Line-of-sight RF signal classification apparatus and method

Abstract

Described here is a mechanism (apparatus and method) that leverages histogram characteristic of a signal phase in common signal modulation methods like quadrature amplitude modulation (QAM) and phase shift keying (PSK) to differentiate between line-of-sight (LOS) and non-line-of-sight (NLOS) conditions. The apparatus comprises an array of antennas and a processor coupled to the array. In at least one embodiment, the processor applies a beamforming scheme to an output of the array to generate a beamformed signal. In at least one embodiment, the processor extracts phase information from the beamformed signal. In at least one embodiment, the processor generates a histogram of the phase information. In at least one embodiment, the processor classifies line-of-sight and non-line-of-sight conditions based on the histogram.

Description

OSU02P052PCT (OSU-24-37) LINE-OF-SIGHT RF SIGNAL CLASSIFICATION APPARATUS AND METHOD CLAIM OF PRIORITY

[0001] This application claims priority to U.S. Provisional Patent Application No.63 / 708,202, filed on October 16, 2024, titled “LINE-OF-SIGHT RF SIGNAL CLASSIFICATION APPARATUS AND METHOD,” which is incorporated by reference in its entirety for all purposes. BACKGROUND

[0002] Indoor positioning systems (IPS) have gained significant attention in research due to the growing need for reliable location-based services in indoor settings where global navigation satellite systems are unreliable or unavailable. A major challenge in developing dependable IPS lies in accurately classifying line-of-sight (LOS) and non-line-of-sight (NLOS) propagation conditions, which heavily influence positioning accuracy. Existing solutions often demand high computational resources or are constrained by specific assumptions about the environment, limiting their broad applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0003] The examples will be understood more fully from the detailed description given below and from the accompanying drawings, which, however, should not be taken to limit the disclosure to the specific examples, but are for explanation and understanding only.

[0004] FIG. 1A is a plot of a 64- quadrature amplitude modulation (QAM) constellation, in accordance with at least one embodiment.

[0005] FIG.1B is a plot of a histogram of the phase of 64-(QAM) signals, in accordance with at least one embodiment.

[0006] FIG. 2 is a plot of a histogram of the phase of 64-QAM signal in an Additive white Gaussian noise (AWGN) channel, in accordance with at least one embodiment.

[0007] FIG.3A is a flowchart of a method of classifying LOS and NLOS conditions in an RF communication system, in accordance with at least one embodiment.

[0008] FIG.3B is a flowchart of a method of signal separation and characteristic extraction, in accordance with at least one embodiment.

[0009] FIG.4A is a plot of a histogram of received signal phases in non-line-of-sight (NLOS),OSU02P052PCT (OSU-24-37) in accordance with at least one embodiment.

[0010] FIG. 4B is a plot of a histogram of received signal phases in line-of-sight (LOS), in accordance with at least one embodiment.

[0011] FIGS.5A-D are plots showing phase histograms at signal-to-noise ratio (SNR) levels at 20 dB, 15 dB, 10 dB, and 5 dB, respectively, in accordance with at least one embodiment.

[0012] FIG. 6A is a plot of histograms of signal phase in LOS environments, in at least one embodiment.

[0013] FIG. 6B is a plot of histograms of signal phase in NLOS environments, in at least one embodiment.

[0014] FIG. 6C is a plot showing a portion of DFT of xq[n] given in Eq. (22), in at least one embodiment.

[0015] FIG.7A is a flowchart of a method for LOS / NLOS classification, in accordance with at least one embodiment.

[0016] FIG. 7B is a plot showing performance of a mechanism (apparatus and method) that leverages histogram characteristic of a signal phase in common signal modulation to differentiate between LOS and NLOS conditions, in accordance with at least one embodiment.

[0017] FIG.8 is a schematic of an apparatus that leverages histogram characteristic of a signal phase in common signal modulation to differentiate between LOS and NLOS conditions, in accordance with at least one embodiment.

[0018] FIG.9 is a layout of an experiment, in accordance with at least one embodiment.

[0019] FIG.10 is a plot showing performance of the mechanism, in accordance with at least one embodiment.

[0020] FIG. 11 is a plot showing attenuation vs. F1-score, in accordance with at least one embodiment.

[0021] FIG. 12 is a layout showing point selection for experiment on distance impact, in accordance with at least one embodiment.

[0022] FIG. 13 is a layout showing point selection for an experiment on signal-to-noise ratio (SNR) impact, in accordance with at least one embodiment.

[0023] FIG.14 is a schematic of a processor system with a machine-readable storage medium having machine-readable instructions that when executed cause a processor to execute machine- readable instructions according to the method summarized by the flowcharts of FIGS.3A-B, FIG.OSU02P052PCT (OSU-24-37) 7A and other examples and algorithms, in accordance with at least one embodiment. DETAILED DESCRIPTION

[0024] Described here is a mechanism (apparatus and method) that leverages histogram characteristic of a signal phase in common signal modulation methods like quadrature amplitude modulation (QAM) and phase shift keying (PSK) to differentiate between line-of-sight (LOS) and non-line-of-sight (NLOS) conditions. Phase modulation forms a distinctive pattern in the histogram, serving as a reliable identifier for LOS / NLOS classification. The mechanism of various embodiments demonstrates up to 100% real-time accuracy in LOS / NLOS identifications, resilience to noise, extensive operational range, and free of computing-intensive machine learning models.

[0025] Advancements in the Internet and communication technologies have enabled people to conveniently access information at their fingertips with smart devices. Consequently, people have grown increasingly dependent on Internet-based smart devices to gather information and incorporate it into their routine. Today, location-based services (LBS) are one of the most utilized applications on those devices, for example, to use navigation apps to determine real-time positions and directions. Location- based services have also played a crucial role in propelling the logistics industry forward, where accurate location information is paramount. While outdoor environments can rely on the global navigation satellite systems (GNSS) for location services, GNSS does not work in most indoor environments because of the heavy signal loss and multipath effect. This has resulted in a significant demand for indoor positioning systems (IPS) capable of providing location-based services (LBS) in indoor settings. The growing requirement for LBS within locales like warehouses, shopping malls, museums, hospitals, and educational institutions has intensified research interests in this field. However, the complexity of indoor environments characterized by substantial obstacles and fluctuating propagation conditions leads to significant challenges in designing and implementing efficacious IPS.

[0026] IPS techniques include time-of-arrival (TOA), time-difference-of-arrival (TDOA), angle-of-arrival (AoA), and fingerprinting. These methods typically assume the existence of a line-of-sight (LOS) path between the transmitter and receiver. In practice, this assumption may not always hold true. Identifying whether the system operates in a LOS or NLOS condition is highly valuable for most IPS, as NLOS conditions can significantly reduce system accuracy.OSU02P052PCT (OSU-24-37) However, this characterization also presents a significant technical challenge due to the sheer complexity of the problem.

[0027] Various examples herein provide an algorithm that acts as a “universal switch” for IPS; instead of directly improving accuracy. The apparatus and method of various examples helps IPS identify whether they are in LOS or NLOS conditions. This knowledge enables IPS to optimize their strategies accordingly, drastically minimizing errors caused by mismatched assumptions. The apparatus and method of various examples boosts IPS accuracy by ensuring they always operate under the right assumptions in dynamically changing indoor environments; that is, it empowers IPS to adapt intelligently to optimize performance.

[0028] Here, an NLOS scenario happens when obstacles block the direct path between the transmitter and receiver, leaving mostly reflected signals. In these situations, the channel behaves like a Rician fading model with a very low K-factor (approaching a Rayleigh fading channel). This definition emphasizes the absence of an LOS component rather than the severity of multipath dispersion, which makes it tricky to distinguish between LOS and NLOS conditions accurately. Note that this definition is tailored to the methods and explorations in our study.

[0029] Machine learning (ML), signal processing, and the use of additional sensor data can distinguish between LOS / NLOS conditions. Machine learning may use channel impulse response (CIR) data. ML-based methods with CIR can provide near-perfect classification rates but face significant challenges. First, obtaining CIR data is too slow for real-time IPS applications. Second, these methods are sensitive to changes in the environment, since they rely on neural networks to establish relationships between channel conditions and signal parameters, which often vary dynamically. Machine-learning-based LOS / NLOS identification methods using received signal strength indicator (RSSI), CIR, and channel state information (CIS) data requires an up-to- date dataset of environment-specific measurements, which compromises the system’s robustness.

[0030] To address these challenges, at least one embodiment describes a new approach to LOS / NLOS classification. Some embodiments exploit a distinct and consistent characteristic in common modulation schemes like quadrature amplitude modulation (QAM) and phase shift keying (PSK). Unlike metrics such as received RSSI, CIR, or other power / time-based measurements which fluctuate with environmental variations, the phase distribution patterns explored herein change merely with channel conditions (LOS or NLOS) when employing beamforming techniques. This inherent stability, rooted in the statistical nature of the modulations’OSU02P052PCT (OSU-24-37) symbol mapping and their native integration in the baseband signal, renders the embodiments robust to environmental changes, offering a reliable and an efficient method for LOS / NLOS classification.

[0031] Some embodiments provide RF module-equipped devices with the ability to determine LOS / NLOS conditions, boasting high accuracy, minimal latency, great scalability, and seamless integration into existing infrastructure. These advancements carry profound implications for both the Internet-of-Things (IoT) and indoor localization domains. For IoT, this functionality supports devices in critical decision-making scenarios where maintaining LOS is desired. It also assists IPS in optimizing operation according to the current LOS / NLOS conditions, and the enhanced IPS capabilities further enrich the IoT landscape.

[0032] The LOS / NLOS classification apparatus and method described herein exhibits resilience against noise and multipath effects, potentially ensuring more reliable classification in various wireless environments. The LOS / NLOS classification apparatus and method enables LOS / NLOS classification with a single raw signal waveform. This avoids the need for complex processes such as machine learning training using large number of measurements, enhancing its practicality for real-time applications. The LOS / NLOS classification apparatus and method is simple and easily deployable since it uses the phase information and may not use complex hardware or extensive calibration. The LOS / NLOS classification apparatus and method advances RF signal LOS / NLOS classification for noise resilience, real-time processing, and versatility. It is suitable for a variety of indoor positioning and communications applications. While it can effectively distinguish between LOS and NLOS conditions within, for example, 40 ms using commercial processors, it works with phase or amplitude and phase modulated signals, in accordance with at least one embodiment.

[0033] FIG.1A is a plot 100 of a 64- quadrature amplitude modulation (QAM) constellation, in accordance with at least one embodiment. FIG.1B is a plot 120 of a histogram of the phase of 64-(QAM) signals, in accordance with at least one embodiment. One objective of digital communications is to maximize the data rate by encoding multiple bits into a single transmission symbol, thereby increasing the spectral efficiency. To achieve this, two-dimensional modulation methods such as PSK and QAM are commonly employed. The constellations of these schemes typically exhibit multiple instances in the diagonal direction, resulting in a unique distribution in the signal’s phase. This distribution is observable as a special pattern in a histogram. Here, theOSU02P052PCT (OSU-24-37) extracted phase is an array of numbers.

[0034] To demonstrate this characteristic, a square 64-QAM modulation scheme is used as an example. In this scheme, the constellation diagram includes 64 distinct symbols. Assuming a normalized average symbol energy (Es) of 1, the energy per bit (Eb) can be calculated as:

[0035] ^^ ൌாೞ ൌ ^^^^^మ^ெ^^ (1)

[0036] of symbols in the constellation (M = 64). In 64-QAM, both thein-phase components have eight equally spaced amplitude levels, symmetrically arranged around zero. These levels are ±1, ±3, ±5, and ±7. To scale these amplitude levels while maintaining the average symbol energy at E = 1, the amplitude values for I and Qs components are scaled as follows:

[0037] ^^ ଶ^^ᇱൌ ଷ^2 െ 1^ ൌ 42 (2)the average symbol energy is given as:ᇱ

[0039] ^^ ൌ 1 / ^^ (3)^ ^

[0040] The scaled amplitude values are given as: ^^

[0041] ^^ ^^ ൌ ^^ ൈ ^^ (4a)ூ^ ^

[0042] ^^ ^^ ൌ ^^ ൈ ^^ (4b)ொ

[0043] where i, j take values from {±1, ±3, ±5, ±7}, representing the actual amplitude levels used in the modulation.

[0044] The phase ϕ(i, j) of each symbol is given by: ^^^ ^ೂ^ ^

[0045] ^^ ^^, ^^ ൌ ^^^^^^^^^^^^ ^^ (5)^^^^^of each constellation point is calculated as:ଶଶ^ ^ ^ ^ ^ ^

[0047] ^^ ^^, ^^ ൌ ^^ ^^ ^ ^^ ^^ (6) ^ ூ ொ

[0048] These results are shown in FIG. 1A, where for constellation points are aligned along the angles [−135°, −45°, 45°, 135°]. In contrast, one or two points are located along other angular lines. Consequently, there exists a significant statistical disparity in the phase information of the transmitted signal: symbols corresponding to these four angles occur with a frequency four times greater than those at other angles, assuming an equal probability of occurrence for each symbol.

[0049] Consider the transmitted signal s(t) over the Gaussian channel. The received signal is given as:OSU02P052PCT (OSU-24-37)

[0050] ^^^^^^ ൌ ^^^^^^ ^ ^^^^^^ (7)

[0051] where n(t) represents the AWGN. For QAM, s(t) can be written as:

[0052] ^^^^^^ ൌ ^^ூ^^^^ cos^2^^^^^^^^ െ ^^ொ^^^^ sin^2^^^^^^^^ (8)and the AWGN can be written as:Thus, the received signal is

[0056] ^^^^^^ ൌ ൫^^ூ^^^^ ^ ^^ூ^^^^൯ cos^2^^^^^^^^ െ ^^^ொ^^^^ ^ ^^ொ^^^^^ sin^2^^^^^^^^ (10)

[0058] ^^^ ൌ ^^^^^^^^^^^^ ೂ^^ା^^(11) nI ∼ N (0, σ2) and nQ ∼ N (0, σ2). Thisto a the formation of a Lorentzian

[0060] FIG.2 is a plot 200 of a histogram of the phase of 64-QAM signal in an Additive white Gaussian noise (AWGN) channel, in accordance with at least one embodiment. A 64-QAM communication system can be modeled subject to AWGN. A random binary sequence can be generated (e.g., using MATLAB’s wireless waveform generator) with a setup of 60,000 bits of binary input, a common benchmark in digital communication simulations. This sequence follows a uniform distribution and serves as the input to the 64-QAM module.

[0061] Symbol rate is calculated to ensure a 20 MHz bandwidth. A raised cosine filter with a roll-off factor of 0.2 can be used to generate the symbol rate. In this example, the symbol rate of 16.67 × 106symbols / s meets the bandwidth requirement while maintaining signal integrity and minimizing inter-symbol interference. The modulation and filtering process results in a complex signal array, S0, with 40,000 samples.

[0062] In at least one embodiment, signal S0 is added by AWGN with a 20 dB SNR to simulate realistic noise conditions, resulting in signal S1. The phase of the signal is extracted for both S0 (denoted as ϕ0) and S1(denoted as ϕ1) and their histograms are calculated using 360 bins. These histograms, depicted in FIG.1B and FIG.2, visualize the phase distribution under clean and noisy conditions, providing insights into the signal’s behavior in different environments.

[0063] FIG.3A is a flowchart 300 of a method of classifying LOS and NLOS conditions in an RF communication system, in accordance with at least one embodiment. The various blocks ofOSU02P052PCT (OSU-24-37) flowchart 320 can be performed by hardware, software, or a combination of them. Flowchart 300 utilizes techniques to differentiate the histogram characteristics of signal phase in common modulation schemes such as Quadrature Amplitude Modulation (QAM) and Phase Shift Keying (PSK). These histogram characteristics are inherent in a baseband signal. Note, without beamforming, both LOS and NLOS conditions exhibit similar characteristics. In at least one embodiment, the beamformer preserves these characteristics in LOS signals while distorting them in NLOS signals, thereby enabling effective classification.

[0064] At block 301, input RF signal is received by a uniform linear array (ULA) of antennas. This RF signal is a modulated signal. For instance, the RF signal is modulated by QAM or PSK. After, the modulated RF signal is collected by the antenna elements of the ULA, beamforming is applied. The ULA receives a frequency-shifted version of a baseband signal. For example, if the transmitter's carrier frequency is set to 5.2 GHz, the ULA will receive the signal centered at 5.2 GHz. The receiver down converts the signal, bringing it back to baseband. In at least one embodiment, the beamforming is a delay-and-sum beamforming. The delay-and-sum beamforming is applied to align the phase of the signals received at each antenna element of the ULA.

[0065] At block 302, a histogram of the phase information is constructed from the beamformed signal from block 301. In at least one embodiment, for LOS conditions, phase peaks are preserved in the histogram. In at least one embodiment, for NLOS conditions, the applied delays, and sums (or combinations) of signals disrupt these peaks in the histogram. In at least one embodiment, operation of block 302 can be performed by a digital signal processor (DSP), a central processing unit (CPU), a graphics processor unit (GPU), a field programmable grid array (FPGA), an application specific integrated circuit (ASIC), and the like coupled to the ULAs.

[0066] At block 303, the histogram of the phase information from block 302 is analyzed to distinguish between LOS and NLOS conditions based on the presence or absence of the phase peaks. The preserved peaks in the histogram indicate LOS while absence of phase peaks indicates NLOS. As such, LOS and NLOS conditions are classified, in accordance with at least one embodiment. In at least one embodiment, the classification operation of block 303 is performed by the DSP.

[0067] Accurately distinguishing between LOS and NLOS conditions is useful for optimizing performance and reliability in modern RF communication systems. In the context of evolvingOSU02P052PCT (OSU-24-37) communication technologies, such as 6G millimeter-wave communications and Internet-of-Things (IoT) systems, the ability to effectively identify and adapt to LOS and NLOS conditions is useful for several reasons. For instance, millimeter-wave signals in 6G communication are highly susceptible to blockages and reflections, making NLOS conditions a significant issue. The method and apparatus discussed herein including flowchart 300 can help 6G communication systems dynamically identify and select base stations with LOS channels, ensuring higher data throughput and lower latency by avoiding NLOS-induced degradation.

[0068] Many IoT applications, such as smart home devices and industrial automation, require reliable and low-latency communication, which is often compromised in NLOS conditions. The method and apparatus discussed herein including flowchart 300 enable IoT devices to perform real-time LOS / NLOS classification, allowing them to make informed decisions about communication paths and improving overall system reliability and performance.

[0069] Accurate indoor positioning is hindered by NLOS conditions, leading to significant errors in location estimates. By accurately identifying LOS / NLOS conditions, the method and apparatus discussed herein including flowchart 300 enhances the accuracy of IPS, leading to more reliable location-based services.

[0070] In dense urban environments, signals often encounter obstacles, leading to frequent NLOS conditions that degrade network performance. The method and apparatus discussed herein including flowchart 300 allow wireless networks and mobile communication systems to adaptively manage resources and optimize signal paths, improving connectivity and user experience.

[0071] FIG.3B is a flowchart 320 of a method of signal separation and characteristic extraction, in accordance with at least one embodiment. The various blocks of flowchart 320 can be performed by hardware, software, or a combination of them. The signal is received by an antenna array (e.g., a multi-input multi-output MIMO, a uniform linear antenna array, etc.). The signal is then amplified for processing. For example, a low noise amplifier (LNA) amplifies the received signal from the antenna. The amplified received signal is then down converted by one or more mixers, filtered, converted from analog to digital domain, and then processed in the digital domain. In at least one embodiment, beamforming is performed in the digital domain as indicated by block 321. One such process of beamforming is delay-and-sum beamforming.

[0072] In at least one embodiment, the delay-and-sum beamforming scheme improves the signal-to-noise ratio (SNR) of the received signal in a uniform linear array (ULA). In at least oneOSU02P052PCT (OSU-24-37) embodiment, the beamforming scheme introduces a time delay or carrier phase shift in the received signal at each antenna element to adjust the phase of the signal, steering the beam in a specific direction. The delayed signals from each antenna element are then combined (summed) to produce the beamformed output.

[0073] For purposes of discussion, assume the antenna to be a ULA of N elements, and received signal is r(t). At each antenna element i, a carrier phase shift is introduced to the received signal ri(t), given by:

[0074] ^^^^^^^ ൌ ^^^^^^^ ∙ ^^^థ^ (12)

[0075] where r0 is the received signal at the first antenna and ϕi is the carrier phase shift between the i-th antenna and the first antenna due to the extra distance that signal has traveled.

[0076] Consider each sample of the received signal as a complex scalar, which can be written as:

[0077] ^^ ൌ ^^ொ ^ ^^^^ூ (13)

[0078] When a carrier phase shift is introduced to the signal, it becomes:

[0079] ^^ᇱ ൌ ൫^^ொ ^ ^^^^ூ൯ ൈ ^cos^^^^^ ^ ^^ sin^^^^^^ (14)

[0080] The phase of the received signal from (11) is then rotated, denoted as ϕʹr, which is given as:

[0081] ^^′ ^ೂ ^୧୬^థ^^ା^^ ୡ୭^^థ^^^ ൌ ^^^^^^^^^^^^ ୡ୭^^థ^^ା^^ ^୧୬^థ^^(15)is changed to ϕʹrby the carrier phase shift accordingly.

[0083] When beamforming at block 321 is applied to the received signals, the delayed signals from each antenna element are then multiplied by a complex exponential to adjust the carrier phase of the signal:

[0084] ^^^^^^ ∙ ^^ି^థ^ ^^ఏ^(16)first and the i-th antenna when considering the signal originating from angle θ, which is expressed as:

[0086] ^^ ൌଶగൈ^ൈௗ ^୧୬^ఏ^^(17)adjacent antennas, i is the antenna index, and λ is the wavelength of the signal. The beamformed output signal is then obtained by summing the delayed signals from each antenna element as:OSU02P052PCT (OSU-24-37)

[0088] ^^^^^^ ൌ ∑ேି^ ^ୀ^ ^^^^^^^ ∙ ^^ି^థ^^ఏ^(18) where N is the number of antenna elements in the ULA. If θ is selected accurately, thebe offset. Thus, all channels will align perfectly in phase. Superposition of all channels leads to an enhanced SNR and preserve the pattern in the histogram.

[0090] Consider an array of N antenna elements and let s(t) be the transmitted signal. In indoor environments, the signal received at each antenna comprises of signals from various propagation paths, which can be expressed as:

[0091] ^^^^థ^^^^^ ൌ ∑ ^ୀ^ ^^^^ ∙ ^^൫^^ െ ^^^൯ ∙ ^^ ^ೕ ^ ^^^^^^^ (19)of propagation paths, αijis theis the relative phase shift between the first and the i-th antenna of the j-th propagation path, ni(t) is the noise at the i-th antenna, τj is delay of the j-th path when the first path (direct path) is used as the reference, i.e., τ1= 0.

[0093] The delay-and-sum beamformer introduces time delays to account for the differences in the arrival times of the signals at each antenna element. In the LOS scenario, although both the direct path and the multipath components are present in the received signals, the carrier phase shift from the direct path dominates and the multipath components are negligible. Thus, the received signal in the LOS scenario can be simplified as:

[0094] ^^ ^థ^^^^^ ൌ ^^^^^^^^^ ∙ ^^ ^ ^ ^^^^^^^ (20)applying beamforming to signals, the correct θ can be determined to compensate for the carrier phase shift in all channels. Thus, the rotation of the phase is restored. Upon combining all channels, peaks in the histogram of the phase are preserved.

[0096] In the NLOS scenario, the absence of a direct path component means that none of the phases of the data symbols dominates because the signal is received from different paths, each introducing an additional random phase shift. The received signal at each antenna is given as:

[0097] ^^^^థ^^^^^ ൌ ∑ ^ୀଶ ^^^^ ∙ ^^^^^^ ∙ ^^ ^ೕ ^ ^^^^^^^ (21)not be present for NLOS scenarios. In this case, the amplitude aij and carrier phase shift ϕij are identically distributed for different antenna elements.OSU02P052PCT (OSU-24-37)

[0099] When beamforming is applied to an NLOS signal, the phase adjustments made to each channel cannot effectively reverse the phase rotation. Hence, when all channels are combined, these phase discrepancies contribute to filling the troughs in the phase histogram, leading to a much more evenly dispersed pattern histogram than LOS cases.

[0100] Block 322 is a computational process, which involves angle function. Mathematically, block 322 is equivalent to equation (5). Note this is a very generic function and can find equivalent function in all coding languages. In at least one embodiment, block 323 is also a computational process, which is related to generating histograms based on the phase information (see FIG.1B, FIG.2, and FIGS 4-6).

[0101] To illustrate this, two sets of data discussed herein can be recorded in real-world scenarios, one in an LOS scenario and another in an NLOS scenario.

[0102] FIG. 4A is a plot 400 of a histogram of received signal phases in non-line-of-sight (NLOS), in accordance with at least one embodiment. FIG. 4B is a plot 420 of a histogram of received signal phases in line-of-sight (LOS), in accordance with at least one embodiment. As depicted in FIGS. 4A-B, the histogram of the LOS scenario (plot 420) exhibits clear peaks at [−135°, −45°, 45°, 135°], whereas the histogram of the NLOS scenario (plot 400) shows a flat pattern across all phases.

[0103] In at least one embodiment, if the received signal corresponds to a LOS scenario, the histogram of the phase of the received QAM signal exhibits the following characteristics: (a) The phase histogram has four cluster-like peaks, equally spaced by 90°; (b) The intensity of peaks is at least two times higher than the median. In at least one embodiment, the phase histogram of NLOS signals will exhibit a flat pattern. One such LOS / NLOS classification algorithm as described in Algorithm1, in accordance with at least one embodiment.

[0104] Algorithm1 Basic LOS / NLOS Classification

[0105] S ^ Separated signal matrix;

[0106] for angle ^ from 0 to 180 do

[0107] P ^ Histogram of signal’s phase from ^;

[0108] [B,I] ^ maxk(P,4);

[0109] I ^ sort(I,’ascend’);

[0110] Iʹ ^ [I(1) I(1)+90 I(1)+180 I(1)+270]pOSU02P052PCT (OSU-24-37)

[0111] M ^ median(P)

[0112] if Iʹ = I and B[1] ≥ 2M then

[0113] Is LOS ^ True;

[0114] else if Iʹ ≠ I and ^ ≤ 181 then

[0115] continue;

[0116] else

[0117] Is LOS ^ False;

[0118] end if

[0119] end for

[0120] Algorithm1 may use extensive searching in different directions. In some examples, peak detection may become challenging under low SNR conditions.

[0121] In Algorithm1, the system first acquires data matrix S from the beamformer. The system iteratively generates a phase histogram P in each iteration. Then, the index I and intensity B of 4 highest value within P are found using maxk function. The four highest values in the histogram P are indexed as I and reordered in ascending order. The smallest index is then used to compute theoretical indexes I′, based on the assumption that four peaks should be spaced by 90 degrees. M is the median intensity of the histogram. If I match I’, and the minimum intensity in B is higher or equal to two times of M, then it should be LOS. If in any iteration the system can’t find LOS indication, then output is NLOS.

[0122] In some examples, the maxk function may identify points near the peaks instead of the actual peak values. In those cases, I and I’ may not match. This mismatch may occur when peak intensities differ significantly, causing errors in peak identification and subsequent misclassification of LOS as NLOS.

[0123] FIGS. 5A-D are plots 500, 520, 530, and 540 showing phase histograms at signal-to- noise ratio (SNR) levels at 20 dB, 15 dB, 10 dB, and 5 dB, respectively, in accordance with at least some examples. As illustrated in FIGS.5A-D, in the high SNR scenarios, peaks are easy to detect by Algorithm1. In the low-SNR scenario, however, weak peaks may be less noticeable because of interference in the phase, resulting in possible erroneous peak detection. To address this issue, in at least one embodiment, an algorithm that uses a Fourier transform to optimize weak peak detection is used.

[0124] When the received signal has a low SNR, peaks in the histogram are less noticeable dueOSU02P052PCT (OSU-24-37) to noise. In such cases, the Algorithm1’s ability to identify the peaks may be limited to amplitude comparison. In at least one embodiment, the peaks remain quite discernible when examined using Fourier transformation, in which the histogram data needs to be treated as a discrete signal xθ[n], which represents the counts at each phase bin, expressed as:

[0125] ^^ఏ^^^^ ൌ ^^^^, ^^ଶ,^^ଷ,⋯ , ^^ଷହଽ^ (22)

[0126] where θ denotes the steering angle used in beamforming, n is histogram bin index, Nn, n = 0, · · · , 359, is the measured counts of the phase falling in the corresponding bin. Since the peaks are equally spaced by 90◦, the discrete Fourier transform (DFT) of xθ[n] will reveal period of fundamental frequency of the histogram data. The DFT of xθ can be expressed as:[00 ଷହଽ ି^ మഏ127] Χఏ^^^^ ൌ ∑^ୀ^ ^^ఏ^^^^ ∙ ^^యలబ^^(23) LOS environments, in at least onea in NLOS environments, in at least one embodiment. FIG.6C is a plot 630 showing a portion of DFT of xq[n] given in Eq. (22), in at least one embodiment. The representation of Xθ(k) is shown in FIG.6C.

[0129] The Fourier transform is useful under low SNR conditions, where the peaks may not be visible in the time domain. By transforming the histogram data xθ[n], the periodicity of the signal becomes evident at, for example, 4 Hz in the spectrum. A strong amplitude at this frequency confirms the presence of the expected periodicity. If this amplitude exceeds a certain threshold, it clearly indicates a LOS signal.

[0130] By utilizing the distinct peak at 4 Hz, for example, the processing speed and accuracy of LOS / NLOS classification are significantly enhanced. In at least one embodiment, the amplitude at this frequency, after applying the DFT to the histogram, is directly compared with a predefined threshold of 22; that is, if Xθ(4) ≥ 22 for any values of θ that the beamformer could scan, then the signal is classified as LOS; otherwise it is classified as NLOS. In at least one embodiment, a threshold of 22 is optimized empirically through extensive amount of experiments and can be optimized for particular localization applications. This scheme is captured by Algorthim2 which can by summarized as:

[0131] Algorithm2 Optimized LOS / NLOS classification

[0132] S ^ Separated signal matrix;

[0133] Is_LOS ^ False;OSU02P052PCT (OSU-24-37)

[0134] for ^ from 0:step: ScanRange do

[0135] ^ ^ Histcounts(S(:,^), 360);

[0136] ^ ^ normalize(^);

[0137] X ^ FFT (^);

[0138] if X[5] ≥ A then

[0139] Is_LOS ^ True;

[0140] break;

[0141] else

[0142] Continue

[0143] end if

[0144] end for

[0145] FIG. 7A is a flowchart 700 of a method for LOS / NLOS classification, in accordance with at least one embodiment. While various blocks are shown in a particular order, the order can be modified. For example, some blocks may be performed simultaneously. The blocks can be performed by hardware (e.g., digital signal processor, a CPU, a GPU, an FPGA, an ASIC, etc.), software, or a combination of them.

[0146] At block 701, after input signal is received by the ULA and beamforming is applied on the received signals, a signal matrix S is generated. The size of the matrix may depend on ScanRange and scan step of the beamformer. The terms ScanRange and step are discussed below. As discussed herein, beamforming is a delay-and-sum beamforming process to align the phase of the signals received at each antenna element of the ULA. In at least one embodiment, signal matrix S is a two-dimensional matrix with columns organized by the steering angle ^ (or scanned angle) used in beamforming and rows containing complex data. In one example, the rows or columns of signal matrix S correspond to scanned angle and the other dimension are the signal waveform data associated with the scanned angle. Note, rows and columns can be flipped while maintaining the information in the matrix. For instance, the steering angle ^ can be in the rows and the signals’ complex data can be in the columns.

[0147] At block 702, LOS register Is_LOS is set to 0 (e.g., false). In at least one embodiment NLOS register is also set to 0 (e.g., false). In at least one embodiment, LOS register is enough to conclude LOS and NLOS conditions. For instance, by initializing the LOS register to 0, NLOSOSU02P052PCT (OSU-24-37) condition is assumed as default and when LOS register is set to 1 based on an analysis, NLOS condition automatically becomes invalid. Note, the same can be achieved by two district registers, one for LOS condition and one for NLOS condition.

[0148] At block 703, a determination is made whether all columns of matrix S are processed. Starting with the first column with steering angle ^ as 0 and counting by a programmable-degree increments, at block 704 a histogram of the phase information from the beamformed signal is generated for each column (or steering angle) of signal matrix S. In at least one embodiment, the programmable-degree is set to 10 for faster analysis. For finer analysis, programmable-degree can be set to a number lower than 10.

[0149] At block 705, the histogram is normalized. The peak intensity in a histogram is influenced by several factors, including data length, modulation scheme, and other variables. These factors cause fluctuations in the frequency domain intensities, scaled by the histogram's peak intensities, making it difficult to establish a universal threshold for detecting periodicity. By applying normalization, the peak intensity of all histograms is standardized to 1, eliminating the influence of these factors and enabling adaptive detection of periodicity.

[0150] The process of block 703 continues for a scan range which goes through all the columns of signal matrix S. Here, “step” is a configurable variable, which is determined based on the resolution of ULA. This variable serves as a computing acceleration function. For ULA, “step” has a maximum resolution based on the design of the ULA. Within this resolution, carrier phase shift may not be rapid. Thus, if there is LOS indicator (e.g., periodic peaks), it is likely observable within the resolution range. In at least one embodiment, skipping half of the resolution has a very low chance of missing the indicator. In at least one embodiment, by utilizing this step, the algorithm can rapidly accelerate. Here, “ScanRange” is a configurable variable for controlling an effective range of Algorithm2. By default, this variable is set to 180, to cover the plane in front of the antenna array. In at least one embodiment, setting ScanRange to 360 can enable omnidirectional ability. In at least one embodiment, “ScanRange” variable is less or equal to scan range of beamforming (e.g., if the beamformer is set to 180, the maximum of ScanRange is also 180).

[0151] At block 706, fast Fourier transform (FFT) is applied on the normalized histogram. An output of FFT is saved as variable X. As discussed herein, histogram of the phase information from the beamformed signal provides insights into LOS and NLOS conditions. For LOSOSU02P052PCT (OSU-24-37) conditions, phase peaks are preserved. For NLOS conditions, the applied delays and combination of signals during beamforming disrupts these phase peaks.

[0152] At block 707, a determination is made regarding the phase peak. If the output of the FFT at a particular frequency component [5] on the histogram is greater or equal to a threshold A, LOS condition is identified and the LOS register is set to 1 (e.g., True) at block 708. If the output of the FFT at the particular frequency component [5] on the histogram is less than the threshold A, an NLOS condition is inferred, and the process proceeds to block 703 and repeated for all columns of signal matrix S until LOS condition is identified. Here, “break” is a flag that instructs Algorithm2 to exit the loop once LOS is determined. Removing the break flag may not affect the functionality of Algorithm2, but it may result in redundant computations, leading to increased processing time. Here, “[5]” refers to a 4 Hz frequency component in the frequency domain X, where index 1 corresponds to the DC component (0 Hz). Note that different programming languages may use different indexing systems (e.g., in C / C++ / C#, indexing starts at 0, and in this case, 4 Hz would correspond to [4]). In at least one embodiment, depending on the modulation scheme, new indicators may appear at different frequencies (e.g., if a new modulation results in six distinct regular peaks, the corresponding frequency index might be [7]). Here, threshold “A” is used for classification. In at least one embodiment, value of threshold A is a tunable parameter that can be customized by a user. In at least one embodiment, threshold A is a predefined value and stored in memory or a register. In at least one embodiment, threshold A is a programmable value that can be programmed or changed by software, hardware, or a combination of them.

[0153] While block 707 compares an intensity of a specific frequency component (output of FFT) with a threshold A, block 707 can be modified based on a type of modulation scheme used for the received signal.

[0154] FIG.7B is a plot 720 showing performance of a mechanism (apparatus and method) that leverages histogram characteristic of a signal phase in common signal modulation to differentiate between LOS and NLOS conditions, in accordance with at least one embodiment.

[0155] In at least one embodiment, the LOS signal is generated using the same method outlined herein, but with QPSK modulation. This approach ensures that the peak and periodic characteristics critical for the analysis are pronounced and clear. The resulting LOS signal is denoted as S2. In at least one embodiment, signal S2is then passed through an AWGN channel with varying SNR levels. This step simulates different noise conditions that the signal mightOSU02P052PCT (OSU-24-37) encounter in real-world scenarios. For each SNR level, the optimized and basic algorithms go through 1000 rounds of Monte-Carlo simulations. The success rate of classification as LOS is recorded for each SNR level, defined as the number of determined true LOS conditions divided by the total number of runs. Given the nature of Algorithm2, any signal that does not meet the classification criteria for LOS is determined as NLOS by default. Consequently, explicit NLOS condition testing is not included in this simulation, as failure to classify as LOS automatically implies NLOS classification.

[0156] Compared to Algorithm1, this simulation aims to demonstrate Algorithm2’s precision in identifying LOS conditions and its robustness under various SNR levels. Plot 720 of FIG. 7B shows that Algorithm2 maintains a 100% success rate in identifying LOS conditions down to an SNR of 8 dB without degradation, showcasing its effectiveness in low-SNR conditions prevalent in real-world environments. This robustness is useful for practical applications, particularly in challenging indoor scenarios where signals are often weakened or distorted.

[0157] In Algorithm2, any signal lacking a distinct direct path is classified as NLOS by default. Consequently, simulating explicit NLOS conditions may be unnecessary, as the absence of these characteristics guarantees a 100% NLOS classification by Algorithm2 at any SNR levels.

[0158] FIG. 8 is a schematic of an apparatus 800 that leverages histogram characteristic of a signal phase in common signal modulation to differentiate between LOS and NLOS conditions, in accordance with at least one embodiment. Apparatus 800 implements flowcharts 300, 320, and / or 700, in accordance with at least one embodiment.

[0159] Apparatus 800 comprises antenna array 801, low noise amplifier (LNA) 802, RF mixer 803, I and Q mixers 804a and 804b, respectively, filters 805a and 805b, analog-to-digital converter (ADC) 806, and digital signal processor (DSP) 807. Any suitable frontend architecture can be used from LNA 802 to input of DSP 807. In at least one embodiment antenna array 801 is an N- element antenna array (where N is a number e.g., 4). In at least one embodiment, Algorithm1 and / or Algorithm2 are executed by DSP 807. Output from antenna elements is amplified by LNA 802 and down converted using local oscillator (LO) signal by RF mixer 803. In at least one embodiment, to get the in-phase (I) and quadrature phases (Q), additional down converting may be performed using LO signal by mixers 804a and 804b, respectively. In at least one embodiment, capacitors C1 and C2 are coupled between RF mixer 803 and mixers 804a and 804b. In at least one embodiment, output I from mixer 804a is filtered by filter 805a and provided to ADC 806. InOSU02P052PCT (OSU-24-37) at least one embodiment, output Q from mixer 804b is filtered by filter 805b and provided to ADC 806. ADC 806 converts the analog signal to a digital representation which is then provided to DSP 807. One output of DSP 807 may indicate LOS / NLOS condition.

[0160] Analyzing the computational complexity of LOS / NLOS classification algorithm (Algorithm2) is useful for understanding its performance, particularly in terms of processing speed and resource efficiency, important for real-time indoor positioning. Computational time complexity is generally denoted as O and is used to describe the number of operations or the amount of computational work an algorithm performs relative to the size of its input data. In terms of time complexity, it provides an upper limit on the time required to execute an algorithm as the input size grows. In at least one embodiment, Algorithm2 involves scanning angles from 0 to 180 degrees at 10-degree intervals, resulting in a total of 18 iterations for a full cycle. In each iteration, it constructs a histogram with 360 bins for the baseband signal’s phase, a choice dictated by the range of the phase within a complex signal spanning from −179 to 180 degrees, for example. This ensures full angle coverage for the subsequent FFT application.

[0161] The computational complexity of constructing the histogram can be described generally as O(n + k), where n is the number of data points and k is the number of bins. For this specific application, where k = 360, the complexity of this operation becomes O(n+360). Note, k = 360 because the range of the phase within the complex signal spans from -179 degree to 180 degree.

[0162] The FFT assesses the frequency components of the histogram, which is typically characterized by a computational complexity of O(n log n). In at least one embodiment, the histogram’s fixed number of bins is 360, thereby fixing the input size for the FFT. Consequently, the specific computational complexity for our scenario is O(360 log 360) which remains constant due to the predetermined number of samples. Note, here the fixed number of bins is 360 because the histogram has 360 bins which causes the size of the histogram to be 360*1 resulting in n=360 for FFT.

[0163] When combined, the overall computational complexity of both operations can be described as:

[0164] ^^^^^ ^ 360^ ^ ^^^360^^^^^^360^ (24)

[0165] Since 360 and 360 log 360 are constants, for large n, the dominant term in the overall complexity is O(n).

[0166] However, the algorithm’s time consumption is affected by the iteration counts. In LOSOSU02P052PCT (OSU-24-37) scenarios, where the signal’s direct path is unobstructed, Algorithm2 may identify LOS conditions early, generally requiring fewer than 18 iterations. In NLOS scenarios with no direct path, Algorithm2 may use 18 iterations to confirm the absence of LOS. Although the computation required per iteration remains unchanged, the total time consumption varies depending on the signal condition, with a maximum of 18 iterations needed for NLOS cases. In at least one embodiment, parallel computing can be performed to improve computation time.

[0167] Algorithm2’s computational efficiency highlights its suitability across a diverse range of computational environments, spanning from resource-rich servers to constrained IoT devices. The algorithm’s constant time complexity is essential for its deployment in real-time systems where processing speed is critical. Consequently, its integration into indoor positioning systems.

[0168] FIGS.9-12 show an experimental setup to verify operation of Algorithm2. FIG.9 is a layout 900 of an experiment, in accordance with at least one embodiment. Here, the environment dimensions are (9.6m × 9.3m). FIG.10 is a plot 1000 showing performance of the mechanism, in accordance with at least one embodiment. FIG. 11 is a plot 1100 showing attenuation vs. F1- score, in accordance with at least one embodiment. FIG. 12 is a layout 1200 showing point selection for experiment on distance impact, in accordance with at least one embodiment. FIG. 13 is a layout 1300 showing point selection for experiment on signal-to-noise ratio (SNR) impact, in accordance with at least one embodiment.

[0169] In this experiment, NLOS condition is simulated by positioning an adult directly between a receiver and a transmitter. This method serves two purposes: it creates a realistic indoor scenario where the LOS is obstructed, like everyday environments, and it provides a relevant test to evaluate the algorithm’s ability to accurately classify LOS and NLOS conditions. Here, the dimension of antenna array 801 is (12cm × 5cm) and beamwidth is 29.4°. The compact size of antenna array 801 together with the human body’s ability to absorb RF signals, primarily due to its high water content, ensure that an average adult can effectively obstruct the direct path when positioned close to the transmitter, thus creating NLOS scenarios. Moreover, this setup effectively simulates realistic scenarios, as human presence indoors often contributes to dynamically changing NLOS conditions in various applications. In at least one embodiment, NLOS scenarios created by doorways and walls are experimented as well.

[0170] The following LOS / NLOS identification performance metrics are assessed via experiments: success rate (for both LOS and NLOS) and F 1-score. The F 1-score is a metric forOSU02P052PCT (OSU-24-37) evaluating the accuracy of a test. It is calculated from the precision and recall values, where precision is the ratio of the number of true positives to all positive results (including incorrect identification), and recall is the ratio true positives to all samples that should have been identified as positives. Precision is also known as positive predictive value, and recall is known as sensitivity in diagnostic binary classification. The F 1-score is defined as:

[0171] ^^1 ൌ 2 ∙ ^∙ோ^ାோ (25) recall of the test, which are given as:ൌ ்^ାி^

[0174] ^^ ൌ ்^்^ାிே

[0175] where TP is the number of true positives, FN is the number of false negatives.

[0176] In the drawings, like reference numerals refer to like elements throughout, and the various features are not necessarily drawn to scale. Here, the same reference numbers or other reference designators are used in the drawings to designate the same or similar (either by function and / or structure) features.

[0177] For this experiment, ten locations are selected. To create NLOS conditions an adult is positioned in the direct path between the transmitter and receiver. Fifty samples are collected at each location under both the LOS and NLOS scenarios, resulting in a total of 7000 samples for validation. This experiment aims to assess the accuracy, efficiency, and versatility of the proposed algorithm across common modulations.

[0178] As Table 1 and FIG. 10 show, Algorithm2 exhibits perfect accuracy with QPSK modulation and generally achieves an F 1 score over 0.98 for higher-order modulations. Also, Algorithm2 can achieve an update rate over 26 Hz, highlighting its capability of real-time processing. Table 1: Performance vs. modulations Modulation LOS NLOS F 1-Score Update Rate (Hz)OSU02P052PCT (OSU-24-37) 64-QAM 100% 97.1% 0.986 28 128-QAM 100% 98.1% 0.991 26es. Using a QPSK- modulated signal, and four selected locations at the same angle relative to the receiver, and the NLOS scenario can be created the same way as the first experiment. For each location and scenario (LOS and NLOS), 200 samples are collected, resulting in a total of 1600 samples.

[0180] As shown in Table 2, the F 1 score of Algorithm2 remains high as the distance between transmitter receiver increases. Table 2: Distance vs F1-Score Distance (cm) LOS NLOS F1-Score

[0181] In a third experiment, Algorithm2 s performance is evaluated in low-SNR scenarios obtained by adjusting the attenuation of transmitting power of a QPSK- modulated signal with the transmitter placed at a fixed location. The NLOS scenario can be created in a similar manner to the previous experiments. For each scenario at each attenuation level, 200 samples are collected, resulting in a total of 4800 samples for validation.

[0182] As shown in Table 3 and FIG.11, Algorithm exhibits resilience for low-SNR operation. Furthermore, the trend observed in FIG. 11 aligns with that in FIG. 7B, further validating algorithm’s theoretical framework. Table 3: Attenuation vs. F1-score Attenuation (dB) LOS NLOS F 1-ScoreOSU02P052PCT (OSU-24-37) 1 100 99 0.99 2 100 100 1 [00 83] s s own n ab e , gort m e ect ve y detects b oc ages caused by uman bodies and sheet metal. Table 4: Success rates of Algorithm2 vs. different obstacle materials m %

[0184] For blockage by plywood and foam, a strong direct path is present since RF signals can effectively penetrate them. For such cases, while a direct path may seem obstructed, it does not pose a genuine NLOS issue that necessitates special processes (e.g., NLOS bias reduction or mitigation) from a positioning system.

[0185] The performance of LOS / NLOS classification algorithm can be further tested using structural elements of a laboratory to create clear LOS and NLOS scenarios. In at least one embodiment, a doorway of a laboratory is used to establish LOS and NLOS conditions. PlacingOSU02P052PCT (OSU-24-37) the transmitter and receiver at the doorway, we conducted 100 trials for each case. For LOS, the door was left open; for NLOS the door was closed. Algorithm2 achieved a 99% success rate in identifying the LOS (open door) scenarios and a perfect 100% for NLOS (closed door) conditions.

[0186] In at least one embodiment, Algorithm2 is tested by placing the transmitter and receiver at two sides of a wall of a lab, creating a clear NLOS condition. The experiment achieved a 100% success rate in recognizing such NLOS conditions.

[0187] FIG. 14 is a schematic of a processor system 1400 with a machine-readable storage medium having machine-readable instructions that when executed cause a processor to execute machine-readable instructions according to the method summarized by the flowcharts of FIGS. 3A-B, FIG.7A and other examples and algorithms, in accordance with at least one embodiment.

[0188] In at least one embodiment, processor system 1400 comprises memory 1401, processor 1402, machine-readable storage medium 1403 (also referred to as tangible machine-readable medium), communication interface 1404 (e.g., wireless or wired interface), and network bus 1405 coupled together as shown. In at least one embodiment, processor system 1400 may be part of apparatus 800. In at least one embodiment, processes described herein may be stored in machine readable medium 1403 as computer-executable instructions. In at least one embodiment, a machine-readable storage medium may be random access memory (RAM).

[0188] In at least one embodiment, processor 1402 is a digital signal processor (DSP), an application specific integrated circuit (ASIC), a general-purpose central processing unit (CPU), a field programmable gate array (FPGA), a graphics processing unit (GPU), an application specific integrated circuit (ASIC), or a low power logic implementing a simple finite state machine to perform various processes described herein. In at least one embodiment, processor 1402 is equivalent to DSP 807 shown in FIG.8.

[0189] In at least one embodiment, various logic blocks of processor system 1400 are coupled together via network bus 1405. Any suitable protocol may be used to implement network bus 1405. In at least one embodiment, machine-readable storage medium 1403 includes instructions (also referred to as program software code / instructions) for classifying LOS and NLOS conditions, coded into software stored in machine-readable storage medium 1403.

[0190] In at least one embodiment, machine-readable storage media 1403 is a machine- readable storage media with instructions for classifying LOS and NLOS conditions. In at leastOSU02P052PCT (OSU-24-37) one embodiment, machine-readable medium 1403 has machine-readable instructions, that when executed, cause processor 1402 to perform the method discussed herein.

[0191] In at least one embodiment, program software code / instructions associated with various embodiments may be implemented as part of an operating system or a specific application, component, program, object, module, routine, or other sequence of instructions or organization of sequences of instructions referred to as "program software code / instructions," "operating system program software code / instructions," "application program software code / instructions," or simply "software" or firmware embedded in processor. In some embodiments, program software code / instructions associated with processes of various embodiments are executed by processor system 1400.

[0192] In at least one embodiment, machine-readable storage media 1403 is a computer executable storage medium. In at least one embodiment, program software code / instructions associated with various embodiments are stored in computer executable storage medium 1403 and executed by processor 1402. Here, computer executable storage medium 1403 is a tangible machine-readable medium 1403 that can be used to store program software code / instructions and data that, when executed by a computing device, causes one or more processors (e.g., processor 1402) to perform a process.

[0193] In at least one embodiment, tangible machine-readable medium 1403 may include storage of executable software program code / instructions and data in various tangible locations, including for example, ROM, volatile RAM, non-volatile memory, and / or cache, and / or other tangible memory as referenced in present application. Portions of this program software code / instructions and / or data may be stored in any one of these storage and memory devices. In some embodiments, program software code / instructions can be obtained from other storage, including, e.g., through centralized servers or peer to peer networks and the like, including the Internet. Different portions of software program code / instructions and data can be obtained at different times and in different communication sessions or in the same communication session.

[0194] In at least one embodiment, software program code / instructions associated with various embodiments can be obtained in their entirety prior to execution of a respective software program or application. Alternatively, portions of software program code / instructions and data can be obtained dynamically, e.g., just in time, when needed for execution. Alternatively, some combination of these ways of obtaining software program code / instructions and data may occur,OSU02P052PCT (OSU-24-37) e.g., for different applications, components, programs, objects, modules, routines, or other sequences of instructions or organization of sequences of instructions, by way of example. Thus, it is not required that data and instructions be on a tangible machine-readable medium 1403 in entirety at a particular instance of time.

[0195] In at least one embodiment, tangible machine-readable medium 1403 include but are not limited to recordable and non-recordable type media such as volatile and non-volatile memory devices, read only memory (ROM), random access memory (RAM), flash memory devices, floppy and other removable disks, magnetic storage media, optical storage media (e.g., Compact Disk Read-Only Memory (CD ROMs), Digital Versatile Disks (DVDs), etc.), among others. In at least one embodiment, software program code / instructions may be temporarily stored in digital tangible communication links while implementing electrical, optical, acoustical, or other forms of propagating signals, such as carrier waves, infrared signals, digital signals, etc., through such tangible communication links.

[0196] In at least one embodiment, the software program code / instructions can be compiled and integrated into firmware as a library, which is executed directly by the processor when the firmware is loaded. The software program code / instructions can be stored in a static library or dynamic library form, and compiled into the firmware of a device, allowing direct execution from the firmware.

[0197] In at least one embodiment, the method or apparatus described herein may be implemented in a cloud computing environment, where the processing of beamforming scheme, phase extraction, histogram generation, and LOS / NLOS classification is performed on a remote server or in a cloud infrastructure. In at least one embodiment, the system delivers the results to a client device without locally executing the code or storing the software. The use of this remote or cloud-based processing for classification is also within the scope of various examples herein.

[0198] The following are additional examples provided in view of the above-described implementations. Here, one or more features of example, in isolation or in combination, can be combined with one or more features of one or more other examples to form further examples also falling within the scope of the disclosure. As such, one implementation can be combined with one or more other implementations without changing the scope of disclosure.

[0199] Example 1 is an apparatus comprising: an array of antennas; and a processor coupled to the array, the processor to: apply a beamforming scheme to an output of the array to generate aOSU02P052PCT (OSU-24-37) beamformed signal; extract phase information from the beamformed signal; generate a histogram of the phase information; and classify line-of-sight and non-line-of-sight conditions based on the histogram.

[0200] In at least one embodiment, the processor is to identify distinct regular peaks formed in the phase histogram, arising from statistical differences in phase angles generated by phase modulation. Such peaks exhibit regularity and persist under line-of-sight (LOS) conditions, but diminish or disappear under non-line-of-sight (NLOS) conditions due to the effect of beamforming, as a result of the absence of the direct signal path and the combined phase shifts from multipath signals, independent of the specific modulation scheme employed.

[0201] Example 2 is an apparatus according to any example herein, in particular example 1, wherein the array of antennas includes N antenna elements, and wherein the beamforming scheme is to align phases of signals received at individual elements of the N antenna elements.

[0202] Example 3 is an apparatus according to any example herein, in particular example 1, wherein the beamforming scheme is a delay-and-sum beamforming scheme.

[0203] Example 4 is an apparatus according to any example herein, in particular example 1, wherein the processor is to classify the line-of-sight and non-line-of-sight conditions by comparison of periodicity of the histogram with a threshold, or wherein the processor is to classify the line-of-sight and non-line-of-sight conditions by comparison of a phase peak of the histogram with a threshold.

[0204] Example 5 is an apparatus according to any example herein, in particular example 4, wherein the threshold is predetermined or programmable.

[0205] Example 6 is an apparatus according to any example herein, in particular example 1, wherein the array of antennas is a uniform antenna array.

[0206] Example 7 is an apparatus according to any example herein, in particular example 1, wherein the beamformed signal is a signal matrix, wherein the processor is to generate an individual histogram of each column or row of the signal matrix, wherein the processor is to apply fast Fourier transform to the individual histogram, and wherein the processor is to classify the line- of-sight and non-line-of-sight conditions by comparison of an output of the fast Fourier transform against a threshold.

[0207] Example 8 is a method comprising: applying a beamforming scheme to an output of an array of antennas to generate a beamformed signal; extracting phase information from theOSU02P052PCT (OSU-24-37) beamformed signal; generating a histogram of the phase information; and classifying line-of-sight and non-line-of-sight conditions based on the histogram.

[0208] Example 9 is a method according to any example herein, in particular example 8, wherein the array of antennas includes N antenna elements, and wherein applying the beamforming scheme comprises aligning phases of signals received at individual elements of the N antenna elements.

[0209] Example 10 is a method according to any example herein, in particular example 8, wherein the beamforming scheme is a delay-and-sum beamforming scheme.

[0210] Example 11 is a method according to any example herein, in particular example 8, wherein classifying the line-of-sight and non-line-of-sight conditions comprises comparing periodicity of the histogram with a threshold, or wherein classifying the line-of-sight and non-line- of-sight conditions comprises comparing a phase peak of the histogram with a threshold..

[0211] Example 12 is a method according to any example herein, in particular example 11, wherein the threshold is predetermined or programmable.

[0212] Example 13 is a method according to any example herein, in particular example 8, wherein the array of antennas is a uniform antenna array.

[0213] Example 14 is a method according to any example herein, in particular example 8, wherein the beamformed signal is a signal matrix, wherein generating the histogram comprises generating an individual histogram of each column or row of the signal matrix, wherein classifying the line-of-sight and non-line-of-sight conditions comprises applying fast Fourier transform to the individual histogram and comparing an output of the fast Fourier transform against a threshold.

[0214] Example 15 is a machine-readable storage media having machine-readable instructions stored thereon, that when executed, cause one or more machines to perform a method comprising: applying a beamforming scheme to an output of an array of antennas to generate a beamformed signal; extracting phase information from the beamformed signal; generating a histogram of the phase information; and classifying line-of-sight and non-line-of-sight conditions based on the histogram.

[0215] Example 16 is a machine-readable storage media according to any example herein, in particular example 15, wherein the array of antennas includes N antenna elements, and wherein applying the beamforming scheme comprises aligning phases of signals received at individual elements of the N antenna elements.

[0216] Example 17 is a machine-readable storage media according to any example herein, inOSU02P052PCT (OSU-24-37) particular example 15, wherein the beamforming scheme is a delay-and-sum beamforming scheme.

[0217] Example 18 is a machine-readable storage media according to any example herein, in particular example 15, wherein classifying the line-of-sight and non-line-of-sight conditions comprises comparing periodicity of the histogram with a threshold or wherein classifying the line- of-sight and non-line-of-sight conditions comprises comparing a phase peak of the histogram with a threshold.

[0218] Example 19 is a machine-readable storage media according to any example herein, in particular example 18, wherein the threshold is predetermined or programmable.

[0219] Example 20 is a machine-readable storage media according to any example herein, in particular example 18, wherein the beamformed signal is a signal matrix, wherein generating the histogram comprises generating an individual histogram of each column or row of the signal matrix, and wherein classifying the line-of-sight and non-line-of-sight conditions comprises applying fast Fourier transform to the individual histogram and comparing an output of the fast Fourier transform against a threshold.

[0220] Here, some methods and devices may be shown in block diagram form, rather than in detail, to avoid obscuring present disclosure. Reference throughout this specification to “an embodiment” or “one embodiment” or “some embodiments” means that a particular feature, structure, function, or characteristic described in connection with an embodiment is included in at least one embodiment of disclosure. Thus, appearances of phrase “in an embodiment” or “in one embodiment,” “in at least one embodiment,” or “some embodiments” in various places throughout this specification are not necessarily referring to same embodiment of disclosure. Furthermore, particular features, structures, functions, or characteristics can be combined in any suitable manner in one or more embodiments. For example, a first embodiment can be combined with a second embodiment anywhere particular features, structures, functions, or characteristics associated with two embodiments are not mutually exclusive. A list of definitions follows, whereby following definitions may provide or augment literal support for claims.

[0221] As used in herein, singular forms “a,” “an,” and “the” are intended to include plural forms as well, unless context clearly indicates otherwise. It will also be understood that term “and / or” as used herein refers to and encompasses all possible combinations of one or more of associated listed items.OSU02P052PCT (OSU-24-37)

[0222] Here, “coupled” and “connected,” along with their derivatives, may be used herein to describe functional or structural relationships between components. These terms are not intended as synonyms for each other. Rather, in particular embodiments, “connected” may be used to indicate that two or more elements are in direct physical, optical, or electrical contact with each other. “Coupled” may be used to indicated that two or more elements are in either direct or indirect (with other intervening elements between them) physical, electrical or in magnetic contact with each other, and / or that two or more elements co-operate or interact with each other (e.g., as in a cause an effect relationship). Coupled may also have the meaning of non-mechanical contact or connection. Coupling may also have the meaning of thermal connectivity, where one object may be a heat source and another object may be a heat sink, either in thermal equilibrium with each other or subject to a common conductive, convective or radiative heat flow between them; electrically coupled, where objects may be connected electrically in an electric or electronic circuit and a current flow may be induced by application of a voltage between the electrically interconnected objects or by an electric field between mechanically coupled or isolated objects; magnetically, where two mechanically coupled or isolated objects mutually share a common magnetic field flux; and fluidically, where objects such as vessels and conduits may share a common gas or liquid fluid that is static or flowing.

[0223] Here, a device that is “configured to” perform a task or function may be configured (e.g., programmed and / or hardwired) at a time of manufacturing by a manufacturer to perform the function. In at least one embodiment, the device may be configurable (or reconfigurable) by a user after manufacturing to perform the function and / or other additional or alternative functions. In at least one embodiment, the configuring may be through firmware and / or software programming of the device, through a construction and / or layout of hardware components and interconnections of the device, or a combination thereof.

[0224] Here, “adjacent” can generally refer to a position of a thing being next to (e.g., immediately next to or close to with one or more things between them) or adjoining another thing (e.g., abutting it).

[0225] Unless otherwise specified in explicit context of their use, terms “substantially equal,” “about equal,” and “approximately equal” can generally mean that there is no more than incidental variation between two things so described. In at least one embodiment, such variation is no more than + / -10% of referred value.OSU02P052PCT (OSU-24-37)

[0226] Besides what is described herein, various modifications can be made to disclosed embodiments and embodiments thereof without departing from their scope. Therefore, illustrations of embodiments herein should be construed as examples, and not restrictive to scope of present disclosure.

Claims

OSU02P052PCT (OSU-24-37) CLAIMS What is claimed is:

1. An apparatus comprising: an array of antennas; and a processor coupled to the array, the processor to: apply a beamforming scheme to an output of the array to generate a beamformed signal; extract phase information from the beamformed signal; generate a histogram of the phase information; and classify line-of-sight and non-line-of-sight conditions based on the histogram.

2. The apparatus of claim 1, wherein the processor is to identify distinct regular peaks in the histogram to classify the line-of-sight condition.

3. The apparatus of claim 1, wherein the processor is to identify lack of peaks in the histogram to classify the non-line-of-sight condition.

4. The apparatus of claim 1, wherein the array of antennas includes N antenna elements, and wherein the beamforming scheme is to align phases of signals received at individual elements of the N antenna elements.

5. The apparatus of claim 1, wherein the beamforming scheme is a delay-and-sum beamforming scheme.

6. The apparatus of claim 1, wherein the processor is to classify the line-of-sight and non- line-of-sight conditions by comparison of periodicity of the histogram with a threshold.

7. The apparatus of claim 6, wherein the threshold is predetermined or programmable.

8. The apparatus of claim 1, wherein the array of antennas is a uniform antenna array.OSU02P052PCT (OSU-24-37) 9. The apparatus of claim 1, wherein the beamformed signal is a signal matrix, wherein the processor is to generate an individual histogram of each column or row of the signal matrix, wherein the processor is to apply fast Fourier transform to the individual histogram, and wherein the processor is to classify the line-of-sight and non-line-of-sight conditions by comparison of an output of the fast Fourier transform against a threshold.

10. A method comprising: applying a beamforming scheme to an output of an array of antennas to generate a beamformed signal; extracting phase information from the beamformed signal; generating a histogram of the phase information; and classifying line-of-sight and non-line-of-sight conditions based on the histogram.

11. The method of claim 10, wherein the array of antennas includes N antenna elements, and wherein applying the beamforming scheme comprises aligning phases of signals received at individual elements of the N antenna elements.

12. The method of claim 10, wherein the beamforming scheme is a delay-and-sum beamforming scheme.

13. The method of claim 10, wherein classifying the line-of-sight and non-line-of-sight conditions comprises comparing periodicity of the histogram with a threshold.

14. The method of claim 13, wherein the threshold is predetermined or programmable.

15. The method of claim 10, wherein the array of antennas is a uniform antenna array.

16. The method of claim 10, wherein the beamformed signal is a signal matrix, wherein generating the histogram comprises generating an individual histogram of each column or row of the signal matrix, wherein classifying the line-of-sight and non-line-of-sight conditions comprises applying fast Fourier transform to the individual histogram and comparing an output of the fast Fourier transform against a threshold.OSU02P052PCT (OSU-24-37) 17. A machine-readable storage media having machine-readable instructions stored thereon, that when executed, cause one or more machines to perform a method comprising: applying a beamforming scheme to an output of an array of antennas to generate a beamformed signal; extracting phase information from the beamformed signal; generating a histogram of the phase information; and classifying line-of-sight and non-line-of-sight conditions based on the histogram.

18. The machine-readable storage media of claim 17, wherein the array of antennas includes N antenna elements, and wherein applying the beamforming scheme comprises aligning phases of signals received at individual elements of the N antenna elements.

19. The machine-readable storage media of claim 17, wherein classifying the line-of-sight and non-line-of-sight conditions comprises comparing periodicity of the histogram with a threshold.

20. The machine-readable storage media of claim 19, wherein the beamformed signal is a signal matrix, wherein generating the histogram comprises generating an individual histogram of each column or row of the signal matrix, and wherein classifying the line-of- sight and non-line-of-sight conditions comprises applying fast Fourier transform to the individual histogram and comparing an output of the fast Fourier transform against a threshold.

Images