016, entitled “ANGLE OF ARRIVAL (AOA) POSITIONING METHOD AND SYSTEM FOR POSITIONAL FINDING AND TRACKING OBJECTS USING REDUCED ATTENUATION RF TECHNOLOGY”, each of which is incorporated herein by reference in its entirety.

[0003] This application is a related application of U.S. Patent Application No. 13 / 566,993, filed August 3, 2012, entitled "MULTI-PATH MITIGATION IN RANGEFINDING AND TRACKING OBJECTS USING REDUCED ATTENUATION RF TECHNOLOGY," which is a continuation-in-part of U.S. Patent Application No. 13 / 109,904, filed May 17, 2011, issued November 6, 2012, entitled "MULTI-PATH MITIGATION IN RANGEFINDING AND TRACKING OBJECTS USING REDUCED ATTENUATION RF TECHNOLOGY," now U.S. Patent No. 8,305,215, filed January 18, 2011, entitled "METHODS AND SYSTEM FOR MULTI-PATH MITIGATION IN No. 7,969,311 entitled "METHODS AND SYSTEM FOR REDUCED ATTENUATION IN TRACKING OBJECTS USING RF TECHNOLOGY," which is a continuation-in-part of U.S. patent application Ser. No. 12 / 502,809, filed Jul. 14, 2009, issued Jan. 18, 2011, now U.S. Patent No. 7,872,583 entitled "METHODS AND SYSTEM FOR REDUCED ATTENUATION IN TRACKING OBJECTS USING RF TECHNOLOGY," which is a continuation-in-part of U.S. patent application Ser. No. 11 / 610,595, filed Dec. 14, 2006, issued Jul. 14, 2009, now U.S. Patent No. 7,561,211 entitled "METHODS AND SYSTEM FOR REDUCED ATTENUATION IN TRACKING OBJECTS USING RF TECHNOLOGY," which is a continuation-in-part of U.S. patent application Ser. No. 11 / 610,595, filed Dec. 14, 2006, issued Jul. 14, 2009, now U.S. Patent No. 7,872,583 entitled "METHODS AND SYSTEM FOR REDUCED ATTENUATION IN TRACKING OBJECTS USING RF TECHNOLOGY," which is a continuation-in-part of U.S. patent application Ser.This application is a continuation of U.S. Provisional Patent Application No. 60 / 597,649, filed December 15, 2005, entitled "METHOD AND SYSTEM FOR REDUCED ATTENUATION IN TRACKING OBJECTS USING MULTI-BAND RF TECHNOLOGY," which claims the benefit under 35 U.S.C. §119(e) of U.S. Provisional Patent Application No. 60 / 597,649, filed December 15, 2005, and which is hereby incorporated by reference in its entirety.

[0004] U.S. patent application Ser. No. 12 / 502,809, filed Jul. 14, 2009, now U.S. Patent No. 7,872,583, issued Jan. 18, 2011, entitled "METHODS AND SYSTEM FOR REDUCED ATTENUATION IN TRACKING OBJECTS USING RF TECHNOLOGY," also claims the benefit under 35 U.S.C. §119(e) of U.S. Provisional Patent Application Ser. No. 61 / 103,270, filed Oct. 7, 2008, entitled "METHODS AND SYSTEM FOR MULTI-PATH MITIGATION IN TRACKING OBJECTS USING REDUCED ATTENUATION RF TECHNOLOGY," which are incorporated herein by reference in their entirety. [Background technology]

[0005] RF-based identification and location-finding systems for determining the relative or geographical location of an object are typically used for tracking a single object or a group of objects, and for tracking an individual. Conventional location-finding systems have been used for location determination in open outdoor environments. RF-based Global Positioning System (GPS) and its complement GPS are commonly used. However, conventional location-finding systems suffer from certain inaccuracies when locating objects in closed (i.e., indoor) environments as well as outdoors. Although cellular wireless communication systems provide excellent data coverage in urban and most indoor environments, the location accuracy of these systems is limited by self-interference, multipath and non-line-of-sight propagation.

[0006] Indoor and outdoor location inaccuracies are primarily due to the physics of RF propagation, in particular RF signal loss / attenuation, signal scattering and reflection. The loss / attenuation and scattering problems can be solved by using narrowband ranging signal(s) and operating at low RF frequencies, for example in the VHF range or below (see related patent document 1). [Prior art documents] [Patent documents]

[0007] [Patent Document 1] U.S. Patent No. 7,561,048 [Patent Document 2] U.S. Patent Application Publication No. 2011 / 0124347A1 ("Method and Apparatus for UE Positioning in LTE Networks") [Patent Document 3] U.S. Pat. No. 6,788,199 [Patent Document 4] U.S. Patent No. 7,872,583 [Patent Document 5] U.S. Pat. No. 5,525,967 [Patent Document 6] U.S. Patent Application Publication No. 2008 / 0285505 ("SYSTEM AND METHOD FOR NETWORK TIMING RECOVERY IN COMMUNICATIONS NETWORKS") [Patent Document 7] International Publication No. 2010 / 104436 ("METHOD ANDARRANGEMENT FOR DL-OTDOA (DOWNLINK OBSERVED TIME DIFFERENCE OF ARRIVAL) POSITIONING IN A LTE (LONG TERM EVOLUTION) WIRELESS COMMUNICATIONS SYSTEM") [Patent Document 8] US Patent Application Publication No. 2003 / 008156, US Patent No. 7,167,456 ("APPARATUS FOR ESTIMATING PROPAGATION PATH CHARACTERISTICS") [Patent Document 9] US Patent Application Publication No. 2010 / 0091826 [Patent Document 10] U.S. Pat. No. 7,969,311 [Patent Document 11] U.S. Patent No. 8,305,215 [Non-patent literature]

[0008] [Non-Patent Document 1] S. Salous, "Indoor and Outdoor UHF Measurements with a 90 MHz Bandwidth", IEEE Colloquium on Propagation Characteristics and Related System Techniques for Beyond Line-of-Sight Radio, pp. 8 / 1-8 / 6, 1997 [Non-Patent Document 2] Polaris Wireless, "Polaris Wireless Location technology overview", July 29 [Non-Patent Document 3] Tsung-Han Lin, et al. "Microscopic Examination of an RSSI-Signature-Based Indoor Localization System" [Non-Patent Document 4] Ray H. Hashemi, William G. Bradley, "MRI: the basics", 2003 [Non-Patent Document 5] Dan Dobkin, "Indoor Propagation and Wavelength", WJ Communications, V 1.4 V / 10 / 02 [Non-Patent Document 6] Kei Sakaguchi et al., "Influence of the Model Order Estimation Error in the ESPRIT Based High Resolution Techniques" [Non-Patent Document 7] Jakub Marek Borkowski, "Performance of Cell ID+RTT Hybrid Positioning Method for UMTS" [Non-Patent Document 8] Jim Zyren, "Overview of the 3GPP Long Term Evolution Physical Layer, white paper" Summary of the Invention [Problem to be solved by the invention]

[0009] Although multipath phenomena (e.g., RF energy reflection) are less severe at VHF and lower frequencies than at UHF and higher frequencies, the impact of multipath phenomena on location finding makes location determination less reliable and accurate than required by the industry. Thus, there is a need for methods and systems for mitigating the effects of RF energy reflection (i.e., multipath phenomena) on RF-based identification and location-finding systems that use narrowband ranging signal(s).

[0010] Generally, conventional RF-based identification and location finding systems mitigate multipath by using wide bandwidth ranging signals, for example, by taking advantage of wideband signal properties for multipath mitigation (see Non-Patent Document 1). Also see Patent Document 2 by Chen et al., where location accuracy versus required PRS bandwidth is shown in Table 1. From this table, for an accuracy of 10 meters, a bandwidth of 83 MHz is required. In addition, spatial diversity and / or antenna diversity techniques are sometimes used.

[0011] However, spatial diversity may not be an option for many location tracking applications because it leads to an increased infrastructure requirement. Similarly, antenna diversity is of limited value because at lower operating frequencies, e.g., VHF, the physical dimensions of the antenna subsystem become too large. A case in point is US Pat. No. 5,399,433, which describes a system and method for location detection of objects, people, pets and personal items.

[0012] The proposed system uses an antenna array to mitigate multipath. Optionally, the system operates in UHF, 902-926 MHz band. It is well known that the linear dimension of an antenna is proportional to the wavelength of the operating frequency. Also, the area of ​​an antenna array is proportional to the square of the linear dimension ratio, and the volume is proportional to the cube of the linear dimension ratio, since in an antenna array, the antennas are usually separated by 1 / 4 or 1 / 2 wavelength. Thus, at frequencies below VHF, the size of the antenna array significantly impacts the portability of the device.

[0013] On the other hand, due to the very limited frequency spectrum, narrow bandwidth ranging signals are not suitable for multipath mitigation techniques currently used by conventional RF-based identification and location-finding systems, because the ranging signal distortion (i.e., signal change) induced by multipath is too small for reliable detection / processing in the presence of noise. Also, due to the limited bandwidth, narrow bandwidth receivers cannot distinguish between the ranging signal direct line-of-sight (DLOS) path and the delayed ranging signal path when they are separated by a small delay, because narrow bandwidth receivers lack the required time resolution proportional to the receiver bandwidth (e.g., narrow bandwidth has an integration effect on the input signal).

[0014] Therefore, there is a need in the art for a multipath mitigation method and system for object identification and location finding that operates at VHF and below frequencies as well as UHF band frequencies and above using narrow bandwidth ranging signal(s).

[0015] The need for tracking and location capabilities is found primarily in wireless networks. The multipath mitigation method and system for object identification and location finding described in related U.S. Pat. No. 6,399,433 can be utilized in most available wireless networks. However, certain wireless networks have communication standards / systems that require integration of the technology into the wireless network to fully benefit from the various ranging and positioning signals described in related U.S. Pat. No. 6,399,433. Generally, these wireless systems can provide excellent data coverage over wide areas and most indoor environments. However, the available position accuracy of these systems is limited by auto-interference, multipath and non-line-of-sight propagation. As an example, the recent 3GPP Release 9 standardized positioning technology for the Long Term Evolution (LTE) standard has the following: 1) A-GNSS (Complementary Global Navigation Satellite System) or A-GPS (Complementary Global Positioning System) as a first method, and 2) OTDOA (Observed Time Difference of Arrival), including Enhanced Cell ID (ECID) and DL-OTDOA (Downlink OTDOA) as fallback methods. While these methods may meet the current mandatory FCC E911 emergency location requirements, the accuracy, reliability and availability of these location methods fall short of the needs of LBS (Location Based Services) or RTLS system users who require highly accurate location within buildings, shopping malls, city corridors, etc. Furthermore, the upcoming FCC 911 requirements may be more stringent than the existing ones and may exceed the positioning capabilities of existing techniques / methods with the exclusion of A-GNSS (A-GPS). It is well known that the accuracy of A-GNSS (A-GPS) is very good in open space but very unreliable in urban / indoor environments.

[0016] At the same time, the accuracy of other techniques / methods is severely affected by the effects of multipath and other radio propagation phenomena, thereby making it impossible to meet certain regulatory requirements such as FCC 911 and LBS requirements. Listed below are additions to DL-OTDOA and ECID location techniques / methods. The U-TDOA concept is similar to OTDOA, but uses a Location Measurement Unit (LMU) installed on radio towers to calculate the phone's location. It is (was) designed for the original 911 requirements. LMUs have only been deployed in 2G GSM networks and require a major hardware upgrade for 3G UMTS networks. U-TDOA is not standardized for support in 4G LTE or WiMAX. Also, LMUs are not used in LTE deployments. Like other methods, U-TDOA accuracy is affected by multipath. The LTE standardization group may abandon the LMU-additional hardware and model U-TDOA after DL-OTDOA, for example, with UL-OTDOA. Of note, DL-OTDOA is standardized in Release 9.

[0017] Another contender for the upcoming FCC 911 requirements is RF fingerprinting(s). This technique is based on the principle that every location has a unique radio frequency (RF) signature, like a fingerprint pattern, and a location can be identified by a unique set of values, including measurements such as neighboring cell signal strength. Fingerprinting does not require additional hardware. However, this technique suffers from the fact that it requires a large database and a long training phase. Unlike human fingerprints, which are also truly unique, RF signatures are repeated at several different locations due to RF propagation phenomena. Furthermore, the database becomes stale, for example signatures become outdated quickly as the environment, including the weather, changes. This makes the task of maintaining the database burdensome. The number of radio towers that can be heard has a significant impact on the accuracy, and to obtain a reasonable accuracy (60 meters, as required by Polaris Wireless), readings must be obtained from a large number of towers (eight or more). Thus, in suburban environments, the accuracy drops to 100 meters (see Non-Patent Document 2). Also, there is a significant variation (up to 140%) in the estimated position depending on the handset antenna orientation (see Non-Patent Document 3).

[0018] There are several causes of RF fingerprint database instability, but one of the major ones is multipath. Multipath is very dynamic and can change the RF signature instantly. Specifically, in heavy multipath environments such as indoors, movement of people and elevators, changes in the location of furniture, cabinets, equipment, etc. result in different multipath distributions, severely impacting the RF signature. Also, in indoor and similar environments, small changes in physical location (in three dimensions) cause significant changes in the RF signature. This is a result of the combination of multipath, which makes the RF signature short wavelength in three dimensions, resulting in significant RF signature changes over quarter wave distances. Therefore, in such environments, the points based data must be exponentially multiplied.

[0019] Less accurate location methods exist, e.g. RTT, RTT+CID, including those based on received signal strength. However, in the latter case, RF propagation phenomena cause signal strength to vary by 30 dB to 40 dB over the distance of a wavelength, which in wireless networks can be significantly shorter than one meter. This severely impacts the accuracy and / or reliability of methods based on received signal strength. Also, the accuracy of all these methods is subject to multipath.

[0020] Therefore, there is a need for techniques for more accurate and reliable tracking and location capabilities for wireless networks, which can be achieved through multipath mitigation techniques.

[0021] Positioning reference signals (PRS) were added in LTE 3GPP Release 9 and are meant to be used by user equipment (UE) for OTDOA positioning (a type of multilateration). The TS36.211 Release 9 technical specification is titled "Evolved Universal Terrestrial Radio Access (E-UTRA); Physical Channels and Modulation".

[0022] As mentioned above, the PRS can be used by the UE for Downlink Observed Time Difference of Arrival (DL-OTDOA) positioning. The Release 9 specifications also require neighboring base stations (eNBs) to be synchronized. This removes the last barrier for OTDOA methods. The PRS also improves UE hearability for UEs across multiple eNBs. It should be noted that the Release 9 specifications did not specify eNB synchronization accuracy, with some proposals advocating 100ns. UL-TDOA is currently in the study phase and is expected to be standardized in 2011.

[0023] The DL-OTDOA method according to the Release 9 specification is detailed in US Pat. No. 5,399,363 by Chen et al. Release 9 DL-OTDOA suffers from multipath phenomena. Some multipath mitigation can be achieved by increased PRS signal bandwidth. However, this results in increased scheduling complexity and longer times between UE position fixes. In addition, for a network with limited operating bandwidth, e.g., 10 MHz, the best possible accuracy is about 100 meters as shown in Table 1 of Chen et al. These figures are the results in the best case scenario. In other cases, it results in significantly larger (2 to 4 times) location / ranging errors, especially when the DLOS signal strength is significantly lower (10 to 20 dB) compared to the reflected signal(s) strength.

[0024] describe a variant of UL-TDOA positioning that is also PRS-based, called Uplink Positioning Reference Signal (UL-PRS). Although Chen et al. propose improving neighbor cell hearability and / or reducing scheduling complexity, Chen et al. do not teach addressing multipath mitigation. As a result, the accuracy of Chen et al. is comparable to the Release 9 accuracy of the DL-OTDOA method.

[0025] According to Chen et al., DL-OTDOA and UL-TDOA methods are suitable for outdoor environments. Chen et al. further state that DL-OTDOA and UL-TDOA methods do not work well in indoor environments, such as buildings, campuses, etc. Several reasons are mentioned by Chen et al. that explain the poor performance of these methods in indoor environments. For example, in distributed antenna systems (DAS) commonly used indoors, each antenna does not have a unique ID.

[0026] According to Chen, the end result is that in both Release 9 and tower-based systems such as Chen et al.'s UL-TDOA, the UE device is unable to distinguish between multiple antennas. This phenomenon precludes the use of the multilateration method used in Release 9 and Chen et al.'s UL-OTDOA systems. To solve this problem, Chen et al. add hardware and new network signals to the existing indoor wireless network system. Furthermore, for an operational DAS, the best accuracy (error floor) is limited to 50 meters. Finally, Chen et al. do not address the impact of multipath on positioning accuracy in indoor environments, which are the most challenging environments (compared to outdoors), and often result in positioning errors that are significantly larger (by a factor of 2-4) than claimed.

[0027] The modifications for indoor wireless networks taught by Chen et al. are not always possible because upgrading existing systems requires a great deal of effort and high cost. Furthermore, for an operational DAS, the best theoretical accuracy is only 50 meters, and in practice this accuracy is much lower due to RF propagation phenomena including multipath. At the same time, in a DAS system, signals produced by multiple antennas appear as reflections, e.g., multipath. Thus, if all antenna positions are known, and if the signal paths from the individual antennas can be resolved, e.g., using multilateration and position alignment algorithms, it is possible to provide a position fix in a DAS environment without additional hardware and / or new network signals. Thus, there is a need for techniques for accurate and reliable multipath resolution for wireless networks. [Means for solving the problem]

[0028] The present invention relates to a method and system for radio frequency (RF) based identification, tracking and location of objects, including real-time location services (RTLS), that substantially obviates one or more of the disadvantages of the related art. The proposed (exemplary) method and system uses narrow bandwidth ranging and location signal(s). According to the embodiment, the RF based tracking and location is performed in the VHF band, but can also be performed in lower bands (HF, LF and VLF) as well as in the UHF band and above. It uses multipath mitigation methods including techniques and algorithms. The proposed system can use software-based digital signal processing and software defined radio techniques. Digital signal processing can be used as well.

[0029] The system of the embodiment can be built using standard FPGAs and standard signal processing hardware and software with very low incremental cost to the device and the overall system. At the same time, the accuracy of RF-based identification and location finding systems using narrowband ranging signals / second can be significantly improved.

[0030] A signal (e.g., VHF) transmitter and receiver for narrow bandwidth ranging / location signals are used to identify the location of a person or object. Digital signal processing (DSP) and software defined radio (SDR) technology can be used to generate, receive, and process the narrow bandwidth ranging signal(s) and to execute multipath mitigation algorithms. The narrow bandwidth ranging signals are used to identify, locate, and track people or objects in half-duplex, full-duplex, or simplex communication operating modes. Digital signal processing (DSP) and software defined radio (SDR) technology are used in a multipath mitigation processor to implement the multipath mitigation algorithms.

[0031] The method described herein uses a multipath mitigation processor and multipath mitigation techniques / algorithms to increase the accuracy of tracking and location systems performed by wireless networks, as described in related U.S. Patent Application Publication No. 2009 / 0133991. The embodiments of the present invention may be used in all wireless systems / networks and may include unidirectional, half-duplex and full-duplex modes of operation. The embodiments described below operate in wireless networks using various modulation formats, including OFDM modulation and / or its derivatives. Thus, the embodiments described below operate in LTE networks, and may be applied to other wireless systems / networks.

[0032] The methods described herein are based on one or more reference / pilot signal(s) and / or synchronization signals of the network and are also applicable to other wireless networks including WiMax, WiFi and white space. Other wireless networks that do not use reference and / or pilot / synchronization signals can use one or more of the following types of alternative embodiments as described in related U.S. Pat. No. 6,399,436: 1) a portion of a frame is dedicated to ranging signals / ranging signal elements as described in related U.S. Pat. No. 6,399,436; 2) ranging signal elements (see related U.S. Pat. No. 6,399,436) are embedded in the transmit / receive signal frame(s); and 3) ranging signal elements (described in related U.S. Pat. No. 6,399,436) are embedded with data.

[0033] These alternative embodiments may be used in all operational modes, simplex, half-duplex and full-duplex, using the multi-path mitigation processor and multi-path mitigation techniques / algorithms described in related US Pat. No. 6,399,633. Effect of the Invention

[0034] The multipath mitigation processor and multipath mitigation techniques / algorithms described in related U.S. Patent No. 6,399,436 can be incorporated into OFDM-based wireless networks and other wireless networks having reference / pilot and / or synchronization signals with little or no incremental cost to the devices and the overall system. At the same time, the location accuracy of the network and system is significantly improved. As described in this embodiment, RF-based tracking and location detection implemented in 3GPP LTE cellular networks will greatly benefit from the localization of the multipath mitigation methods / techniques and algorithms described in related U.S. Patent No. 6,399,436. The proposed system can use digital signal processing performed by software or hardware.

[0035] Additional features and advantages of the embodiments will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the embodiments. The advantages of the embodiments will be realized and attained by the structure particularly pointed out in the written description and claims of the invention as well as the appended drawings.

[0036] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are intended to provide further explanation of the embodiments as claimed.

[0037] The accompanying drawings, which are included to provide a further understanding of the embodiments and are incorporated in and constitute a part of this specification, illustrate the embodiments and, together with the description, serve to explain the principles of the embodiments. [Brief description of the drawings]

[0038] [Figure 1] FIG. 1 illustrates narrow bandwidth ranging signal frequency components according to an embodiment. [Figure 1A] FIG. 1A illustrates narrow bandwidth ranging signal frequency components according to an embodiment. [Diagram 2] FIG. 2 shows exemplary wide bandwidth ranging signal frequency components. [Figure 3A] FIG. 3A shows a block diagram of a master and slave unit of an RF mobile tracking and locating system according to an embodiment. [Figure 3B] FIG. 3B shows a block diagram of the master and slave units of the RF mobile tracking and locating system, according to an embodiment. [Figure 3C] FIG. 3C shows a block diagram of the master and slave units of the RF mobile tracking and locating system, according to an embodiment. [Figure 4] FIG. 4 shows an example composite wideband base-band ranging signal. [Diagram 5] FIG. 5 illustrates the elimination of a signal precursor by cancellation, according to an embodiment. [Figure 6] FIG. 6 illustrates precursor cancellation with fewer carriers according to an embodiment. [Figure 7] FIG. 7 shows the transfer function phase in one direction. [Figure 8] FIG. 8 illustrates a position detection method according to an embodiment. [Figure 9] FIG. 9 shows the LTE reference signal mapping. [Figure 10] FIG. 10 illustrates an exemplary enhanced Cell ID+RTT location technique. [Figure 11] FIG. 11 illustrates an exemplary OTDOA location technique. [Figure 12] FIG. 12 illustrates the operation of a Time Monitoring Unit (TMO) installed in an operator's eNB functionality. [Figure 13] FIG. 13 illustrates a diagram of an exemplary wireless network location device. [Figure 14] FIG. 14 illustrates an exemplary wireless network location downlink ecosystem for enterprise applications. [Figure 15] FIG. 15 illustrates an exemplary wireless network location downlink ecosystem for network-spanning applications. [Figure 16] FIG. 16 illustrates an exemplary wireless network location uplink ecosystem for enterprise applications. [Figure 17] FIG. 17 illustrates an exemplary wireless network location uplink ecosystem for network-wide applications. [Figure 18] FIG. 18 illustrates an example wireless network DAS and / or femto / small cell uplink location environment. [Figure 19] FIG. 19 illustrates an exemplary wireless network tower location environment. [Figure 20] FIG. 20 illustrates an exemplary radio tower of a wireless network. [Figure 21] FIG. 21 conceptually illustrates arrival angle determination based on phase difference. [Figure 22] FIG. 22 illustrates a process flow for the implementation of an embodiment. [Diagram 23] FIG. 23 is a perspective view of a group unit in one embodiment. [Figure 24] FIG. 24 shows the process flow performed by the group unit in one embodiment. [Diagram 25] FIG. 25 illustrates a process flow for the implementation of one embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0039] Reference will now be made in detail to the preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

[0040] An embodiment of the present invention relates to a method and system for RF-based object identification, tracking and location, including RTLS. According to one embodiment, the method and system uses narrow bandwidth ranging signals. This embodiment operates in the VHF band, but can also be used in the HF, LF and VLF bands, as well as the UHF band and higher frequencies. It uses a multipath mitigation processor. The use of the multipath mitigation processor improves the accuracy of tracking and location performed by the system.

[0041] The present embodiment includes a small, highly portable base unit that allows a user to track, locate and monitor multiple people and objects. Each unit has its own ID. Each unit broadcasts an RF signal with its ID, and each unit can send back a return signal that can include its ID as well as voice, data and additional information. Each unit processes the returned signals from the other units to continuously determine their relative and / or actual positions depending on triangulation or trilateration and / or other methods used. The preferred embodiment can also be easily integrated into products such as GPS devices, smartphones, two-way radios and PDAs. The resulting product has all of the functionality of a standalone device while leveraging the existing displays, sensors (e.g., altimeter, GPS, accelerometer and compass) and processing power of its host. For example, a GPS device with the device technology described herein can present the user's position on a map as well as map the positions of other members of the group.

[0042] The size of the preferred embodiment based on FPGA implementation is between approximately 5.1 x 10.2 x 2.5 cm and 5.1 x 5.1 x 1.3 cm, becoming smaller as integrated circuit technology improves. Depending on the frequency used, the antenna is either integrated into the device or protrudes from the device housing. An ASIC (Application Specific Integrated Circuit) based version of the device can incorporate the functionality of the FPGA and most other electronic components into the unit or tag. ASIC based standalone versions of the product result in device sizes of 2.5 x 1.3 x 1.3 cm or less. The antenna size is determined by the frequency used, and part of the antenna can be integrated into the housing. The ASIC based embodiment is designed to be integrated into a product that can consist of only the chipset. There should not be any substantial physical size difference between Master Units or Tag Units.

[0043] The device can use standard system components (off-the-shelf components) operating in multiple frequency ranges (bands) for the processing of multipath mitigation algorithms. It can use software for digital signal processing and software-defined radio. The signal processing software combined with minimal hardware makes it possible to build radios with transmit and receive waveforms defined by the software.

[0044] Related US Patent No. 5,999,333 discloses a narrow bandwidth ranging signal system whereby the narrow bandwidth ranging signal is designed to fit into a low bandwidth channel, for example using a voice channel which is only a few kHz wide (although some of the low bandwidth channels can reach tens of kHz), in contrast to conventional location finding systems which use channels that are hundreds of kHz to tens of MHz wide.

[0045] The advantages of this narrow bandwidth ranging signal scheme are as follows: 1) At low operating frequencies / bands, the ranging signal bandwidth of conventional location-finding systems exceeds the carrier (operating) frequency value. Thus, such systems cannot be deployed in other low frequency bands, including LF / VLF and HF. Unlike conventional location-finding systems, the narrow bandwidth ranging signal scheme described in the related patent document 1 can be successfully deployed in LF, VLF and other bands, since its ranging signal bandwidth is much lower than the carrier frequency value. 2) For example, at the lower end of the RF spectrum (several VLF, LF, HF and VHF bands) up to the UHF band, conventional location-finding systems cannot be used because the FCC strictly limits the allowable bandwidth (12-25 kHz), and therefore conventional ranging signals cannot be used. Unlike conventional location-finding systems, the ranging signal bandwidth of the narrow bandwidth ranging signal scheme is fully compliant with FCC rules and other international spectrum regulatory bodies. 3) It is well known that narrow bandwidth signals have inherently higher SNR (signal to noise ratio) compared to wide bandwidth signals, independent of the operating frequency / band (see Non-Patent Document 4). This increases the operating range of a narrow bandwidth ranging signal location finding system, independent of the frequency / band it operates at, including the UHF band.

[0046] Thus, unlike conventional location-finding systems, the narrow-bandwidth ranging signal location-finding system can be deployed at the lower end of the RF spectrum, for example, from the VHF and lower frequency bands to the LF / VLF bands where multipath phenomena are less pronounced. At the same time, the narrow-bandwidth ranging signal location-finding system can also be deployed in the UHF bands and above, improving the ranging signal SNR and thus increasing the location-finding system operating range.

[0047] To minimize multipath, e.g., RF energy reflection, it is desirable to operate in the VLF / LF bands. However, at these frequencies, the efficiency of portable / mobile antennas is very small (approximately 0.1% or less due to the small antenna length (size) relative to the RF wavelength). In addition, at these low frequencies, the noise level from natural and man-made sources is much higher than at higher frequencies / bands, e.g., VHF. Together, these two phenomena can limit the applicability of a location-finding system, e.g., its operating range and / or mobility / portability. Thus, for certain applications where operating range and / or mobility / portability are very important, higher RF frequencies / bands, e.g., HF, VHF, UHF and UWB, can be used.

[0048] In the VHF and UHF bands, noise levels from natural and man-made sources are significantly lower compared to the VLF, LF and HF bands, and in the VHF and HF frequencies, multipath phenomena (e.g., RF energy reflection) are not as severe as in the UHF and higher frequencies. Also, in the VHF band, antenna efficiency is significantly better than in the HF and lower frequencies, and in the VHF band, RF penetration capability is much better than in the UHF band. Thus, the VHF band offers a good compromise for mobile / portable applications. On the other hand, in some special cases, such as GPS, where VHF frequencies (or lower frequencies) cannot penetrate the ionosphere (or are deflected / refracted), UHF can be a good option. However, in either case (and in all cases / applications), narrow bandwidth ranging signal systems have advantages over conventional wide bandwidth ranging signal location finding systems.

[0049] The actual application(s) will determine the exact technical specifications (e.g. power, radiation, bandwidth and operating frequency / band). Narrow-bandwidth ranging allows users to either obtain a license or obtain an exemption from license, or, since narrow-bandwidth ranging allows operation over many different bandwidths / frequencies, including the most stringent narrow-bandwidths, 6.25kHz, 11.25kHz, 12.5kHz, 25kHz and 50kHz as specified by the FCC, users can use unlicensed bands as specified by the FCC and comply with the corresponding technical requirements of the appropriate section. As a result, multiple FCC sections and exemptions within such sections can apply. The main applicable FCC regulations are 47 CFR Part 90- Private Land Mobile Radio Services, 47 CFR Part 94 personal Radio Services, 47 CFR Part 15 - Radio Frequency Devices. (By comparison, wideband signals in this context are from a few hundred KHz to 10-20 MHz.)

[0050] Typically, for Part 90 and Part 94, VHF implementations allow users to operate devices up to 100mW under certain exemptions (Low Power Radio Service is an example). For certain applications, the allowable transmit power in the VHF band is between 2 and 5 watts. For 900MHz (UHF band), it is 1W. For frequencies between 160kHz and 190kHz (LF band), the allowable transmit power is 1 watt.

[0051] It complies with many, if not all, of the different spectrum allowances and allows for accurate ranging while complying with the most stringent regulatory requirements, not only for the FCC but also for other international agencies that regulate spectrum use around the world, including Europe, Japan, Korea, etc.

[0052] Below is a list of common frequencies in use along with typical power usage and the distance over which a tag can communicate with another reader in a real-world environment (see Non-Patent Document 5):

[0053] 915MHz 100mW 45.7m 2.4GHz 100mW 30.5m 5.6GHz 100mW 22.9m

[0054] The proposed system operates at VHF frequencies and uses proprietary methods of transmitting and processing RF signals. More specifically, the system uses DSP technology and software-defined radio (SDR) to overcome the limitations of narrow bandwidth requirements at VHF frequencies.

[0055] Operating at a lower (VHF) frequency reduces scattering and provides significantly better wall penetration. The net result is an approximately 10x increase in range over commonly used frequencies. For example, compare the measured range of the prototype to that of the RFID technology mentioned above.

[0056] 216MHz 100mW 213.4m

[0057] Utilizing narrowband ranging techniques, the range of commonly used frequencies in typical power usage and the distance over which a tag communication range can communicate with another reader in a real world environment is significantly increased.

[0058] 915MHz 100mW 45.7m to 152.4m 2.4GHz 100mW 30.5m to 137.2m 5.6GHz 100mW 22.9m to 121.9m

[0059] Battery consumption is a function of the device design, transmission power and duty cycle, e.g., the time interval between two successive distance (position) measurements. In many applications the duty cycle is large, 10x to 1000x. In applications with a large, e.g., 100x, duty cycle, an FPGA version transmitting 100mW of power has an up time of nearly 3 weeks. An ASIC-based version is expected to increase the up time by a factor of 10. ASICs also have an inherently lower noise level. Thus, an ASIC-based version can also increase the operating range by 40%.

[0060] Those skilled in the art will appreciate that the present embodiment significantly improves location-finding accuracy in RF challenging environments (eg, buildings, urban corridors, etc.) while not compromising the long operating range of the system.

[0061] Typically, tracking and location systems use Track-Locate-Navigate methods. These methods include Time-Of-Arrival (TOA), Differential-Time-Of-Arrival (DTOA), and a combination of TOA and DTOA. Time-Of-Arrival (TOA) as a distance measurement technique is generally described in U.S. Pat. No. 5,399,633. TOA / DTOA based systems measure the time-of-flight, e.g., time delay, of an RF ranging signal Direct Line-of-Sight (DLOS), which is then converted to distance range.

[0062] In case of RF reflection (e.g., multipath), multiple copies of the RF ranging signal with various delay times are superimposed on the DLOS RF ranging signal. A tracking and locating system using a narrow bandwidth ranging signal cannot distinguish between the DLOS signal and the reflected signal without multipath mitigation. As a result, these reflected signals cause errors in the DLOS time-of-flight of the estimated ranging signal, which in turn affects the range estimation accuracy.

[0063] The present embodiment advantageously uses a multipath mitigation processor to separate the DLOS signal and the reflected signal. Thus, the present embodiment significantly reduces the error in the estimated ranging signal DLOS time of flight. The proposed multipath mitigation method can be used for all RF bands. The method can also be used with wide bandwidth ranging signal location finding systems. And the method can support various modulation / demodulation techniques including spread spectrum techniques such as DSS (Direct Spread Spectrum) and FH (Frequency Hopping).

[0064] In addition, noise reduction methods can be applied to further improve the accuracy of the method. These noise reduction methods can include, but are not limited to, coherent summation, non-coherent summation, matched filtering, time diversity techniques, etc. Residual multipath interference errors can be further reduced by applying post-processing techniques such as maximum likelihood estimation (e.g., Viterbi algorithm), minimum variance estimation (Kalman filter), etc.

[0065] This embodiment can be used in systems with simplex, half-duplex and full-duplex modes of operation. Full-duplex operation is very demanding in terms of RF transceiver complexity, cost and logistics, which limits the system operating range in portable / mobile device implementations. In half-duplex mode of operation, the reader (often called "master") and tags (sometimes also called "slave" or "target") are controlled by a protocol that only allows the master or slave to transmit at a given time.

[0066] The alternating transmission and reception allows a single frequency to be used in the distance measurement. Such a configuration reduces system cost and complexity compared to full duplex. The simplex mode of operation is conceptually simpler, but requires tighter synchronization of events between the master and target units, including the initiation of the ranging signal sequence.

[0067] In an embodiment of the present invention, the multipath mitigation processor of the narrow bandwidth ranging signal does not increase the ranging signal bandwidth. It advantageously uses different frequency components to allow the propagation of the narrow bandwidth ranging signal. Further ranging signal processing can be performed in the frequency domain by using super-resolution spectrum estimation algorithms (MUSIC, rootMUSIC, ESPRIT) and / or statistical algorithms such as RELAX, or in the time domain by assembling a synthetic ranging signal with a relatively large bandwidth and applying further processing to this signal. The different frequency components of the narrow bandwidth ranging signal can be selected pseudo-randomly, which may also be continuous or spaced in frequency, with uniform and / or non-uniform spacing in frequency.

[0068] This embodiment extends the multipath mitigation technique. The signal model for narrowband ranging is a complex exponential function (introduced elsewhere in this specification) whose frequency is directly proportional to the delay defined by the range plus a similar term that includes the delay defined by the time delay associated with the multipath. The model is independent of the actual implementation of the signal structure, e.g., stepped frequency, linear frequency modulation, etc.

[0069] The frequency separation between the direct path and multipath is nominally very small, and normal frequency domain processing is not sufficient to estimate the direct path range. For example, at a range of 30 meters, a stepped frequency ranging signal with a 100 KHz step rate over 5 MHz (100.07 nanosecond delay) results in a frequency of 0.062875 rad / sec. Multipath reflections at a path length of 35 meters result in a frequency of 0.073355. The separation is 0.0104792. The frequency resolution of an observable of 50 samples has an inherent frequency resolution of 0.12566 Hz. Therefore, it is not possible to use conventional frequency estimation techniques to separate the direct path from the reflected paths and accurately estimate the direct path range.

[0070] To overcome this limitation, the present embodiment uses a unique combination of a subspace decomposition high-resolution spectral estimation methodology and an implementation of multimodal cluster analysis. The subspace decomposition technique relies on decomposing the estimated covariance matrix of the observed data into two orthogonal subspaces, a noise subspace and a signal subspace. The theory behind the subspace decomposition methodology is that the projection of an observable onto the noise subspace consists of only noise, and the projection of an observable onto the signal subspace consists of only signal.

[0071] The super-resolution spectrum estimation algorithm and the RELAX algorithm are able to distinguish closely spaced frequencies (sinusoids) in the spectrum in the presence of noise. The frequencies do not need to have harmonic relationships and, unlike the Digital Fourier Transform (DFT), the signal model does not introduce any artificial periodicity. For a given bandwidth, these algorithms provide a significantly higher resolution than the Fourier Transform. Thus, direct line of sight (DLOS) can be reliably distinguished from other multipaths (MP) with high accuracy. Similarly, applying the thresholding method described below to an artificially generated synthetic wider bandwidth ranging signal allows for reliably distinguishing DLOS from other paths with high accuracy.

[0072] According to this embodiment, digital signal processing (DSP) is used by the multi-path mitigation processor to reliably distinguish DLOS from other MP paths. Various super-resolution algorithms / techniques exist in the spectrum analysis (spectrum estimation) techniques. Examples include subspace-based methods, such as the Multiple Signal Characterization (MUSIC) or root-MUSIC algorithms, the Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) algorithm, the Pisarenko Harmonic Decomposition (PHD) algorithm, and the RELAX algorithm.

[0073] In all of the above mentioned super-resolution algorithms, the incoming (i.e. received) signal is modeled as a linear combination of complex exponential functions of frequency and their complex amplitude. In the multipath case, the received signal is:

[0074]

number

[0075] Where:

number

number

[0076] Parameter α K and τ K are random time-varying functions reflecting the movement of people and devices in and around the building. However, since their rate of change is very slow compared to the measurement time interval, these parameters can be considered as time-invariant random variables within a certain measurement period.

[0077] All these parameters are frequency dependent since they are related to radio signal properties such as transmission and reflection coefficients. However, in this embodiment, the operating frequency hardly changes. Therefore, the above mentioned parameters can be assumed to be frequency independent.

[0078] Equation (1) can be expressed in the frequency domain as follows:

[0079]

number

[0080] where A(f) is the complex amplitude of the received signal, (2π×τ K ) is the artificial "frequency" estimated by the super-resolution algorithm, the operating frequency f is the independent variable, and α K is the Kth path amplitude.

[0081] In equation (2), (2π×τ K ) and a super-resolution estimate of τ K The values ​​are based on continuous frequencies. In reality, there are a finite number of measurements. Thus, the variable f is not a continuous variable, but rather a discrete variable. Therefore, the complex amplitude A(f) can be calculated as follows:

[0082]

number

[0083] Where:

number

[0084] In formula (3),

number

[0085] In some cases, for example, real signal data, e.g.

number

[0086] These low "frequencies" present problems for the implementation of the Hilbert transform (or other methods). In addition, the amplitude values ​​(e.g.

number

[0087] The following shows the complex amplitude in the presence of multipath:

number

[0088] In half-duplex mode of operation, the reader (often called "master") and the tag (also called "slave" or "target") are controlled by a protocol in which only the master or the slave can transmit at a given time. In this mode of operation, the tag (target device) serves as a transponder. It receives a ranging signal from the reader (master device), stores it in memory, and then retransmits the signal to the master after a certain time (delay).

[0089] An example of a ranging signal is shown in Figure 1 and Figure 1A. The exemplary ranging signal uses consecutive and different frequency components. Other waveforms including pseudo-randomly spaced or orthogonal in frequency and / or time can also be used as long as the ranging signal bandwidth remains narrow. In Figure 1, the duration TF of all frequency components is long enough to obtain the narrow bandwidth characteristic of the ranging signal.

[0090] Another variation of the ranging signal with different frequency components is shown in FIG. 2. This variation consists of multiple frequencies (f 1 , f 2 , f 3 , f 4 , f n) to make the individual frequencies narrowband. Although such a signal is more efficient, it occupies a large bandwidth and the wide bandwidth ranging signal impacts the SNR, thereby reducing the operating range. Also, such a wide bandwidth ranging signal goes against the FCC requirements for the VHF band and below. However, in some applications, this wide bandwidth ranging signal can be more easily integrated into existing signal and transmission protocols. Also, such a signal reduces the track-locate time.

[0091] These multiple frequencies (f 1 , f 2 , f 3 , f 4 , f n ) The bursts may also be contiguous and / or pseudo-randomly spaced in frequency and / or time, orthogonal, etc.

[0092] The narrow-bandwidth ranging mode produces accuracy in the form of instantaneous wideband ranging while increasing the range over which this accuracy can be achieved compared to wideband ranging. This performance is achieved because, at a fixed transmit power, the SNR of the narrow-bandwidth ranging signal (in the appropriate signal bandwidth) at the receiver is greater than the SNR of the wideband ranging signal at the receiver. This SNR gain is on the order of the ratio of the total bandwidth of the wideband ranging signal to the bandwidth of each channel of the narrowband ranging signal. This provides a good tradeoff when very fast ranging is not required, for example, for stationary targets or targets moving slowly, such as a person walking or running.

[0093] The master device and the tag device are identical and can operate in either master or transponder mode. All devices include a data / remote control communication channel. These devices can exchange information and the master device can remotely control the tag device. In this example shown in Figure 1, while the operation of the multipath mitigation processor of the master (i.e., reader) emits a ranging signal to the tag, and after a certain delay, the master / reader receives a repeated ranging signal from the tag.

[0094] The master's multipath mitigation processor then compares the received ranging signal with that originally transmitted by the master and calculates the amplitude and phase for all frequency components FN.

number

number

[0095]

number

[0096] For example, the matched filter

number

number

number

number

number

number

[0097] In this embodiment, the ranging base-band signal is the same as that shown in Figure 1. However, for simplicity, in this specification, the ranging base-band signal consists of only two frequency components, each of which has a different frequency: F 1 and F 2 It includes multiple periods of a cosine wave or sine wave of F 1 =f 1 , F 2 =f 2 Note that the number of periods of the first frequency component is L and the number of periods of the second frequency component is P. fNote that for F = constant, L may or may not be equal to P, since each frequency component can have a different number of periods. Also, there is no time difference between each frequency component, and F 1 , F 2 Both start with an initial phase of zero.

[0098] 3A, 3B and 3C show block diagrams of a master or slave unit (tag) of an RF mobile tracking and locating system. OSC refers to the frequency of the device's system clock (crystal oscillator 20 in FIG. 3A). All frequencies generated within the device are generated from this system clock crystal oscillator. The following definitions are used: M is the master device (unit); AM is the tag (target) device (unit). The tag device is operating in transponder mode and is referred to as the transponder (AM) unit.

[0099] In a preferred embodiment, the device consists of an RF front-end, an RF back-end, a baseband, and a multi-path mitigation processor. The RF back-end, baseband, and multi-path mitigation processor are implemented in an FPGA 150 (see Figures 3B and 3C). The system clock generator 20 (see Figure 3A) is OSC = 20MHz, or ω OSC = 2π × 20 × 106. This is the ideal frequency because in real devices, the system clock frequency is not always equal to 20 MHz.

[0100]

number

[0101] In addition,

number

[0102] F other than 20MHz OSC Note that the frequencies can be used without any impact on system performance.

[0103] The electronic configuration of both units (master and tag) is identical and different operation modes are programmable by software. The base-band ranging signal is generated in digital form by the FPGA 150 blocks 155-180 (see Fig. 2B) of the master. It consists of two frequency components, each containing multiple periods of a cosine or sine wave of different frequency. First, at t=0, the FPGA 150 of the master device (Fig. 3B) outputs the digital base-band ranging signal to the up-converter 50 via the I / Q DACs 120 and 125. The FPGA 150 outputs the F 1 Start at frequency and time T 1 After that, T 2 During the period 2 Start generating frequencies.

[0104] Since the crystal oscillator frequency may differ from 20MHz, the actual frequency generated by the FPGA is F 1 Gamma M and F 2 Gamma M In addition, time T 1 is T 1 β M and time T 2 is T 2 β M Also, F 1 Gamma M *T 1 β M =F 1 T 1 and F 1 Gamma M *T 2 β M =F 2 T 2 So that, T 1 , T 2 , F 1 , F 2 is set, where F 1 T 1 And, F 2 T 2Both are integers. This means that F 1 and F 2 This means that the initial phase of is equal to zero.

[0105] Since all frequencies are generated from the system crystal oscillator 20 clock, the output of the master baseband I / Q DAC(s) 120 and 125 is:

[0106]

number

number

number

number

[0107] The master (M) and transponder (AM) operate in half-duplex mode. The master's RF front-end uses a quadrature up-converter (i.e. mixer) 50 to up-convert the base-band ranging signal generated by the multi-path mitigation processor and transmits the up-converted signal. After the base-band signal is transmitted, the master switches from TX to RF mode using the RF front-end's TX / RX switch 15. The transponder receives the signal and down-converts it using the RF front-end's mixer 85 (to generate a first IF) and ADC 140 (to generate a second IF).

[0108] This second IF signal is then digitally filtered in the transponder's RF back-end processor using digital filter 190 and further down-converted to a base-band ranging signal using RF back-end quadrature mixer 200, digital I / Q filters 210 and 230, digital quadrature oscillator 220, and summer 270. This base-band ranging signal is stored in the transponder's memory 170 using RAM data bus controller 195 and control logic 180.

[0109] The transponder then switches from RX to TX mode using the RF front-end switch 15 and waits a certain delay t RTX After that, it starts retransmitting the stored baseband signal. Note that the delay is measured in the AM (transponder) system clock. Therefore,

number

[0110] The master then uses the arctangent block 250 and the phase comparison block 255 of the multipath mitigation processor to calculate F 1 and F 2 The amplitude value is derived from the RF back-end RSSI block 240.

[0111] To improve the estimation accuracy, it is always desirable to improve the SNR of the amplitude estimate from block 240 and the SNR of the phase difference estimate from block 255. In a preferred embodiment, the multipath mitigation processor derives the SNR of the frequency components of the ranging signal duration T fCalculate amplitude and phase difference estimates for many time instances over a period of time. When averaged, these values ​​improve the SNR. The improvement in SNR is

number

[0112] Another approach to improving the SNR is to determine the amplitude and phase difference values ​​by applying matched filter techniques over a period of time. However, another approach is to sample the frequency components of the received (i.e., recovered) base-band ranging signal and match the frequency components of the original (i.e., transmitted by the master / reader) base-band ranging signal in I / Q format for a period T≦T f , where .times. ...

number

number

[0113] Assume that at t=0, under the control of the master multipath processor, the master base-band processor (both in FPGA 150) initiates a base-band ranging sequence.

[0114]

number

[0115] Here, T f ≧T 1 β M It is.

[0116] The phase of the master's DAC 120 and 125 outputs is:

[0117]

number

[0118] The DAC120 and 125 have an internal propagation delay that is independent of the system clock.

number

[0119] Similarly, the transmitter circuit components 15, 30, 40 and 50 provide additional delays,

number

[0120] As a result, the phase of the RF signal transmitted by the master can be calculated as follows:

[0121]

number

[0122] The RF signal from the master (M) experiences a phase shift φ, which is a function of multipath phenomena between the master and the tag. MULT This goes through.

[0123] φ MULT The value is the transmission frequency, e.g. F 1 , F 2 The receiver cannot cancel each path because the RF part of the receiver is bandwidth limited (i.e. narrow). Therefore, after a certain time, say 1 microsecond (corresponding to a flight of -300 meters), when all the reflected signals reach the receiver antenna, the following formula applies:

[0124]

number

[0125] At the AM (transponder) in the first downconverter (element 85), the phase of the signal at the output, for example the first IF, is:

[0126]

number

[0127] Propagation delay within the RF section of the receiver (elements 15 and 60-85)

number

number

[0128] In FPGA 150, the second IF signal (from the ADC output) is filtered by RF back-end digital filter 190 and further downconverted to a base-band ranging signal by the third downconverter (i.e., quadrature mixer 200, digital filters 230 and 210, and digital quadrature oscillator 220), summed in summer 270, and stored in memory 170. At the third downconverter output (i.e., quadrature mixer):

number

[0129] Propagation delay through FIR section 190

number

[0130] After the RX->TX delay, the base-band ranging signal from the master (M) that is stored (in the memory 170) is retransmitted. The RX->TX delay is

number

[0131]

number

[0132] By the time the signal from the transponder reaches the master (M) receiver antenna, the RF signal from the transponder (AM) experiences another phase shift, φ, which is a function of multipath. MULT As discussed above, this phase shift occurs a certain amount of time after all the reflected signals have arrived at the master's receiving antenna.

[0133]

number

[0134] In the master receiver, the signal from the transponder undergoes the same down-conversion process as in the transponder receiver. The result is the recovered base-band ranging signal originally transmitted by the master.

[0135] First frequency component F 1 For:

number

[0136] For the second frequency component F2:

number

[0137] Replaced by:

number

[0138] Here, T D_M-AM is the propagation delay through the master (M) and transponder (AM) circuits.

[0139]

number

[0140] where: φ BB_M-AM (0) is the LO phase shift from the master (M) and transponder (AM) frequency mixers, including the ADC(s), at time t=0.

[0141] Also:

number

[0142] First frequency component F1:

number

[0143] The first frequency component F1 continues:

number

[0144] Second frequency component F2:

number

[0145] The second frequency component F2 continues:

number

[0146] Further replacing:

number

[0147] Here, α is a constant.

[0148] The final phase equation is then:

[0149]

number

[0150] From equation (5),

number

[0151] where i=2, 3, 4...; and

number

number

[0152] For example, the difference between time instances t1 and t2

number

number

[0153]

number

[0154]

number

[0155] Here, T LB_M and T LB_AM is the propagation delay through the TX and RX circuits of the master (M) and transponder (AM), and this delay is measured by placing the devices in loopback mode. The master and transponder devices LB_M and T LB_AM Note that we can automatically measure t RTX The value is also known.

[0156] The above formula and t RTX From the value, T D_M_AM Thus, for given t1 and t2,

number

[0157]

number

[0158] Or, β M =β AM If we set =1, then:

number

[0159] From equation (6), it can be concluded that at the operating frequency, the complex amplitude value of the ranging signal can be found from processing the returned base-band ranging signal.

[0160] The subspace algorithm is not sensitive to constant phase offsets, so the initial phase value

number

number

number

[0161]

number

number

[0162] In a preferred embodiment, the phase value of the returned base-band ranging signal

number

number

number

[0163] From equations (5) and (6), it is clear that the recovered (i.e., received) base-band ranging signal has the same frequency as the original base-band signal transmitted by the master. Therefore, there is no frequency translation, despite the fact that the system clocks of the master (M) and transponder (AM) may be different. Since the base-band signal consists of several frequency components, and each component consists of multiple periods of a sine wave, the individual component frequencies of the received base-band signal are sampled with the corresponding individual frequency components of the original (i.e., transmitted by the master) base-band signal, and the frequency of the sampled signal is calculated as T≦T. f It is also possible to estimate the phase and amplitude of the received ranging signal by integrating the resulting signal over a period of

[0164] This operation (calculation) is the complex amplitude value of the received ranging signal

number

number

number

[0165] Sampling, T≦T f Integrate for a period and convert it to I / Q form

number

number

[0166] Despite the narrow ranging signal bandwidth, the frequency difference f n -f 1 can be relatively large, for example on the order of several megahertz. As a result, the receiver bandwidth is 1 :f n The wideband receiver bandwidth must be kept wide enough to pass all the frequency components of the ranging signal. This wideband receiver bandwidth impacts the SNR. To reduce the effective bandwidth of the receiver and improve the SNR, the received ranging signal baseband frequency components are filtered by the RF back-end processor in the FPGA 150 by digital narrow bandwidth filters tuned to the individual frequency components of the received baseband ranging signal. However, this large amount of digital filters (individual frequency components, number of filters equal to n) puts additional burden on the FPGA resources and increases its cost, size and power consumption.

[0167] In the preferred embodiment, only two narrow bandwidth digital filters are used: one filter, f 1 One filter is always tuned to the frequency component f 2 :f n Multiple instances of the ranging signal are transmitted by the master. Each instance consists of only two frequencies: 1 :f 2 ;f 1 :f 3 ....A similar approach is also possible.

[0168] Also, adjust the frequency synthesizer (e.g., K SYN Note that it is entirely possible to keep the baseband ranging signal components to only two (or even one) that generate the remaining frequency components by changing the LO signal (VHF) of the upconverter mixer and the downconverter mixer. The LO signals for the upconverter mixer and the downconverter mixer are preferably generated using Direct Digital Synthesis (DDS) techniques. This may present an undesirable burden on the transceiver / FPGA hardware for high VHF frequencies. However, for lower frequencies this may be a useful approach. Analog frequency synthesizers can also be used, but may require additional time to stabilize after the frequency is changed. Also, in case of analog synthesizers, two measurements must be made at the same frequency to cancel the phase offset that may occur after changing the frequency of the analog synthesizer.

[0169] The actual T used in the above formula D_M-AM is measured on both the master (M) and transponder (AM) system clocks, e.g., T LB_AM and t RTX is counted in the transponder (AM) clock, and T LB_M is counted in the master's (M) clock. However,

number

[0170]

number

[0171] The phase estimation error (7) affects the accuracy. Therefore, it is necessary to minimize this error. In other words, βM =β AM In the case of t, if the system clocks of all masters and transponders (tags) are synchronized, RTX The contribution from time is removed.

[0172] In a preferred embodiment, the master and transponder units (devices) can synchronize with the clock of any device. For example, the master device serves as a reference. Clock synchronization is achieved by using a remote control communication channel, where the frequency of the temperature compensated crystal oscillator TCXO 20 is adjusted under the control of the FPGA 150. While the selected transponder device is transmitting a carrier signal, the frequency difference is measured at the output of the adder 270 of the master device.

[0173] The master device then sends a command to the transponder to increase / decrease the TCXO frequency. This procedure can be repeated several times to achieve higher accuracy by minimizing the frequency at the output of the adder 270. Note that in the ideal case, the frequency at the output of the adder 270 should be equal to zero. An alternative method is to measure the frequency difference and make a correction to the estimated phase without adjusting the transponder's TCXO frequency.

[0174] β M -β AM can be significantly reduced, but β M If ≠ 1, there is a phase estimation error. In this case, the margin of error depends on the long-term stability of the clock generator of the reference device (usually the master (M)). Also, the process of clock synchronization takes a significant amount of time, especially since there are many units in the technical field. During the synchronization process, the tracking and locating system is partially or completely inoperable, which negatively impacts system readiness and performance. In this case, the above-mentioned method, which does not require adjustment of the transponder's TCXO frequency, is preferred.

[0175] Commercially available (off-the-shelf) TCXO components have high accuracy and stability. In particular, TCXO components for GPS commercial applications are very accurate. When used in these devices, the phase error impact on position detection accuracy can be less than one meter without the need for frequent clock synchronization.

[0176] The narrow bandwidth ranging signal multipath mitigation processor is

number

[0177] The super-resolution algorithm is (2π×τ K ) "frequency" (e.g. τ K In the final step, the multi-path mitigation processor selects the τ that has the smallest value (i.e., the DLOS delay time).

[0178] In some cases where the ranging signal narrow bandwidth requirement is somewhat relaxed, the DLOS path can be separated from the MP path by using a continuous (in time) chirp. In the preferred embodiment, this continuous chirp is Linear Frequency Modulated (LFM). However, other chirp waveforms can also be used.

[0179] Suppose a chirp with bandwidth B and duration T is transmitted under the control of a multipath mitigation processor. It is

number

[0180] In a multipass processor, each received single chirp is positionally adjusted so that the feedback chirp is from the center of the target area.

[0181] The equation of the chirp waveform is as follows.

[0182]

Number

[0183] Here, for 0 < t < T, ω0 is the initial frequency.

[0184] For a single delay round trip τ, for example one that is not multipass, the feedback signal (chirp) is s(t - τ).

[0185] Subsequently, the multipass relaxation processor "deramps" s(t - τ) by performing a complex conjugate mix using the originally transmitted chirp. The resulting signal is a complex sine wave.

[0186]

Number

[0187] Here,

Number

[0188] In the multipath case, the composite deramped signal consists of multiple complex sine waves.

[0189]

number

[0190] where L is the number of ranging signal paths including the DLOS path, and 0≦t≦T.

[0191] Multiple chirps are transmitted and processed. As described above, each chirp is treated / processed individually. A multi-path mitigation processor then assembles the results of the individual chirp processing.

[0192]

number

[0193] where N is the number of chirps,

number

[0194] In equation (10),

number

[0195]

number

[0196] Therefore, the number of samples can be a multiple of N (eg, αN; α=1, 2, . . . ).

[0197] From equation (10), the multi-path mitigation processor generates αN complex amplitude samples in the time domain that are used in further processing (i.e., execution of the super-resolution algorithm). This further processing is performed in a software component that is part of the multi-path mitigation processor. This software component can be executed by the CPU of the master (reader) host computer and / or by a microprocessor embedded in the FPGA 150 (not shown), or both. In a preferred embodiment, the multi-path mitigation algorithm software is executed by the CPU of the master host computer.

[0198] The super-resolution algorithm is 2βτ K "Frequency" (e.g. τ K In the final step, the multi-path mitigation processor selects the τ that has the smallest value (i.e., the DLOS delay time).

[0199] A special processing method called "THRESHOLD TECHNIQUE" is described. This threshold technique serves as an alternative to super-resolution algorithms. In other words, it is used to improve the reliability and accuracy of distinguishing DLOS paths from other MP paths using synthetic, artificially generated, wider bandwidth ranging signals.

[0200] The frequency domain base-band ranging signal shown in FIG. 1 and FIG. 1A can be converted to a time domain base-band signal s(t).

[0201]

number

[0202] It is easy to verify that s(t) is periodic, with period 1 / Δt, and for any integer k, s(k / Δt)=2N+1 is the peak value of the signal, where n=N in Figures 1 and 1A.

[0203] Figure 4 shows two periods of s(t) for N=11 and Δf=250 kHz. The signal appears as a sequence of pulses of height 2N+1=23 separated by 1 / Δf=4 microseconds. Between the pulses it is a sinusoidal waveform with variable amplitude and 2N zeros. The wide bandwidth of the signal can be attributed to the narrowness of the tall pulses. It can also be seen that the bandwidth extends from zero frequency to NΔf=2.75 MHz.

[0204] The basic idea of ​​the threshold method used in the preferred embodiment is to improve the reliability and accuracy of synthetically generated wider bandwidth ranging in distinguishing DLOS paths from other MP paths. The threshold method detects when the leading part of the rising edge of the wideband pulse reaches the receiver. Due to filtering at the transmitter and receiver, the rising edge does not rise instantaneously but rises out of the noise with a smoothly increasing slope. The TOA of the rising edge is measured by detecting when the rising edge crosses a predefined threshold T.

[0205] A small threshold is desirable because the threshold is crossed sooner and the error delay τ between the exact start of the pulse and the threshold crossing is small. Thus, a pulse replica arriving due to multipath will have no effect if the beginning of the replica has a delay larger than τ. However, the presence of noise places a limit on how small the threshold T can be. One way to reduce the delay τ is to use the derivative of the received pulse instead of the pulse itself, because the derivative rises faster. The second derivative rises even faster. Higher order derivatives could be used, but they would raise the noise level to unacceptable values ​​in practice, so the second derivative of the threshold is used.

[0206] The 2.75 MHz wideband signal shown in Figure 4 has a reasonably wide bandwidth, but it is not suitable for measuring range by the method described above. This method requires transmitted pulses, each with a zero signal precursor. However, the goal can be achieved by modifying the signal so that the sinusoidal waveforms between the pulses are essentially cancelled. In the preferred embodiment, this is done by constructing a waveform that closely approximates the signal in selected intervals between the high pulses, and then subtracting it from the original signal.

[0207] This technique can be demonstrated by applying it to the signal in Figure 1. The two black dots shown on the waveform are the endpoints of interval I, which is between the first two pulses. The left and right endpoints of interval I, experimentally determined to provide the best results, are:

[0208]

number

[0209] An attempt is made to generate a function g(t) that essentially cancels the signal s(t) in this interval, but does little harm outside the interval. Since equation (11) shows that s(t) is a sinusoidal sinπ(2N+1)Δft modulated by 1 / sinπΔft, first a function h(t) is found that closely approximates 1 / sinπΔft in the interval I, and then the resulting g(t) is formed.

[0210]

number

[0211] h(t) is generated by the sum of

[0212]

number

[0213] Where:

number

[0214] And the coefficient a k is chosen to minimize the least square error of

[0215]

number

[0216] This solution is a k This is easily obtained by partially differentiating J with respect to and setting them equal to zero. The result is a system of linear simultaneous equations containing M+1 equations.

[0217]

number

[0218] This is a k where

[0219]

number

[0220] At this time,

number

[0221] The function φ given by (12) k Using the definition of (t),

number

[0222] Subtracting g(t) from s(t) gives a function r(t) that should essentially cancel s(t) in the interval I. As shown in the Appendix, a good choice for the upper bound M on the summation in equation (20) is M=2N+1. Using this value and the results from the Appendix,

number

number

[0223] From equation (17), we see that a total of 2N+3 frequencies (including the zero-frequency DC term) are required to obtain the desired signal r(t). Figure 5 shows the resulting signal r(t) for the original signal s(t) shown in Figure 1, where N=11. In this case, 25 carriers (including the DC term b 0 (including

[0224] Important properties of r(t) constructed above are:

[0225] 1. As can be seen from (14), the lowest frequency is 0Hz and the highest frequency is (2N+1)ΔfHz. Therefore, the total bandwidth is (2N+1)ΔfHz.

[0226] 2. All carrier waves have a frequency

number

[0227] 3. The original signal s(t) has a period 1 / Δf, while r(t) has a period 2 / Δf. The first half of each period of r(t) is a full period of s(t) but contains the cancelled part of the signal, and the second half of r(t) is a segment that oscillates significantly. Thus, precursor cancellation occurs only every other period of s(t).

[0228] This occurs because the cancellation function g(t) actually reinforces s(t) in every other period of s(t), because g(t) reverses its polarity at every peak in s(t), but s(t) does not. A method is described below that causes every period of s(t) to include a cancelled portion, thus increasing the processing gain by 3 dB.

[0229] 4. The length of the cancelled portion of s(t) is approximately 80-90% of 1 / Δf. Therefore, Δf needs to be small enough to make this length long enough to eliminate residual signals due to multipath from the previous non-zero portion of r(t).

[0230] 5. Immediately following each zero portion of r(t) is the first cycle of the oscillatory portion. In a preferred embodiment, in the TOA measurement method as described above, the first half of this cycle is used to measure the TOA, and in particular the beginning of its rise. It is interesting to note that the peak value of this first half cycle (called the main peak) is somewhat larger than the corresponding peak of s(t) at approximately the same point in time. The width of this first half cycle is approximately inversely proportional to NΔf.

[0231] 6. Massive processing gains can be achieved by: (a) By repeatedly using the signal r(t), since r(t) is periodic, with period 2 / Δf, an additional 3 dB of processing gain is possible by the method described below. (b) Narrowband filtering. Each of the 2N+3 carriers is a narrowband signal, so that the occupied bandwidth of the signal is much smaller than that of a wideband signal that spans the entire allocated frequency band.

[0232] For the signal r(t) shown in FIG. 5, with N=11 and Δf=250 kHz, the length of the cancelled portion of s(t) is about 3.7 microseconds, or 1110 meters. This is more than enough to remove residual signals due to multipath from the previous non-zero portion of r(t). The main peak has a value of approximately 35, and the maximum magnitude in the precursor (i.e., cancelled) region is about 0.02, which is 65 dB lower than the main peak. This is desired for good performance using the TOA measurement threshold technique as described above.

[0233] The use of fewer carriers is illustrated in Figure 6, which shows a signal generated with Δf=850 kHz, N=3, and M=2N+1=7 for a total of only 2N+3=9 carriers. In this case, the period of the signal is only 2 / Δf≈2.35 microseconds, compared to the signal in Figure 5, which has a period of 8 microseconds. Since this example has more periods per unit time, it is expected that a larger processing gain can be achieved.

[0234] However, because fewer carriers are used, the amplitude of the main peak is about 1 / 3 as large as before, which tends to cancel the expected additional processing gain. Also, the length of the precursor segment of the zero signal is shorter, about 0.8 microseconds, or 240 meters. This should still be enough to remove residual signals due to multipath from the previous non-zero portion of r(t). Note that the total bandwidth of (2N+1)Δf=5.95 MHz is about the same as before, and the width of the half cycle of the main peak is also roughly the same. Because fewer carriers are used, there should be some additional processing gain as each carrier is narrowband filtered at the receiver. Furthermore, the maximum magnitude in the precursor (i.e., cancellation) region is about 75 dB lower than the main peak, a 10 dB improvement from the previous example.

[0235] Transmission at RF Frequencies: Up to this point, r(t) has been described as a baseband signal for simplicity. However, r(t) can be converted up to RF, transmitted, received, and then reconstructed as a baseband signal at the receiver. For illustration purposes, consider a frequency component ω of a baseband signal r(t) traveling through one of the multipath propagation paths with index j (rad / sec frequency is used for notational simplicity). k Think about what happens to one of them.

[0236]

number

[0237] Here, it is assumed that the transmitter and receiver are frequency synchronized. k is the Kth coefficient in equation (21) for r(t). The parameters τJ and φJ are the path delay and phase shift of the Jth propagation path (due to the dielectric properties of the reflector), respectively. The parameter θ is the phase shift introduced upon downconversion to baseband at the receiver. A similar set of functions can be presented for the sinusoidal components of equation (21).

[0238] It is important to note that the final baseband signal in equation (20) will still have zero signal predecessors as long as the zero signal predecessors of r(t) have a length significantly greater than the largest propagation delay. Of course, when all frequency components (index K) across all paths (index J) are combined, the baseband signal at the receiver will be a distorted version of r(t), including all of the phase shifts.

[0239] Sequential carrier transmission and signal reconstruction are shown in Figures 1 and 1A. The transmitter and receiver are synchronized in time and frequency, and the 2N+3 transmitted carriers do not have to be transmitted simultaneously. As an example, consider the transmission of a signal whose baseband is that of Figures 1A and 6.

[0240] In Figure 6, assume that N=3 and that each of the nine frequency components of 1 ms is transmitted consecutively. The start and end times of each transmitted frequency are known at the receiver. Thus, the receiver can sequentially start and end reception of each frequency component at each time. Since the signal propagation time is very short compared to 1 ms (which is typically less than a few microseconds in the intended application), a portion of each received frequency component should be ignored and the receiver can easily turn it off.

[0241] The entire step of receiving the nine frequency components can be repeated with additional reception of 9 ms blocks to increase the processing gain. In a total receive time of 1 second, there will be approximately 111 such 9 ms blocks available for processing gain. Furthermore, within each block there will be an additional processing gain available from the main peak of 0.009 / (2 / Δf) ≈ 383.

[0242] It is worth noting that in general, signal reconstruction can be done very economically, essentially allowing for any possible processing gain. For each of the 2N+3 receive frequencies: 1. Measure the phase and amplitude of each 1 millisecond of reception of that frequency to form a sequence of stored vectors (phasors) corresponding to that frequency. 2. Average the stored vector of frequencies. 3. Finally, use the average of 2N+3 vectors of 2N+3 frequencies to reconstruct one period of the baseband signal with duration 2 / Δf, and use the reconstruction to estimate the TOA of the signal.

[0243] This method is not limited to 1 millisecond transmissions, the length of the transmissions may be increased or decreased, however the total time of all transmissions must be short enough to stop any motion of the receiver or transmitter.

[0244] Obtaining cancellation at alternate half periods of r(t): By simply reversing the polarity of the cancellation function g(t), cancellation between peaks in s(t) is possible where r(t) previously oscillated. However, to obtain cancellation between all peaks in s(t), the function g(t) and its polarity-reversed version must be applied at the receiver, which involves weighting the coefficients at the receiver.

[0245] Weighting of coefficients at the receiver: When desired, the coefficient b k may be used to construct r(t) at the transmitter or introduced at the receiver. This is k This can easily be seen by considering the sequence of signals in equation (20) where the final signal is the same when is introduced at the last step instead of the first. Neglecting noise, the values ​​are

[0246]

number

[0247] The transmitter can then transmit all frequencies with the same amplitude, thereby simplifying the design. Note that this method also weights the noise at each frequency, the effect of which must be taken into account. Note also that a weighting of the coefficients must be done at the receiver to affect the polarity reversal of g(t) to obtain twice the usable main peak.

[0248] Scaling Δf to channel center frequency: To meet FCC requirements at VHF and below, channelized transmission with constant channel spacing is required. In channelized transmission bands with constant channel spacing that is small compared to the total allocated bandwidth (as is the case for VHF and below frequency bands), small adjustments to Δf, if necessary, allow all transmitted frequencies to be centered in channels without significantly changing performance from the original design value. In the two examples of baseband signals given above, the frequency components are all multiples of Δf / 2, so if the channel spacing is divided by Δf / 2, the lowest RF transmitted frequency will be centered in one channel and all other frequencies will fall in the center of the channel.

[0249] In some radio frequency (RF) based identification, tracking and locating systems, in addition to performing distance measurement functions, both the Master Unit and the Tag Unit also perform voice, data and control communication functions. Similarly, in a preferred embodiment, both the Master Unit and the Tag Unit perform voice, data and control communication functions in addition to performing distance measurement functions.

[0250] In accordance with a preferred embodiment, the ranging signals are subjected to a wide range of sophisticated signal processing techniques, including multipath mitigation. However, these techniques cannot serve the voice, data and control signals. As a result, the operating range of the proposed system (as well as other existing systems) may be limited not by its ability to reliably and accurately measure distance, but by being out of range during voice and / or data and / or control communications.

[0251] In other radio frequency (RF) based identification, tracking and location systems, the distance measurement function is separated from the voice, data and control communication functions. In these systems, separate RF transceivers are used to perform the voice, data and control communication functions. The drawbacks of this approach are the increased cost, complexity, size, etc. of the system.

[0252] In a preferred embodiment, to avoid the above mentioned drawbacks, several individual frequency components of the narrow bandwidth ranging signal or baseband narrow bandwidth ranging signal are modulated with the same data / control signal, in case of voice, with digital voice packet data. At the receiver, the individual frequency component with the highest signal strength is demodulated and the obtained information reliability is further improved by "voting" or other signal processing techniques that exploit the information redundancy.

[0253] This method can avoid the "null" phenomenon, where incoming RF signals from multiple paths combine destructively with the DLOS path and with each of the RF signals, thereby significantly reducing the strength of the received signal and the SNR associated therewith. Furthermore, such a method allows finding a set of frequencies where incoming signals from multiple paths combine constructively with the DLOS path and with each of the incoming signals, thus increasing the strength of the received signal and the SNR associated therewith.

[0254] As mentioned above, super-resolution algorithms based on spectral estimation generally use the same model: a linear combination of complex exponentials of frequency and their complex amplitudes, given by equation (3) above.

[0255] All super-resolution algorithms based on spectral estimation require a priori knowledge of multiple complex exponentials, i.e., multiple multipaths. This multiple complex exponentials is called the model size and is determined by the number of multipath components L, as shown in equations (1) to (3). However, when estimating the path delays (which is the case for RF tracking-location applications), this information is not available. This adds another dimension, i.e., the estimation of the model size, to the spectrum estimation process by super-resolution algorithms.

[0256] It has been shown that if the model size is underestimated, the accuracy of frequency estimation is affected, and if the model size is overestimated, the algorithm generates spurious frequencies, e.g., frequencies that do not exist (Non-Patent Document 6). Existing methods of model size estimation, such as AIC (Akaike Information Criterion), MDL (Minimum Description Length), etc., are highly sensitive to correlations between signals (complex exponentials). However, in the case of RF multipath, this is always the case. For example, even after a forward-backward smoothing algorithm is applied, there is always a correlation residual.

[0257] In Non-Patent Document 6, it is suggested to use an overestimated model to distinguish real frequencies (signals) from spurious frequencies (signals) by estimating these signal powers (amplitudes) and rejecting signals with very low power. This method is an improvement over existing methods, but is not guaranteed. The inventors implemented the method of Non-Patent Document 6 and performed simulations of more complicated cases with larger model sizes. It was observed that in some cases the spurious signals could have amplitudes very close to the real signal amplitudes.

[0258] All super-resolution algorithms based on spectral estimation work by splitting the complex amplitude data of the incoming signal into two subspaces: the noise subspace and the signal subspace. If these subspaces are properly defined (separated), the model size is equal to the signal subspace size (dimension).

[0259] In one embodiment, the estimation of the model size is achieved using the "F" statistic. For example, for the ESPRIT algorithm, the singular value decomposition of the estimate of the covariance matrix (by forward / backward correlation smoothing) is sorted in ascending order. Then, a division is performed whereby the (N+1)th eigenvalue is divided by the Nth eigenvalue. This ratio is the "F" random variable. The worst case is an "F" random variable with (1,1) degrees of freedom. The 95% confidence interval for an "F" random variable with (1,1) degrees of freedom is 161. Setting that value as the threshold determines the model size. Note also that for the noise subspace, the eigenvalues ​​represent an estimate of the noise power.

[0260] This method of applying the "F" statistic to the ratio of eigenvalues ​​is a more accurate method of estimating model size. Note that other degrees of freedom in the "F" statistic can be used for the threshold calculation and, therefore, for model size estimation.

[0261] Nevertheless, in some cases, two or more very closely spaced (in time) signals may degenerate to one signal due to imperfect real-world measurements. As a result, the above method underestimates the number of signals (i.e., the model size). Since an underestimation of the model size leads to a lower accuracy of the frequency estimation, it is prudent to increase the model size by adding a number. This number can be determined experimentally and / or from simulations. However, if the signals are not closely spaced, the model size will be overestimated.

[0262] In such cases, spurious frequencies, i.e. non-existent frequencies, may appear. As mentioned above, in some cases, spurious signals have been observed to have amplitudes very close to the amplitude of the actual signal, so using signal amplitudes for spurious signal detection does not always work. Therefore, in addition to amplitude discrimination, filters can be implemented to improve the probability of eliminating spurious frequencies.

[0263] The frequencies estimated by the super-resolution algorithm are artificial frequencies (equation (2)). In reality, these frequencies are the individual path delays of a multipath environment. As a result, negative frequencies cannot exist, and all negative frequencies generated by the super-resolution algorithm are spurious frequencies that should be rejected.

[0264] Furthermore, the DLOS distance range is determined by a method other than the super-resolution method, which uses the complex amplitudes acquired during the measurement.

number

number

[0265] This is the signal amplitude

number

[0266] In this embodiment, the ranging signal goes round trip, in other words, it travels in both directions: from the master / reader to the target / slave and from the target / slave to the master / reader.

[0267] The master is the tone

number

[0268] At the target receiver, the received signal (one-way) is:

[0269]

number

[0270] where N is the number of signal paths in a multipath environment; K0 and τ 0 are the amplitude and time of flight of the DLOS signal.

number

[0271]

number

[0272] where:

number

[0273] The target retransmits the received signal.

[0274]

number

[0275] At the master receiver, the round trip signal is:

[0276]

number

number

[0277] On the one hand, from equations (26) and (28),

number

[0278] where:

number

[0279] From equation (29), the round-trip multipath channel has more paths than the one-way channel multipath because τ 0 ÷τ N In addition to the path delay,

number

[0280] These combinations dramatically increase the number of signals (complex exponentials). Therefore, the probability of very closely spaced signals (in time) also increases, which can lead to a large underestimation of the model size. Therefore, it is desirable to obtain a unidirectional multipath RF channel transfer function.

[0281] In a preferred embodiment, the unidirectional amplitude value:

number

number

[0282]

number

number

[0283] However, for each value of ω, there are two values ​​of the phase α(ω), so that

[0284]

number

[0285] A detailed description to resolve this ambiguity is given below. When the different frequency components of the ranging signal are close to each other, for the most part, the one-way phase can be found by dividing the round trip phase by two. Exceptions include the region close to "zero", where the phase can undergo large changes even with small frequency steps. Note: The "zero" phenomenon is where the RF signals coming from multiple paths are destructively combining with the DLOS path and with each of the RF signals, thereby significantly weakening the received signal strength and the associated SNR.

[0286] Let h(t) be the one-way impulse response of a communication channel. The corresponding transfer function in the frequency domain is

[0287]

number

[0288] where A(ω)≧0 is the magnitude and α(ω) is the phase of the transfer function. If the one-way impulse response is retransmitted back through the same channel as it was received, the resulting two-way transfer function is:

[0289]

number

[0290] where β(ω) ≥ 0. The bidirectional transfer function G(ω) is 1 ,ω 2 ) for all ω in G(ω). 1 ,ω 2 We consider whether it is possible to determine the one-way transfer function H(ω) defined in

[0291] The magnitude of the bidirectional transfer function is the square of the unidirectional magnitude, so

number

[0292] However, when trying to recover the phase of a one-way transfer function from observations about G(ω), the situation is more subtle. For each value of ω, there are two values ​​of phase α(ω), so

[0293]

number

[0294] Multiple different solutions may be generated by independently selecting one of two possible phase values ​​for each different frequency ω.

[0295] The following theorem, which assumes that any one-way transfer function is continuous at all frequencies, helps to resolve this situation.

[0296] Theorem 1: Let I be a bidirectional transfer function:

number

number

[0297] Proof: One solution to the one-way transfer function is a function that is continuous in I.

number

number

[0298] H(ω) and J(ω) are continuous and nonzero in I, and their ratio is continuous in I. Thus, the right-hand side of (34) is continuous in I. The condition β(ω)=2α(ω)=2γ(ω) implies that for each ω∈I, α(ω)-γ(ω) is either 0 or π. However, α(ω)-γ(ω) cannot switch between these two values ​​without causing a discontinuity in the right-hand side of (34). Thus, either α(ω)-γ(ω)=0 for all ω∈I, or α(ω)-γ(ω)=π for all ω∈I. In the first case, we have J(ω)=H(ω), and in the second case, we have J(ω)=-H(ω).

[0299] This theorem states that the transfer function

number

number

[0300] An alternative procedure for finding a one-way solution is based on the following theorem.

[0301] Theorem 2:

number

[0302] Proof: ω 0 Let be the frequency in interval I. In FIG. 7, the complex value H(ω 0 ) are plotted as points in the complex plane, and the assumption is that H(ω 0 ) ≠ 0. Let ε>0 be an arbitrarily small real number, and consider the two angles of the measurement ε shown in FIG. 7, as well as the tangents of the two lines OA and OB, and H(ω 0 ) is centered. The assumption is that H(ω) is continuous for all ω. Therefore, if ω is 0 If the complex value H(ω) is close enough to

number

[0303] Theorem 3: Let I be the bidirectional transfer function

number

number

[0304] Proof: The proof is similar to that of Theorem 1. One of the solutions for the one-way transfer function is the function

number

number

[0305] Assumption is that γ(ω) is continuous in I, and by Theorem 2, α(ω) is also continuous in I. Hence, α(ω)-γ(ω) is continuous in I. The condition β(ω)=2α(ω)=2γ(ω) implies that for each ω∈I, α(ω)-γ(ω) is either 0 or π. However, α(ω)-γ(ω) cannot switch between these two values ​​without introducing a discontinuity in I. Thus, either α(ω)-γ(ω)=0 for all ω∈I, or α(ω)-γ(ω)=π for all ω∈I. In the first case, we have J(ω)=H(ω), and in the second case, we have J(ω)=-H(ω).

[0306] Theorem 3 states that the transfer function

number

number

[0307] The above theorem shows how to reconstruct the two one-way transfer functions that generate the two-way transfer function G(ω), but they are only useful in the frequency interval I where G(ω) is nonzero. In general, it is observed in the frequency interval (ω1, ω2) where G(ω) is nonzero. Here is a way to get around this problem, by using a few finite numbers to find the exact solution for (ω 1 ,ω 2 ), G(ω) becomes zero, and the one-way transfer function is (ω 1 ,ω 2 ), not all of which are zero at any given frequency ω.

[0308] H(ω) in the interval (ω 1 ,ω 2 ) to generate G(ω), where G(ω) is a one-way transfer function 1 ,ω 2 ) has at least one zero. If G(ω) is zero, then (ω 1 ,ω 2 ) into a finite number of adjacent open frequency intervals J 1 , J 2 , ...J.N. n For each such interval, a solution H(ω) or -H(ω) can be found using either Theorem 1 or Theorem 3. These solutions need to be "stitched together" so that the stitched solution is (ω 1 ,ω 2 ), the solution will be either H(ω) or -H(ω). To do this, we need to know how to pair up solutions in two adjacent subintervals such that we don't switch from H(ω) to -H(ω) or -H(ω) to H(ω) when moving from one subinterval to the next.

[0309] The first two adjacent open subintervals J 1 and J. 2 These subintervals are the frequencies ω where G(ω) is zero. 1 (Obviously, ω 1(is not included in any of the subintervals.) The above assumptions about the properties of the one-way transfer function mean that H (n) (ω 1 ) ≠ 0, where the superscript (n) denotes the nth derivative. Then, from the left, ω → ω 1 As J 1 The limit of the nth derivative of the one-way solution at J 1 Depending on whether the solution at is H(ω) or -H(ω), H (n) (ω 1 ) or -H (n) (ω 1 ) from the right. Similarly, ω → ω 1 As J 2 The limit of the nth derivative of the one-way solution at J 2 Depending on whether the solution at is H(ω) or -H(ω), H (N) (ω 1 ) or -H (n) (ω 1 ) H (n) (ω 1 ) ≠ 0, so J 1 and J. 2 The two limits are equal if and only if the solutions in are both H(ω) or both -H(ω). If the left and right limits are unequal, then the subinterval J 2 Invert the solution in , otherwise do not.

[0310] (If necessary) subinterval J 2 After inverting the solution in the subinterval J 2 and J. 3 Perform the same procedure for and (if necessary) the subinterval J 3 Continuing in this manner, we finally invert the solution in the interval (ω 1 ,ω 2 ) to develop a complete solution.

[0311] It is desirable that higher order derivatives of H(ω) are not required in the above reconstruction procedure, because they are difficult to compute accurately in the presence of noise. This is unlikely to happen if G(ω) is zero, since the first derivative of H(ω) is very likely to be nonzero, and if it is not, the second derivative is very likely to be nonzero.

[0312] In a practical scheme, the bidirectional transfer function G(ω) is measured at discrete frequencies that must be close enough together to allow reasonably accurate calculation of the near-zero derivative of G(ω).

[0313] For RF based distance measurement, it is necessary to resolve the unknown number of closely spaced, overlapping and noisy echoes of a ranging signal with an a priori known shape. Given that the ranging signal is narrowband, in the frequency domain this RF phenomenon can be described (modeled) as the sum of many sinusoids, each one for a multipath component, with the complex attenuation and propagation delay of the path.

[0314] Taking a Fourier transform of the above sum represents this multipath model in the time domain. By exchanging the roles of variable time and variable frequency in this time domain representation, this multipath model becomes a harmonic signal spectrum, where the propagation delays of the paths are transformed into harmonic signals.

[0315] Super-resolution spectral estimation methods are designed to identify closely spaced frequencies in the spectrum and are used to estimate the individual frequencies of multiple harmonic signals (e.g., path delays), so that the path delays can be accurately estimated.

[0316] Super-resolution spectrum estimation solves the underlying estimation of individual frequencies, e.g., path delays, by exploiting the inherent structure of the covariance matrix of the base-band ranging signal samples and the inherent properties of the covariance matrix. One of the inherent structure properties is that the eigenvalues ​​can be combined and, as a result, separated into orthogonal noise and signal eigenvectors (also called subspaces). Another inherent structure property is the subspace property of rotationally invariant signals.

[0317] Subspace decomposition techniques (MUSIC, rootMUSIC, ESPRIT, etc.) rely on classifying the estimated covariance matrix of the observed data into two orthogonal subspaces: the noise subspace and the signal subspace. The theory behind the subspace decomposition methodology is that the projection of the observables onto the noise subspace consists of only noise, and the projection of the observables onto the signal subspace consists of only signal.

[0318] Spectral estimation methods assume that the signal is narrowband and the number of harmonic signals is also known, i.e., the size of the signal subspace needs to be known. The size of the signal subspace is called the model size. In general, this size is not known in detail and can change rapidly when the environment changes, especially indoors. One of the most difficult and subtle problems in applying any subspace decomposition algorithm is the size of the signal subspace that can be obtained as the number of frequency components present, which number is the number of direct paths and multipath reflections. Since real-world measurements are imperfect, there will always be errors in the estimation of the model size, which will in turn result in a loss of accuracy in the frequency estimation (i.e., distance).

[0319] To improve distance measurement accuracy, one embodiment includes six features that advance state-of-the-art methodologies for high-resolution estimation of subspace decomposition, including combining two or more algorithms that estimate individual frequencies by using different inherent structural properties that further reduce the ambiguity of delay path decisions.

[0320] RootMusic finds the individual frequencies that minimize the energy of the projection when the observable is projected onto the noise subspace. The Esprit algorithm determines the individual frequencies from a rotation operator. Moreover, in many ways this operation is the conjugate of Music, in that it finds the frequencies that maximize the energy of the projection when the observable is projected onto the signal subspace.

[0321] Model size is key for both of these algorithms, and in fact in complex signal environments such as those found in indoor ranging, and the model sizes that provide the best performance for Music and Esprit are generally not equivalent, for reasons discussed below.

[0322] For Music, it is desirable to err on the side of identifying the building blocks of the decomposition as "signal eigenvalues" (type I error). This minimizes the amount of signal energy projected onto the noise subspace, improving accuracy. For Esprit, the opposite is true: it is desirable to err on the side of identifying the building blocks of the decomposition as "noise eigenvalues". Again, this is a type I error. This minimizes the effect of noise on the energy projected onto the signal subspace. Thus, model sizes for Music will generally be somewhat larger than those for Esprit.

[0323] Second, in a complex signal environment, there are strong reflections and the possibility exists that the direct path is in fact much weaker than some of the multipath reflections, creating a situation where it becomes difficult to estimate the model size with sufficient statistical confidence. This problem is addressed by estimating "base" model sizes for both Music and Esprit, and processing the observable data using Music and Esprit in windows of model sizes defined by their respective base model sizes. This results in multiple measurements for each measurement.

[0324] The first feature of this embodiment is to estimate the model size by using the F statistic (see above). The second feature is to use different type I error rates in the F statistic for Music and Esprit. This introduces a type I error between Music and Esprit as described above. The third feature is to use a base model size and window to maximize the possibility of detecting the direct path.

[0325] Due to potentially rapidly changing physical and electronic environments, not all measurements will provide a robust answer. This is addressed by using cluster analysis on multiple measurements to provide robust range estimates. A fourth feature of the present embodiment is the use of multiple measurements.

[0326] Since there are multiple signals, the probability distribution of the multiple answers, each derived from multiple measurements using multiple model sizes from both Music and Esprit runs, is multimodal. Traditional cluster analysis is not sufficient for this application. The fifth feature is to extend the multimodal cluster analysis to estimate the direct range and equivalent range of the reflected multipath components. The sixth feature is to analyze the statistics of the range estimates provided by the cluster analysis (range and standard deviation) and combine these estimates to be statistically identical. This results in a more accurate range estimate.

[0327] The above described method can also be used in wide bandwidth ranging signal location finding systems.

[0328] For the derivation of r(t) in the threshold method, starting from equation (20), we get:

[0329]

number

[0330] Here, the trigonometric formula

number

[0331] a 0 Except for coefficient a k is zero for even K. The reason for this is that in the interval I, the function 1 / SinπΔft that we are trying to approximate by h(t) is even near the center of I. However, the basis function SinKπΔft for even k, k≠0, is odd near the center of I and is therefore orthogonal to 1 / SinπΔft in I. Therefore, we can make the substitution k=2n+1, making M an odd positive integer. In practice, we take M=2N+1. This choice has been experimentally determined to provide a significant amount of cancellation of oscillations in the interval I.

[0332]

number

[0333] Now, by making the substitution k=Nn in the first sum and K=N+n in the second sum,

number

[0334] Subtracting g(t) from s(t) gives us the following.

[0335]

number

[0336] Here we do the following:

[0337]

number

[0338] Then, (A4) can be written as follows:

[0339]

number

[0340] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention relates to a method for positioning / locating in wireless communications and other wireless networks that substantially obviates one or more of the shortcomings of the prior art. The present embodiments advantageously improve the accuracy of tracking and locating functions in multiple types of wireless networks by utilizing multipath mitigation processes, techniques and algorithms described in related U.S. Pat. No. 6,399,363. These wireless networks include Wireless Personal Area Networks (WPGANs) such as ZigBee and Bluetooth, Wireless Local Area Networks (WLANs) such as WiFi and UWB, Wireless Metropolitan Area Networks (WMANs) that typically consist of multiple WLANs, of which WiMax is a representative example, Wireless Wide Area Networks (WANs) such as white space TV bands, and Mobile Device Networks (MDNs) that are typically used to transmit voice and data. MDNs are typically based on the Global System for Mobile Communications (GSM) and Personal Communications Service (PCS) standards. More recent MDNs are based on the Long Term Evolution (LTE) standard. These wireless networks are typically made up of a combination of devices including base stations, desktop computers, tablet computers, laptop computers, handsets, smartphones, actuators, specialized tags, sensors, and other communication and data devices (collectively, all of these devices are referred to as "wireless network devices").

[0341] Existing location and positioning solutions use multiple technologies and networks including GPS, AGPS, cell tower triangulation and Wi-Fi. Some of the methods used to derive this location information include RF fingerprinting, RSSI, and TDOA. While meeting the requirements of current E911, existing location detection and ranging methods do not have the reliability and accuracy required to support the requirements of upcoming E911 as well as LBS and / or RTLS applications, especially in indoor and urban environments.

[0342] The method described in related U.S. Patent No. 6,399,633 significantly improves the ability to accurately locate and track a device of interest within a single wireless network or a combination of multiple wireless networks. This embodiment is a significant improvement over existing implementations of tracking and location methods used by wireless networks that use Enhanced Cell-ID and OTDOA (with Time Difference of Arrival) including DL-OTDOA (Downlink OTDOA), U-TDOA, UL-TDOA, and others.

[0343] The Cell ID location technique is able to estimate the location of the user (UE - User Equipment) with an accuracy of the coverage area of ​​a particular sector. The achievable accuracy therefore depends on the cell (base station) sectoring scheme and the antenna beam width. To improve the accuracy, the enhanced Cell ID technique adds a measurement of the RTT (round trip time) from the eNB. Note: Here, the RTT constitutes the difference between the transmission of a frame of the downlink DPCH (Dedicated Physical Channel (DPDCH / DPCCH: Dedicated Physical Data Channel / Dedicated Physical Control Channel)) and the start of the corresponding uplink physical frame. In this example, the above mentioned frame acts as a ranging signal. Based on the knowledge of the length of time this signal propagates from the eNB to the UE, the distance from the eNB can be calculated (see Figure 10).

[0344] In the Measured Time Difference of Arrival (OTDOA) technique, the arrival time of signals coming from neighboring base stations (eNBs) is calculated. When signals from three base stations are received, the UE's location can be estimated at the handset (UE-based method) or at the network (NT-based, UE-assisted method). The measured signal is the CPICH (COMMON PILOT CHANNEL). The propagation time of the signal is correlated with a locally generated replica. The correlation peak indicates the observed propagation time of the measured signal. The value of the time difference of arrival between the two base stations determines the hyperbola. At least three reference points are required to define the two hyperbolae. The UE's location is at the intersection of these two hyperbolae (see Figure 11).

[0345] IDLE PERIOD DOWNLINK (IPDL) is a further enhancement of OTDOA. OTDOA-IPDL technology is based on the same measurements as regular OTDOA. Time measurements are performed during idle periods during which the serving eNB stops its transmission and allows UEs in the coverage of this cell to HEAR pilots coming from distant base stations. The serving eNB provides the idle periods in continuous or burst mode. In continuous mode, one idle period is inserted in every downlink physical frame (10 MS). In burst mode, the idle periods occur in a pseudo-random manner. A further improvement is obtained via TIME ALIGNED IPDL (TA-IPDL). Time alignment creates a common idle period. During this common idle period, each base station stops its transmission or transmits a common pilot. Measurements of the pilot signals are performed during the idle period. There are several other techniques that can further enhance the DL OTDOA-IPDL method, e.g., CUMULATIVE VIRTUAL BLANKING, UTDOA (Uplink TDOA), etc. All these techniques improve the ability to hear other (non-serving) eNBs.

[0346] One significant drawback of OTDOA-based techniques is that the timing relationships of the base stations need to be known or measured (synchronized) to make the method viable. For unsynchronized UMTS networks, the 3GPP standard provides proposals on how this timing can be recovered. However, network operators have not implemented such a solution. As a result, a variant has been proposed that uses RTT measurements instead of CPICH signal measurements (see US Pat. No. 5,399,323 by John Carlson et al.).

[0347] All the above mentioned methods / techniques are based on the measurement of the arrival time of terrestrial signals and / or the measurement of the difference of arrival time (RTT, CPICH, etc.). The problem with such measurements is that they are severely affected by multipath, which significantly reduces the accuracy of the above mentioned methods / techniques, location detection / tracking (see Non-Patent Document 7).

[0348] One multipath mitigation technique uses detection / measurements from redundant eNBs or radio base stations (RBSs). For multipath mitigation, the number of RBSs required is at least six to eight, although the minimum number is three (see US Pat. No. 5,399,633). However, the probability that the UE will hear from this many eNBs is much lower than from three eNBs. This is because with many RBSs (eNBs), some RBSs (eNBs) are far away from the UE, and the received signals from these RBSs (eNBs) may be below the sensitivity level of the UE receiver or the received signals may have low signal-to-noise ratio.

[0349] In case of RF reflections (e.g., multipath), multiple copies of the RF signal with different delay times are superimposed on the DLOS (straight line of sight) signal. Because the CPICH, uplink DPCCH / DPDCH, and other signals used in various cell ID and OTDOA methods / techniques, including measurements of RTT, are bandwidth limited, the DLOS signal and the received signal cannot be distinguished without proper multipath processing / mitigation. Without this multipath processing, these reflected signals induce errors in the estimated time difference of arrival (DTOA) values ​​and time of arrival (TOA), including measurements of RTT.

[0350] For example, the 3G TS 25.515 v.3.0.0(199-10) standard defines RTT as "the difference between the transmission of a downlink DPCH frame (signal) and the reception of the start of the corresponding uplink DPCCH / DPDCH frame (signal) from the UE (first critical path)". The standard does not define what constitutes this "first critical path". The standard goes on to note that "the definition of the first critical path requires further elaboration". For example, in heavy multipath environments, it often occurs that the first critical path, the DLOS signal, is severely attenuated (10dB-20dB) compared to one or more reflected signals. If the first critical path is determined by measuring signal strength, the first critical path may not be the DLOS signal, but one of the reflected signals. This results in erroneous TOA / DTOA / RTT measurements and loss of location accuracy.

[0351] In previous wireless network generations, location characteristic accuracy was also affected by the low sampling rate of the frames (signals) used by the location methods - RTT, CPICH and other signals. Current 3G and later wireless network generations have much higher sampling rates. As a result, in these networks, the real impact on location accuracy comes from terrestrial RF propagation phenomena (multipath).

[0352] This embodiment can be used in all wireless networks that use reference and / or pilot signals and / or synchronization signals, including simplex, half-duplex and full-duplex modes of operation. For example, this embodiment works in wireless networks that use OFDM modulation and / or its derivatives. Thus, this embodiment works in LTE networks.

[0353] This embodiment can also be applied to other wireless networks including WiMax, WiFi, white space. Other wireless networks that do not use reference and / or pilot signals and / or synchronization signals may use one or more of the following types of alternative modulation aspects as described in co-pending application Ser. No. 12 / 502,809: 1) a portion of the frame is dedicated to the ranging signal / ranging signal element as described in related U.S. Pat. No. 6,399,414; 2) the ranging signal element (related U.S. Pat. No. 6,399,414) is embedded in the frame of the transmit / receive signal; 3) the ranging signal element (related U.S. Pat. No. 6,399,414) is embedded with data.

[0354] These alternative embodiments utilize the multi-path mitigation processor and multi-path mitigation techniques / algorithms described in related US Pat. No. 6,399,323 and can be used in all operational modes: simplex, half-duplex, and full-duplex.

[0355] Also, multiple wireless networks may be utilized simultaneously in the preferred and / or alternative embodiments. As an example, a smart phone may have BlueTooth, WiFi, GSM and LTE capabilities such that it has the ability to operate on multiple networks simultaneously. Depending on application requirements and / or network availability, different wireless networks may be utilized to provide positioning / location information.

[0356] The method and system of the proposed embodiment leverages the reference / pilot and / or synchronization signals of the wireless network. Furthermore, the measurement of the reference / pilot / synchronization signals may be combined with the measurement of RTT (round trip time) or system timing. According to one embodiment, RF-based tracking and location is performed on a 3GPP LTE cellular network, but may also be performed on other wireless networks using various signaling technologies, such as WiMax, WiFi, LTE, sensor networks, etc. Both the exemplary embodiment and the above-mentioned alternative embodiment use the multipath mitigation methods / techniques and algorithms described in related US Pat. No. 6,399,433. The proposed system may use software-implemented digital signal processing.

[0357] The system of the present embodiment utilizes the hardware / software of the user equipment (UE), e.g., mobile phone or smartphone, and the hardware / software of the base station (eNB) / enhanced base station (eNB). The base station typically consists of a transmitter and a receiver in a cabin or cabinet connected to an antenna by a feeder. These base stations include microcells, picocells, macrocells, umbrella cells, cell towers, routers, and femtocells. As a result, there is little or no additional cost for the UE device and the whole system. At the same time, the location accuracy is significantly improved.

[0358] The improved accuracy is due to the multipath mitigation provided by this embodiment and related US Pat. No. 6,399,633. The embodiment(s) use multipath mitigation algorithms, reference / pilots and / or synchronization signals of the network and the network N (eNB). These embodiments may be complemented with RTT (round trip time) measurements. The multipath mitigation algorithms are implemented in the UE and / or the base station (eNB), or both the UE and the eNB.

[0359] This embodiment advantageously uses a multipath mitigation processor / algorithm that is able to separate DLOS and reflected signals even when the DLOS signal is significantly attenuated (10-20 dB or less) compared to one or more reflected signals (see related patent document 4). Thus, this embodiment significantly reduces the error in the estimated DLOS time of flight of the ranging signal and, as a result, the errors in the TOA, RTT and DOTA measurements. The proposed multipath mitigation and DLOS differentiation (recognition) method can be used on all RF bands and wireless systems / networks. And, the method can support various modulation / demodulation techniques including spread spectrum techniques such as DSS (Direct Spread Spectrum) and FH (Frequency Hopping).

[0360] Additionally, noise reduction methods can be applied to further improve the accuracy of the method. These noise reduction methods can include, but are not limited to, coherent summation, non-coherent summation, matched filtering, time diversity techniques, etc. Residual multipath interference errors can be further reduced by applying post-processing techniques such as maximum likelihood estimation (e.g., Viterbi algorithm), minimum variance estimation (Kalman filter), etc.

[0361] In this embodiment, the multipath mitigation processor and multipath mitigation techniques / algorithms do not modify the RTT, CPICH, other signals and / or frames. This embodiment leverages wireless network reference, pilot and / or synchronization signals that are used to obtain channel responses / estimates. The present invention uses channel estimation statistics generated by the UE and / or eNB (see U.S. Patent No. 6,333,393 to Iwamatsu et al.).

[0362] LTE networks use certain (non-data) reference / pilot and / or synchronization signals (known signals) transmitted in every downlink and uplink subframe, which may span the entire bandwidth of the cell. For ease of explanation, hereafter, the reference / pilot and synchronization signals are referred to as reference signals. An example of LTE reference signals is shown in FIG. 9 (these signals are scattered among LTE resource elements). From FIG. 2, a reference signal (symbol) is transmitted every sixth subcarrier. Furthermore, the reference signals (symbols) are staggered in both time and frequency. In total, the reference signals cover every third subcarrier.

[0363] These reference signals are used by the UE in initial cell search, downlink signal strength measurements, scheduling, handover, etc. Included among the reference signals are UE-specific reference signals for channel estimation (response decision) and for coherent demodulation. In addition to UE-specific reference signals, other reference signals may also be used for channel estimation (see U.S. Pat. No. 6,399,323 to Chen et al.).

[0364] LTE uses OFDM (Orthogonal Frequency Division Multiplexing) modulation. In LTE, ISI (Inter-Symbol Interference) due to multipath is handled by inserting a Cyclic Prefix (CP) at the beginning of each OFDM symbol. The CP provides enough delay so that the delayed reflected signal of the previous OFDM symbol disappears before arriving at the next OFDM symbol.

[0365] An OFDM symbol consists of multiple subcarriers spaced very closely together. Within an OFDM symbol, time-staggered copies of the current symbol (due to multipath) introduce inter-carrier interference (ICI). In LTE, ICI is handled (mitigated) by determining the multipath channel response and compensating for the channel response at the receiver.

[0366] In LTE, the multipath channel response (estimate) is calculated at the receiver from the subcarriers with reference symbols. Interpolation is used to estimate the channel response of the remaining subcarriers. The channel response is calculated (estimated) in terms of the channel amplitude and phase. Once the channel response is determined (by periodic transmission of a known reference signal), the channel distortion due to multipath is mitigated by applying amplitude and phase shifts on a subcarrier-by-subcarrier basis (see non-patent document 8).

[0367] LTE multipath mitigation is designed to remove ISI and ICI (by inserting a cyclic prefix), but not to separate the DLOS signal from the reflected signal. For example, a time-staggered copy of the current symbol spreads each modulated subcarrier signal in time, thereby causing ICI. Correcting the multipath channel response using the LTE techniques described above shrinks the modulated subcarrier signal in time, but this type of correction does not guarantee that the resulting modulated carrier signal (within an OFDM symbol) is the DLOS signal. If the DLOS modulated subcarrier signal is significantly attenuated compared to the delayed reflected signal, the resulting output signal is the delayed reflected signal and the DLOS signal is lost.

[0368] In an LTE-compliant receiver, further signal processing includes DFT (Digital Fourier Transform). It is well known that DFT techniques can only cancel (remove) copies of a signal that are delayed by a time equal to or greater than the inverse proportionality of the signal and / or channel bandwidth. The accuracy of this method may be sufficient for efficient data transfer, but may not be accurate enough for accurate distance measurements in heavy multipath environments. For example, to achieve an accuracy of 30 meters, the signal and receiver channel bandwidth must be 10 MHz (1 / 10 MHz = 100 ns) or greater. For better accuracy, the signal and receiver channel bandwidth must be wider, 100 MHz at 3 meters.

[0369] However, the CPICH, uplink DPCCH / DPDCH, and other signals used for various Cell ID and OTDOA methods / techniques, including RTT measurements, and LET receiver signal subcarriers have bandwidths significantly smaller than 10 MHz. As a result, currently used methods / techniques (in LTE) result in location errors in the 100 meter range.

[0370] To overcome the above-mentioned limitations, the present embodiment uses a unique combination of a subspace decomposition high-resolution spectral estimation approach implementation and multi-modal cluster analysis. This analysis, together with the associated multipath mitigation methods / techniques and algorithms described in related U.S. Pat. No. 6,399,433, enables reliable and accurate separation of DLOS paths from other reflected signal paths.

[0371] Compared to the methods / techniques used in LTE, in heavy multipath environments, the methods / techniques and algorithms (related patent document 4) provide a 20-50 times improvement in accuracy in distance measurements by reliably and accurately separating DLOS paths from other multipath (MP) paths.

[0372] The method / technique and algorithm described in related U.S. Patent No. 5,993,366 requires the estimation of the complex amplitude of the ranging signal. Therefore, the LTE reference signal and other reference signals (including pilots and / or synchronization signals) used for channel estimation (response determination) can also be interpreted as the ranging signal in the method / technique and algorithm described in related U.S. Patent No. 5,993,366. In this case, the complex amplitude of the ranging signal is the channel response calculated (estimated) by the LTE receiver in the form of amplitude and phase. In other words, the statistics of the channel response calculated (estimated) by the LTE receiver can provide the complex amplitude information required by the method / technique and algorithm described in related U.S. Patent No. 5,993,366.

[0373] In an ideal open space RF propagation environment where there is no multipath, the phase change of the received signal (ranging signal), i.e. the channel response phase, is directly proportional to the frequency of the signal (linear), and the time of flight (propagation delay) of the RF signal in such an environment can be calculated directly from the phase vs frequency dependence by calculating the first derivative of the phase with respect to the frequency dependence, resulting in a constant propagation delay.

[0374] In this ideal environment, the absolute phase value of the first (or any) frequency is not important because the derivative is not affected by the absolute phase value.

[0375] In heavy multipath environments, the phase change vs. frequency of the received signal is a complex curve (non-linear) and the first derivative does not provide information that can be used to accurately separate the DLOS path from other reflected signal paths. This is why the multipath mitigation processor, method / technique and algorithm described in related US Pat. No. 6,399,363 is used.

[0376] When the phase and frequency synchronization (phase coherence) achieved in a wireless network / system is very good, the multipath mitigation processor, method / technique and algorithm described in related patent document 4 accurately separates the DLOS path from other reflected signal paths and determines the length (time of flight) of this DLOS path.

[0377] In this phase-coherent network / system, no additional measurements are required, i.e., one-way ranging (unidirectional ranging) can be realized.

[0378] However, if the degree of synchronization (phase coherence) achieved in a wireless network / system is not accurate enough, in a heavy multipath environment, the phase and amplitude change vs. frequency of a received signal may be very similar for measurements performed at two or more different locations (distances). This phenomenon may lead to ambiguity in determining the DLOS distance (time of flight) of a received signal.

[0379] To resolve this ambiguity, it is necessary to know the actual (absolute) phase value of at least one frequency.

[0380] However, the amplitude and phase vs. frequency dependence calculated by the LTE receiver does not include the actual phase value, because all amplitude and phase values ​​are calculated from downlink / uplink reference signals, e.g., signals relative to each other. Therefore, the amplitude and phase of the channel response calculated (estimated) by the LTE receiver requires the actual phase value at at least one frequency (carrier frequency).

[0381] In LTE, this actual phase value can be determined from one or more RTT measurements, TOA measurements, or timestamps of one or more received reference signals, provided that 1) the timestamps of these signals transmitted by the eNB are also known at the receiver (and vice versa); 2) the receiver and eNB clocks are well synchronized in time; and / or 3) multilateration techniques are used.

[0382] All of the above methods provide time-of-flight values ​​of one or more reference signals from which the actual phase values ​​at one or more frequencies can be calculated.

[0383] The present embodiment achieves high accuracy DLOS distance determination / location in heavy multipath environments by combining the multipath mitigation processors, methods / techniques and algorithms described in related patent document 4 with 1) amplitude and phase vs. frequency dependence calculated by the LTE UE and / or eNB receiver, or 2) amplitude and phase vs. frequency dependence calculated by the LTE UE and / or eNB receiver with actual phase values ​​of one or more frequencies obtained via RTT and / or TOA, and / or with timestamp measurements.

[0384] In these cases the actual phase value is subject to multipath effects, however this does not affect the performance of the methods / techniques and algorithms described in related US Pat.

[0385] In LTE RTT / TOA / TDOA / OTDOA, including DL-OTDOA, U-TDOA, UL-TDOA, etc., measurements can be made with a resolution of 5 meters. RTT measurements are performed during a dedicated connection. Thus, multiple simultaneous measurements are possible when the UE is in handover and the UE periodically collects measurements and reports them back to the UE. Here, DPCH frames are exchanged between the UE and different networks (base stations). Similar to RTT, TOA measurements provide the time of flight (propagation delay) of the signal, but TOA measurements cannot be made simultaneously (Non-Patent Document 7).

[0386] To locate a UE on a plane, the DLOS distances from / to at least three eNBs must be determined. To locate a UE in three-dimensional space, a minimum of four DLOS distances from / to four eNBs must be determined (assuming at least one eNB is not co-planar).

[0387] An embodiment of a UE positioning method is shown in FIG.

[0388] If the synchronization is very good, then measuring the RTT is not necessary.

[0389] If the degree of synchronization is not accurate enough, other methods such as OTDOA, Cell ID+RTT, and other methods, such as AOA (Angle of Arrival) and combinations of AOA and other methods, can be used to locate the UE.

[0390] The accuracy of the Cell-ID+ RTT tracking-location method is affected by multipath (RTT measurements and the eNB (base station) antenna beamwidth, which is between 33 and 65 degrees. These wide beamwidths result in location errors of 50 to 10 meters in urban areas (Non-Patent Document 7). Considering that in heavy multipath environments, the average error of current LTE RTT distance measurements is about 100 meters, the overall expected average location error currently used by the LTE Cell-ID+ RTT method is about 150 meters.

[0391] One of the embodiments is locating a UE based on the AOA method, where one or more reference signals from the UE are used to locate the UE. This involves locating an AOA determination device to determine the AOA of the DLOS. This device can be used together with a base station and / or be installed at one or more separate locations independent of the base station location. The coordinates of these locations are assumed to be known. No changes are required on the UE side.

[0392] This device includes a small antenna array and is based on a variation of the same multipath mitigation processor, methods / techniques and algorithms described in related U.S. Patent No. 6,399,433. This one possible embodiment has the advantage of accurately determining the AOA of DLOS RF energy from the UE unit (very narrow beamwidth).

[0393] In another option, this added device is just a receiving device, so that its size / weight and cost are very low.

[0394] The combination of the embodiment that provides accurate DLOS distance measurements and the embodiment that can perform accurate DLOS AOA determination significantly improves the accuracy of the Cell-ID+RTT tracking location method, by a factor of 10 or more. Another advantage of this approach is that the UE's location can be determined at any time using one tower (without requiring the UE to be in soft handover mode). Since an accurate location fix can be obtained using one tower, there is no need to synchronize multiple cell towers. Another option for determining the DLOS AOA is to use the existing eNB antenna array and eNB equipment. This option may further reduce the cost of implementing the enhanced Cell-ID+RTT method. However, positioning accuracy may be degraded since eNB antennas are not designed for location applications. Also, network operators may be reluctant to implement the necessary changes in base stations (software / hardware).

[0395] In LTE (Evolved Universal Terrestrial Radio Access (E-UTRA); Physical channels and modulation; 3GPP TS 36.211 Release 9 technical Specification), Positioning Reference Signals (PRS) were added. These signals are used by the UE for DL-OTDOA (Downlink OTDOA) positioning. Also in this Release 9, it is necessary to synchronize the eNBs, thus clearing the last hurdle of the OTDOA method (see paragraph 0258 above). The PRS improves UE hearability for UEs of multiple eNBs. Note: Release 9 does not specify the synchronization accuracy of the eNBs (some proposals: 100ns).

[0396] U-TDOA / UL-TDOA is in the research stage and will be standardized in 2011.

[0397] The DL-OTDOA method (Release 9) is detailed in Patent Document 2 (Chen et al., "Method and Apparatus for UE positioning in LTE networks"). Release 9 DL-OTDOA suffers from multipath. Some multipath mitigation can be achieved through increased PRS signal bandwidth. However, this tradeoff increases scheduling complexity and increases the time between UE position fixes. Furthermore, in networks with limited operating bandwidth, e.g., 10 MHz, the best possible accuracy is 100 meters (see Table 1 by CHEN).

[0398] The above figures are the best possible case. Other cases, especially when the DLOS signal strength is significantly lower (10-20 dB) compared to the reflected signal strength, will result in significantly larger (2-4 times) location / ranging errors as mentioned above.

[0399] The embodiments described herein can improve the ranging / location accuracy of the performance achieved by the Release 9 DL-OTDOA and UL-PRS methods of Chen et al. described in the Background by up to 50 times for a given signal bandwidth. Thus, applying embodiments of the methods described herein to Release 9 PRS processing reduces location error to 3 meters or less in 95% of all possible cases. This accuracy gain also reduces the scheduling complexity and the time between UE location fixes.

[0400] Further improvements to the OTDOA method are possible using the embodiments described herein: For example, ranging to the serving cell can be determined from other serving cell signals, thereby improving hearability of neighboring cells and reducing scheduling complexity and time between UE position corrections.

[0401] The present embodiment can also improve the accuracy of the U-TDOA and UL-TDOA methods from Chen et al. (described in the Background section) by up to 50 times. Applying the embodiment to the UL-TDOA variant of Chen reduces the location detection error to less than 3 meters in 95% of all possible cases. Furthermore, this accuracy gain further reduces the scheduling complexity and the time between UE location fixes.

[0402] Using this embodiment, the accuracy of Chen's UL-TDOA method can also be improved by up to 50 times. Thus, applying this embodiment to Chen's U-TDOA variant reduces the location detection error to less than 3 meters in 95% of all possible cases. Furthermore, this accuracy gain further reduces the scheduling complexity and the time between UE location fixes.

[0403] The DL-TDOA and U-TDOA / UL-TDOA methods described above rely on one-way measurements (ranging). This embodiment and virtually all other ranging techniques require that the PRS and / or other signals used in the one-way ranging process are frequency and phase coherent. In OFDM-based systems such as LTE, the frequency is coherent. However, the UE units and eNBs are not phase and time synchronized to within a few nanoseconds by a common source such as UTC, e.g., there is a random phase adder.

[0404] To avoid the impact of phase coherence on ranging accuracy, the multi-processor embodiment calculates the phase difference between ranging signals, e.g., between reference signals, and between individual components (subcarriers), which eliminates the need for a random phase term adder.

[0405] As identified above in the discussion of Chen et al., applying the embodiments described herein provides significant accuracy improvements in indoor environments compared to the performance achieved by Chen et al. For example, according to Chen et al., DL-OTDOA and / or U-TDOA / UL-TDOA are primarily for outdoor environments, and DL-OTDOA and U-TDOA techniques may not work well indoors (buildings, campuses, etc.). Some reasons cited include distributed antenna systems (DAS) commonly used indoors (see Chen, #161-164), where each antenna does not have a unique ID.

[0406] The embodiments described below operate in wireless networks that use OFDM modulation and / or its derivatives, as well as wireless networks that use reference / pilot / and / sync signals. Thus, the embodiments described below operate in LTE networks, but the embodiments may also be applied to other wireless systems and other wireless networks that include other types of modulation, with or without reference / pilot / and / sync signals.

[0407] The approach described herein can also be applied to other wireless networks, including WiMax, WiFi, white space, etc. Other wireless networks that do not use reference / pilot and / or synchronization signals may use one or more of the following types of alternative modulation aspects as described in related U.S. Patent Application Publication No. 2009 / 0233999: 1) a portion of a frame is dedicated to ranging signals / ranging signal elements; 2) the ranging signal elements are embedded in the transmit / receive signal frame; 3) the ranging signal elements are embedded with data.

[0408] The embodiment of the multipath mitigation range estimation algorithm described herein (also described in U.S. Pat. No. 6,399,433 and U.S. Pat. No. 6,499,433) operates by providing an estimate of the ranging of an ensemble (ENSEMBLE) of multipath reflections and straight line paths (DLOS) of the signal.

[0409] LTE DAS systems generate multiple copies of the same signal seen at various time offsets at the mobile receiver (UE). The delays are used to uniquely determine the geometric relationship between the antennas and the mobile receiver. The signal seen by the receiver resembles the signal seen in a multipath environment, except that the major "multipath" components result from the summation of offset signals from multiple DAS antennas.

[0410] The signal constellation seen by the receiver is the same type of signal constellation that the embodiment is designed to utilize, except that in this case the dominant multipath components are not traditional multipath. The multipath mitigation processor (algorithm) can determine the DLOS and each path, e.g., reflection attenuation and propagation delay (see Equations 1-3 and related text). While multipath may exist due to a dispersive RF channel (environment), the dominant multipath components in this signal constellation are associated with transmissions from multiple antennas. An embodiment of the multipath algorithm can estimate these multipath components, isolate the range from the DAS antenna to the receiver, and provide range data to a location processor (implemented in software). Depending on the antenna placement, the solution can provide both X and Y and X, Y and Z location coordinates.

[0411] As a result, the present embodiment does not require additional hardware and / or new network signals. Furthermore, the positioning accuracy can be significantly improved by 1) mitigating multipath and 2) in the case of active DAS, the lower limit of positioning error can be significantly reduced, e.g., from about 50 meters to about 3 meters.

[0412] It is assumed that the position (location) of each antenna of the DAS is known. The signal propagation delay of each antenna (relative to the other antennas) must also be determined (known).

[0413] In an active DAS system, the signal propagation delay can be determined automatically using a loopback technique in which a known signal is sent round trip and the round trip time is measured. This loopback technique also eliminates changes (drift) in the signal propagation delay with temperature, time, etc.

[0414] Using multiple macrocells and associated antennas, picocells and microcells further enhance the resolution by providing additional reference points. The above embodiment of individual range estimates in a collection of multiple copies of signals from multiple antennas can be further improved by modifying the signal transmission structure in two ways: the first is time multiplexing of the transmissions from each antenna; the second approach is frequency multiplexing of each antenna; using both improvements, time multiplexing and frequency multiplexing simultaneously, further improves the ranging and location accuracy of the system; another approach is to add a propagation delay to each antenna. The delay value is selected to be large enough to exceed the delay spread in the particular DAS environment (channel), but smaller than the length of the cyclic prefix (CP), so that the multipath caused by the additional delay does not result in ISI (Inter-Symbol Interference).

[0415] Adding a unique ID or unique identifier to each antenna improves the efficiency of the resulting solution, for example by eliminating the need for the processor to make full range estimates from the signal from each antenna.

[0416] In one embodiment utilizing the LTE downlink, one or more reference signal subcarriers, including pilot and / or synchronization signal subcarriers, are used to determine the phase and amplitude of the subcarriers, which are then applied to a multipath processor to compile wild points using multilateration and position alignment algorithms to reduce multipath interference and generate range-based position observables and position estimates.

[0417] Another embodiment takes advantage of the fact that the LTE uplink also delivers a reference signal including a reference subcarrier from the mobile device to the base station. In practice, there is more than one mode including a reference carrier, from a full sounding mode used by the network to allocate frequency bands to the uplink devices, to a mode in which the reference carrier is used to generate a channel impulse response to aid in the modulation of the uplink signal, etc. Also, similar to the DL PRS added in Release 9, additional UL reference signals may be added in upcoming and future standard releases. In this embodiment, the uplink signal is processed into phase by multiple base units (eNBs) using the same range and multipath mitigation processing to generate range related observables. In this embodiment, a location alignment algorithm is used to compile wild point observables as established by a multilateration algorithm to generate a location estimate.

[0418] Yet another embodiment is where one or more associated reference (including pilot and / or sync) subcarriers on both the LTE downlink and LTE uplink are collected, range-to-phase mapping is applied, multipath mitigation is applied, and range-related observables are estimated. These data are then fused using multilateration and position alignment algorithms to provide a more robust set of observables for location detection. The advantage of this is redundancy resulting in improved accuracy due to the presence of two different frequency bands for downlink and uplink, or in the case of TDD (Time Division Duplexing), improved system coherence.

[0419] In a distributed antenna system (DAS) environment where multiple antennas transmit the same downlink signal from a microcell, the location alignment algorithm is extended to separate the range of the DAS antenna from the observables generated by multipath processing from the subcarriers of the reference signal (including pilots and / or synchronization) to obtain location estimates from the range of multiple DAS emitters (antennas).

[0420] In a DAS system (environment), it is only possible to obtain an accurate location estimate if the signal paths from each individual antenna can be resolved with high accuracy, where the path error is a small fraction of the distance between the antennas (accuracy of 10 meters or less). Since all existing techniques / methods are unable to provide such accuracy in a heavy multipath environment (where signals from multiple DAS antennas appear to have heavy multipath induced), existing techniques / methods cannot utilize this location method / technology in a DAS environment, utilizing an extension of the location alignment algorithm described above.

[0421] The InvisiTrack multipath mitigation method and system for object identification and location described in related U.S. Patent No. 5,993,333 utilizes subcarriers of one or more LTE downlink, uplink and / or both (downlink and uplink) reference signals for range to signal phase mapping, multipath interference mitigation, and processing to generate range related position observables, and generate position estimates using multilateration and position consistency.

[0422] In all the above embodiments, a trilateration positioning algorithm may also be used.

[0423] DL-OTDOA location is specified in LTE Release 9 "Evolved Universal Terrestrial Radio Access (E-UTRA); Physical channels and modulation; 3GPP TS 36.211 Release 9 technical Specification". However, this is not implemented by wireless carriers. Downlink location, on the other hand, can be implemented in current, e.g., unmodified, LTE network environments by using existing physical layer measurement operations.

[0424] In LTE, UEs and eNBs are required to make physical layer measurements of radio characteristics. Measurement definitions are specified in 3GPP TS 36.214. These measurements are performed periodically, reported to higher layers, and used for various purposes. These purposes include intra- and inter-frequency handovers, inter-RAT handovers, timing measurements, and other purposes to support Radio Resource Management (RRM).

[0425] For example, RSRP (Reference Signal Received Power) is the average of the power of all resource elements carrying a cell-specific reference signal over the entire bandwidth.

[0426] Another example is the measurement of RSRQ (Reference Signal Received Quality), which provides additional information (RSRQ combines signal strength and interference level).

[0427] The LTE network provides the UE with an eNB neighbor (neighbor eNBs relative to the serving eNB) list. Based on the network knowledge configuration, the (serving) eNB provides the UE with the identities of the neighboring eNBs, etc. The UE then measures the signal quality of the neighboring eNBs that the UE can receive. The UE reports the result back to the eNB. Note: The UE also measures the signal quality of the serving eNB.

[0428] According to the specification, the RSRP is defined as the linear average value of the power contributions (in W) of resource elements carrying cell-specific reference signals within the considered measurement frequency band. The measurement bandwidth used by the UE to determine the RSRP is left to the UE implementation with the restriction that the corresponding measurement accuracy requirements must be met.

[0429] Considering the accuracy requirements of the measurement bandwidth, which is quite large, the cell-specific reference signals used in the RSRP measurement can be further processed to determine the subcarrier phase and amplitude of these reference signals, which are then applied to a multipath processor to reduce multipath interference and generate range-based position observables. Other reference signals used in the RSRP measurement, such as SSS (Secondary Synchronization Signal), can also be used.

[0430] Thereafter, based on range observables from three or more cells, a position fix can be estimated using multilateration and position matching algorithms.

[0431] As mentioned above, there are several sources of RF fingerprint database instability, one of the main ones being multipath (RF signatures are very sensitive to multipath). As a result, the location accuracy of RF fingerprinting methods / techniques is heavily affected by multipath dynamics, including changes over time, changes in the environment (e.g., weather), changes in people's movements, and / or changes in object movements, including vertical uncertainties with variability greater than 100% depending on the Z-height of the device and / or the orientation of the antenna (see Non-Patent Document 3).

[0432] This embodiment can greatly improve the location accuracy of RF Fingerprinting because it has the ability to find and characterize each individual path, including the highly attenuated DLOS (multipath processor). As a result, RF Fingerprinting determination of location correction can be complemented by real-time multipath distribution information.

[0433] As mentioned above, a location fix requires that the location references are synchronized in time. In wireless networks, these location references may include access points, macro / mini / pico and femto cells, as well as so-called small cells (eNBs). However, wireless operators do not implement the synchronization accuracy required for accurate location fixes. For example, in the case of LTE, the standard does not require time synchronization between eNBs for FDD (Frequency Division Duplex) networks. In LTE TDD (Time Division Duplex), the limit of this time synchronization accuracy is + / - 1.5 microseconds. This corresponds to a location detection uncertainty of more than 400 meters. Although not required, LTE FDD networks are also synchronized, but with an even larger limit (greater than 1.5 microseconds).

[0434] Wireless LTE operators use GPS / GNSS signals to synchronize eNBs in frequency and time. Note: LTE eNBs must maintain very accurate carrier frequencies: 0.05 ppm for macrocells / minicells, a little less accurate (0.1-0.25 ppm) for other types of cells. GPS / GNSS signals also enable the required time synchronization accuracy (for location detection) of better than 10 nanoseconds. However, network operators and network equipment manufacturers are trying to reduce the costs associated with GPS / GNSS units in favor of Internet / Ethernet networking time synchronization using packet transport, e.g. NTP (Network Time Protocol) and / or PTP (Precision Time Protocol) (e.g. IEEE 1588V2 PTP).

[0435] Synchronization based on IP networks has the potential to meet the minimum frequency and time requirements, but lacks the GPS / GNSS accuracy required for position fix.

[0436] The approach described herein is based on GPS / GNSS signals, signals generated by eNBs and / or APs, or signals generated by other wireless network equipment. The approach is also based on IP networking synchronization signals, protocols, signals generated by eNBs and / or APs, or signals generated by other wireless network equipment. The approach can also be applied to other wireless networks, including WiMax, WiFi, and white space.

[0437] The eNB signals are received by a Time Observation Unit (TMO) installed in the operator's eNB premises (Figure 12). The TMO also includes an external synchronization source input.

[0438] The eNB's signal is processed by the TMO and time-stamped using a clock synchronized to the external synchronization source input.

[0439] The external synchronization source can be from GPS / GNSS and / or Internet / Ethernet networking, e.g., PTP, NTP, etc.

[0440] The time-stamped signal, e.g. start of frame in LTE (but in other networks it may be other signals), also contains the eNB (cell) location and / or cell ID and is sent back via Internet / Ethernet to a central TMO server which creates, maintains and updates a database of all eNBs.

[0441] The UEs and / or eNBs involved in the process of ranging and obtaining a location fix request the TMO server, which returns time synchronization offsets between the involved eNBs, which are used by the UEs and / or eNBs involved in the process of obtaining a location fix to adjust the location fix.

[0442] Alternatively, the position fix calculation and adjustment can be done by the TMO server, which also provides the obtained ranging information to the TMO server by the UE and / or eNB(s) involved in the ranging process. The TMO server then returns the accurate (adjusted) position fix.

[0443] In the case where multiple cell eNB equipment is co-located, a single TMO may process and timestamp signals from all eNB(s).

[0444] RTT (round trip time) measurements (ranging) can be used for location. The drawback is that RTT ranging is subject to multipath which significantly impacts location accuracy.

[0445] On the other hand, RTT location detection does not generally require location-based synchronization (time synchronization), and in the case of LTE, does not require an eNB in ​​particular.

[0446] Also, when operating with pilot reference signals and / or other signals of a wireless network, the multipath mitigation processors, methods / techniques and algorithms described in related US Patent Application Publication No. 2007 / 0139999 can determine the channel response of the RTT signal, e.g., identify the multipath channel through which the RTT signal passes. This allows the measurement of the RTT to be corrected so that the actual DLOS time can be determined.

[0447] If the DLOS time is known, it is possible to obtain a position fix using trilateration and / or similar location methods without requiring time synchronization of the eNB or the position reference.

[0448] Even with the TMO and TMO server in place, InvisiTrack technology integration requires changes in the macro / mini / pico and small cells and / or UEs (mobile phones). These changes are limited to SW / FW (software / firmware) only, but require a lot of effort to retrofit the existing infrastructure. Also, in some cases network operators and / or UE / mobile phone manufacturers / suppliers are resistant to equipment changes. Note: UE is user equipment of a wireless network.

[0449] If the TMO and TMO server functionality is extended to support InvisiTrack location technology, this SW / FW modification can be avoided entirely. In other words, the alternative embodiments described below operate with wireless network signals but do not require modifications to the wireless network equipment / infrastructure. Thus, the embodiments described below operate with LTE networks and may also be applied to other wireless systems / networks, including Wi-Fi.

[0450] Essentially, this embodiment creates a parallel wireless location infrastructure that obtains location fixes by using wireless network signals.

[0451] Similar to the TMO and TMO Server, InvisiTrack's location infrastructure consists of one or more wireless Network Signals Acquisition Units (NSAUs) and one or more Locate Server Units (LSUs). The Locate Server Units collect data from the NSAUs, analyze it, determine range and location, and convert it into a table of, for example, phone / UE instantaneous ID and location. The LSUs interface to the wireless network via a network API.

[0452] Multiple of these units may be placed at various locations in a large infrastructure, and if the NSAUs are timing coherent, then all results may be used, resulting in greater accuracy.

[0453] The coherent timing may be derived from a GPS clock and / or other stable clock source.

[0454] The NSAU communicates with the LSU via a local area network (LAN), a metro area network (MAN), and / or the Internet.

[0455] In some installations / cases, the NSAU and LSU may be combined / integrated into a single unit.

[0456] To support location services using LTE or other wireless networks, transmitters need to be clock and event synchronized within tight tolerances. Typically, this is achieved by locking onto the GPS 1PPS signal, which results in time synchronization within 3 nanoseconds 1-sigma in the local area.

[0457] However, this type of synchronization is often not practical. The present embodiment provides time offset estimation between downlink transmitters and tracking of the time offset to provide delay compensation to the location process. The location process then proceeds as if the transmitters were clock and event synchronized. This is achieved with a priori knowledge of the transmitting antennas (required for location services) and a receiver with the antenna positions already known. This receiver, called a synchronization unit, collects data from all downlink transmitters and, given position information, calculates offset timing from preselected base antennas. These offsets are tracked by the system through the use of a tracking algorithm that compensates for clock drift for the downlink transmitters. Note: The process of deriving pseudoranges from the received data utilizes InvisiTrack's multipath mitigation algorithm (described in related US Pat. No. 5,399,433). Therefore, synchronization is not affected by multipath.

[0458] These offset data are used by the location processor (Location Server, LSU) to properly align the data from each downlink transmitter, so that the data appears to be generated by synchronized transmitters. Time accuracy is comparable to the best 1-PPS tracking and supports 3 meter location accuracy (1 sigma).

[0459] Synchronous receivers and / or receiver antennas are positioned based on optimal GDOP for best performance. Large installations may utilize multiple synchronous receivers to provide a synchronization offset equivalent to 3 nanoseconds (1 sigma) across the network. Utilizing synchronous receivers eliminates the requirement for downlink transmitter synchronization.

[0460] The synchronous receiver unit may be a stand-alone unit that communicates with the NSAU and / or the LSU, or the synchronous receiver may be integrated with the NSAU.

[0461] A diagram of an exemplary location device for a wireless network is shown in FIG.

[0462] An embodiment of a fully autonomous system utilizing LTE signals and requiring no customer network investments operates in the following modes: 1. Uplink Mode - Uses the uplink (UL) signal of the wireless network for location detection (Figure 16 and Figure 17). 2. Downlink Mode - Uses wireless network downlink (DL) signals for location (Figures 14 and 15). 3. Bidirectional mode – uses both UL and DL signals for location. In uplink mode, multiple antennas are connected to one or more NSAUs. The locations of these antennas are independent of the radio network antennas. The antenna locations of the NSAUs are selected to minimize the Geometric Dilution of Precision (GDOP).

[0463] Network RF signals from the UE / cellular device are collected by the NSAU's antenna and processed by the NSAU to generate time-stamped samples of the processed network RF signals during a time interval appropriate to capture one or more instances of all signals of interest.

[0464] Optionally, the NSAU also receives, processes and timestamps samples of the downlink signal to obtain additional information, such as for determining UE / phone ID.

[0465] From the captured time-stamped samples, the time-stamped wireless network signals of interest associated with each UE / cell phone ID and the UE / cell phone ID number (ID) are determined. This operation can be performed by either the NSAU or the LSU.

[0466] The NSAU periodically provides data to the LSU. When unscheduled data is required for one or more UE / mobile phone IDs, the LSU requests additional data.

[0467] No changes / modifications are required in the radio network infrastructure and / or existing UEs / mobile phones for UL mode operation.

[0468] Downlink (DL) mode requires an InvisiTrack enabled UE and if a mobile phone is used to obtain a position fix, the mobile phone's FW must be modified.

[0469] In some cases, operators can make available baseband signals from a BBU (Baseband Unit). In such cases, the NSAU can also process these available baseband wireless network signals instead of RF wireless network signals.

[0470] In DL mode, there is no need to associate the identity of the UE / mobile phone with one or more wireless network signals, since these signals are processed in the UE / mobile phone, or the UE / mobile phone periodically generates time-stamped samples of the processed network RF signals and transmits these to the LSU, which transmits the results back to the UE / mobile phone.

[0471] In DL mode, the NSAU processes time-stamped RF or baseband (if available) wireless network signals. From the captured time-stamped samples, the DL frame start of the wireless network signal associated with the network antenna is determined and the difference (offset) between those frame starts is calculated. This operation can be performed by either the NSAU or the LSU. The frame start offset of the network antenna is stored in the LSU.

[0472] In DL mode, if the UE / phone device processes / determines its location fix using InvisiTrack technology, the frame start offset of the network antenna is sent from the LSU to the UE / phone device. Otherwise, if the UE / phone device periodically sends time-stamped samples of the processed network RF signal to the LSU, the LSU determines the location fix of the device and sends the location fix data back to the device.

[0473] In DL mode, the RF signal of the wireless network comes from one or more wireless network antennas. To avoid the effect of multipath on the resulting accuracy, the RF signal should be intercepted from the antenna or from the antenna connection to the wireless network device.

[0474] The bidirectional mode involves determining a location fix from both UL and DL operation, which can further improve location accuracy.

[0475] Some enterprise setups use one or more BBUs feeding one or more remote radio heads (RRHs), and then each RRH feeds the same ID to multiple antennas. In such environments, depending on the wireless network configuration, it may not be necessary to determine the frame start offset of the network antennas in downlink mode. This includes single BBU setups as well as multiple BBUs, where each BBU antenna is assigned a specific zone and adjacent zone coverage overlaps.

[0476] On the other hand, configurations where antennas fed by multiple BBUs are interleaved in the same zone require determining the frame start offset of the network antennas in DL mode.

[0477] In DL mode operation in a DAS environment, multiple antennas can share the same ID.

[0478] In an embodiment of the present invention, the position alignment algorithm is extended / deployed to separate the range of DAS antennas from observables generated by multipath mitigation processing from subcarriers of reference signals (including pilots and / or synchronization) and to obtain position estimates from the ranges of multiple DAS emitters (antennas).

[0479] However, these matching algorithms have a limit on the number of antennas that emit the same ID. The number of antennas that emit the same ID can be reduced by: 1. In a coverage zone, alternate antennas fed by different sectors of a sectorized BBU (a BBU can support up to 6 sectors). 2. In a given coverage zone, antennas fed from different sectors of a sectorized BBU are alternated with antennas fed from different BBUs. 3. Add a propagation delay element to each antenna. The delay value is chosen to be large enough to exceed the delay spread in the particular DAS environment (channel) but smaller than the length of the cyclic prefix (CP) so that multipath due to the additional delay does not result in ISI (Inter-Symbol Interference). Attaching a unique delay ID to one or more antennas further reduces the number of antennas emitting the same ID.

[0480] In one embodiment, a fully autonomous system may be provided that does not require any network investment by the customer. In such an embodiment, the system may operate in bands other than the LTE bands. For example, the ISM (Industrial Scientific And Medical) bands and / or white space bands may be used in locations where LTE service is not available.

[0481] This embodiment can also be integrated with macro / mini / pico / femto base stations and / or UE (mobile phone) devices. Although integration may require network investment by customers, it can reduce overhead and significantly improve TCO (total cost of ownership).

[0482] As described above in this specification, the PRS can be used by the UE for downlink observed time difference of arrival (DL-OTDOA) positioning. Regarding neighboring base station (eNB) synchronization, 3GPP TS 36.305 (Stage 2 functional specification of User Equipment (UE) positioning in E-UTRAN) specifies the timing of transfer to the UE. This timing is relative to the eNB serving the candidate cell (e.g., the neighboring cell). 3GPP TS 36.305 also specifies the Physical Cell Identity (PCI) and Global Cell Identity (GCI) of the candidate cell for measurement.

[0483] According to 3GPP TS 36.305, this information is distributed by an Enhanced Serving Mobile Location Centre (E-MLC) server. It should be noted that TS 36.305 does not specify the timing accuracy mentioned above.

[0484] 3GPP TS 36.305 also specifies that the UE returns downlink measurements to the E-MLC, including measurements of Reference Signal Time Difference (RSTD).

[0485] RSTD is a measurement taken between a pair of eNBs (see TS 36.214 Evolved Universal Terrestrial Radio Access (E-UTRA); Physical Layer measurements; Release 9). This measurement is defined as the relative timing difference between a subframe received from a neighboring cell J and the corresponding subframe of the serving cell I. Positioning reference signals are used to make these measurements. The results are reported back to the location server which calculates the position.

[0486] In one embodiment, a hybrid method may be defined to accommodate both the newly introduced PRS and existing reference signals, in other words, the hybrid method may be used / operate with PRS signals, with other reference signals (e.g., cell-specific reference signals (CRS)), or with both types of signals.

[0487] Such a hybrid approach offers the advantage that network operators can dynamically select the operating mode depending on the situation and network parameters. For example, PRS has better audibility than CRS but may result in up to 7% reduction in data throughput. CRS signals, on the other hand, do not cause a reduction in throughput. Also, CRS signals are backward compatible with all previous LTE releases (e.g., Release 8 and below). Thus, the hybrid approach offers network operators the ability to balance the trade-offs between audibility, throughput, and compatibility.

[0488] Furthermore, the hybrid method may be transparent to the LTE UE positioning architecture, for example, the hybrid method may operate within the 3GPP TS 36.305 framework.

[0489] In one embodiment, RSTD may be measured and transferred from the UE to the E-SMLC according to 3GPP TS 36.305.

[0490] UL-TDOA (U-TDOA) is currently in the research stage and is expected to be standardized in the upcoming Release 11.

[0491] An embodiment of UL-TDOA (uplink) is described herein above and shown in Figures 16 and 17. Figures 18 and 19 described herein below provide examples of alternative embodiments of UL-TDOA.

[0492] FIG. 18 illustrates an environment that may include one or more DAS and / or femto / small cell antennas. In this exemplary embodiment, each NSAU is equipped with one antenna. As illustrated, at least three NSAUs are required. However, additional NSAUs can be added to improve hearability, since each UE must be "heard" by at least three NSAUs.

[0493] Further, the NSAUs may be configured as receivers. For example, each NSAU receives but does not transmit information over the air. In operation, each NSAU may listen for wireless uplink network signals from UEs, each of which may be a mobile phone, a tag, and / or other UE devices.

[0494] Further, the NSAU can be configured to communicate with a Locate Server Unit (LSU) via an interface such as wired services, LAN, etc., and the LSU can communicate with a wireless or LTE network. The communication can be via a network API, where the LSU can communicate with, for example, an E-SMLC of an LTE network and can use wired services such as a LAN and / or a WAN.

[0495] Optionally, the LSU can also communicate directly with the DAS base stations and / or femto / small cells. This communication can use the same or modified network APIs.

[0496] In this embodiment, a Sounding Reference Signal (SRS) can be used for position detection, but other signals can also be used.

[0497] The NSAU may convert the UE's uplink transmit signal into a digital format, e.g., I / Q samples, and may periodically transmit the converted signals to the LSU with time stamps.

[0498] The DAS base station and / or femto / small cell may pass one or all of the following data to the LSU: 1) SRS, I / Q samples, and timestamps; 2) A list of UE IDs being served; and 3) For each UE with a UE ID, an SRS schedule (the SRS schedule includes SRS scheduling request configuration information and SRS-UL configuration information).

[0499] The information passed to the LSU may not be limited to the information described above, and may include any information necessary to correlate the uplink signals (e.g., UE SRS) of each UE device with the UE ID of each UE.

[0500] The functionality of the LSU can include deriving ranging calculations for the UE and position fixes for the UE. These determinations / calculations can be made based on information passed to the LSU from the NSAU, DAS base stations and / or femto / small cells.

[0501] The LSU may also determine the timing offset from available downlink transmission information passed to the LSU from the NSAU.

[0502] The LSU may then provide the wireless or LTE network with a location fix and other calculations and data for the UE. Such information may be communicated via a network API.

[0503] For synchronization purposes, each NSAU may receive, process, and time stamp samples of the downlink signal, and each NSAU may also periodically transmit a number of such samples, including the time stamps, to the LSU.

[0504] Each NSAU may also include an input configured for synchronization with an external signal.

[0505] Figure 19 illustrates another embodiment of UL-TDOA. In addition to the components illustrated in Figure 18, the environment of this embodiment may include one or more cell towers that can be used in place of DAS base stations and / or femto / small cells. Data from the one or more cell towers can be used to obtain a position fix for the UE.

[0506] Thus, advantages of this embodiment include obtaining a position fix using only one cell tower (eNB), and this embodiment can be configured to operate in a similar manner as described in Figure 18, except that one or more eNBs can replace the DAS base stations and / or femto / small cells.

[0507] Embodiments of the present invention relate to wireless communication, wireless network systems and radio frequency (RF) based systems for identification, tracking and location. The disclosed methods and systems use angle of arrival (AOA) positioning. According to one embodiment, RF based tracking and location is performed in cellular networks, but can be performed in any wireless system and RTLS environment. Closed systems can use digital signal processing (DSP) and / or software defined radio technology (SDR) implemented in software and / or hardware.

[0508] The present disclosure improves upon the methods and systems described in the applications incorporated by reference. One embodiment uses AOA positioning. This embodiment provides a significant improvement in location accuracy of current single tower / sector and / or multiple tower / sector location systems without requiring any hardware infrastructure modifications. In an embodiment, location accuracy is improved using TDOA (Time Difference of Arrival) measurements to determine AOA and / or LOB (Line of Bearing). For example, accuracy can be improved from hundreds of meters to less than 10 meters for a single tower / sector.

[0509] The embodiments of the present invention may be used in any wireless system / network and may include simplex, half-duplex and full-duplex modes of operation. The embodiments described below operate in wireless networks using various modulation formats including OFDM modulation and / or its derivatives. Thus, the embodiments described below operate in LTE networks and may also be applied to other wireless systems / networks including WiFi and other wireless technologies. As described herein, RF-based tracking and location implemented in LTE cellular networks can greatly benefit from accurate AOA positioning.

[0510] The present disclosure further introduces a location system that uses multiple receive channels in a cluster that is particularly well suited for indoor environments, introduces the use of DMRS (Demodulation Reference Signal) as a ranging signal, and uses one or more spectral estimation "subspaces" and high resolution non-parametric algorithms. The present disclosure also introduces embodiments that use elevation LOB measurements and determinations to improve location, and addresses problems associated with locating UEs by networks that use UTDOA methods based on LTE Release 11.

[0511] As mentioned above, in LTE environment, the current single tower / sector location technology suffers from lack of location accuracy. More specifically, the UE location error in cross-range length is very large, which may exceed 3,000 meters for a typical 120 degree beamwidth serving sector. At the same time, such U-TDOA requirements are not well suited to LTE network infrastructure.

[0512] For example, more recent U-TDOA location methods (LTE Release 11) require the UE to be detectable by at least three towers, require unambiguous cell locations and well-synchronized relative timing of reference signals between cells, or require precise knowledge of cell timing offsets with an accuracy of less than 20 ns. Neither requirement is guaranteed by typical network infrastructure. However, even meeting these requirements does not guarantee success, because a) in practice detection by four towers is often necessary, and b) a "no signal cone" effect occurs when the handset or UE is close to the serving cell / tower.

[0513] The "no signal cone" effect involves the handset or UE reducing its transmit power when in the vicinity of a serving cell / tower. The reduced transmit power creates an audibility challenge for adjacent non-serving towers that reduces the detectability of the handset by adjacent non-serving towers. Therefore, for all of the above, a solution is needed that does not require precise synchronization of TDOA and that either mitigates the no signal "cone" or provides it with an alternative when it cannot be mitigated. Additionally, there is a need to mitigate the effects of multipath on location accuracy.

[0514] FIG. 20 illustrates an example of a wireless tower 2000 that includes three LTE sectors (i.e., a first sector 2010, a second sector 2020, and a third sector 2030). As noted above, for a given LTE band, a typical macrocell MIMO sector uses a pair of horizontally separated antenna enclosures with each antenna enclosure housing a pair of antenna arrays. For example, the first sector 2010 uses antenna enclosures 2012 and 2014, the second sector 2020 uses antenna enclosures 2022 and 2024, and the third sector 2030 uses antenna enclosures 2032 and 2034. Thus, FIG. 20 illustrates six such antenna array enclosures arranged in a typical dual antenna array enclosure sector configuration.

[0515] Each LTE sector has an antenna azimuth (horizontal) main lobe beamwidth of approximately 120 degrees. The elevation (vertical) main lobe beamwidth is only approximately 10-20 degrees. These beamwidth configurations can improve antenna efficiency and reduce interference. To achieve this configuration, each antenna array housing (single housing) of the two antenna array housing LTE antenna sectors can include multiple antenna elements positioned in a vertical direction (i.e., in a row). For example, a typical LTE antenna housing array can include eight antenna elements per row.

[0516] This configuration of antenna elements would be useful for determining the AOA in the elevation plane if data (signals) could be accessed from each element, but current LTE antenna implementations such as these do not allow signals to pass from each element to the receive channel. In this way, the individual signals are available for determining the AOA in the elevation plane. In other words, each array enclosure is operating (used) as a standalone antenna with an azimuth (horizontal) mainlobe beamwidth of about 120 degrees and an elevation (vertical) mainlobe beamwidth of about 10-20 degrees. Typically, a MIMO sector antenna consists of two of the antenna array enclosures described above, each operating as a standalone antenna. If the network operator controlling the LTE antenna must be passed on the individual signals, they can be used for determining the AOA, as described further below.

[0517] LTE networks use MIMO technology, which improves the capacity of network performance at base stations and other aspects. Since UEs are primarily distributed in a relatively narrow azimuth plane (in elevation), each LTE sector antenna includes two or more of the above-mentioned antenna array enclosures (spaced horizontally), which can still be used for determining the AOA in the azimuth plane. One embodiment uses (utilizes) an eNB sector dual MIMO antenna for each sector. Each LTE antenna sector has a beamwidth of about 120 degrees, and the antenna array enclosures of a sector (e.g. enclosures 2012 and 2014 of the first sector 2010) are separated by a distance "d" between the antenna arrays, where "d" is typically 1.2-1.8 m. Although the embodiment discloses the use of two antennas in an antenna sector, any number of antenna arrays greater than two can be used.

[0518] The disclosed embodiments improve upon the existing U-TDOA technique (LTE 3GPP Release 11) of network-based UE tracking and location. Note that this U-TDOA approach can rely on uplink time difference of arrival techniques and on multilateration methods. The U-TDOA technique can also use SRS (Sounding Reference Signal) as a ranging signal. The described embodiments also improve the accuracy of current single tower / sector location techniques without requiring any infrastructure changes. Current single tower / sector location techniques can use a location process that utilizes RTT (Round Trip Time, also called Timing Advance (TADV)), measured at the serving cell, along with the serving sector beamwidth of the serving cell. See FIG. 10.

[0519] In one embodiment, the SRS symbols are transmitted as ranging signals by the known UE, which leverages the existing infrastructure MIMO sector antenna pairs on the macro / radio tower for Angle of Arrival (AOA) measurements from two MIMO sector antennas. To support MIMO functionality, each sector antenna pair includes a MIMO RF transceiver and has transmit / receive channels that are fully carrier frequency / phase and timing coherent. One embodiment also leverages the evolved NodeB (eNB) of each sector antenna pair receive channel to support MIMO functionality (the eNB of each sector antenna pair receive / transmit channel is fully carrier frequency / phase and timing coherent).

[0520] The AOA system and process described herein is a unique adaptation to conventional location methods depicted in Figure 21. Figure 21 provides a conceptual explanation of conventional AOA location detection of how the AOA is determined when signals from each element 2100 of an antenna array are passed through a receive channel, where the AOA is determined based on the phase difference (θ) of the ranging signals collected by each antenna element 2100. In Figure 21, an antenna array with six antennas is shown. However, the principles can work for any number greater than two, and test results based on the present disclosure confirm that two antennas are sufficient.

[0521] Angle of arrival works very effectively with narrowband emitters with two or more closely spaced antennas feeding a pair of time-, phase-, and frequency-coherent receiving systems that compare the phase difference of the signals collected by each antenna. That phase difference is converted to AOA. AOA methods work best on narrowband signals; in general, they do not work well with instantaneous wideband signals such as SRS. On the other hand, time difference of arrival (TDOA) techniques do not work well for narrowband signals (and do not work at all on a single unmodulated carrier), but work well for instantaneous wideband signals.

[0522] While the MIMO antenna sectors and SRS (for ranging) can be used to determine the AOA as disclosed herein, there are still some inherent challenges associated with using two MIMO sector antennas and an SRS as a ranging signal compared to traditional AOA location concepts, which require mitigation in order to make accurate AOA measurements, as described further herein.

[0523] First, when the SRS symbol signal is used for ranging, there is a challenge because the bandwidth of the SRS symbol signal is very wide, which can reach up to 20 MHz. As described above with respect to FIG. 21, the conventional AOA technique compares the phase differences of the ranging signals collected by each antenna array element 2100 and converts those phase differences into the AOA of radio waves. Since the phase difference is also a function of the ranging signal frequency, the conventional AOA method works effectively with narrow bandwidth ranging signals and cannot work well for instantaneous wideband signals such as SRS signals.

[0524] Second, due to multipath effects, typical AOA solutions (as described above) produce multiple lines of bearing (LOBs), and as a result, the desired LOB associated with the direct line of sight (DLOS) often cannot be determined among the multiple LOBs.

[0525] Third, there is a 2π wraparound of the phase difference in the two antenna arrays for a sector. Conventional AOA techniques may be based on closely spaced antenna arrays, as in FIG. 20. Each antenna array is typically required to be placed less than one wavelength of the carrier frequency relative to the other antenna (optimum spacing is half a wavelength). However, this is not the case for cell-sector MIMO antenna spacing. For example, the wavelength for an LTE Band 12 uplink operating in the 700 MHz band is 0.422 meters or 42.2 cm, but the MIMO sector antennas are typically separated by d, which is 1.2 to 1.8 m, which covers multiple wavelengths. This may result in a 2π wraparound of the phase difference in the two antenna arrays, which may cause AOA ambiguities.

[0526] Fourth, AOA ambiguity due to larger antenna array separation or reflections can arise from the presence of a mirror image of the desired (true) LOB. The antenna array cannot distinguish between the actual angle of wave incidence θ (see FIG. 21) and its mirror image -θ. Thus, there are two LOBs, one of which is the desired LOB where the handset (UE) is located, and the other is the undesired LOB mirror image.

[0527] The AOA system and method described herein utilizes the baseband spectrum of the SRS symbols transmitted by the known UE. Once the known phase encoding is removed, the resulting burst is multiple orthogonal unmodulated subcarriers of an OFDM waveform. At this point, multipath mitigation (as described in the related application to this invention) can be used to estimate the small linear phase shift between the subcarriers that results from the time difference of flight (TDOA) of the radio waves to each antenna.

[0528] These phase shift determinations (across all subcarriers) can then be used to determine the apparent AOA LOB. The embodiments of the systems and methods described herein use all of the information in all of the subcarriers, which is 288 subcarriers for a 10 MHz bandwidth (48 resource blocks) SRS signal.

[0529] The multipath mitigation described in the cross-referenced application can resolve the direct path from the multipath and provide an estimate of the time difference of flight for direct line of sight (DLOS), i.e., the time difference for the shortest path. This method can solve the AOA ambiguity problem associated with the wide bandwidth of the SRS ranging signal and the aforementioned SRS ranging signal multipath. This approach can also address the multiple LOB problem caused by multipath and can create the desired LOB associated with the direct line of sight (DLOS) path.

[0530] Multipath mitigation engines / algorithms use high-resolution spectral estimation and statistical algorithms (analysis), such as Matrix Pencil (MP), Multiple Signal Characterization (MUSIC) or root-MUSIC, Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT), Multitaper Method (MTM), Pisarenko Harmonic Decomposition (PHD), and Relaxation (RELAX) algorithms (analysis).

[0531] The MTM can assemble statistically independent estimates of the power spectrum using a set of orthogonal tapers. An F-test is then used to compare the expected degrees of freedom for each spectral bin to estimates calculated from the individual multitaper spectra. The direct path return is expected to have a single line component after demodulation, while the multipath return contains additional degrees of freedom due to specular reflections. Thresholding of the F-test ratio at a specified confidence level can be used to identify the line components in the spectrum. The identified line component frequencies can then be used to calculate non-parametric estimates of the direct path delays or as an aid in selecting the desired direct path output frequencies from the high-resolution spectral estimates.

[0532] Multipath mitigation operates at the SRS signal baseband and not at the carrier frequency, which allows the above-mentioned 2π wraparound of phase difference AOA ambiguity to be resolved since the highest frequency in the baseband signal spectrum is a fraction of the carrier frequency (e.g. 10 MHz vs. 700 MHz).

[0533] Mitigation of the fourth AOA ambiguity item (LOB mirror image) mentioned above is based on the known radiation pattern of the serving sector. Since the served handset (UE) is always within the serving sector radiation pattern, the desired LOB can be uniquely identified. This knowledge of the true LOB of the serving tower / sector allows the determination of one or more additional LOBs from one or more additional non-serving sectors of the same tower and / or one or more non-serving adjacent tower sectors. For example, depending on exactly where the UE is relative to the serving sector's main beam lobe centerline, one or more of the additional tower sectors can also see the UE and generate a unique LOB(s). Using two or more LOBs from different sectors further enhances the location accuracy. Most location cases fall into this scenario.

[0534] The method and system embodiments improved herein show that for two receivers, the TDOA Line of Bearing (LOB) is a hyperbola, and when the receiver antennas are closely spaced, the LOB approaches the AOA LOB, which is a straight line. The present disclosure starts with TDOA measurements between two or more closely spaced antennas, but unlike previous methods and systems, determines the actual AOA LOB based on the signal structure of the ranging signal (SRS). As a result, location accuracy is greatly improved. Also, the ranging signal detection (aka "triggering") process was not disclosed in the previously disclosed related methods and systems.

[0535] One example of a process flow for implementing the present disclosure will now be described with reference to FIG. 22. Signals from one or more repeater tower / sector sector MIMO antenna pairs 2210 are passed to a dual coherent receive channel 2220. The dual coherent receive channel 2220 generates I / Q samples, and a number of these samples are periodically buffered in an I / Q sample buffer 2230 under the control of a position detection processor 2260. In this manner, the dual coherent receive channel 2220 and I / Q sample buffer buffer or collect I / Q samples from and around the LTE subframe slot symbol that is assigned to the SRS, for example, in the last symbol of an uplink FDD subframe. A ranging signal detection processor 2240 receives these buffered I / Q samples to locate the SRS and establish the start of the SRS symbol.

[0536] The location detection processor 2260 provides the SRS parameters to the ranging signal detection processor 2240, which performs matched filtering on the buffered I / Q samples from each receive channel. The output of this matched filter is peak detected to determine the I / Q sample that is closest to the start of the SRS symbol.

[0537] A 1024-point FFT is applied to the SRS bandwidth having 48 resource blocks to produce the demodulated SRS resource elements. This is an integration process with a processing gain of 30.1 dB. This 30 dB processing gain allows reliable ranging signal detection for SNRs from -10 dB to -12 dB. In one embodiment, the ranging signal detection is reliable above a predefined threshold, i.e., a threshold SNR that is a calculated value based on the desired detection performance, e.g., 15 dB. Those skilled in the art will recognize that such ranging or SRS symbol signal detection by the signal detection processor is also known as "triggering." Alternatively, this ranging signal detection is referred to as "link closure." Reliable AOA observables cannot be determined without link closure. The 30 dB processing gain and matched filtering described above are not available for data / voice communications. As a result, reliable location detection can be achieved even when signals are otherwise unsuitable (too weak and / or too poor) for data / voice communications.

[0538] In one embodiment, if the SRS starting point is the same for each channel, the first sample estimate of the strongest signal is used. This is nominally from the serving cell sector antenna, but this is not always the case. For example, when the UE is located between the azimuth directions of two sectors, either sector may output a stronger signal depending on the propagation path. However, the multipath mitigation processor 2250 can detect each of the above paths and other paths that may also be present, as described further below.

[0539] A pointer to the SRS symbol start is then passed by the ranging signal detection processor 2240 to the multipath mitigation processor 2250. The multipath mitigation processor 2250 utilizes all of the SRS subcarrier data (baseband) to provide sub-sample estimates of the DLOS (direct line of sight) time of flight for each internal sector antenna array 2210, from which the multipath mitigation processor 2250 unambiguously and highly accurately estimates the TDOA between the two internal sector antenna arrays 2210. The TDOA estimates are sent to the location detection processor 2260.

[0540] The location processor 2260 determines a phase difference of arrival from the TDOA estimate. That phase difference of arrival combined with the TDOA estimate creates an unambiguous and highly accurate AOA estimate. The location processor 2260 then determines the AOA for the antenna array sector (i.e., the reliable AOA observable), which is then used (by the location processor 2260) to determine the LOB of the UE from that antenna array sector (i.e., the tower on which the antenna array is mounted). The location processor 2260 can then determine the desired LOB, e.g., by performing LOB ambiguity resolution, which is used to determine the location of the UE.

[0541] To further reduce tracking and location errors and further improve the reliability of location based on weak and poor signals, an embodiment may implement one or more noise reduction and / or post-processing techniques that may include position matching methods, maximum likelihood estimation (e.g., Viterbi algorithm), and / or minimum variance estimation (e.g., Kalman filter). Noise reduction methods may also include coherent summation, non-coherent summation, matched filtering, and / or temporal diversity.

[0542] As previously mentioned, the disclosed system and method utilizes all of the SRS subcarrier data (288 subcarriers in the case of 48 resource blocks). As a result, reliable location can be achieved using signals that are unsuitable for data / voice communication because they are too weak or in poor condition. This therefore eliminates the "no signal cone phenomenon" as the serving cell maintains Es / N0, e.g., SNR, and other signal parameters sufficient for reliable communication. Additionally, in one embodiment, the location reliability of weak and poorly-conditioned signals can be further improved by using noise reduction methods such as coherent summing, non-coherent summing, matched filtering, and temporal diversity.

[0543] The angular error of the AOA estimation is measured by the standard deviation of the AOA estimation, which depends on the SNR (Es / N0). For a single tower / sector, the error ellipse has one axis along the range direction (i.e., the distance from the UE to the tower) and another axis in a cross-range direction, e.g., perpendicular to the range direction. Based on tests performed using simulated data, the standard deviation of the AOA estimation error decreases with a better Es / N0. For example, an Es / N0 of -6 dB correlates with a standard deviation of 0.375 degrees.

[0544] Table 1 below summarizes simulated results for the cross-range AOA estimated angle error standard deviation as a function of Es / N0. In this table, the cross-range multiplier (i.e., multiplication factor) determines the relationship between range and cross-range error standard deviation. The cross-range standard deviation is:

number

[0545] Table 1. Multiplication factors for crossover range errors ---------------------------------------------------------------------------------- Es / N0(dB) Intersection range multiplier (meters per meter): σ f ---------------------------------------------------------------------------------- -9 0.00870 -6 0.00670 -3 0.00440 ----------------------------------------------------------------------------------

[0546] From Table 1, at a 1500 meter range (i.e., distance from the UE to the radio tower), the cross-range (i.e., perpendicular to the range of the UE's location) standard deviation error is 6.5 meters with an Es / N0 of -9 dB, or 3.3 meters with an Es / N0 of -3 dB. Note: These location accuracy levels were achieved in the presence of multipath.

[0547] As described above, embodiments allow for the generation of reliable position determinations even when based on signals that are otherwise unsuitable (too weak and / or too bad) for data / voice communications, e.g., when Es / N0 is too low. This capability can increase the likelihood of link closure and producing reliable AOA observables from more than one tower. In one embodiment, enhanced UE location accuracy can be produced without the use of independent range estimates (e.g., RTT) when AOA observables are available from multiple towers. This improvement in accuracy is possible in part due to multipath mitigation.

[0548] For example, in one case where a UE LOB is available from two towers and the Geometric Dilution of Precision (GDOP) does not exceed 2, the combined cross-range error standard deviation for the two LOBs is:

[0549]

number

[0550] In another case, if the UE location is some distance away from the serving sector beam lobe centerline, there is a chance that one or more other tower sectors may "encounter" the UE and generate a unique LOB(s) as well. The availability of these two or more LOBs from different sectors further enhances location accuracy. Most location decisions are likely to fall into this latter scenario, since it is unlikely that the UE will be on the serving sector beam lobe centerline most of the time, based on simple probability alone.

[0551] The ability to produce reliable AOA observables from two or more sectors / towers also substantially reduces the no signal cone phenomenon, which is further eliminated when using the serving sector AOA+RTT (round trip time) location method since the serving sector keeps Es / N0, e.g., SNR, and other signal parameters sufficient for reliable communication.

[0552] It should be further noted that the current LTE infrastructure supports azimuth AOA estimation but is not suited for estimating elevation angles (i.e., two-dimensional (2D) angles (AOA). At the same time, assuming that the UE is on the ground or its elevation angle is known, e.g., with barometric pressure measurements, knowledge of both azimuth and elevation AOA angles (LOB) can enable the UE location to be determined from a single sector / tower without the need for independent (e.g., RTT) range estimation. Future LTE deployments may have sector antenna array topologies supporting azimuth and elevation communication, which would serve to further enhance current embodiments.

[0553] Furthermore, to achieve a very narrow vertical radiation pattern, all currently used sector antenna array enclosures consist of a vertical antenna array with, for example, eight antenna elements. Currently, these vertical elements are internally integrated into the sector antenna unit and are not externally accessible. However, this sector antenna structure may change in the future, opening up further possibilities for AOA estimation. For example, once signals from both azimuth and elevation antenna arrays become accessible, embodiments of the AOA estimation process of the present invention may be applied to both arrays (or any combination of antenna elements in the arrays) producing azimuth and elevation LOBs from a single tower and / or multiple towers.

[0554] In one embodiment, the system and method of the present disclosure is implemented using existing wireless network infrastructure components and / or resources (e.g., a base station (eNB) for receive channels and / or an Evolved Serving Mobile Location Centre (E-SMLC) in which the location processor 2260 is implemented). In other embodiments, the system and method of the present disclosure is implemented without reliance on existing network infrastructure components and / or resources. In some embodiments, a separate dedicated component, e.g., a location processor server, can communicate with the network infrastructure. Yet another embodiment combines wireless network infrastructure components / resources with separate dedicated components / resources.

[0555] One embodiment utilizes a base station receive channel. Another embodiment can have a separate receive channel sharing the sector antenna array receive signal with the base station receive channel (receiver). These receive channels can include external synchronization inputs, such as GPS, NTP (Network Time Protocol), IEEE1588-PTP (Precision Time Protocol), etc.

[0556] In yet another embodiment, the receive channels can be connected to an independent (from the network infrastructure) antenna array and multiple receive channels can be combined into a cluster. The antenna arrays and receive channels can be configured as multiple antenna array and receiver (receive channel) clusters. Within each cluster, the receivers are frequency / timing / phase coherent. However, the synchronization requirements between each array-receiver cluster do not require coherency.

[0557] In such an embodiment, one or more clusters may be grouped together to form a group unit. Multiple group units may be deployed. The group units may be synchronized, but the synchronization requirements may be relaxed and no coherency is required. Multiple group unit synchronization may be achieved by an external timing input, such as GPS, Network Time Protocol (NTP), Precision Time Protocol (IEEE1588-PTP), etc. The processing software may be implemented on a network infrastructure or an independent server in communication with the network infrastructure. Each group unit may also include additional computational resources to house some of the processing software, e.g., UE location determination processing may be distributed between multiple group units and an independent server or network infrastructure.

[0558] An example location for the group unit 2300 is shown in Figure 23. In this embodiment, the group unit 2300 is housed in a 4U x 19 inch (48.3 cm) enclosure 2310. In the example represented by Figure 23, the group unit 2300 includes two coherent receiver cards 2320, an I / Q sample buffer card 2330, a synchronization module card 2340, an additional computational resource card 2350, a power supply 2360, and multiple external timing inputs 2370.

[0559] An example of a process flow performed by a group unit (e.g., group unit 2300 of FIG. 23) during operation is described below with respect to FIG. 24. Signals from the multiple antenna array 2410 are passed to a coherent receiver cluster 2420, which generates I / Q samples. A synchronization module 2440 controls a number of these I / Q samples, which are periodically buffered in an I / Q sample buffer 2430. A ranging signal detection processor 2460 searches for the SRS and establishes the start of the SRS symbol. An independent location detection processor server 2480 provides the SRS parameters to the ranging signal detection processor 2460 with additional computational resources 2450, which then performs matched filtering of the I / Q samples from each receive channel. The output of this matched filter is peak detected to determine the I / Q sample closest to the start of the SRS symbol.

[0560] The I / Q sample buffer(s) are shared by the ranging signal detection processor 2460 and the multipath mitigation processor 2470. The ranging signal detection processor 2460 searches the I / Q sample buffer(s) at the SRS symbol start and provides the symbol start buffer location (address) to the multipath mitigation processor 2470 (described in the cross-referenced application). The multipath mitigation processor 2470 utilizes the SRS subcarrier data (baseband) to provide sub-sampled estimates of DLOS (direct line-of-sight) time of flight from the UE to each antenna of the antenna array 2410. In one embodiment, the multipath mitigation processor 2470 utilizes all of the SRS subcarrier data (baseband) to provide sub-sampled estimates of DLOS time of flight. Based on these DLOS time of flight estimates, the multipath mitigation processor 2470 or additional computational resources 2450 determine an explicit TDOA estimate between the antenna arrays 2410.

[0561] As part of this process, many decisions may be performed by the location processor 2480 or the additional computational resources 2450 or distributed among these entities / elements. These decisions include: (i) determining the arrival phase difference from the TDOA estimates, (ii) determining the AOA observables for the antenna array 2410, (iii) determining the UE's LOB from the antenna array 2410 using information from the determined AOA observables, and (iv) determining the desired LOB, for example, by removing LOB ambiguity. These decisions are based on the TDOA estimates developed by the multipath mitigation processor 2470 or the additional computational resources 2450, or a combination thereof.

[0562] The results of these calculations may be collected by the location processor 2480. Alternatively, the functionality of the ranging signal detection processor 2460 and the multipath mitigation processor 2470 may be implemented in the additional computational resources 2450.

[0563] To reduce any remaining tracking and location errors and further improve location reliability for weak and poorly-conditioned signals, an embodiment may implement one or more noise reduction and / or post-processing techniques including position matching methods, maximum likelihood estimation (e.g., Viterbi algorithm), and / or minimum variance estimation (e.g., Kalman filter). Noise reduction methods may include coherent summation, non-coherent summation, matched filtering, and / or temporal diversity.

[0564] One embodiment of a process flow performed by the synchronization module 2440 can provide frequency / time / phase coherent clocks and other timing signals to a controller associated with the coherent receiver cluster 2420 and the I / Q sample buffer 2430. The synchronization module 2440 can also synchronize the clocks and other synchronization signals with one or more external signals input from an external time synchronization 2490, such as GPS / GNSS synchronization or Ethernet synchronization. The synchronization module 2440 can also communicate with additional computational resources 2450 and a location process server 2480.

[0565] In one embodiment, the I / Q samples collected (buffered) in the I / Q sample buffer 2430 from each receiver cluster 2420 can be aligned in time order. As shown in FIG. 25, this function can be performed by a controller 2535 that is part of the I / Q sample buffer 2430. The I / Q sample buffer controller 2535 allocates and stores a number of samples from each receiver 2525 of the coherent receiver cluster 2420 in each buffer of the I / Q sample buffer 2430. The synchronization module 2440 can control the number of I / Q samples collected (stored) by I and Q sample command / control signals and data 2543. The synchronization module 2440 can also generate signals to enable (start) the I / Q sample collection process by clock and other timing signals 2545. The process can be automatically stopped once a specified number of I / Q samples have been transferred.

[0566] Many buffer alignment methods / techniques can be used, ranging from a simple timestamp at the beginning of each buffer to inserting a beacon sequence into the I / Q sample stream. These techniques can be used for I / Q sample alignment in real time or (later) offline. Although the embodiment currently uses a beacon sequence mechanism for alignment, various modifications, adaptations and alternative embodiments thereof can be made. For example, the alignment process can be performed by the I / Q sample buffer controller 2535, the ranging signal detection processor 2460 and / or additional computational resources 2450, or some combination thereof.

[0567] For example, the ranging signal detection processor 2460 can search for the SRS and determine the start of the SRS symbol. The search process can be further accelerated if the location of the SRS signal within the buffered frames can be estimated prior to the search. Thus, in one embodiment, one or more of the coherent receiver clusters 2420 are switched from uplink mode to downlink mode under the command of the synchronization module 2440. Thereafter, under the command of the synchronization module 2440, a number of I / Q samples from each receiver 2525 can be stored in individual buffers of the I / Q sample buffer 2430, where the I / Q sample buffer controller 2535 can provide a timestamp at the beginning of each buffer in addition to storing the I / Q samples in the individual buffer. The ranging signal detection processor 2460 and / or another element of the group unit can determine the downlink frame start time of the individual buffer and then estimate each serving cell frame start time relative to the timestamp.

[0568] This location downlink frame start time information is then forwarded to the synchronization module 2440. The synchronization module 2440 may have a timer / counter that is synchronized with a timer / counter of the I / Q sample buffer controller 2535. Based on the timestamp and the location downlink frame start time, the synchronization module 2440 may be able to predict the upcoming downlink frame / subframe time for one or more serving cells.

[0569] Alternatively, the synchronization module 2440 can initiate the I / Q sample collection process. Then, the trigger may be time-stamped, buffer alignment can be requested, and each serving cell frame start time can be estimated relative to the start of the buffer (the trigger). The synchronization module 2440 can then determine the downlink frame start time of each individual buffer and then estimate each serving cell frame / subframe start time relative to the timestamp. Note: Since all receivers in a cluster are coherent, it may not be necessary to time-stamp and / or mark each buffer separately within the cluster.

[0570] The uplink frame / subframe start time is correlated to the downlink frame / subframe start time via a timing advance, which is known to the eNB for each UE, and is the negative offset between the start of the received downlink subframe and the transmitted uplink subframe at the UE. Since the SRS position within the uplink frame / subframe is fixed and known a priori, the synchronization module 2440 may be able to estimate the approaching SRS signal time, and this information is used by the ranging signal detection processor 2460. The eNB(s) also provide the SRS parameters per UE, SFN (system frame number), and time advance information or RTT (round trip time) for each served UE to the location processor server 2480, which passes it on to the synchronization module 2440 and other entities in the group unit.

[0571] Alternatively, the SFN information may be derived by the synchronization module 2440 and / or other entities in the group unit. One or more SRSs may also be located and the SRS and downlink frame relationship determined from using the SRS parameters. During the location detection operation, the actual SRS location may be compared to the location / window estimated by the synchronization module 2440 and / or other entities in the group unit and fed back to the synchronization module 2440 to adjust the SRS location estimate.

[0572] After estimating the uplink SRS timing, the synchronization module 2440 can command the coherent receiver cluster 2420 to switch back to uplink mode. After this command, the location processor server 2480 can be started / restarted. The SRS signal location estimation process described above can be resumed by a command from the location processor server 2480 or the additional computational resources 2450.

[0573] As mentioned above, the standardized TDOA discussed in the network-based location standard has drawbacks including UE hearability / detectability issues and strict network synchronization (<20ns) requirements for accuracy. However, if these issues can be mitigated, joint TDOA and AOA tracking location can be performed. The multipath mitigation engine 2470 can produce accurate TDOA estimates for multiple towers (serving and adjacent towers), and the location processor server 2480 functionality can also support joint AOA and TDOA or single TDOA location and tracking.

[0574] One such example is the embodiment described above that aggregates multiple receive channels of a cluster, which can mitigate the shortcomings of standardized TDOA due to the inherent synchronization capabilities within and between clusters, and the ability to add more clusters, i.e. antenna arrays, to mitigate audibility / detectability and no signal cone problems.

[0575] Since one or more of the above described embodiments use SRS as a ranging signal, these embodiments can (and do) work with all handsets and UE phones because SRS transmission is part of the wireless network control plane and is controlled by the network infrastructure and not by the handset / UE.

[0576] In the same manner, the processes described herein can be applied directly to the demodulation reference signal (DMRS), i.e., used as a ranging signal. The DMRS is used for channel estimation and for coherent demodulation. Like the SRS, the DMRS is part of the uplink transmission from the UE / handset. The DMRS signal structure is identical to the SRS, and the same methods described herein work with the DMRS without requiring network infrastructure or modifications to the devices. The DMRS is transmitted by all handsets / UE phones, since the DMRS transmission is part of the uplink transmission and is transmitted even when the UE is in idle mode.

[0577] Still, there are differences between using SRS or DMRS signals. SRS can be configured to span many resource blocks, i.e., wideband signals, e.g., 10 MHz, and the standard deviation of the AOA error decreases with the larger bandwidth. In contrast, DMRS is usually narrowband, resulting in lower accuracy. Also, DMRS does not have the bandwidth-dependent processing gain of SRS, and produces fewer reliable AOA observations from adjacent non-serving towers. Thus, when SRS is used as a ranging signal, a location accuracy of between 3-5 meters can be achieved at a range of 1500 meters from the tower. For DMRS, the location accuracy drops to 20-40 meters at the same distance from the tower.

[0578] However, unlike SRS, DMRS does not consume any network bandwidth, it is always included in uplink transmissions, even when the UE is in idle mode. It is not optional like SRS, and does not need to be enabled / configured / reconfigured for location detection, for example. Also, DMRS is used for channel estimation and for coherent demodulation. Furthermore, if the DMRS is bad or cannot be properly decoded by the base station for any reason, the uplink transmission will not be decoded.

[0579] In a single tower / sector serving sector AOA+RTT (round trip time) location method described in one of the cross-referenced applications, the UE location cross-range error is determined by the serving sector radiation pattern (typically 120 degrees), e.g., the AOA angle error is equal to 120 degrees, resulting in a cross-range error of over 3000 meters at a distance of 1500 meters from the UE to the radio tower. An embodiment according to the present disclosure dramatically reduces this single tower / sector AOA angle error to fractional degrees (see Table 1), resulting in a cross-range error of a few meters versus 3000 meters (at a range of 1500 meters) and does not require infrastructure changes.

[0580] The distance from the UE to the radio tower is derived from the RTT measurements, which have a resolution of + / - 10 meters, but the RTT measurements are affected by multipath, the effects of which can be mitigated using the multipath mitigation engine described in the cross-referenced application.

[0581] It should also be appreciated that various modifications, adaptations, and alternative embodiments may be made within the scope and spirit of the present disclosure. For example, in another embodiment, one or more super-resolution (subspace) algorithms (methods) and / or one or more multi-taper methods may be used by the multi-path mitigation processor for reliable and accurate separation of DLOS (direct line-of-sight) paths (time of flight) from other reflected (multi-path) signal paths. In addition to SRS and DMRS, other wireless network radio signals, such as random access preamble signal(s), may be used as ranging signals for position determination.

[0582] Thus, although different embodiments of the system and method have been described, it should be clear to one skilled in the art that certain advantages of the described method and device have been achieved. In particular, it should be understood by one skilled in the art that a system for tracking and locating objects can be assembled at very little additional cost using a combination of FGPA or ASIC and standard signal processing software / hardware. Such a system is useful in various applications, for example, for locating people in indoor or outdoor environments, harsh and harsh environments, etc. The invention as described in the claims of the present application as originally filed is set forth below. [C1] 1. A method for determining a location of a user equipment (UE) in a wireless system, comprising: buffering a plurality of in-phase and quadrature (I / Q) samples generated from signals provided by receive channels associated with two or more antennas; receiving reference signal parameters describing a reference signal within the range of signals; identifying the reference signal within the range of signals from each receive channel among the receive channels; determining an angle of arrival between a baseline of the two or more antennas and incident energy from the UE to the two or more antennas based on the identified reference signals from each receive channel; calculating the location of the UE using the angle of arrival; The method includes: [C2] The method of claim 1, wherein the step of identifying the reference signal includes determining a start time of the reference signal using the reference signal parameters and the plurality of buffered I / Q samples. [C3] The method of claim 1, wherein the step of identifying the reference signal includes detecting a peak in a matched filter output associated with each receive channel to identify a particular I / Q sample among the plurality of I / Q samples that is closest to a start time of the reference signal. [C4] The method of C1, wherein the step of determining the angle of arrival includes determining a sub-sample estimate of a direct line-of-sight (DLOS) time of flight from the UE to each antenna among the two or more antennas. [C5] The method of claim 4, wherein the DLOS time of flight from the UE to each antenna is used to determine a time difference of arrival of the reference signal between the two or more antennas. [C6] The method of claim 4, wherein the DLOS time of flight from the UE to each antenna is used to determine a phase shift between individual subcarriers of the reference signal. [C7] The method of claim 1, wherein the two or more antennas are located in one sector of a radio tower. [C8] The method of C1, wherein a distance between the two or more antennas is small relative to a distance from the two or more antennas to the UE. [C9] The method of claim 1, wherein the reference signal includes a sounding reference signal (SRS) or a demodulation reference signal (DMRS). [C10] The method of claim 1, wherein the angle of arrival is associated with a line of bearing (LOB) between a serving sector of a wireless tower and the UE. [C11] determining a second LOB between a non-serving sector of the wireless tower and the UE using the angle of arrival; The method of C10, further comprising: [C12] determining a third LOB between a non-serving sector of a second wireless tower and the UE using the angle of arrival; The method of C11, further comprising: [C13] 1. A method for determining an elevation angle of a user equipment (UE) in a wireless system, comprising: buffering a plurality of in-phase and quadrature (I / Q) samples generated from signals provided by receive channels associated with two or more antenna elements of a sector antenna pair, one antenna element of the two or more antenna elements being vertically displaced with respect to another antenna element of the two or more antenna elements; receiving reference signal parameters describing a reference signal within the range of signals; identifying the reference signal within the range of signals from each receive channel among the receive channels; determining a vertical angle of arrival between a baseline of the sector antenna pair and energy incident on the two or more antenna elements from the UE based on the identified reference signals from each receive channel; calculating the elevation angle of the UE using the vertical angle of arrival; The method includes: [C14] The method of C13, wherein two or more antenna elements are located in a single antenna housing of the sector antenna pair. [C15] The method of C13, wherein two or more antenna elements are located on different antenna housings of the sector antenna pair. [C16] 1. A method for determining a location of a user equipment (UE) in a wireless system, comprising: buffering a plurality of in-phase and quadrature (I / Q) samples generated from signals provided by receive channels associated with two or more antennas including a plurality of antenna elements, one antenna element of the plurality of antenna elements being vertically displaced with respect to another antenna element of the plurality of antenna elements; receiving reference signal parameters describing a reference signal within the range of signals; identifying the reference signal within the range of signals from each receive channel among the receive channels; determining a horizontal angle of arrival between a baseline of the antenna and incident energy from the UE to the two or more antennas and a vertical angle of arrival between the baseline of the antenna and incident energy from the UE to the two or more antennas based on the identified reference signals from each receive channel; calculating the position of the UE using the horizontal angle of arrival and the vertical angle of arrival; The method includes: [C17] The method of C16, wherein the horizontal angle of arrival and the vertical angle of arrival are each associated with a line of bearing (LOB) between a serving sector of a radio tower and the UE. [C18] The method of C17, further comprising: utilizing the horizontal angle of arrival to determine a second LOB between a non-serving sector of the wireless tower and the UE. [C19] The method of C17, further comprising: utilizing the vertical angle of arrival to determine a second vertical angle of arrival between a non-serving sector of a second wireless tower and the UE. [C20] clustering two or more coherent receive channels among the receive channels; forming a group unit from the clusters; synchronizing said group units using an external synchronization input; The method of C16, further comprising:

Claims

1. 1. A method for determining a location of a user equipment (UE) in a wireless system, comprising: buffering a plurality of in-phase and quadrature (I / Q) samples generated from signals provided by receive channels associated with two or more antennas; receiving reference signal parameters describing a reference signal within the signal; identifying a reference signal within the signal from each receive channel among the receive channels, wherein the step of identifying the reference signal includes determining a start time of the reference signal using the reference signal parameters and the plurality of buffered I / Q samples, and the step of identifying the reference signal includes detecting a peak in a matched filter output associated with each receive channel to identify a particular I / Q sample among the plurality of I / Q samples that is closest to a start time of the reference signal. determining an angle of arrival between a baseline of the two or more antennas and incident energy from the UE to the two or more antennas based on the identified reference signals from each receive channel; calculating the position of the UE using the angle of arrival; The method includes:

2. 2. The method of claim 1, wherein the step of determining the angle of arrival comprises determining a sub-sample estimate of a direct line-of-sight (DLOS) time of flight from the UE to each antenna among the two or more antennas.

3. The method of claim 2 , wherein the DLOS time of flight from the UE to each antenna is used to determine a time difference of arrival of the reference signal between the two or more antennas.

4. The method of claim 2 , wherein the DLOS time of flight from the UE to each antenna is used to determine a phase shift between individual subcarriers of the reference signal.

5. The method of claim 1 , wherein the two or more antennas are located in one sector of a radio tower.

6. The method of claim 1 , wherein a distance between the two or more antennas is small relative to a distance from the two or more antennas to the UE.

7. The method of claim 1 , wherein the reference signal comprises a sounding reference signal (SRS) or a demodulation reference signal (DMRS).

8. The method of claim 1 , wherein the angle of arrival is associated with a line of bearing (LOB) between a serving sector of a radio tower and the UE.

9. determining a second LOB between a non-serving sector of the radio tower and the UE utilizing the angle of arrival; The method of claim 8 further comprising:

10. determining a third LOB between a non-serving sector of a second wireless tower and the UE using the angle of arrival; The method of claim 9 further comprising:

11. 1. A method for determining a location of a user equipment (UE) in a wireless system, comprising: buffering a plurality of in-phase and quadrature (I / Q) samples generated from signals provided by receive channels associated with two or more antennas including a plurality of antenna elements, one antenna element of the plurality of antenna elements being vertically displaced with respect to another antenna element of the plurality of antenna elements; receiving reference signal parameters describing a reference signal within the signal; identifying a reference signal within the signal from each receive channel among the receive channels, wherein the step of identifying the reference signal includes determining a start time of the reference signal using the reference signal parameters and the plurality of buffered I / Q samples, and the step of identifying the reference signal includes detecting a peak in a matched filter output associated with each receive channel to identify a particular I / Q sample among the plurality of I / Q samples that is closest to a start time of the reference signal. determining a horizontal angle of arrival between a baseline of the antenna and incident energy from the UE to the two or more antennas and a vertical angle of arrival between the baseline of the antenna and incident energy from the UE to the two or more antennas based on the identified reference signals from each receive channel; calculating the position of the UE using the horizontal angle of arrival and the vertical angle of arrival; The method includes:

12. 12. The method of claim 11, wherein the horizontal and vertical angles of arrival are each associated with a line of bearing (LOB) between a serving sector of a radio tower and the UE.

13. 13. The method of claim 12, further comprising utilizing the horizontal angle of arrival to determine a second LOB between a non-serving sector of the wireless tower and the UE.

14. 13. The method of claim 12, further comprising: utilizing the vertical angle of arrival to determine a second vertical angle of arrival between a non-serving sector of a second wireless tower and the UE.

15. clustering two or more coherent receive channels among the receive channels; forming a group unit from the clusters; synchronizing said group units using an external synchronization input; The method of claim 11 further comprising:

Images