Collision detection and pointing correction for optical inter-satellite links

Abstract

To provide a system and method for collision detection and pointing correction ("IPDC") in optical intersatellite links ("OISL"). [Solution] The system includes an optical terminal, an IPDC system, and a steering module. The optical terminal receives a laser signal from a second satellite via OISL and outputs signal measurement parameters based on the laser signal. The IPDC system determines a number of angle correction values ​​using (i) a mapping between a predetermined collision pointing angle of the optical terminal and a corresponding angle correction value, or (ii) a machine learning model configured to receive signal measurement parameters as input and output angle correction values. The IPDC system determines a pointing correction value based on the signal measurement parameters and angle correction values. The steering module steers the optical terminal based on the pointing correction values.

Description

[Technical Field]

[0001] The following concerns, in general, optical inter-satellite links, and more specifically, systems and methods for improving the connectivity of optical inter-satellite links. [Background technology]

[0002] Optical inter-satellite links (OISLs) are becoming more widespread today due to the increasing number of satellites in low Earth orbit (LEO), the advantage of directly interconnecting satellites (i.e., without the need to relay information through ground stations), and the advantages in data transmission rates compared to inter-satellite radio frequency (RF) links. OISLs establish inter-satellite links using a pointing, acquisition, and tracking process based on pointing, acquisition, and tracking (PAT) techniques.

[0003] In some existing systems, the PAT technique uses an Acquisition and Tracking Sensor (ATS). The ATS is typically a quad-cell photodiode detector. Figure 6 shows an exemplary quad-cell photodetector 600, having cells 602, 604, 606, and 608. The ATS is divided into four identical cells (each cell referred to as A, B, C, and D, respectively) evenly spaced in a rectangular pattern around the center of the ATS. Each cell acquires power measurements from the laser beam used for OISL.

[0004] During the pointing, acquisition, and tracking processes, power measurements from each cell (A, B, C, D) are collected and compared using the controller. The controller calculates feedback azimuth and altitude factors to determine the direction from which the laser beam is coming and how to optimally steer the telescope toward the transmitting satellite. The feedback factors are: Equation (1) Azimuth (horizontal) coefficient, that is,

[0005]

number

[0006] Equation (2) Height (vertical) coefficient, that is,

[0007]

number

[0008] That is the case.

[0009] Due to the extremely narrow beam size and high relative satellite displacement velocity (5-8 kilometers per second), the PAT technique requires the accurate and rapid aggregation of azimuth and altitude factors to properly determine the pointing correction values ​​required to create and / or maintain the link. Pointing errors may arise, in particular, due to link budget fluctuations, satellite attitude errors, thermoelastic strain, ephemeris inaccuracies, and deformation of the laser beam caused by interference (e.g., mild and severe) to the laser beam. Interference to the laser beam may occur due to the presence of collisions within the field of view (FOV) of the two interconnected optical terminals, such as antennas mounted on the satellite. Interference to the laser beam is particularly dominant for OISLs operating at low elevation angles relative to their reference boundary plane. [Overview of the project] [Problems that the invention aims to solve]

[0010] Therefore, there is a need for improved systems and methods for OISL collision detection and pointing correction that overcome at least some of the shortcomings of existing systems and methods. [Means for solving the problem]

[0011] A system for use in a first satellite for collision detection and pointing correction of an optical intersatellite link (OISL) is provided. The system includes an optical terminal for receiving a laser signal from a second satellite via the OISL and outputting signal measurement parameters based on the laser signal; an Impingement Detection and Pointing Correction System (IDPC) configured to determine a number of angle correction values ​​using a machine learning model, which is stored in a data storage device and configured to receive signal measurement parameters as input and output angle correction values, and to determine a pointing correction value based on the signal measurement parameters and angle correction values; and a steering module configured to steer the optical terminal based on the pointing correction value.

[0012] In one embodiment, the optical terminal receives the laser signal via a laser signal acquisition device equipped with an acquisition and tracking sensor (ATS).

[0013] In one embodiment, the ATS is a four-segment photodiode detector.

[0014] In one embodiment, the angle correction value includes an angle correction coefficient for each quadrant in the four-segment photodiode detector.

[0015] In one embodiment, a predetermined collision pointing angle is defined by the azimuth angle and elevation angle of the optical head of the optical terminal with respect to the satellite bodies of the first satellite and the second satellite.

[0016] In one embodiment, a predetermined collision pointing angle in mapping includes an azimuth angle and an elevation angle.

[0017] In one embodiment, a pointing correction value is added to the azimuth signal and altitude signal to bias the controller of the optical terminal to the actual center of the non-deformed power capture pattern.

[0018] In one embodiment, the angle correction value comprises an angle correction coefficient configured based on the angular range of the azimuth angle and elevation angle where the collision exists and the sensitivity of the area obstructed by the collision.

[0019] In one embodiment, the angle correction coefficient for each azimuth angle and elevation angle is estimated through simulation and tested against the efficiency in the static mode or learning mode.

[0020] In one embodiment, a predetermined collision pointing angle is identified through a field of view (FOV) analysis performed during the design of the first satellite and the second satellite.

[0021] In one embodiment, the FOV analysis further includes identifying the location of any collision pointing angle over the full azimuth range and altitude range of the optical terminal, and the collision pointing angle is stored in the mapping together with the corresponding angle correction value.

[0022] In one embodiment, a predetermined collision pointing angle includes the azimuth angle and elevation angle where the collision exists, the shape of the collision, and the impact of the collision.

[0023] In one embodiment, the pointing correction value includes one or more commands to the optical terminal for changing the azimuth and elevation angles of the optical terminal in a manner that reduces the pointing error.

[0024] In one embodiment, the onboard processor receives a communication from the network operating center indicating that the second satellite is an insider satellite, and the onboard processor acquires an angle correction value only when it identifies the second satellite as an insider satellite.

[0025] In one embodiment, the onboard processor uses configuration information from a previous connection between the first satellite and the second satellite, stored in the computer memory on the first satellite, indicating that the second satellite is an insider satellite, and the onboard processor obtains an angle correction value only when identifying the second satellite as an insider satellite.

[0026] In one embodiment, the first satellite and the second satellite are part of the same satellite constellation.

[0027] In one embodiment, the machine learning model includes at least one of least squares mean, recursive least squares, Bayesian, maximum likelihood, correlation, and covariance matching.

[0028] In one embodiment, the machine learning model includes transfer learning.

[0029] In one embodiment, the machine learning model includes an anomaly detection model.

[0030] In one embodiment, the machine learning model includes a time series model.

[0031] In one embodiment, the time-series model includes a long-term short-term memory model.

[0032] In one embodiment, the time series model includes a recurrent neural network-based model.

[0033] A method for collision detection and pointing correction of an optical satellite link (OISL) is also provided. The method includes receiving a laser signal from a second satellite via the OISL at the optical terminal of a first satellite when the optical terminal of the first satellite is pointing at a pointing angle; acquiring signal measurement parameters based on the laser signal by a sensor on the optical terminal; acquiring the azimuth and elevation angles of the optical terminal of the second satellite via an acquisition and tracking sensor (ATS) and outputting the signal measurement parameters by the optical terminal; storing a mapping of the pointing angle to an angle correction value in a data storage device on the first satellite, wherein the pointing angle is a predetermined collision pointing angle; acquiring the angle correction value from the data storage device by an IDPC (Identification and Pointing Correction) system on the first satellite when the optical terminal is at the pointing angle using the mapping; determining a pointing correction value by the IDPC system based on the signal measurement parameters and the angle correction value; and steering the optical terminal based on the pointing correction value.

[0034] A system for use on a first satellite is also provided. The system includes an optical terminal for receiving a laser signal from a second satellite via an optical inter-satellite link (OISL) and outputting signal measurement parameters based on the laser signal; a data storage device for storing a mapping of a predetermined collision pointing angle of the optical terminal to an angle correction value; a collision detection and pointing correction (IDPC) system configured to retrieve an angle correction value from the data storage device when the optical terminal is at a predetermined collision pointing angle using the mapping, and to determine a pointing correction value based on the signal measurement parameters and the angle correction value; and a steering module configured to steer the optical terminal based on the pointing correction value.

[0035] A system for use in a first satellite for collision detection and pointing correction of an optical intersatellite link is also provided. The system includes an optical terminal for receiving a laser signal from a second satellite via an optical intersatellite link (OISL) and outputting signal measurement parameters based on the laser signal; a collision detection and pointing correction (IDPC) system configured to obtain multiple angle correction values ​​for correcting a signal measurement parameter by, if the second satellite is an insider satellite, obtaining an angle correction value from a mapping data structure that maps one or more predetermined collision pointing angles of the optical terminal to respective angle correction values, the mapping data structure being stored in a data storage device; and if the second satellite is an outsider satellite, obtaining an angle correction value from a machine learning model configured to receive a signal measurement parameter as input and output an angle correction value; and a steering module configured to steer the optical terminal based on the pointing correction value.

[0036] In one embodiment, the second satellite is identified as an insider satellite or an outsider satellite based on information provided to the first satellite from the network operating center.

[0037] In one embodiment, the second satellite is identified as an insider satellite or an outsider satellite based on information stored in the computer memory onboard the first satellite from a previous connection between the first satellite and the second satellite.

[0038] A method for collision detection and pointing correction of an optical intersatellite link (OISL) performed on a first satellite is also provided. The method includes storing information in the computer memory or data storage device onboard the first satellite indicating whether the second satellite is an insider satellite with known OISL collisions or an outsider satellite with unknown OISL collisions; if a signal indicates that the second satellite is an insider satellite, obtaining angular correction values ​​from a mapping stored as a data structure in the data storage device on the first satellite by a collision detection and pointing correction (IDPC) system; if a signal indicates that the second satellite is an outsider satellite, obtaining angular correction values ​​from a machine learning model performed by an onboard processor by the IDPC system; and performing pointing correction on the optical terminal of the first satellite used for transmitting or receiving over the OISL based on the angular correction values.

[0039] In one embodiment, information is obtained from signals provided by the network operating center.

[0040] In one embodiment, the information is from a previous connection between the first satellite and the second satellite.

[0041] A method for collision detection and pointing correction of an optical intersatellite link (OISL) is also provided. The method includes receiving a laser signal from a second satellite via an optical intersatellite link (OISL) at an optical terminal of a first satellite; outputting signal measurement parameters from the optical terminal based on the laser signal; providing the signal measurement parameters as input to a machine learning model performed by a collision detection and pointing correction (IDPC) system of the first satellite, wherein the machine learning model is configured to receive the signal measurement parameters as input and output a plurality of angular correction values; determining the angular correction values ​​using the machine learning model; determining a pointing correction value by the IDPC system based on the signal measurement parameters and the angular correction values; and steering the optical terminal based on the pointing correction values.

[0042] A system for collision detection and pointing correction of the optical intersatellite link (OISL) in the first satellite is also provided. The system includes a data storage device that stores a mapping data structure that maps each of one or more predetermined collision pointing angles of an optical terminal to a respective angle correction value; a machine learning model configured to receive signal measurement parameters recorded by the optical terminal as input and to output angle correction values; and one or more processors configured to determine, based on signals received by the first satellite, whether a second satellite communicating with the first satellite through the optical terminal is of the first class or the second class; to determine an angle correction value using the mapping data structure when the second satellite is determined to be of the first class; to determine an angle correction value using the machine learning model when the second satellite is determined to be of the second class; and to determine a pointing correction value based on the signal measurement parameters and the angle correction value, the pointing correction value being used to steer the optical terminal.

[0043] In one embodiment, the signal is provided by the network operating center.

[0044] In one embodiment, the signal is provided by a second satellite.

[0045] In one embodiment, the first class is a characterized satellite class, and the second class is an uncharacterized satellite class.

[0046] Other embodiments and features will become apparent to those skilled in the art upon consideration of the following description of some exemplary embodiments.

[0047] The drawings included herein are for illustrative purposes only and illustrate various examples of the articles, methods, and apparatus described herein. [Brief explanation of the drawing]

[0048] [Figure 1] This is a block diagram of a system for inter-satellite communication using OISL according to one embodiment. [Figure 2] This is a more detailed block diagram of the collision detection and pointing correction system shown in Figure 1, according to one embodiment. [Figure 3] This is a flowchart of a method for OISL collision detection and pointing correction according to one embodiment. [Figure 4] This is a flowchart of a method for OISL collision detection and pointing correction for known collisions according to one embodiment. [Figure 5] This is a flowchart of a method for OISL collision detection and pointing correction for unknown collisions according to one embodiment. [Figure 6] This is a schematic diagram of a four-segment photodetector or sensor, sometimes called a quad cell, according to one embodiment. [Modes for carrying out the invention]

[0049] Various apparatuses or processes are described below to provide examples of each claimed embodiment. The embodiments described below are not limiting to any claimed embodiment, and any claimed embodiment may cover processes or apparatuses different from those described below. The claimed embodiments are not limited to apparatuses or processes having all of the features of any one apparatus or process described below, or to features common to some or all of the apparatuses described below.

[0050] One or more systems described herein may be implemented by computer programs running on a programmable computer having at least one processor, a data storage system (including volatile and non-volatile memory and / or memory elements), at least one input device, and at least one output device. For example, the programmable computer may be a programmable logic unit or any other suitable computing device, but is not limited to these.

[0051] Each program is preferably implemented in a high-level procedural or object-oriented programming language and / or scripting language for communication with a computer system. However, the program may be implemented in assembly language or machine language, if desired. In either case, the language may be a compiled language or an interpreted language. Each such computer program is preferably stored in a storage medium or device that is readable by a general-purpose or dedicated programmable computer for configuring and operating the computer when the storage medium or device is read by the computer to perform the procedures described herein.

[0052] The description of embodiments involving several components communicating with one another does not necessarily imply that all such components are required. On the contrary, various optional components are described in order to illustrate the wide variety of possible embodiments of the present invention.

[0053] Furthermore, process steps, method steps, algorithms, etc., may be described sequentially (in this disclosure and / or in the claims), but such processes, methods, and algorithms may be configured to function in an alternative order. In other words, any sequence or order of steps that may be described does not necessarily indicate a requirement that the steps be performed in that order. The steps of the processes described herein may be performed in any practical order. Furthermore, several steps may be performed simultaneously.

[0054] When a single device or article is described herein, it will be readily apparent that two or more devices / articles may be used instead of a single device / article (whether they work together or not). Similarly, when two or more devices or articles are described herein (whether they work together or not), it will be readily apparent that a single device / article may be used instead of two or more devices or articles.

[0055] The following relates, in general, to optical inter-satellite links, and more specifically, to systems and methods for collision detection and pointing correction for OISLs. In some embodiments, the systems and methods of the present disclosure use correction algorithms and / or machine learning models to obtain angular correction values ​​used to determine pointing correction values ​​for the optical terminals of the OISL. In some embodiments, whether a correction algorithm or a machine learning algorithm is used depends on whether the communicating satellites belong to a first class or category (to which correction algorithms apply) or a second class or category (to which machine learning models apply). Generally, the first and second classes may be distinguished based on any factors that may ensure or benefit from the differential use of correction algorithms or machine learning models. In certain embodiments, the first class or category corresponds to characterized satellites (with known collisions), and the second class or category corresponds to uncharacterized satellites (with unknown collisions).

[0056] The term “collision” and its variations as used herein may be used interchangeably with the term “interference” and its variations.

[0057] As used herein, the term “interference” and its variations (e.g., “interference”) encompass both mild and severe interference (e.g., first-degree and second-degree interference). Whether interference is considered or categorized as mild or severe may be determined with respect to a threshold (e.g., degree of signal attenuation, percentage of interference). In some embodiments, mild and severe interference may be distinguished by the degree of signal attenuation, with mild interference corresponding to relatively small attenuation (e.g., less than about 2 dB) and severe interference corresponding to larger attenuation (e.g., greater than about 2 dB). The concepts and techniques described herein are applicable to scenarios involving either mild or severe interference. Note that severe interference may exclude complete interference (i.e., complete blockage of the laser signal) where the quad cell is in darkness along with the four cells (e.g., A, B, C, D). In such cases, correction (e.g., using a lookup table) may not be helpful.

[0058] Next, referring to Figure 1, a system 100 for inter-satellite communication via OISL according to one embodiment is shown therein.

[0059] System 100 includes satellites 102-1 and 102-2. Satellites 102-1 and 102-2 communicate with each other via OISL 104. Communication via OISL 104 is bidirectional between satellites 102-1 and 102-2.

[0060] Satellites 102-1 and 102-2 may belong to the same satellite constellation or to different constellations. Satellites 102-1 and 102-2 belonging to the same constellation are insider satellites to each other. Satellites 102-1 and 102-2 belonging to different constellations are outsider satellites to each other. In some instances of this disclosure, insider satellites and outsider satellites may be referred to as characterized satellites and uncharacterized satellites, respectively. Characterized satellites include known collision scenarios in which the collision details at each satellite are known to each other, while uncharacterized satellites include unknown collision scenarios in which the collision details at each satellite are unknown to each other.

[0061] Satellites 102-1 and 102-2 each contain OISL units 106-1 and 106-2, respectively. OISL units 106-1 and 106-2 communicate via two laser beams. A laser beam is transmitted from one OISL unit in one direction at a given laser wavelength and received by the other OISL unit. Another laser beam is transmitted in the other direction at a different laser wavelength. The laser beams transport information between one satellite and the other. The laser beams transport information as a modulated optical signal that carries the information.

[0062] OISL units 106-1 and 106-2 are further configured to detect potential laser beams transmitted within the field of view (FOV) range of the OISL units.

[0063] The OISL units 106-1 and 106-2 each include optical terminals 120-1 and 120-2, respectively, which are configured to transmit and receive laser beams bidirectionally.

[0064] Upon receiving the laser beam signal, optical terminals 120-1 and 120-2 point the optical terminals to establish and maintain the OISL 104 using a laser signal acquisition device including an acquisition and tracking sensor (ATS).

[0065] The ATS comprises four individual cells ("quad cells") arranged around the center of the ATS. An exemplary quad cell is shown in Figure 6. Each cell of the ATS is configured to acquire power measurements of the laser beam. The power measurements perform individual power acquisition patterns.

[0066] Power measurements are output individually by optical terminals 120-1 and 120-2 as part of the Received Signals Measurement (RSM) 111 (also called signal measurement parameters). Once acquisition is achieved (i.e., in tracking mode), additional signal information is included in the RSM 111.

[0067] The ATS detects the direction from which the laser beam is coming and calculates the pointing error. In some cases, the incoming laser beam is split, with one portion going to the ATS and the other to the receiver for information detection and extraction. Alignment of the laser beam with the center of the ATS is achieved through calculations performed by the controller, resulting in optimal detection, acquisition, and tracking of the laser beam.

[0068] Alignment of OISL units 106-1 and 106-2 is achieved by a pointing and acquisition process. Alignment of OISL units 106-1 and 106-2 occurs when the orientation of the optical terminals 120-1 and 120-2 of OISL units 106-1 and 106-2 is such that the two optical terminals are aligned on the same optical axis and the bidirectional transmitting laser beam is at the center of the ATS of the optical terminals.

[0069] A laser beam may be disrupted by collisions in space at points along the path of the laser beam's pointing direction. For example, the collision could be an object attached to satellites 102-1 and 102-2 (e.g., an antenna). A laser beam disrupted by a collision will experience a reduction in received power and deformation. This reduction in received power and deformation of the laser beam at certain azimuth and elevation angles may result in a power pattern pointing deviation of the laser beam away from the ATS center that is suboptimal for detection, acquisition, and tracking of the laser beam. The disruption may be mild (sometimes called small or first-degree disruption) or severe (sometimes called large or second-degree disruption). Mild disruption may correspond to relatively small signal attenuation (e.g., less than about 2 dB). Severe disruption may correspond to relatively large signal attenuation (e.g., greater than about 2 dB). Thresholds may be used to categorize or distinguish between mild and severe disruption.

[0070] Collisions can be anticipated when the design details of the communicating satellites 102-1 and 102-2 are known. For example, if satellite 102-1 has knowledge of its own design and that of satellite 102-2 (e.g., shape, distance, azimuth, and elevation), then the details of a collision between satellites 102-1 and 102-2 can be stored in satellite 102-1 and therefore can be considered known ("known collision" scenario). Such details may include, for example, information on the azimuth and elevation of the optical terminal where interference occurs. In contrast, if satellite 102-1 does not have knowledge of its own design and that of satellite 102-2, then the collision between satellites 102-1 and 102-2 is unknown to satellite 102-1 ("unknown collision" scenario). Known collision scenarios can occur when satellites 102-1 and 102-2 belong to the same satellite constellation. Such satellites may be called insider satellites (or satellites characterized by known collision problems). An unknown collision scenario may occur when satellites 102-1 and 102-2 are not in the same constellation (and therefore their designs are not known to each other). Such satellites may be referred to as outsider satellites (or uncharacterized satellites) to each other (because the communicating satellite is an outsider). OISL unit 106-1 receives information from onboard processors such as onboard processor 108-1 whether OISL 104 is accompanied by an insider satellite or an outsider satellite.

[0071] Satellites 102-1 and 102-2 each include onboard processors 108-1 and 108-2, respectively. Onboard processor 108 controls OISL 106 by sending commands 124 to the OISL controller 112 of OISL unit 106 and by receiving telemetry values ​​126 from the OISL controller 112 of OISL unit 106. Onboard processor 108 may be implemented on one or more data processing devices. The connection between OISL unit 106-1 and OISL unit 106-2 may be established according to the connection plan. Onboard processor 108-1 receives both host satellite ephemeris and remote satellite ephemeris over time and provides OISL unit 106-1 with information on whether satellite 102-2 is an insider satellite or an outsider satellite in order to point to OISL unit 106-2. Once a target is acquired, OISL units 106-1 and 106-2 autonomously track each other.

[0072] The OISL units 106-1 and 106-2 each include collision detection and pointing correction (IDPC) systems 110-1 and 110-2, respectively. The IDPC systems 110-1 and 110-2 may also be implemented within the onboard processors 108-1 and 108-2, respectively. The IDPC systems may be implemented in one or more data processing devices.

[0073] IDPC systems 110-1 and 110-2 are configured to compensate for collisions based on RSM111 from OISL units 106-1 and 106-2, and to determine correction values ​​for pointing based on RSM111 and, where applicable, known collision scenarios. IDPC system 110 is further configured to generate azimuthal and altitude coefficients based on RSM111 and, where applicable, known collision scenarios. IDPC system 110 includes a controller such as IDPC controller 206 in Figure 2, and is configured to iteratively use the azimuthal and altitude coefficients until RSM111 converges to a small error that matches the data from either the lookup table 208 and machine learning module 212 shown in Figure 2. The controller may be, for example, a proportional-integral-derivative (PID) controller.

[0074] IDPC systems 110-1 and 110-2 are further configured to carry out either the IDPC process for insider satellites (known collision process) or the process for outsider satellites (unknown collision process). Insider satellites may be called characterized satellites, and outsider satellites may be called uncharacterized satellites (based on known and unknown collision processes, respectively).

[0075] A known collision scenario occurs when a collision is identified with respect to a given optical terminal within a specific combination of azimuth and elevation angles within the field of view (FOV) range for the OISL unit 106.

[0076] Unknown collision scenarios arise when connecting to an "outsider" optical terminal without recognizing the presence of collisions within the azimuth / elevation angle of the outsider optical terminal, and without knowledge of the shape, dimensions, and impact of collisions existing along the OISL with the outsider optical terminal.

[0077] The OISL units 106-1 and 106-2 each include OISL controllers 112-1 and 112-2, respectively. The OISL controllers 112-1 and 112-2 generate control signals for steering the OISL units 106-1 and 106-2. The OISL controllers 112-1 and 112-2 are configured to generate a pointing control signal 116 based on a pointing correction value 114 output by the IDPC systems 110-1 and 110-2.

[0078] OISL units 106-1 and 106-2 each include steering modules 118-1 and 118-2, respectively. Steering modules 118-1 and 118-2 generate motor control signals 122 according to a pointing control signal 116. Steering modules 118-1 and 118-2 are configured to move optical terminals 120-1 and 120-2, respectively, according to the motor control signal 122 in order to perform pointing correction of OISL units 106-1 and 106-2 of satellites 102-1 and 102-2.

[0079] Optical terminals 120-1 and 120-2 may, for example, comprise a telescope, a control module, and a set of optical sensors, mirrors, and lenses. The telescope is configured to be steered using motors at specific azimuth and elevation angles. The control module is configured to send signals to control the motors in order to steer the telescope in a specific direction. The optical sensors, mirrors, and lenses are configured, for example, to guide and amplify the received laser beam of the OISL 104 inside the telescope toward the receiver module of optical terminal 120-1, and further configured, for example, to guide and amplify the transmitted laser beam of the OISL 104 toward the telescope of optical terminal 120-1. Steering modules 118-1 and 118-2 perform the steering of the telescope by sending control signals 122 to the motors of optical terminals 120-1 and 120-2.

[0080] During operation, satellite 102-2 may transmit a laser beam from OISL unit 106-2 to satellite 102-1. Satellite 102-1 receives the laser beam at optical terminal 120-1 of OISL unit 106-1. OISL unit 106-1 outputs the RSM 111 captured at optical terminal 120-1 to IDPC system 110-1. IDPC system 110-1 is configured to determine a pointing correction value 114 based on the RSM 111 and, if applicable, known collision scenarios. The pointing correction value 114 is output to OISL controller 112-1. OISL controller 112-1 generates a pointing control signal 116 based on the pointing correction value 114. The pointing control signal 116 is sent to steering module 118-1. The steering module 118-1 adjusts the motor of the telescope of the optical terminal 120-1 using the motor control signal 122 according to the pointing control signal 116. The steering module 118-1 adjusts the motor of the telescope of the optical terminal 120-1 in order to steer the optical terminal 120-1 and perform pointing correction.

[0081] Next, referring to Figure 2, a more detailed representation of satellite 102-1 of Figure 1, according to one embodiment, is shown therein.

[0082] Satellite 102-1 receives the satellite laser beam 202 in the Acquisition and Tracking Sensor (ATS) 204 of the OISL unit 106-1. The laser beam 202 was transmitted by the OISL unit 106-2 of satellite 102-2 in Figure 1. The ATS 204 is a component of the optical terminal 120-1. The ATS 204 detects the laser beam 202, processes the laser beam 202, and generates an RSM 111. The RSM 111 includes, among other things, information about the direction of arrival of the laser beam 202. The onboard processor 108-1 receives information from satellite 102-2 via telemetry indicating whether the laser beam 202 is from an insider satellite or an outsider satellite. The onboard processor 108-1 provides this information to the OISL unit 106-1. This information may be received, for example, as part of prior communication via OISL 104 between satellites 102-1 and 102-2 before receiving the laser beam 202. The onboard processor stores in memory configuration information from the prior communication indicating whether satellite 102-2 is an insider or outsider satellite. Information on whether the laser beam 202 is from an insider or outsider satellite may also be received by the satellite from the network operating center. The onboard processor stores this information from the network operating center in memory. Accordingly, if no prior connection has been established and satellite 102-1 receives information on whether the laser beam is from an insider or outsider satellite, the satellite's onboard processor may store the signal for future use in subsequent OISL connections with the same satellite.

[0083] The first and second scenarios, in which satellite 102-2 is either an insider (i.e., characterized) satellite or an outsider (i.e., uncharacterized) satellite, are described in further detail below.

[0084] In the first scenario, OISL unit 106-1 recognizes that satellite 102-2 is an insider satellite for which known collision scenarios can be used.

[0085] Satellite 102-1 may have received this information from the network operating center, or it may have stored this information in the memory of its onboard processor 108-1 based on prior communication between satellites 102-1 and 102-2 via OISL 104. The onboard processor 108-1 provides the OISL unit 106-1 with the information (that satellite 102-2 is an insider satellite).

[0086] In this scenario, as will be further explained below, information regarding collisions within the respective fields of view (FOV) of satellites 102-1 and 102-2 is stored on satellite 102-1.

[0087] The IDPC system 110-1 includes an IDPC controller 206 and a lookup table (LUT) 208. The LUT 208 may also be stored in the data storage device of the OBP 108-1. The IDPC controller 206 receives RSM 111 from the ATS 204. The IDPC controller 206 is configured to provide input 213 to the LUT 208. Input 213 is a corresponding set (δ) of angle correction coefficients 210. A , δ B , δ C , δ D Input 213 includes the azimuth and elevation angles of the RSM111 and optical terminals 120-1 and 120-2, respectively, for obtaining the values ​​from LUT208. Input 213 further includes the address of LUT208, where the corresponding set of angle correction coefficients 210 is stored.

[0088] LUT208 is configured to provide angle correction coefficients 210 to the IDPC controller 206. The angle correction coefficients 210 stored in addresses within LUT208 correspond to the azimuth and elevation angles of satellites 102-1 and 102-2 where collisions are present. The angle correction coefficients 210 depend on the nature of the collision, including the range of angular azimuth and altitude where the collisions are present, as well as the sensitivity of the areas affected by the collisions. The angle correction coefficients 210 for each azimuth and elevation angle may be estimated via simulation and then tested for efficiency in either static or learning mode. As an example, LUT208 stores a mapping of the relationships between (a) the azimuth and elevation angles of optical terminals 120-1 and 120-2, (b) collision presence information, and (c) corresponding sets of angle correction coefficients. The collision presence information is known by satellite 102-1. During operation, the azimuth and elevation angles of optical terminals 120-1 and 120-2 relative to the bodies of satellites 102-1 and 102-2 are further known. As an example, information about the presence of a collision may be stored as a set comprising the azimuth, elevation, and angular correction coefficients. To compensate for misleading measurements caused by the collision, four angular correction coefficients are added to the power measurements from the quad cell. Applying such corrections, optical terminal 120-1 is steered in an optimized direction toward the opposite optical terminal 120-2. Thus, the angular correction coefficients of IDPC 110-1 offset the pointing error created by the collision.

[0089] For example, if the collision originates from an object on satellite 102-1 or 102-2, the range of azimuth and elevation angles in which the satellite laser beam 202 will be obstructed by the collision of satellites 102-1 and 102-2 is known from the design process of satellites 102-1 and 102-2. More specifically, FOV analysis may be performed during the design of satellites 102-1 and 102-2 to determine the location of the collision over the full azimuth and elevation range relative to optical terminals 120-1 and 120-2. For any azimuth and elevation position of optical terminals 120-1 and 120-2, an angular correction factor is obtained from LUT 208 and applied to the optical terminal. Known prior information about the collision includes the azimuth and elevation angles in which the collision occurs, the shape of the collision, and the impact of the collision.

[0090] The IDPC controller 206 is configured to determine, based on the RSM 111 and the angle correction coefficient 210, where the laser beam of the OISL unit 106-1 should be pointing so that the optical terminals 120-1 and 120-2 of satellites 102-1 and 102-2 are best directed in the presence of a collision.

[0091] The IDPC controller 206 is configured to generate commands to the optical terminal 120-1 to change its azimuth and elevation angles in a manner that minimizes the impact of the collision. Thus, the IDPC controller 206 reduces misleading power measurements caused by the presence of a collision by applying an angle correction coefficient that better represents the optimal direction of the laser beam toward the other optical terminal 120-2.

[0092] The IDPC controller 206 generates a pointing correction value 114 based on the RSM 111 and the angle correction coefficient 210. The pointing correction value 114 is provided to the OISL controller 112-1 in order to generate a pointing control signal 116.

[0093] The steering module 118-1 generates a motor control signal 122 based on the pointing correction value 114. The motor control signal 122 includes a signal to move the telescope motor of the optical terminal 120-1 of the OISL unit 106-1 in order to perform pointing correction on the OISL unit 106-1.

[0094] In the second scenario, OISL unit 106-1 receives information that satellite 102-2 is an outsider satellite (for example, satellites 102-1 and 102-2 are not in the same constellation and known collision scenarios cannot be used).

[0095] Satellite 102-1 may receive this information from the network operating center, or it may store this information in the memory of its onboard processor 108-1 based on prior communication between satellites 102-1 and 102-2 via OISL 104. Satellite 102-1 provides information from its onboard processor 108-1 to the OISL unit 106-1 (that satellite 102-2 is an outsider satellite or an uncharacterized satellite).

[0096] In this scenario, the IDPC system 110-1 includes a machine learning module 212 to enable the IDPC controller 206 to process the RSM 111 and generate a Pointing correction value 114. The machine learning module 212 may provide the IDPC controller 206 with various machine learning (ML) data 211.

[0097] The IDPC controller 206 is configured to provide input 214 to the ML module 212. Input 214 includes RSM111.

[0098] In one embodiment, the machine learning module 212 includes a machine learning model constructed or trained using any suitable mathematical optimization technique. Examples include, but are not limited to, least squares mean, recursive least squares, Bayesian, maximum likelihood, correlation, and covariance matching.

[0099] In one embodiment, the machine learning module 212 may use transfer learning with a pre-trained model to determine the presence of collisions in the path of the laser beam 202 from satellite 102-2, given a set of RSM 111. In such an embodiment, the ML data 211 contains information about the presence of collisions.

[0100] In another embodiment, the machine learning module 212 may use anomaly detection to determine the presence of collisions in space, given a set of RSM111. In such an embodiment, the ML data 211 contains information about the presence of collisions.

[0101] In another embodiment, the machine learning module 212 may use a time series model to determine angle correction coefficients based on the azimuth and elevation angles of the optical terminal 120-1 and a series of RSM111 stored over some time period. In such an embodiment, the ML data 211 contains information on angle correction coefficients.

[0102] In another embodiment, given a set of RSM111s, reinforcement learning (RL) may be used to develop a specific collision detection and pointing correction algorithm. In this embodiment, the machine learning model may be deployed to the IDPC controller 206. Since RL is a powerful artificial intelligence method, it can perform the same overall role in pointing optimization whether or not there is a collision. In that scenario, RL may be trained over a period of time using a set of azimuth and altitude ranges as well as corresponding received power measurements. These angular ranges can include both clear FOV and collision scenarios so that the algorithm optimizes pointing of the optical terminal 120-1 for both cases. RL may provide pointing correction values ​​114 to the OISL controller 112-1.

[0103] The OISL controller 112-1 receives a pointing correction value 114 and generates a pointing control signal 116. The steering module 118-1 performs pointing correction of the optical terminal 120-1 based on the pointing control signal 116. The steering module 118-1 may perform pointing correction by steering the telescope of the optical terminal 120-1 so that the laser beam 202 is pointing to the center of the ATS 204. The steering module 118-1 steers the optical terminal 120-1 using a motor control signal 122.

[0104] Next, referring to Figure 3, a method 300 for OISL collision detection and pointing correction according to one embodiment is shown therein. Method 300 may be carried out by the system 100 of Figure 1.

[0105] In 302, method 300 includes receiving a laser beam from the transmitting satellite's OISL in a satellite acquisition and tracking sensor (ATS), the ATS comprising a quad-cell photodiode detector with four sensors. The satellite may be satellite 102-1 in Figures 1-2.

[0106] In 304, method 300 includes obtaining power measurements from the laser beam in each of the four cells of the ATS and generating received signal measurements (RSM).

[0107] In 306, method 300 includes receiving information on whether the laser beam is from an insider satellite (and therefore a known collision) or from an outsider satellite (and therefore an unknown collision).

[0108] If, in 308, the satellite is determined to be an insider satellite in 306, method 300 further includes providing an RSM to a known IDPC module to obtain an angle correction factor. The known IDPC module performs a collision detection and pointing correction process for known collision types as described herein. OISL units 106-1 and 106-2 are operating in tracking mode and the azimuth and elevation angles of the insider satellite are available.

[0109] In one embodiment, a known collision detection and pointing correction module operates as follows:

[0110] Within a specific azimuth and elevation angle of the optical terminal's FOV range, a collision is identified relative to a given optical terminal. When the optical terminal is oriented so that it falls within the range where the collision exists, an angular correction factor is included in the azimuth and elevation factors to bias the IDPC controller toward convergence towards the actual center of the undeformed power acquisition pattern. The angular correction factor is added to the corresponding power measurements in Equations 1 and 2 above (i.e., (A, B, C, D) → (A ± δ) A , B±δ B , C±δ C , D±δ D The angle correction coefficients vary depending on the azimuth and elevation angles of the optical terminal, as well as the corresponding projection of the collision in transmit (Tx) and / or receive (Rx), and may be placed in a lookup table (LUT). The collision projection is the 2D region of interference created by the 3D volume. Thus, the IDPC controller "recognizes" the presence of a collision when operating within a specific azimuth and altitude range, and uses the LUT to perform pointing correction (collision detection and pointing correction, to perform IDPC).

[0111] If, in 310, the satellite is determined to be an outsider satellite in 306, method 300 further includes providing an RSM to an unknown IDPC module to obtain an angular correction coefficient or other pointing correction value. The unknown IDPC module then performs a collision detection and pointing correction process for an unknown collision type, as described herein.

[0112] Generally, the Unknown Collision Detection and Pointing Correction module employs artificial intelligence (AI) to assess the received laser beam signal level at each optical terminal of the satellite. The received signal level is assessed via constantly changing telemetry between the two optical terminals, and the transfer function is derived stepwise and loaded into a LUT for adaptive adjustment. The Unknown IDPC module includes a machine learning model trained to run as an IDPC module. The machine learning module includes parameters configured to store information on how collisions should be detected and how the OISL pointing should be corrected.

[0113] When an optical terminal connects to a terminal it has previously connected to, prior link information (such as link duration, explored azimuth and elevation angles, inter-satellite distance, and received signal) is used in a comparative manner to perform further AI learning and improve the quality of the connection between optical terminals (including bitrate and availability).

[0114] Depending on the rate of change of the received signal level between optical terminals, the prior links established between optical terminals, and the amount of available information between optical terminals, system 100 may employ multiple AI models, among other factors, depending on the required convergence rate, potential for divergence, and responsiveness to changes in signal levels.

[0115] Exemplary AI models that can be used include algorithms such as least squares mean, recursive least squares, Bayesian, maximum likelihood, correlation, and covariance matching.

[0116] In one embodiment, an unknown collision detection and pointing correction module operates as follows: This scenario may occur when communicating with an “outsider” terminal without awareness that a collision exists within its azimuth / altitude. An artificial intelligence (AI) process or machine learning (ML) algorithm may be invoked without such knowledge of the azimuth and altitude ranges in which the FOV is affected, or of the size / projection of the collision and its impact on the alignment of each terminal. When there is a lack of a rigorous model / knowledge but there is measurement “data” to learn, the AI ​​may provide a suitable solution to such a problem. To address this problem, various machine learning techniques may be used, but are not limited to: transfer learning to develop a pre-trained ML model, anomaly detection for collision detection, long short-term memory (LSTM) for pointing correction, and reinforcement learning (RL) to develop a more adaptive and optimized model. One or more combinations of these ML algorithms or techniques may be referred to as an “AI-assisted IDPC”. AI-assisted IDPC may improve PAT accuracy, reduce pointing errors, and provide improved link quality and availability.

[0117] Transfer learning is an efficient method for training an ML model to be used in another environment using data from one environment. Transfer learning can be used to train an ML model to be deployed and updated when collisions occur from an outsider satellite, using existing data from a constellation, for example, A, B, C, D measurements versus collisions. Doing so may improve the performance of the trained model and reduce the "learning" phase when the model is transferred to a different environment.

[0118] For anomaly detection, the first target of the ML algorithm is to detect the presence of a collision. When a specific set of power measurements is received by the ATS, one or more anomaly detection ML models can be used to train the model to flag the presence of a collision.

[0119] As an example, the anomaly detection model may employ a classification technique configured to take an input and classify that input into two or more categories. The classification technique may be trained by providing a set of power measurements to the model as an input corresponding to cases where there is no collision along the path of the laser beam, and labeling the corresponding output as not having a collision. The classification technique may be further trained by providing a set of power measurements to the model as an input corresponding to cases where there is a collision along the path of the laser beam, and labeling the corresponding output accordingly. The classification technique may be, for example, a K-nearest neighbor algorithm or a support vector machine.

[0120] In the case of pointing correction using LSTM, as explained, the angular correction coefficients (δ A , δ B , δ C , δ D ) should vary with the azimuth and elevation angles and thus vary over time during the PAT process. Different series of events in which power measurements (A, B, C, D) will be stored over some time period and processed accordingly to reflect the variation of the collision projection (masking) over time will provide different results (δ A , δ B , δ C , δ D ), and a time-series ML model may be trained to optimize the angular correction coefficients accordingly. LSTM is an example of an ML model that may be used to perform this function.

[0121] Other examples of time-series ML models include recurrent neural network (RNN) based models that enable learning of input data and the order or sequence of input data. Such models may be configured to capture variations in input data over time and to provide outputs specific to those variations. Such models may correlate variations in power measurements over time, along with the presence of collisions, while steering in azimuth and elevation.

[0122] Reinforcement learning (RL) is a more powerful alternative to time-series ML models. RL enables a continuous learning and adaptation process, as well as better case-specific performance optimization. RL may replace the entire controller job, and a carefully crafted IDPC algorithm may be developed for each scenario. In detail, RL may be used to provide input-specific outputs and a sequence of decisions made. As an example, one RL model may be trained to optimize pointing accuracy in scenarios where there are no collisions, where collisions exist and are from insider satellites, and where collisions exist and are from outsider satellites.

[0123] In 312, method 300 further includes performing OISL pointing correction using an angle correction coefficient.

[0124] Next, referring to Figure 4, a method 400 for OISL collision detection and pointing correction for known collisions, according to one embodiment, is shown therein. Method 400 may be carried out by the system 100 of Figure 1.

[0125] In 402, method 400 includes receiving a laser beam from the OISL of a transmitting satellite in a receiving satellite acquisition and tracking sensor (ATS), the ATS comprising a quad-segment photodiode detector having four cells.

[0126] In 404, method 400 includes obtaining power measurements from a laser beam in each of the four cells of the ATS.

[0127] In 406, method 400 includes obtaining an angle correction coefficient from a lookup table (LUT).

[0128] In 408, method 400 includes correcting power measurements using an angle correction coefficient.

[0129] In 410, method 400 includes determining a Poynting correction value based on the corrected power measurement.

[0130] In 412, method 400 includes performing pointing correction by steering an optical terminal.

[0131] Next, referring to Figure 5, a method 500 for OISL collision detection and pointing correction for an unknown collision, according to one embodiment, is shown therein. Method 500 may be carried out by the system 100 of Figure 1.

[0132] In 502, method 500 includes receiving a laser beam from the OISL of a transmitting satellite in a receiving satellite acquisition and tracking sensor (ATS). The ATS comprises a quad-segment photodiode detector having four cells.

[0133] In 504, method 500 includes obtaining power measurements from the laser beam in each of the four cells of the ATS and outputting received signal measurements (RSM).

[0134] In 506, method 500 includes providing an RSM as input to a machine learning model trained on existing data, including prior power measurements and prior angle correction coefficient calculations.

[0135] In 508, method 500 includes detecting the presence of a collision using a machine learning model.

[0136] In 510, method 500 includes predicting various angle correction coefficients (based on fluctuating azimuth and elevation angles) using a machine learning model.

[0137] In 512, method 500 includes determining a Poynting correction value by adding an angle correction coefficient to a power measurement.

[0138] In 514, method 500 includes performing pointing correction by steering an optical terminal.

[0139] While the above description provides examples of one or more devices, methods, or systems, it will be understood that other devices, methods, or systems may fall within the scope of the claims, as interpreted by those skilled in the art. [Explanation of Symbols]

[0140] 100 Systems 102 satellite 104 Optical Intersatellite Link (OISL) 106 Optical Intersatellite Link (OISL) Unit 108 Processors 110 Collision Detection and Pointing Correction (IDPC) System 111 Received signal measurement (RSM) 112 OISL Controllers 114 Pointing Correction Value 116 Pointing control signal 118 Steering Module 120 Optical Terminal 122 Motor control signal 124 Commands 126 Remote measurement values 202 Satellite laser beam 204 Capture and Tracking Sensor (ATS) 206 IDPC Controller 208 Look-up Table (LUT) 210 Angle correction coefficient 211 Machine Learning (ML) Data 212 Machine Learning Modules 213 Inputs 214 inputs 600 Quad-segment photodetector Cells 602, 604, 606, 608

Claims

1. A system for use in a first satellite for collision detection and pointing correction of optical intersatellite links (OISL), An optical terminal for receiving a laser signal from a second satellite via the OISL and outputting signal measurement parameters based on the laser signal, Collision detection and pointing correction system (IDPC), (i) A mapping between a predetermined collision pointing angle of the optical terminal and a corresponding angle correction value, which is stored as a data structure in a data storage device, or (ii) A machine learning model stored in the data storage device and configured to receive the signal measurement parameters as input and output the angle correction value. Use this to determine multiple angle correction values, The pointing correction value is determined based on the signal measurement parameters and the angle correction value. A collision detection and pointing correction system (IDPC) is configured as follows: A steering module configured to steer the optical terminal based on the pointing correction value, A system equipped with these features.

2. The system according to claim 1, wherein the optical terminal receives the laser signal via a laser signal acquisition device equipped with an acquisition and tracking sensor (ATS).

3. The system according to claim 2, wherein the ATS is a four-segment photodiode detector.

4. The system according to claim 3, wherein the angle correction value includes an angle correction coefficient for each quadrant in the four-segment photodiode detector.

5. The system according to claim 1, wherein the predetermined collision pointing angle is defined by the azimuth and elevation angles of the optical head of the optical terminal with respect to the satellite bodies of the first and second satellites.

6. The system according to claim 1, wherein the predetermined collision pointing angle in the mapping includes an azimuth angle and an elevation angle.

7. The system according to claim 1, wherein the pointing correction value is added to the azimuth signal and altitude signal in order to bias the controller of the optical terminal to the actual center of the undeformed power intake pattern.

8. The system according to claim 1, wherein the angle correction value comprises an angle correction coefficient configured based on the angular range of the azimuth and elevation angles in which the collision occurs and the sensitivity of the region obstructed by the collision.

9. The system according to claim 8, wherein the angle correction coefficients for each azimuth and elevation angle are estimated via simulation and tested for efficiency in static mode or learning mode.

10. The system according to claim 1, wherein the predetermined collision pointing angle is identified through field of view (FOV) analysis performed during the design of the first and second satellites.

11. The system according to claim 10, further comprising the FOV analysis identifying the location of any collision pointing angle over the full azimuthal and altitude range of the optical terminal, wherein the collision pointing angle is stored in the mapping along with a corresponding angle correction value.

12. The system according to claim 1, wherein the predetermined collision pointing angle includes the azimuth and elevation angles in which the collision occurs, the shape of the collision, and the effect of the collision.

13. The system according to claim 1, wherein the pointing correction value includes one or more commands to the optical terminal for changing the azimuth and elevation angles of the optical terminal in a manner that reduces the pointing error.

14. The system according to claim 1, wherein the onboard processor receives a communication from the network operating center indicating that the second satellite is an insider satellite, and the onboard processor acquires the angle correction value only when it identifies the second satellite as an insider satellite.

15. The system according to claim 1, wherein the onboard processor uses configuration information from a previous connection between the first satellite and the second satellite, stored in the computer memory on the first satellite, indicating that the second satellite is an insider satellite, and the onboard processor obtains the angle correction value only when it identifies the second satellite as an insider satellite.

16. The system according to claim 14, wherein the first satellite and the second satellite are part of the same satellite constellation.

17. The system according to claim 1, wherein the machine learning model includes at least one of least squares mean, recursive least squares, Bayesian, maximum likelihood, correlation, and covariance matching.

18. The system according to claim 1, wherein the machine learning model includes one or more of the following: transfer learning, anomaly detection models, time series models, long short-term memory models, and recurrent neural network-based models.

19. A method for collision detection and pointing correction of optical intersatellite links (OISLs), When the optical terminal of the first satellite is pointing at the pointing angle, the optical terminal receives a laser signal from the second satellite via the OISL. The steps include acquiring signal measurement parameters based on the laser signal using the sensor of the optical terminal, The steps include acquiring the azimuth and elevation angles of the optical terminal of the second satellite via an acquisition and tracking sensor (ATS), and outputting the signal measurement parameters via the optical terminal, A step of storing the mapping of the Pointing angle to an angle correction value in a data storage device on the first satellite, wherein the Pointing angle is a predetermined collision Pointing angle; When the optical terminal is at the pointing angle using the mapping, the angle correction value is obtained from the data storage device by the first satellite collision detection and pointing correction (IDPC) system; The steps include: determining a pointing correction value by the IDPC system based on the signal measurement parameters and the angle correction value; A step of steering the optical terminal based on the pointing correction value. A method that includes [a certain feature].

20. A system for use in a first satellite, An optical terminal for receiving a laser signal from a second satellite via an optical intersatellite link (OISL) and outputting signal measurement parameters based on the said laser signal, A data storage device that stores the mapping of a predetermined collision pointing angle of the optical terminal to an angle correction value, Collision detection and pointing correction (IDPC) system, When the optical terminal is at the predetermined collision pointing angle using the mapping, the angle correction value is obtained from the data storage device. The pointing correction value is determined based on the signal measurement parameters and the angle correction value. A collision detection and pointing correction (IDPC) system configured as follows: A steering module configured to steer the optical terminal based on the pointing correction value, A system equipped with these features.

Images