Cloud-based multi-vehicle path planner

Abstract

The aspects presented herein relate to the collection and compilation of sensed data. In one aspect, a network entity receives a first set of sensed data collected by a first set of user equipment (UEs). Based on the first set of sensed data, the network entity configures at least one of the following for a second set of UEs: a set with embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data. The network entity receives a second set of sensed data from the second set of UEs based on at least one of the configured set of planned routes and time plans or the configured set with embedded centroids.

Description

[0001] Cross-references to related applications

[0002] This application claims the benefit of U.S. Non-Provisional Patent Application Serial No. 18 / 404,625, entitled “CLOUD BASED MULTI VEHICLE PATH PLANNER”, filed January 4, 2024, the entire contents of which are expressly incorporated herein by reference. Technical Field

[0003] This disclosure relates generally to communication systems, and more specifically to wireless communication concerning data collection. Background Technology

[0004] Wireless communication systems are widely deployed to provide a variety of telecommunications services, such as telephone, video, data, messaging, and broadcasting. Typical wireless communication systems may employ multiple access technologies capable of supporting communication with multiple users by sharing available system resources. Examples of such multiple access technologies include Code Division Multiple Access (CDMA) systems, Time Division Multiple Access (TDMA) systems, Frequency Division Multiple Access (FDMA) systems, Orthogonal Frequency Division Multiple Access (OFDMA) systems, Single Carrier Frequency Division Multiple Access (SC-FDMA) systems, and Time Division Synchronous Code Division Multiple Access (TD-SCDMA) systems.

[0005] These multiple access technologies have been adopted in various telecommunications standards to provide a common protocol that enables different wireless devices to communicate at the city, national, regional, and even global levels. An example telecommunications standard is 5G New Radio (NR). 5G NR is part of the Continuous Evolution of Mobile Broadband (CEM) program issued by the 3rd Generation Partnership Project (3GPP) to meet new requirements associated with latency, reliability, security, scalability (e.g., with the Internet of Things (IoT),) and other requirements. 5G NR includes services associated with enhanced mobile broadband (eMBB), massive machine-type communications (mMTC), and ultra-reliable low-latency communications (URLLC). Some aspects of 5G NR can be based on the 4G Long Term Evolution (LTE) standard. Further improvements to 5G NR technology are needed. Furthermore, these improvements can also be applied to other multiple access technologies and telecommunications standards that adopt these technologies. Summary of the Invention

[0006] The following is a simplified summary of one or more aspects to provide a basic understanding of these aspects. This summary is not a comprehensive overview of all conceived aspects. It neither identifies key or essential elements of all aspects nor describes the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed descriptions that follow.

[0007] In one aspect of this disclosure, a method, computer-readable medium, and apparatus are provided. The apparatus receives a first set of sensed data collected by the first set of user equipment (UEs) from the first set of UEs. Based on the first set of sensed data, the apparatus configures at least one of the following for a second set of UEs: a set with embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data. The apparatus receives a second set of sensed data from the second set of UEs based on at least one of the configured set of planned routes and time plans or the configured set of embedded centroids.

[0008] In one aspect of this disclosure, a method, computer-readable medium, and apparatus are provided. The apparatus transmits a first set of sensed data collected by a UE to a network entity. Based on the first set of sensed data, the apparatus receives from the network entity a configuration for at least one of: a set of embedded centroids or a set of planned routes and time plans for collecting a second set of sensed data. The apparatus collects the second set of sensed data based on this configuration.

[0009] To achieve the foregoing and related objectives, one or more aspects may include the features fully described below and specifically pointed out in the claims. The following description and drawings set forth some exemplary features of one or more aspects in detail. However, these features indicate only a few of the various ways in which the principles of the various aspects may be employed. Attached Figure Description

[0010] Figure 1 This is a diagram illustrating an example of a wireless communication system and an access network.

[0011] Figure 2A This is an illustration of an example of the first frame according to various aspects of this disclosure.

[0012] Figure 2B This is a diagram illustrating examples of downlink (DL) channels within a subframe according to various aspects of this disclosure.

[0013] Figure 2C This is an illustration of an example of a second frame according to various aspects of this disclosure.

[0014] Figure 2D This is a diagram illustrating examples of uplink (UL) channels within a subframe according to various aspects of this disclosure.

[0015] Figure 3 This is a diagram illustrating examples of base stations and user equipment (UEs) in an access network.

[0016] Figure 4This is an illustration of examples of road object detection using machine learning (ML) / neural network (NN) models according to various aspects of this disclosure.

[0017] Figure 5 These are illustrations illustrating examples of data collection and compilation according to various aspects of this disclosure.

[0018] Figure 6 This is a diagram illustrating an example architecture of a centralized cloud storage system that, according to various aspects of this disclosure, can effectively filter out samples to be collected in order to reduce the cost of data storage, increase data diversity, and / or reduce redundant samples.

[0019] Figure 7 This is a diagram illustrating an example of an embedded centroid distributor according to various aspects of this disclosure.

[0020] Figure 8 This is a diagram illustrating examples of online task path planners according to various aspects of this disclosure.

[0021] Figure 9A This is an illustration of examples of vehicles collecting sensing data based on planned / proposed tracks, according to various aspects of this disclosure.

[0022] Figure 9B This is an illustration of examples of vehicles collecting sensing data based on planned / proposed tracks, according to various aspects of this disclosure.

[0023] Figure 9C This is an illustration of examples of vehicles collecting sensing data based on planned / proposed tracks, according to various aspects of this disclosure.

[0024] Figure 10 This is a diagram illustrating examples of online file selectors according to various aspects of this disclosure.

[0025] Figure 11 This is an example of a communication flow that illustrates a set of planned / designed trajectories configured by a network entity (e.g., a centralized cloud storage) to a UE (e.g., a vehicle) for the purpose of collecting perception data, according to various aspects of this disclosure.

[0026] Figure 12 This is a flowchart of a wireless communication method.

[0027] Figure 13 This is a flowchart of a wireless communication method.

[0028] Figure 14 This is a diagram illustrating an example of a hardware implementation used for an example network entity.

[0029] Figure 15This is a flowchart of a wireless communication method.

[0030] Figure 16 This is a flowchart of a wireless communication method.

[0031] Figure 17 These are illustrations illustrating specific hardware implementations of example devices and / or network entities. Detailed Implementation

[0032] The aspects presented in this paper can improve the efficiency of perception data collection and organization, thereby improving the overall performance of machine learning (ML) / neural network (NN) model training based on perception data (e.g., for autonomous / assisted driving systems). Autonomous perception data can specify large amounts of multi-sensor perception data to be collected across multiple countries, conditions, and scenarios. This can involve data collection and data organization processes. To date, data collection and data organization (dataset design) are separate steps, and therefore frequently generate redundant and undiversified data. The aspects presented in this paper provide a best-in-class cloud-based solution for collecting and optimizing data for route planning. Data collected globally from a pool of vehicles is configured to be uploaded to a centralized cloud-based entity for processing and optimization. The aspects presented in this paper may include the following aspects / features: online file selector, embedded storage, track storage, attribute storage, high-definition (HD) maps, online distribution allocator, embedded centroid allocator, online route planner, etc.

[0033] This paper presents a general approach that simultaneously satisfies evolving perception specifications and task-specific specifications, as well as logistical costs, in an online setting, and leverages the cloud to enhance data collection and selection across multiple vehicles. An online route planner can be configured to plan optimal routes for multiple vehicles to meet desired data distribution and geographic separation constraints. An online file selector can be configured to pick the most diverse and informative multi-model perception data for training convolutional neural network (CNN) / deep neural network (DNN) models for various autonomous driving perception tasks.

[0034] The detailed descriptions following, illustrated with reference to the accompanying drawings, describe various configurations and do not represent the only configurations in which the concepts described herein can be practiced. To provide a thorough understanding of the various concepts, the detailed descriptions include specific details. However, these concepts can be practiced without these specific details. In some instances, well-known structures and components are shown in block diagram form to avoid obscuring such concepts.

[0035] Various apparatuses and methods are presented with reference to several aspects of a telecommunications system. These apparatuses and methods are described in detail below and illustrated in the accompanying drawings by various blocks, components, circuits, processes, algorithms, etc. (collectively, “elements”). These elements may be implemented using electronic hardware, computer software, or any combination thereof. Whether such elements are implemented as hardware or software depends on the specific application and the design constraints imposed on the system as a whole.

[0036] As an example, an element, any part of an element, or any combination of elements may be implemented as a "processing system" including one or more processors. When multiple processors are implemented, the multiple processors may perform functions individually or in combination. Examples of processors include microprocessors, microcontrollers, graphics processing units (GPUs), central processing units (CPUs), application processors, digital signal processors (DSPs), reduced instruction set computing (RISC) processors, system-on-a-chip (SoCs), baseband processors, field-programmable gate arrays (FPGAs), programmable logic devices (PLDs), state machines, gate logic, discrete hardware circuits, and other suitable hardware configured to perform the various functionalities described throughout this disclosure. One or more processors in the processing system can execute software. Whether referred to as software, firmware, middleware, microcode, hardware description language, or other terms, software should be broadly interpreted as instructions, instruction sets, code, code segments, program code, programs, subroutines, software components, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, or any combination thereof.

[0037] Therefore, in one or more example aspects, specific implementations, and / or use cases, the described functionality may be implemented in hardware, software, or any combination thereof. If implemented in software, the functionality may be stored or encoded as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media. Storage media may be any available medium accessible to a computer. By way of example, such computer-readable media may include random access memory (RAM), read-only memory (ROM), electrically erasable programmable ROM (EEPROM), optical disc storage devices, magnetic disk storage devices, other magnetic storage devices, combinations of these types of computer-readable media, or any other medium that can be used to store computer-executable code in the form of instructions or data structures accessible to a computer.

[0038] While aspects, implementations, and / or use cases are described herein by way of example, additional or different aspects, implementations, and / or use cases may arise in many different arrangements and scenarios. The aspects, implementations, and / or use cases described herein can be implemented across many different platform types, devices, systems, shapes, sizes, and package arrangements. For example, aspects, implementations, and / or use cases may arise via integrated chip implementations and other devices based on non-modular components (e.g., end-user devices, vehicles, communication devices, computing devices, industrial equipment, retail / purchasing devices, medical devices, AI-enabled devices, etc.). While some examples may or may not be specific to a use case or application, the described examples may exhibit broad applicability. Aspects, implementations, and / or use cases can range from chip-level or modular components to non-modular, non-chip-level implementations, and further to aggregated, distributed, or original equipment manufacturer (OEM) devices or systems incorporating one or more of the technologies described herein. In some practical settings, devices incorporating the described aspects and features may also include additional components and features for implementing and practicing the claimed and described aspects. For example, the transmission and reception of wireless signals necessarily involve multiple components for analog and digital purposes (e.g., hardware components including antennas, RF chains, power amplifiers, modulators, buffers, processors, interleavers, adders / summers, etc.). The techniques described herein can be practiced in a wide variety of devices, chip-level components, systems, distributed arrangements, aggregated or decomposed components, end-user equipment, etc., of various sizes, shapes, and configurations.

[0039] The deployment of communication systems such as 5G NR systems can be arranged in a variety of ways using various components or parts. In a 5G NR system or network, network nodes, network entities, network mobility elements, radio access network (RAN) nodes, core network nodes, network elements or network equipment (such as base stations (BS)), or one or more units (or components) performing base station functionality can be implemented in aggregated or decomposed architectures. For example, BSs (such as Node B (NB), evolved NB (eNB), NR BS, 5G NB, access point (AP), transmit / receive point (TRP), or cell, etc.) can be implemented as aggregated base stations (also known as standalone BS or monolithic BS) or decomposed base stations.

[0040] Aggregated base stations can be configured to utilize a radio protocol stack that is physically or logically integrated within a single RAN node. Decentralized base stations can be configured to utilize a protocol stack that is physically or logically distributed across two or more units, such as one or more central or centralized units (CUs), one or more distributed units (DUs), or one or more radio units (RUs) (i.e., central or distributed units). In some respects, the CU may be implemented within a RAN node, and one or more DUs may co-located with the CU, or alternatively, may be geographically or virtually distributed across one or more other RAN nodes. DUs may be implemented to communicate with one or more RUs. Each of the CU, DU, and RU may be implemented as a virtual unit, namely a virtual central unit (VCU), a virtual distributed unit (VDU), or a virtual radio unit (VRU).

[0041] Base station operation or network design can take into account the aggregation characteristics of base station functionality. For example, decomposed base stations can be utilized in Integrated Access Backhaul (IAB) networks, Open Radio Access Networks (O-RAN (such as network configurations initiated by the O-RAN Alliance)), or Virtualized Radio Access Networks (vRAN, also known as Cloud Radio Access Networks (C-RAN)). Decomposition can include distributing functionality across two or more units in various physical locations, as well as virtually distributing the functionality of at least one unit, which enables flexibility in network design. The various units of a decomposed base station or decomposed RAN architecture can be configured to communicate wirelessly with at least one other unit.

[0042] Figure 1 Figure 100 illustrates an example of a wireless communication system and access network. The illustrated wireless communication system includes a decomposed base station architecture. The decomposed base station architecture may include one or more CUs 110, which may communicate directly with the core network 120 via a backhaul link, or indirectly with the core network 120 via one or more decomposed base station units, such as a near real-time (near-RT) RAN Intelligent Controller (RIC) 125 via an E2 link, or a non-real-time (non-RT) RIC 115 associated with a Service Management and Orchestration (SMO) framework 105, or both. CUs 110 may communicate with one or more DUs 130 via a corresponding midhaul link (such as an F1 interface). DUs 130 may communicate with one or more RUs 140 via a corresponding fronthaul link. RUs 140 may communicate with a corresponding UE 104 via one or more radio frequency (RF) access links. In some implementations, a UE 104 may be served simultaneously by multiple RUs 140.

[0043] Each of these units (i.e., CU 110, DU 130, RU 140, and near-RT RIC 125, non-RT RIC 115, and SMO frame 105) may include or be coupled to one or more interfaces configured to receive or transmit signals, data, or information (collectively, signals) via wired or wireless transmission media. Each of the units, or an associated processor or controller providing instructions to the communication interfaces of these units, may be configured to communicate with one or more other units via transmission media. For example, these units may include wired interfaces configured to receive signals or transmit signals to one or more other units via wired transmission media. Additionally, these units may include wireless interfaces that may include receivers, transmitters, or transceivers (such as RF transceivers) configured to receive signals or transmit signals to one or more other units via wireless transmission media.

[0044] In some aspects, the CU 110 can host one or more higher-level control functions. Such control functions may include Radio Resource Control (RRC), Packet Data Convergence Protocol (PDCP), Serving Data Adaptation Protocol (SDAP), etc. Each control function can be implemented using an interface configured to signal to other control functions hosted by the CU 110. The CU 110 can be configured to handle user plane functionality (i.e., Central Unit-User Plane (CU-UP)), control plane functionality (i.e., Central Unit-Control Plane (CU-CP)), or a combination thereof. In some implementations, the CU 110 can be logically divided into one or more CU-UP units and one or more CU-CP units. When implemented in an O-RAN configuration, the CU-UP units can communicate bidirectionally with the CU-CP units via an interface such as an E1 interface. The CU 110 can be implemented to communicate with the DU 130 for network control and signaling, as needed.

[0045] DU 130 may correspond to a logic unit that includes one or more base station functions for controlling the operation of one or more RU 140s. In some aspects, DU 130 may at least partially host one or more of the Radio Link Control (RLC) layer, Media Access Control (MAC) layer, and one or more high physical (PHY) layers (such as modules for forward error correction (FEC) encoding and decoding, scrambling, modulation and demodulation, etc.) according to functional splits (such as those defined by 3GPP). In some aspects, DU 130 may further host one or more low PHY layers. Each layer (or module) may be implemented using an interface configured to communicate signaling with other layers (and modules) hosted by DU 130 or with control functions hosted by CU 110.

[0046] Lower-layer functionality can be implemented by one or more RU 140s. In some deployments, an RU140 controlled by a DU 130 may correspond to a logical node that is at least partially based on functional decomposition, such as lower-layer functional decomposition, to host RF processing functions or low-PHY layer functions (such as performing Fast Fourier Transform (FFT), Inverse FFT (iFFT), digital beamforming, Physical Random Access Channel (PRACH) extraction and filtering, or both). In such architectures, the RU 140 may be implemented to handle over-the-air (OTA) communications with one or more UEs 104. In some specific implementations, the real-time and non-real-time aspects of control plane and user plane communications with the RU 140 may be controlled by the corresponding DU 130. In some scenarios, this configuration allows the DU 130 and CU 110 to be implemented in a cloud-based RAN architecture, such as a vRAN architecture.

[0047] SMO framework 105 can be configured to support RAN deployment and provisioning of both non-virtualized and virtualized network elements. For non-virtualized network elements, SMO framework 105 can be configured to support the deployment of dedicated physical resources for RAN coverage requirements, which can be managed via operation and maintenance interfaces such as the O1 interface. For virtualized network elements, SMO framework 105 can be configured to interact with a cloud computing platform such as Open Cloud (O-Cloud) 190 to perform network element lifecycle management (such as instantiating virtualized network elements) via a cloud computing platform interface such as the O2 interface. Such virtualized network elements may include, but are not limited to, CU 110, DU 130, RU 140, and near-RT RIC 125. In some implementations, SMO framework 105 can communicate with hardware aspects of the 4G RAN, such as Open eNB (O-eNB) 111, via the O1 interface. Additionally, in some implementations, SMO framework 105 can communicate directly with one or more RU 140s via the O1 interface. SMO framework 105 may also include a non-RT RIC 115 configured to support the functionality of SMO framework 105.

[0048] The non-RT RIC 115 can be configured to include logical functions enabling non-real-time control and optimization of RAN elements and resources, including artificial intelligence (AI) / machine learning (ML) workflows for model training and updates, or policy-based guidance for applications / features in the near-RT RIC 125. The non-RT RIC 115 can be coupled to or communicate with the near-RT RIC 125, such as via an A1 interface. The near-RT RIC 125 can be configured to include logical functions enabling near real-time control and optimization of RAN elements and resources via data collection and actions through an interface such as an E2 interface, connecting one or more CU 110s, one or more DU 130s, or both, and O-eNBs to the near-RT RIC 125.

[0049] In some implementations, to generate AI / ML models to be deployed in the near-RT RIC 125, the non-RT RIC 115 may receive parameters or external enrichment information from an external server. This information can be utilized by the near-RT RIC 125 and may be received from non-network data sources or network functions at the SMO framework 105 or the non-RT RIC 115. In some examples, the non-RT RIC 115 or the near-RT RIC 125 may be configured to tune RAN behavior or performance. For example, the non-RT RIC 115 may monitor long-term trends and patterns in performance and employ AI / ML models to perform corrective actions via the SMO framework 105 (such as reconfiguration via O1) or by creating RAN management policies (such as A1 policies).

[0050] At least one of CU 110, DU 130, and RU 140 may be referred to as base station 102. Therefore, base station 102 may include one or more of CU 110, DU 130, and RU 140 (each component is indicated by a dashed line to indicate that each component may or may not be included in base station 102). Base station 102 provides UE 104 with an access point to core network 120. Base station 102 may include macro cells (high-power cellular base stations) and / or small cells (low-power cellular base stations). Small cells include femtocells, picocells, and microcells. A network that includes both small cells and macro cells may be referred to as a heterogeneous network. A heterogeneous network may also include an evolved home node B (eNB) (HeNB), which can provide service to a restricted group referred to as a closed subscriber group (CSG). The communication link between RU 140 and UE 104 may include uplink (UL) transmission (also known as reverse link) from UE 104 to RU 140 and / or downlink (DL) transmission (also known as forward link) transmission from RU 140 to UE 104. The communication link may utilize multiple-input multiple-output (MIMO) antenna techniques, including spatial multiplexing, beamforming, and / or transmit diversity. The communication link may use one or more carriers. For each carrier allocated in a carrier aggregation of up to Yx MHz (x component carriers) for transmission in each direction, base station 102 / UE 104 may use a spectrum with a bandwidth of up to Y MHz (e.g., 5MHz, 10MHz, 15MHz, 20MHz, 100MHz, 400MHz, etc.). Carriers may be adjacent to each other or may not be adjacent to each other. Carrier allocation may be asymmetric with respect to DL and UL (e.g., more or fewer carriers may be allocated to DL compared to UL). Component carriers may include primary component carriers and one or more secondary component carriers. The primary component carrier can be referred to as the primary cell (PCell) and the secondary component carrier can be referred to as the secondary cell (SCell).

[0051] Some UEs 104 can communicate with each other using device-to-device (D2D) communication link 158. D2D communication link 158 can use DL / UL wireless wide area network (WWAN) spectrum. D2D communication link 158 can use one or more sidelink channels, such as Physical Sidelink Broadcast Channel (PSBCH), Physical Sidelink Discovery Channel (PSDCH), Physical Sidelink Shared Channel (PSSCH), and Physical Sidelink Control Channel (PSCCH). D2D communication can be performed through various wireless D2D communication systems, such as Bluetooth. ™ (Bluetooth is a trademark of the Bluetooth Special Interest Group (SIG), and is based on the IEEE 802.11 standard for Wi-Fi.) ™(Wi-Fi is a trademark of the Wi-Fi Alliance), LTE, or NR.

[0052] The wireless communication system may also include a Wi-Fi AP 150, which communicates with the UE 104 (also referred to as a Wi-Fi station (STA)) via a communication link 154, for example, in an unlicensed spectrum such as 5 GHz. When communicating in unlicensed spectrum, the UE 104 / AP 150 may perform a free channel assessment (CCA) to determine whether a channel is available before communication.

[0053] The electromagnetic spectrum is typically subdivided into various categories, bands, channels, etc., based on frequency / wavelength. In 5G NR, two initial operating bands have been designated as frequency ranges FR1 (410MHz to 7.125GHz) and FR2 (24.25GHz to 52.6GHz). Although a portion of FR1 is greater than 6GHz, in various documents and articles, FR1 is often (interchangeably) referred to as the "sub-6GHz" band. Similar naming issues sometimes occur with FR2, which is often (interchangeably) referred to as the "millimeter wave" band in documents and articles, although this is distinct from the Extremely High Frequency (EHF) band (30GHz to 300GHz) designated as "millimeter wave" by the International Telecommunication Union (ITU).

[0054] The frequencies between FR1 and FR2 are generally referred to as mid-band frequencies. Recent 5G NR studies have identified the operating bands used for these mid-band frequencies as the frequency range designation FR3 (7.125 GHz to 24.25 GHz). Bands falling within FR3 can inherit FR1 and / or FR2 characteristics, thus effectively extending the features of FR1 and / or FR2 to mid-band frequencies. Furthermore, higher frequency bands are currently being explored to extend 5G NR operation beyond 52.6 GHz. For example, three higher operating bands have been identified as the frequency range designations FR2-2 (52.6 GHz to 71 GHz), FR4 (71 GHz to 114.25 GHz), and FR5 (114.25 GHz to 300 GHz). Each of these higher frequency bands falls within the EHF band.

[0055] In view of the above, unless otherwise specified, the term "below 6 GHz" as used herein can broadly refer to frequencies less than 6 GHz, within FR1, or including intermediate frequency band frequencies. Furthermore, unless otherwise specified, the term "millimeter wave" as used herein can broadly refer to frequencies that can include intermediate frequency band frequencies, within FR2, FR4, FR2-2 and / or FR5, or within the EHF band.

[0056] Base station 102 and UE 104 may each include multiple antennas (such as antenna elements, antenna panels, and / or antenna arrays) to facilitate beamforming. Base station 102 may transmit beamformed signals 182 to UE 104 in one or more transmit directions. UE 104 may receive beamformed signals from base station 102 in one or more receive directions. UE 104 may also transmit beamformed signals 184 to base station 102 in one or more transmit directions. Base station 102 may receive beamformed signals from UE 104 in one or more receive directions. Base station 102 / UE 104 may perform beamforming training to determine the optimal receive and transmit directions for each of base station 102 / UE 104. The transmit and receive directions of base station 102 may be the same or different. The transmit and receive directions of UE 104 may be the same or different.

[0057] Base station 102 may include and / or be referred to as gNB, Node B, eNB, access point, transceiver base station, radio base station, radio transceiver, transceiver function, basic service set (BSS), extended service set (ESS), TRP, network node, network entity, network equipment, or some other suitable terminology. Base station 102 may be implemented as an integrated access and backhaul (IAB) node, relay node, sidelink node, aggregated (monolithic) base station with baseband units (BBU) (including CU and DU) and RU, or may be implemented as a decomposed base station including one or more of CU, DU, and / or RU. A collection of base stations that may include decomposed base stations and / or aggregated base stations may be referred to as Next Generation (NG) RAN (NG-RAN).

[0058] The core network 120 may include Access and Mobility Management Function (AMF) 161, Session Management Function (SMF) 162, User Plane Function (UPF) 163, Unified Data Management (UDM) 164, one or more location servers 168, and other functional entities. AMF 161 is the control node that handles signaling between UE 104 and the core network 120. AMF 161 supports registration management, connection management, mobility management, and other functions. SMF 162 supports session management and other functions. UPF 163 supports packet routing, packet forwarding, and other functions. UDM 164 supports authentication and key agreement (AKA) credential generation, user identity processing, access authorization, and subscription management. One or more location servers 168 are exemplified as including a Gateway Mobile Location Center (GMLC) 165 and a Location Management Function (LMF) 166. However, generally, one or more location servers 168 may include one or more location / positioning servers, which may include one or more of GMLC 165, LMF 166, Position Determination Entity (PDE), Serving Mobile Location Center (SMLC), Mobile Location Center (MPC), etc. GMLC 165 and LMF 166 support UE location services. GMLC 165 provides an interface for clients / applications (e.g., emergency services) to access UE location information. LMF 166 receives measurement and auxiliary information from NG-RAN and UE 104 via AMF 161 to calculate the location of UE 104. NG-RAN may use one or more positioning methods to determine the location of UE 104. Positioning UE 104 may involve signal measurement, location estimation, and optional rate calculation based on these measurements. Signal measurement may be performed by UE 104 and / or base station 102 serving UE 104. The measured signals may be based on one or more of the following: Satellite Positioning System (SPS) 170 (e.g., one or more of Global Navigation Satellite System (GNSS), Global Positioning System (GPS), Non-Terrestrial Network (NTN) or other satellite positioning / location systems), LTE signals, Wireless Local Area Network (WLAN) signals, Bluetooth signals, Terrestrial Beacon System (TBS), sensor-based information (e.g., barometric pressure sensor, motion sensor), NR Enhanced Cell ID (NR E-CID) method, NR signals (e.g., multiple round-trip time (multiple RTT), DL departure angle (DL-AoD), DL time difference of arrival (DL-TDOA), UL time difference of arrival (UL-TDOA), and UL angle of arrival (UL-AoA) positioning) and / or other systems / signals / sensors.

[0059] Examples of UE 104 include cellular phones, smartphones, Session Initiation Protocol (SIP) phones, laptops, personal digital assistants (PDAs), satellite radios, GPS devices, multimedia devices, video devices, digital audio players (e.g., MP3 players), cameras, game consoles, tablet devices, smart devices, wearable devices, vehicles, one or more components or computing platforms / devices implemented in vehicles, electricity meters, air pumps, large or small kitchen appliances, healthcare devices, implants, sensors / actuators, displays, or any other similarly functional devices. Some UEs in UE 104 may be referred to as IoT devices (e.g., parking timers, air pumps, toasters, vehicles, heart monitors, etc.). UE 104 may also be referred to as a station, mobile station, subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, mobile phone, user agent, mobile client, client, or some other suitable terminology. In some scenarios, the term UE may also be applied to one or more companion devices, such as in a device constellation arrangement. One or more of these devices may access the network together and / or individually.

[0060] Refer again Figure 1 In some aspects, UE 104 may have a sensing data collection component 198, which can be configured to send a first set of sensing data collected by the UE to a network entity; receive configuration from the network entity based on the first set of sensing data for at least one of: a set of embedded centroids or a set of planned routes and time plans for collecting a second set of sensing data; and collect a second set of sensing data based on the configuration. In some aspects, base station 102 or one or more location servers 168 may have a trajectory planning component 199, which can be configured to receive the first set of sensing data collected by the first set of user equipment (UE) from the first set of UE; configure for a second set of UEs at least one of: a set of embedded centroids or a set of planned routes and time plans for collecting a second set of sensing data; and receive a second set of sensing data from the second set of UEs based on at least one of the configured set of planned routes and time plans or the configured set of embedded centroids.

[0061] Figure 2A Figure 200 illustrates an example of the first subframe within a 5G NR frame structure. Figure 2B Figure 230 illustrates an example of a DL channel within a 5G NR subframe. Figure 2CFigure 250 is an example of a second subframe within a 5G NR frame structure. Figure 2D Figure 280 illustrates an example of a UL channel within a 5G NR subframe. The 5G NR frame structure can be Frequency Division Duplex (FDD) (where subframes within a specific set of subcarriers (carrier system bandwidth) are dedicated to either DL or UL), or Time Division Duplex (TDD) (where subframes within a specific set of subcarriers (carrier system bandwidth) are dedicated to both DL and UL). Figure 2A , Figure 2C In the provided example, the 5G NR frame structure is assumed to be TDD, where subframe 4 is configured using slot format 28 (most of which are DL), where D is DL, U is UL, and F is flexible between DL / UL, and subframe 3 is configured using slot format 1 (all of which are UL). Although subframes 3 and 4 are shown as having slot formats 1 and 28 respectively, any particular subframe can be configured using any of the various available slot formats 0-61. Slot formats 0 and 1 are both DL and UL, respectively. Other slot formats 2-61 include a mixture of DL, UL, and flexible symbols. The slot format is configured for the UE via the received Slot Format Indicator (SFI) (dynamically configured via DL Control Information (DCI) or semi-statically / statically configured via Radio Resource Control (RRC) signaling). Note that the following description also applies to the 5G NR frame structure as TDD.

[0062] Figures 2A to 2D The frame structure is illustrated, and aspects of this disclosure are applicable to other wireless communication technologies that may have different frame structures and / or different channels. A frame (10 ms) can be divided into 10 equal-sized subframes (1 ms). Each subframe may include one or more time slots. Subframes may also include micro-time slots, which may include 7, 4, or 2 symbols. Each time slot may include 14 or 12 symbols, depending on whether the cyclic prefix (CP) is normal or extended. For normal CP, each time slot may include 14 symbols, and for extended CP, each time slot may include 12 symbols. Symbols on the DL can be CP Orthogonal Frequency Division Multiplexing (OFDM) (CP-OFDM) symbols. Symbols on the UL can be CP-OFDM symbols (for high-throughput scenarios) or Discrete Fourier Transform (DFT) Extended OFDM (DFT-s-OFDM) symbols (for power-constrained scenarios; limited to single-stream transmission). The number of time slots within a subframe is based on the CP and parameter set. The parameter set defines the subcarrier spacing (SCS) (see Table 1). The symbol length / duration can be scaled by 1 / SCS.

[0063]

[0064] Table 1: Parameter Set, SCS, and CP

[0065] For a normal CP (14 symbols / slot), different parameter sets µ 0 through 4 allow 1, 2, 4, 8, and 16 slots per subframe, respectively. For an extended CP, parameter set 2 allows 4 slots per subframe. Therefore, for a normal CP and parameter set µ, there are 14 symbols / slot and 2... µ One time slot / subframe. Subcarrier spacing can be equal to ,in The parameter sets are 0 to 4. Therefore, the subcarrier spacing is 15 kHz for parameter set µ=0 and 240 kHz for parameter set µ=4. The symbol length / duration is negatively correlated with the subcarrier spacing. Figures 2A to 2D Examples of a normal frequency division multiplexing (CP) with 14 symbols per time slot and a parameter set of µ=2 with 4 time slots per subframe are provided. The time slot duration is 0.25 ms, the subcarrier spacing is 60 kHz, and the symbol duration is approximately 16.67 μs. Within the frame set, there may be one or more distinct bandwidth portions (BWPs) of frequency division multiplexing (see [link to relevant documentation]). Figure 2B Each BWP can have a specific set of parameters and CP (normal or extended).

[0066] A resource grid can be used to represent the frame structure. Each time slot consists of a resource block (RB) extending for 12 consecutive subcarriers (also known as a physical RB (PRB)). The resource grid is divided into multiple resource elements (REs). The number of bits carried by each RE depends on the modulation scheme.

[0067] like Figure 2A As illustrated, some of the REs carry reference (pilot) signals (RS) for the UE. RS may include demodulation RS (DM-RS) (indicated as R for a particular configuration, but other DM-RS configurations are possible) and channel state information reference signals (CSI-RS) for channel estimation at the UE. RS may also include beam measurement RS (BRS), beam refinement RS (BRRS), and phase tracking RS (PT-RS).

[0068] Figure 2BExamples of various DL channels within a subframe of a frame are illustrated. The Physical Downlink Control Channel (PDCCH) carries the DCI within one or more Control Channel Elements (CCEs) (e.g., 1, 2, 4, 8, or 16 CCEs), each CCE comprising six RE Groups (REGs), each REG comprising 12 consecutive REs in the OFDM symbol of the RB. A PDCCH within a BWP can be referred to as a Control Resource Set (CORESET). The UE is configured to monitor PDCCH candidates in the PDCCH search space (e.g., the common search space, the UE-specific search space) during PDCCH monitoring timing on the CORESET, where the PDCCH candidates have different DCI formats and different aggregation levels. Additional BWPs may be located at higher and / or lower frequencies on the channel bandwidth. The Primary Synchronization Signal (PSS) may be located within symbol 2 of a specific subframe of the frame. The PSS is used by the UE 104 to determine subframe / symbol timing and physical layer identification. The Secondary Synchronization Signal (SSS) may be located within symbol 4 of a specific subframe of the frame. The SSS is used by the UE to determine the Physical Layer Cell Identifier Group Number and radio frame timing. Based on the Physical Layer Identifier and the Physical Layer Cell Identifier Group Number, the UE can determine the Physical Cell Identifier (PCI). Based on the PCI, the UE can determine the location of the DM-RS. The Physical Broadcast Channel (PBCH), carrying the Master Information Block (MIB), can be logically grouped with the PSS and SSS to form a Synchronization Signal (SS) / PBCH block (also known as an SS block (SSB)). The MIB provides the System Frame Number (SFN) and the number of Restricted Blocks (RBs) in the system bandwidth. The Physical Downlink Shared Channel (PDSCH) carries user data, broadcast system information not transmitted via the PBCH (such as System Information Blocks (SIBs)), and paging messages.

[0069] like Figure 2C As illustrated, some REs in the REs carry DM-RS (indicated as R for one particular configuration, but other DM-RS configurations are possible) for channel estimation at the base station. The UE can transmit DM-RS for the Physical Uplink Control Channel (PUCCH) and DM-RS for the Physical Uplink Shared Channel (PUSCH). The PUSCH DM-RS can be transmitted in the first or first two symbols of the PUSCH. Depending on whether a short or long PUCCH is transmitted and depending on the specific PUCCH format used, the PUCCH DM-RS can be transmitted in different configurations. The UE can transmit a Sounding Reference Signal (SRS). The SRS can be transmitted in the last symbol of a subframe. The SRS can have a comb structure, and the UE can transmit the SRS on one of the comb teeth. The SRS can be used by the base station for channel quality estimation to enable frequency-dependent scheduling of the UL.

[0070] Figure 2DExamples of various UL channels within a subframe of a frame are illustrated. The PUCCH may be located as indicated in one configuration. The PUCCH carries uplink control information (UCI), such as scheduling requests, channel quality indicators (CQI), pre-decoding matrix indicators (PMI), rank indicators (RI), and hybrid automatic repeat request (HARQ) acknowledgment (ACK) (HARQ-ACK) feedback (i.e., one or more HARQ ACK bits indicating one or more ACKs and / or negative ACKs (NACKs)). The PUCCH carries data and may additionally be used to carry buffer status reports (BSR), power clearance reports (PHR), and / or UCIs.

[0071] Figure 3 This is a block diagram illustrating communication between base station 310 and UE 350 in the access network. In the DL, Internet Protocol (IP) packets can be provided to controller / processor 375. Controller / processor 375 implements Layer 3 and Layer 2 functionality. Layer 3 includes the Radio Resource Control (RRC) layer, and Layer 2 includes the Service Data Adaptation Protocol (SDAP) layer, Packet Data Convergence Protocol (PDCP) layer, Radio Link Control (RLC) layer, and Media Access Control (MAC) layer. The controller / processor 375 provides RRC layer functionality associated with broadcasting system information (e.g., MIB, SIB), RRC connection control (e.g., RRC connection paging, RRC connection establishment, RRC connection modification, and RRC connection release), inter-Radio Access Technology (RAT) mobility, and measurement configuration for UE measurement reporting; PDCP layer functionality associated with header compression / decompression, security (encryption, decryption, integrity protection, integrity verification), and handover support functions; RLC layer functionality associated with the delivery of upper-layer packet data units (PDUs), error correction via ARQ, concatenation, segmentation, and reassembly of RLC service data units (SDUs), resegmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functionality associated with mapping between logical channels and transport channels, multiplexing of MAC SDUs onto transport blocks (TBs), demultiplexing of MAC SDUs from TBs, scheduling information reporting, error correction via HARQ, priority handling, and logical channel priority ordering.

[0072] Transmit (TX) processor 316 and receive (RX) processor 370 implement Layer 1 functionality associated with various signal processing functions. Layer 1 (which includes the physical (PHY) layer) may include error detection on the transport channel, forward error correction (FEC) decoding / decoding of the transport channel, interleaving, rate matching, mapping to the physical channel, modulation / demodulation of the physical channel, and MIMO antenna processing. TX processor 316 processes the mapping to the signal constellation based on various modulation schemes (e.g., binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), M-order phase shift keying (M-PSK), M-order quadrature amplitude modulation (M-QAM)). The decoded and modulated symbols can then be divided into parallel streams. Each stream can then be mapped to OFDM subcarriers, multiplexed with a reference signal (e.g., a pilot) in the time and / or frequency domains, and subsequently combined using inverse fast Fourier transform (IFFT) to produce a physical channel carrying a stream of time-domain OFDM symbols. The OFDM stream undergoes spatial pre-decoding to generate multiple spatial streams. A channel estimate from channel estimator 374 can be used to determine the decoding and modulation scheme, as well as for spatial processing. This channel estimate can be derived from a reference signal transmitted by UE 350 and / or channel condition feedback. Each spatial stream can then be provided to a different antenna 320 via a separate transmitter 318Tx. Each transmitter 318Tx can use the corresponding spatial stream to modulate a radio frequency (RF) carrier for transmission.

[0073] At UE 350, each receiver 354Rx receives signals via its corresponding antenna 352. Each receiver 354Rx recovers the information modulated onto the RF carrier and provides that information to the receive (RX) processor 356. The TX processor 368 and RX processor 356 implement Layer 1 functionality associated with various signal processing functions. The RX processor 356 can perform spatial processing on the information to recover any spatial stream destined for UE 350. If multiple spatial streams are destined for UE 350, the RX processor 356 can combine them into a single OFDM symbol stream. The RX processor 356 then uses a Fast Fourier Transform (FFT) to transform the OFDM symbol stream from the time domain to the frequency domain. The frequency domain signal consists of a separate OFDM symbol stream for each subcarrier of the OFDM signal. The symbols on each subcarrier, along with the reference signal, are recovered and demodulated by determining the most probable signal constellation points transmitted by base station 310. These soft decisions can be based on a channel estimate calculated by channel estimator 358. The soft decision is then decoded and deinterleaved to recover the data and control signals originally transmitted by base station 310 on the physical channel. The data and control signals are then provided to controller / processor 359, which implements layer 3 and layer 2 functionality.

[0074] The controller / processor 359 may be associated with at least one memory 360 storing program code and data. The at least one memory 360 may be referred to as a computer-readable medium. In the UL, the controller / processor 359 provides demultiplexing, packet reassembly, decryption, header decompression, and control signal processing between transport and logical channels to recover IP packets. The controller / processor 359 is also responsible for error detection using ACK and / or NACK protocols to support HARQ operation.

[0075] Similar to the functionality described in conjunction with DL transmission performed by base station 310, controller / processor 359 provides RRC layer functionality associated with system information (e.g., MIB, SIB) acquisition, RRC connectivity, and measurement reporting; PDCP layer functionality associated with header compression / decompression and security (encryption, decryption, integrity protection, integrity verification); RLC layer functionality associated with upper-layer PDU delivery, error correction via ARQ, concatenation, segmentation, and reassembly of RLC SDUs, resegmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functionality associated with mapping between logical channels and transport channels, multiplexing of MAC SDUs onto TBs, demultiplexing of MAC SDUs from TBs, scheduling information reporting, error correction via HARQ, priority handling, and logical channel priority ordering.

[0076] The TX processor 368 can use the reference signal transmitted from the base station 310 or the channel estimate derived from feedback by the channel estimator 358 to select an appropriate decoding and modulation scheme and facilitate spatial processing. The spatial stream generated by the TX processor 368 can be provided to different antennas 352 via individual transmitters 354Tx. Each transmitter 354Tx can use the corresponding spatial stream to modulate an RF carrier for transmission.

[0077] UL transmission is processed at base station 310 in a manner similar to that described in conjunction with the receiver function at UE 350. Each receiver 318Rx receives signals via its corresponding antenna 320. Each receiver 318Rx recovers the information modulated onto the RF carrier and provides that information to RX processor 370.

[0078] The controller / processor 375 may be associated with at least one memory 376 storing program code and data. The at least one memory 376 may be referred to as a computer-readable medium. In the UL, the controller / processor 375 provides demultiplexing, packet reassembly, decryption, header decompression, and control signal processing to recover IP packets between transport and logical channels. The controller / processor 375 is also responsible for error detection using ACK and / or NACK protocols to support HARQ operation.

[0079] At least one of the TX processor 368, RX processor 356, and controller / processor 359 can be configured to perform coupling. Figure 1 The various aspects of the perception data collection component 198.

[0080] At least one of the TX processor 316, RX processor 370, and controller / processor 375 can be configured to perform coupling. Figure 1 The various aspects of the trajectory planning components 199.

[0081] In recent years, vehicle manufacturers have developed vehicles with assisted driving and / or autonomous driving capabilities. Assisted driving (also known as Advanced Driver Assistance Systems (ADAS)) refers to a set of technologies designed to enhance vehicle safety and improve the driving experience by providing assistance and automation to the driver. These technologies use various sensors (such as cameras, radar, light detection and ranging (LiDAR or LiDAR sensors)) and other components to monitor the vehicle's surroundings and assist the driver in performing certain driving tasks. For example, some features of driver assistance systems may include: (1) Adaptive Cruise Control (ACC) (e.g., a system that automatically adjusts the speed of a vehicle to maintain a safe following distance from the vehicle in front), (2) Lane Keeping Assist (LKA) (e.g., a system that uses cameras to detect lane markings and help keep the vehicle centered within the lane and provides steering input to prevent unintended lane departure), (3) Autonomous Emergency Braking (AEB) (e.g., a system that detects a potential collision with an obstacle or pedestrian and automatically applies braking to avoid or mitigate the impact), (4) Blind Spot Monitoring (BSM) (e.g., a system that uses sensors to detect vehicles in the driver's blind spots and provides visual or audible alerts to avoid potential collisions during lane changes), (5) Parking Assist (e.g., a system that assists drivers in parking their vehicles by using cameras and sensors to help with parallel parking or maneuvering into tight spaces), and / or traffic sign recognition (e.g., cameras and image processing used to recognize and display traffic signs, such as speed limits, stop signs, and other road regulations on the vehicle's dashboard).

[0082] Autonomous driving (also known as self-driving or driverless technology) refers to the ability of a vehicle to navigate and operate itself without designated human intervention (e.g., moving from one location to another without a person controlling the vehicle). The goal of autonomous driving is to create vehicles capable of perceiving their surroundings, making decisions, and controlling their movement without the direct involvement of a human driver. To achieve or improve autonomous driving, vehicles may be required to use maps (or map data) with detailed information, such as high-definition (HD) maps. HD maps can refer to highly detailed and accurate digital maps designed for autonomous driving and ADAS. In one example, an HD map may typically include one or more of the following: (1) geometric information (e.g., precise road geometry, including lane boundaries, curvature, slope, and a detailed 3D model of the surrounding environment), (2) lane-level information (e.g., information about individual lanes on a road, such as lane width, lane type (e.g., driving, turning, or parking lanes), and lane connectivity), (3) road attributes (e.g., data about road features such as traffic signs, signals, traffic lights, speed limits, and road markings), (4) topology (e.g., information about the relationships between different roads, intersections, and connectivity patterns), (5) static objects (e.g., the location and details of fixed objects along the road, such as buildings, traffic barriers, and poles), (6) dynamic objects (e.g., real-time or frequently updated data about moving objects, such as other vehicles, pedestrians, and cyclists), and / or (7) location determination and positioning: precise reference points and landmarks that help determine the accurate location of vehicles on the map.

[0083] To enable vehicles to provide assisted driving and / or autonomous driving, vehicles can be configured to use various machine learning (ML) and / or neural network (NN) frameworks. ML / NN frameworks can refer to a collection of tools, libraries, and / or software components configured to provide a structured approach for designing, building, and deploying ML / NN models and applications. These frameworks can simplify the development of ML / NN algorithms and applications by providing a foundation of pre-built functions, algorithms, and utilities. These frameworks typically include features for data preprocessing, model training, evaluation, and / or deployment. ML / NN frameworks can appear in various programming languages, and these ML / NN frameworks can be configured to suit different types of machine learning tasks, including supervised learning, unsupervised learning, and / or reinforcement learning. ML / NN models can refer to mathematical representations of real-world processes or problems created using ML / NN algorithms and techniques. These ML / NN models can be configured to make predictions, classify data, and / or solve specific tasks based on patterns and relationships learned from input data. Deep learning frameworks can refer to specialized software libraries or toolsets that provide specified components and abstractions for building, training, and deploying deep neural networks. Deep learning frameworks are designed to facilitate the development of complex neural network models, especially deep neural networks with multiple layers. These frameworks offer a wide range of pre-implemented layers, optimizers, loss functions, and other components, making it easier for researchers and developers to utilize deep learning models.

[0084] Figure 4 Figure 400 illustrates examples of road object detection using ML / NN models according to various aspects of this disclosure. In some embodiments, ADAS or autonomous driving systems may be configured to perform object detection using one or more ML / NN models. For example, as shown at 402, a first ML / NN model (ML / NN model 1) may be trained / used to detect and track polylines based on sensor outputs (e.g., images captured by a vehicle's camera, point clouds generated from radar / LiDAR, etc.), while, as shown at 404, a second ML / NN model (ML / NN model 2) may be trained / used to detect and track objects in three-dimensional (3D) space (e.g., to perform a 3D object detection (3DOD) task). The ADAS or autonomous driving system may then process and use the outputs of both ML / NN models (e.g., for assisted / autonomous driving). In some embodiments, the ML / NN models may also be configured to perform multiple types of object detection (e.g., performing both polyline detection and 3D object detection). A point cloud may refer to a discrete set of data points in space, where these points may represent 3D shapes or objects. In some implementations, each point location can be associated with a set of Cartesian coordinates (X, Y, Z). Point clouds can be generated by radar / LiDAR by detecting multiple points on the external surface of an object.

[0085] For the purposes of this disclosure, “perception data” or “autonomous perception data” can refer to information acquired by the sensors and systems of a vehicle to understand and interpret the vehicle’s surroundings (e.g., for the purpose of providing assisted / autonomous driving). For example, an autonomous vehicle may be configured to rely on a variety of sensors to perceive its environment and make informed decisions. These sensors may typically include one or more of the following: (1) lidar / lidar sensors that use laser beams to measure distances and create detailed 3D maps of the environment; (2) radar that uses radio waves to detect the presence, distance, and speed of objects around the vehicle; (3) cameras that capture visual data, allowing the vehicle to identify and recognize objects, road signs, lane markings, and other important visual cues; (4) ultrasonic sensors that use sound waves to detect objects near the vehicle; (5) Global Navigation Satellite Systems (GNSS) that provide information about the vehicle’s position, speed, and heading, thus contributing to overall situational awareness; and / or (6) inertial measurement units (IMUs) that measure the vehicle’s acceleration and angular rate, thus helping to determine the vehicle’s position and orientation. These sensors may collectively acquire data about the vehicle’s surroundings and create a comprehensive perception system. Then, a set of software algorithms can be implemented to analyze and interpret the perceived data to make decisions, such as navigating vehicles, avoiding obstacles, obeying traffic rules, and ensuring overall safety. Perceived data can be a crucial component of the sensor fusion process, where information from different sensors is combined to create a more accurate and reliable representation of the environment for autonomous vehicles.

[0086] In some implementations, autonomous sensing data can specify a large volume of multi-sensor sensing data to be collected across multiple countries, conditions, and / or scenarios (e.g., sensing data obtained from multiple / various sensors). Therefore, logistical and energy constraints can be factors to consider during data collection design. For example, a supplier may expect to reduce mileage / miles and energy footprint for its overall system (e.g., the data collection system) and design the system efficiently. The data collection planner can specify a convenient system for planning and tracking the status of the data collector (e.g., the vehicle collecting the data), including data coverage and road objects collected. Data collection can also be specified to adapt to changes based on data selection steps and provide new planning tasks to collect data in new scenarios, geographic areas, and / or sensing tasks. In some examples, data collection can be configured as an open-loop system where the collected data may not meet customer specifications, data diversity, and / or annotation specifications, and may not be able to build a good dataset to train sensing (ML / NN) models (e.g., 3DOD, polylines, lanes, segmentation, etc.).

[0087] After the perception data is collected, it can be processed, a process known as "data processing." Data processing is the systematic collection, organization, validation, and management of data to ensure its quality, relevance, and suitability for training and deploying ML / NN models. Data processing may involve preparing and maintaining high-quality datasets that are critical (e.g., important) to the success of ML / NN projects. The goal of data processing can be to enhance the overall performance, accuracy, and reliability of ML / NN models by providing them with clean and representative datasets. Therefore, data processing can be a crucial intermediate step that provides feedback on data collection and measures the feasibility of using samples for dataset creation. For example, data processing can enable / allow the generation of optimal datasets (e.g., regarding data diversity, reduction of similar samples, reduction of the amount of data used for annotation, and selection of more relevant samples based on purpose / task to achieve optimality).

[0088] In typical data collection and curating implementations, data collection and curating (or dataset design) can be configured as separate steps, often resulting in redundant and undiversified data. Furthermore, storing sensing data collected from a large number of vehicles can be expensive, and a universal approach / process may not exist that simultaneously satisfies both evolving dataset specifications and cost constraints for the evolution of data collection and curating. Even after a sensing system is deployed in vehicles, the sensing data can be specified (e.g., by the vehicles) for continuous mining to improve sensing. For example, a typical sensing data collection system might specify a large amount of multimodal (camera / LiDAR) data to be collected across multiple vehicles (e.g., K different vehicles), which the sensing data collection system then uses subsequent ingestion processes and relies on cumbersome data curating processes in later steps to reduce the dataset to achieve / meet certain objectives.

[0089] The aspects presented herein can improve the efficiency of perception data collection and organization, thereby improving the overall performance of ML / NN model training based on perception data (e.g., for autonomous / assisted driving systems). The aspects presented herein can provide a mechanism that can effectively filter samples (e.g., collected perception data of a specific type) to reduce data storage costs, increase data diversity (e.g., geographically, in terms of context), and / or avoid redundant samples (e.g., similar samples collected at different times (during day / night), in different locations (urban / highway / rural), and / or under different environmental conditions (weather and lighting conditions), etc.), which can benefit perception data collection and organization. For example, in one aspect of this disclosure, a network entity (e.g., a data collection server, a centralized server / storage unit, etc.) can have the ability to configure a set of UEs (e.g., a set of vehicles, a set of data collectors, etc.) to optimally select samples (e.g., perception data, embeddings, etc.) online from perception data locally collected by the set of UEs to meet global collection criteria (e.g., dataset / task specifications, cost specifications, storage, vehicle deployment, logistical costs, etc.). The aspects presented in this paper also provide a multi-sensor perception data compiling pipeline, which includes a multi-vehicle collection setup with large data bandwidth resulting from intensive data collection due to large sensor suites and cross-country iterations, a vehicle-to-cloud distributed data compiling pipeline, and collaborative online file selection for vehicle pools.

[0090] Figure 5 Figure 500 illustrates examples of data collection and compilation according to various aspects of this disclosure. As shown at 502, a server (e.g., a sensing data collection server, cloud storage, etc.) may be configured to receive sensing data collected by a collection of vehicles (which may be referred to as a vehicle pool). For example, each vehicle in the vehicle pool may receive information in a heuristic, manual manner (e.g., as shown at 506) to determine where to collect sensing data and what kind of sensing data to collect (e.g., what type of sensing data to collect). Consequently, the yield or availability of the collected sensing data may decrease or be lower.

[0091] Then, as shown at 504, after the server receives perception data from the vehicle pool, a significant amount of data can be stored in each perception data collection step, which can specify a substantial amount of storage space and cost. Then, if some perception data is found to be redundant or does not contain the correct / appropriate metadata attributes, image features, and / or driving scenarios, this perception data can be removed / filtered during a manual filtering step. As shown at 506, after filtering the perception data, the server can (manually) determine whether (e.g., for ML / NN training) to specify additional perception data, and the server can instruct another set of vehicles (which may include the same / different vehicles as the previous vehicle pool) where to collect additional perception data and what additional perception data to collect. The server can repeat the steps described in combination of 502, 504, and 506 until sufficient perception data has been collected.

[0092] As shown at 508, after collecting sufficient perceptual data, the collected perceptual data can be sent and stored in a perceptual dataset (e.g., a database). A server or annotation entity can then be configured / requested to annotate the perceptual dataset to obtain a set of annotated data. In other words, the collected perceptual data can be annotated for ML / NN training purposes.

[0093] Figure 6 Figure 600 illustrates an example architecture of a centralized cloud storage unit that, according to various aspects of this disclosure, can effectively filter out samples to be collected in order to reduce data storage costs, increase data diversity, and / or reduce redundant samples. At a high level, as shown at 630, servers, clouds, or centralized storage units (hereinafter collectively referred to as network entities 602) may be configured to collect sensed data (which may be referred to as “samples”) from a set of vehicles 604 (in a pool of vehicles) based on a time-evolving set of specifications (such as tasks / objectives, costs, available storage, energy constraints / specifications, etc., based on sensed data collection).

[0094] Network entity 602 may include various functions, modules, and / or entities (collectively referred to as “storages”) that can be used to assist network entity 602 in planning routes to be traveled by a collection of vehicles 604 (e.g., for sensing data collection purposes) and / or configuring the collection of vehicles 604 regarding what types of sensing data to generate, collect, and / or upload. For example, in one aspect of this disclosure, network entity 602 may include at least attribute storage 606, track storage 608, embedded storage 610, and map storage 612 (e.g., an HD map database), wherein outputs from these storages and databases may be used by online task path planner 614 to plan / optimize sensing data to be collected (e.g., by the collection of vehicles 604). Online task path planner 614 may also include, or be associated with, embedded centroid allocator 616 and / or online data distribution allocator 618.

[0095] In one example, attribute storage 606 may be configured to include frequency of attributes (e.g., a histogram / chart using rectangles to show the frequency of numerical data), co-occurrence frequency (e.g., a combination of attributes and annotation objects), and / or priority regarding attributes that attract the most attention (e.g., attributes with low frequency). For the purposes of this disclosure and / or in the context of perceptual data collection, an attribute may refer to a characteristic or property of an object, event, or phenomenon that is observed, measured, or recorded. An attribute may be a specific piece of information or variable that network entity 602 (e.g., or a researcher / data collector) is interested in collecting and analyzing. For example, if the set of vehicles 604 is configured to collect perception information while these vehicles are moving on the road (e.g., while these vehicles are being driven), attributes may include road conditions, road topology (e.g., intersections, tunnels, roundabouts, junctions, etc.), road metadata (e.g., hilly / winding roads), designated / special accident-prone areas (e.g., toll plazas, railway crossings, etc.), the country / region where the set of vehicles 604 is located, time of day (e.g., daytime, nighttime, etc.), environmental conditions (e.g., weather, lighting conditions, pollution, etc.), and / or traffic information (e.g., traffic density, road construction, etc.). In some implementations, attributes may also include dynamic objects (e.g., vehicles, cyclists, pedestrians, etc.) and traffic scenarios (e.g., overtaking scenarios, lane-changing scenarios, etc.).

[0096] The track storage 608 can be configured to track and store different coordinates (e.g., GPS / GNSS coordinates) of the tracks traveled by vehicles in the collection of vehicles 604, and maintain these coordinates of the tracks on a map representation. For the purposes of this disclosure and / or in the context of driving path planning, a track can refer to the path followed by a vehicle over time. A track can be a series of states (position and speed) describing the movement of a vehicle from its starting point to its destination. Track planning can be an important aspect of autonomous driving systems because it involves determining safe and efficient paths for vehicles to navigate through their environment. Tracks can also include locations across different countries and / or (e.g., within a country) different regions.

[0097] Embedding memory 610 can be configured to employ a convolutional neural network (CNN) backbone (such as a deep neural network (DNN) or a transformer-based DNN) to compute frame- and object-level feature-sensitive embeddings for the collected images and point clouds. The computed embeddings can then be clustered using a density-based noisy spatial clustering application (DBSCAN), and one or more centroids of the clusters can be evaluated and then assigned to different vehicles in the collection of vehicles 604 (e.g., via embedding centroid assigner 616). For the purposes of this disclosure, "embedding" can refer to a representation of data in a lower-dimensional space. This mechanism is commonly used in machine learning and computer vision to transform complex data such as images or sensor readings (e.g., point clouds) into a more compact and meaningful format. Centroid can refer to the center or average point of a cluster of data points in feature space, and can be used interchangeably with "embedding centroid". For example, in machine learning, data points can represent individual instances or examples in a dataset, where each data point can have multiple features or attributes. A feature space can be a multidimensional space defined by the features of data points, where each axis corresponds to a specific feature. A cluster can refer to a group of data points in the feature space that share similarities or are close to each other. Clustering algorithms (e.g., DBSCAN) typically aim to group data points into clusters based on certain criteria. The centroid of a cluster can refer to the point in the feature space that serves as the representative or central location of the cluster. This centroid can be calculated as the average of the feature values ​​of all data points in the cluster. In the context of machine learning or perceptual data collection / compilation, a feature can refer to a single measurable property or characteristic of the observed phenomenon. Features can be the inputs used by machine learning models to make predictions or perform tasks. These features can represent different dimensions or aspects of the data used by machine learning models to learn patterns and relationships.

[0098] Map storage 612 may be an HD map database or a database that includes HD maps (and non-HD maps). As discussed above, HD maps may contain high-precision and detailed semantic attributes such as: road conditions, road topology (e.g., intersections, tunnels, roundabouts, junctions, etc.), road metadata (e.g., hilly / winding roads), designated / special accident-prone points (e.g., toll plazas, railway crossings, etc.), lane types, intersection topology, map objects (e.g., traffic signs, lights), and / or road / structure information (e.g., tunnels, merging, diverging, etc.). These attributes may not be spatially identifiable in open street maps and / or navigation maps. Depending on the implementation, the output of map storage 612 may provide map database access to an online task route planner 614 (or to a dynamic route planning module associated with the online task route planner 614) so ​​that the dynamic route planning module can search the map database to perform route planning based on information obtained from static (HD) map information (e.g., as discussed below in conjunction with 808).

[0099] An online task route planner 614 (e.g., a cloud-based online route planner) can be configured to take dataset specifications, existing track data, and / or cost constraints as input to optimally compute routes (tracks) for the entire pool of vehicles (e.g., for the collection of vehicles 604). For example, given a desired / defined attribute distribution, the online task route planner 614 can use HD maps and other external sources (e.g., online traffic information / predictions) to plan new tracks for each vehicle in the collection of vehicles 604 (or the pool of vehicles) for data collection. Furthermore, the online task route planner 614 can take embedding centroids from embedding storage 610 as input and then assign these embedding centroids to each vehicle in the collection of vehicles 604 to diversify the collection of samples across geographic locations and joint attribute embedding spaces.

[0100] Figure 7 Figure 700 illustrates an example embedded centroid allocator according to various aspects of this disclosure. In one example, as shown at 702, an embedded centroid allocator 616 associated with an online task path planner 614 may be configured to receive M centroids representing the distribution of features of the perceived data. Then, as shown at 704, the embedded centroid allocator 616 may output subsets of different embedded centroids to K different vehicles (e.g., a set of vehicles 604). As shown at 706, visualization of the image and its embedded features may be shown by a two-dimensional (2D) drawing, where clusters may be evaluated to extract images that are naturally similar to each other, such as forest images, bridge images, snow scene images, rain scene images, city images, highway images, etc.

[0101] Return to reference Figure 6 The online data distribution allocator 618 associated with the online task path planner 614 can be configured to calculate the desired distribution to be collected by evaluating the difference between the specifications in the attribute store 606 and the existing state.

[0102] Figure 8 Figure 800 is an example illustrating an online task path planner according to various aspects of this disclosure. (In conjunction with...) Figure 6 The online task path planner 614 discussed can be configured to provide dynamic path planning on the cloud for multiple vehicles (e.g., a collection of vehicles 604) to collect perception data specified by network entity 602. For example, for K vehicles distributed with different environmental parameters (e.g., as combined...) Figure 7 (As discussed), the online task path planner 614 can assign routes to these vehicles based on the expected attribute distribution.

[0103] For example, as shown at 802, the online data distribution allocator 618 associated with the online task path planner 614 may receive a set of target distributions (e.g., a specification set of sensing data to be collected) and a set of current distributions (e.g., currently available sensing data) from attribute store 606. Then, based on analysis / comparison of the target distribution set and the current distribution set, the online data distribution allocator 618 may generate a set of desired distributions that may include ratios and / or weighted attributes. In other words, the online data distribution allocator 618 may use the current and target distributions to select the attributes and ratios that the network entity 602 configures / expects to collect with the highest priority. Then, as shown at 804, the set of desired distributions may be added to a set of constraints that serve as input to the online task path planner 614. Constraints may include desired attribute ratios, the number of vehicles and their starting locations, and / or predicted traffic density, predicted weather (e.g., for driving time), etc.

[0104] As shown at 808, based on the output from track storage 608 (e.g., a set of tracks traveled by a set of vehicles), constraints, and map storage 612 (e.g., a set of HD map data outputs), the online task path planner 614 can plan / propose tracks (e.g., routes) for each of the K vehicles, such that each of the K vehicles can attempt to prioritize using the planned / proposed tracks and collect perception data while traveling along these planned / proposed tracks. For example, the online task path planner 614 can sample the set of candidate tracks and select the track that best satisfies the constraints (e.g., using attributes generated from static perception data from the HD map), and / or select tracks that geographically do not intersect with tracks already collected from the track storage 608.

[0105] As shown at 810, since the planned / proposed route may have a new attribute distribution, the online data distribution allocator 618 can use this new attribute distribution to update the desired distribution for the next set of route planning for the vehicle. In other words, after the planned / proposed route, network entity 602 or online data distribution allocator 618 can recalculate the desired ratio using the updated distribution (e.g., similar to a closed-loop system where the desired distribution can continue to change based on the data distribution regarding the planned / proposed route).

[0106] Return to reference Figure 6 As shown at 640, based on the planned / proposed trajectories from network entity 602 (e.g., online task path planner 614 from network entity 602), the collection of vehicles 604 can collect sensing data while these vehicles are on the planned / proposed trajectories (and during a specified time period, if configured).

[0107] Figure 9A , Figure 9B and Figure 9C Figures 900A, 900B, and 900C illustrate examples of vehicles collecting sensing data based on planned / proposed tracks according to various aspects of this disclosure. Figure 9A As shown in Figure 900A, network entity 602 can be a vehicle 904 (e.g., combined with...). Figure 6 The vehicles within the described set of vehicles 604 provide a set of planned / proposed tracks to collect sensing data, which may include planned / proposed track 906. For example, network entity 602 may lack sensing data for areas approaching planned / proposed track 906 and the set of planned / proposed tracks.

[0108] like Figure 9BAs shown in Figure 900B, if vehicle 904 is requested / configured (e.g., based on user input) to travel to a destination, vehicle 904 (or its autonomous / assisted driving system or navigation system) may attempt to / plan a navigation route 908 using a planned / proposed track 906 / overlapping with the planned / proposed track (e.g., for the purpose of perception data collection). Vehicle 904 may then collect perception data while traveling on the planned / proposed track 906 and transmit the collected perception data to network entity 602 (e.g., as...). Figure 6 (As shown at 622). In some examples, the vehicle 904 may also include indications of the tracks (such as planned / proposed tracks 906) in which sensing data is collected (e.g., also as shown at 622). Figure 6 (As shown at point 624). In some scenarios, navigation route 908 may not be the best / most efficient / fastest route to the destination. Therefore, once vehicle 904 has collected perception data from the planned / proposed track 906, vehicle 904 can resume using the best / most efficient / fastest route to reach the destination for future navigation.

[0109] like Figure 9C As shown in Figure 900C, in some scenarios, vehicle 904 may not be able to travel to its destination along navigation route 908 (all portions thereof). For example, the driver of vehicle 904 may intervene and decide to take an alternative route, or a portion of the navigation route may be unavailable (e.g., due to road construction or closure). As shown at 910, in such scenarios, vehicle 904 can still collect perception data while traveling along a portion of navigation route 908. Vehicle 904 can then upload the collected perception data to network entity 602 (e.g., as shown in Figure 900C). Figure 6 (as shown at 622), and also includes indications of the trajectories in which sensing data is collected (e.g., as shown at 622), Figure 6 (As shown at position 624).

[0110] In some implementations, the set of planned / proposed tracks for vehicle 904 can also be configured to be associated with at least one other factor / condition (such as time of day, weather, lighting conditions, and / or traffic conditions). For example, network entity 602 can specify that vehicle 904 collects sensing data while traveling on the set of planned / proposed tracks, and also during specified time periods (e.g., 10:00 AM to 1:00 PM), during specified weather periods (e.g., during rainy days), during specified traffic conditions (e.g., the number of detected vehicles exceeds a threshold), and / or during specified lighting conditions (e.g., during dim / low-light conditions). Therefore, if the associated factor / condition is not met, vehicle 904 can be configured to skip collecting sensing data while traveling on the set of planned / proposed tracks.

[0111] Return to reference Figure 6 As shown at 626 (in 640), an online file selector 620 (e.g., configured by network entity 602) can be configured for each vehicle in the set of vehicles 604 (e.g., vehicle 904). The online file selector 620 enables the set of vehicles 604 to determine / select which sensed data from the collected sensed data will be uploaded to network entity 602. This significantly reduces the amount of sensed data to be stored at network entity 602, as well as the signaling overhead between the set of vehicles 604 and network entity 602.

[0112] Figure 10 Figure 1000 illustrates examples of online file selectors according to various aspects of this disclosure. In one example, as shown at 1010, vehicle 1004 (e.g., from a set of vehicles 604) may include a set of embedding models, which may be configured by network entity 602 such that the same set of embedding models is used by both the set of vehicles 604 and network entity 602. Based on images and / or point clouds collected by sensors of vehicle 1004, vehicle 1004 may use the set of embedding models to compute / generate a set of samples (e.g., embeddings). For example, vehicle 1004 may use embedding models (e.g., a transformer-based DNN) to compute / generate a first sample / embedding (sample 1) and a second sample / embedding (sample 2), and output / provide these samples (e.g., embeddings) to online file selector 620.

[0113] In one aspect, as shown at 1012, the online file selector 620 may include an embedding distance sampler 1006 configured to perform distance-based file sampling for embeddings (e.g., between newly collected embeddings and embedding centroids) using an embedding centroid assigner 616. For example, as shown at 1014, the goal of the embedding distance sampler 1006 may be to find the most relevant / interesting samples or perceptual data to retain. For instance, if a sample (e.g., sample 1) is located far from a target cluster, this may indicate that the sample is scarcer / rare compared to samples in or near the target embedding centroid (e.g., sample 2) and may be worth retaining (e.g., to help increase dataset diversity). In other words, if the distance between a sample / data and the target cluster is greater than a distance threshold, the sample / data is suitable / interesting to upload (e.g., as shown at 1016). This mechanism can reduce the amount of sample / perceptual data uploaded by vehicle 1004 to network entity 602.

[0114] In some specific implementations, as shown at 1018, the vehicle 1004 may also include a set of perception models that can be used to perform various detections, such as polyline detection, 3D object detection (3DOD), etc. For example, a convolutional neural network (CNN) may be used by the vehicle 1004 to generate the output of the online file selector 620 on new sample images / point clouds to predict, for example, 3D objects, polylines, pedestrians, etc. As shown at 1012, in some examples, the online file selector 620 may include an active learning sampler 1008 configured to use an active learning loop to compute model predictions (e.g., 3DOD / polyline output, etc.) and to use uncertainty to evaluate whether samples (e.g., first sample, second sample, etc.) are informative and whether they should be uploaded (or suitable for uploading), as shown at 1016.

[0115] Return to reference Figure 6 As shown at 650, after collecting sufficient perceptual data, the collected perceptual data can be sent and stored in a perceptual dataset (e.g., a database). Then, network entity 602 or the annotation entity can be configured to / request annotation of the perceptual dataset to obtain a set of annotated data. In other words, the collected perceptual data can be annotated for ML / NN training purposes.

[0116] Figure 11This is an example of a communication flow 1100 illustrating a set of planned / designed trajectories configured by a network entity (e.g., a centralized cloud storage) to a UE (e.g., a vehicle) for the purpose of collecting perception data. The numbers associated with the communication flow 1100 do not specify a particular time sequence and are used only as a reference to the communication flow 1100.

[0117] At 1110, network entity 1102 (e.g., network entity 602, server, cloud, or centralized storage, etc.) can receive a set of perception data collected by the first set of UEs 1104 (e.g., a first set of vehicles, a first set of autonomous driving systems, etc.), such as combined with Figure 5 , Figure 6 and Figure 10 As described. The first set of UEs 1104 may belong to the UE pool (e.g., combined with...). Figure 6 The pool of vehicles or the set of vehicles 604 discussed. The perception data may include embeddings computed by the first set of UEs 1104 based on images and point clouds captured by these UEs (e.g., using embedding models such as CNNs, DNNs, transformer-based DNNs, etc.) and / or outputs from a set of prediction models (e.g., polyline detection, 3DOD, etc.). Furthermore, the perception data may also include a set of centroids representing the distribution of features of the perception data.

[0118] At position 1112, if combined Figure 6 and Figure 8 As described, network entity 1102 can analyze currently available sensing data / distributions (including sensing data received at 1110 from a first set of UEs 1104) and compare this currently available sensing data / distribution with a specification set (e.g., a set of target distributions / tasks, etc.) to determine specified additional sensing data (e.g., a set of desired distributions to be collected). Then, based on the specified additional sensing data, network entity 1102 can (e.g., by sending configuration to a second set of UEs 1106) configure a set of embedded centroids and / or a set of planned routes and time schedules for collecting the additional sensing data for the second set of UEs 1106. The second set of UEs 1106 may belong to the same UE pool (e.g., a transportation pool) as the first set of UEs 1104, and there may be overlapping and / or non-overlapping UEs between the first set of UEs 1104 and the second set of UEs 1106.

[0119] At 1114, the second set 1106 of the UE can collect sensing data based on a configuration (e.g., a configured set based on additional embedded centroids and / or planned routes and time schedules for collecting additional sensing data), such as combining... Figure 6 , Figure 9A , Figure 9B , Figure 9C and Figure 10 As described. Then, at 1116, the second set 1106 of the UE can send the sensed data collected based on the configuration received at 1112 to the network entity 1102.

[0120] In some specific implementations, at 1118, network entity 1102 may also configure a list of sensing data to be filtered for a second set 1106 of UEs (e.g., based on sensing data received from the first set 1104 of UEs). Note that although the diagram shows this configuration differently from the configuration received at 1112, it is for illustrative purposes only. The configuration associated with the list of sensing data to be filtered may be included in the same configuration as the set of additional embedded centroids and / or route and time plans used to collect additional sensing data, or it may be a different and separate configuration. Then, based on the configuration associated with the list of sensing data to be filtered, the second set 1106 of UEs may filter the collected sensing data, such as combining... Figure 10 As described. For example, the second set 1106 of the UE may use an online file selector (or an embedded distance sampler) to select samples closer to the target cluster for uploading, and / or use an online file selector (or an active learning sampler) to select perceptual model outputs with informational value for uploading (e.g., by filtering out outputs that have no informational value).

[0121] In some specific implementations, at 1120, the second set 1106 of UEs may also send to network entity 1102 an indication of a portion of a configured set of planned routes and time plans for travel by the second set 1106 of UEs (e.g., each UE in the second set 1106 of UEs), such as combining Figure 9A , Figure 9B and Figure 9C As described. Note that this instruction may be included together with the sensory data collected at 1116.

[0122] As shown at 1122, network entity 1102 may repeat one or more of the above steps (e.g., in conjunction with the steps described at 1110 to 1120) until sufficient sensed data is collected or the specified task is met. Therefore, network entity 1102 may continue to update and modify planned / proposed routes for different UEs (e.g., vehicles) based on time-evolving specifications. For example, network entity 1102 may configure embedded centroids and / or planned routes and time plans for another set of UEs in the UE pool (e.g., which may have UEs overlapping / non-overlapping with the first set of UEs 1104 and / or the second set of UEs 1106) based on the sensed data received from the second set of UEs at 1116, and / or for collecting additional sensed data. Similarly, network entity 1102 may receive additional sensed data from the third set of UEs based on the configured set of planned routes and time plans and / or the configured set of embedded centroids.

[0123] Such as combination Figure 6 As discussed in section 650, network entity 1102 may also output perceived data collected from a first set 1104 of UEs and / or a second set 1106 of UEs to a perceived dataset, and receive a set of annotation data based on this output. In some examples, annotation may be performed by network entity 1102 (or an entity associated with network entity 1102). In some examples, network entity 1102 may be performed by an external entity (e.g., an annotation body (company)).

[0124] The aspects presented in this paper can improve the efficiency of perception data collection and organization, thereby improving the overall performance of ML / NN model training based on perception data (e.g., for autonomous / assisted driving systems). Autonomous perception data can specify large amounts of multi-sensor perception data to be collected across multiple countries, conditions, and scenarios. This can involve data collection and data organization processes. To date, data collection and data organization (dataset design) are separate steps, and therefore frequently generate redundant and unsustainable data. The aspects presented in this paper provide a best-in-class cloud-based solution for collecting and optimizing data for route planning. Data collected globally from a pool of vehicles is uploaded to a centralized cloud-based entity for processing and optimization. The aspects presented in this paper include the following features: online file selector, embedded storage, track storage, attribute storage, HD map, online distribution allocator, embedded centroid allocator, online route planner, etc.

[0125] This paper presents a general approach that allows for the simultaneous satisfaction of evolving perception specifications and task-specific specifications, as well as logistical costs, in an online setting, and leverages the cloud to enhance data collection and selection across multiple vehicles. An online path planner plans optimal routes for multiple vehicles to meet desired data distribution and geographic separation constraints. An online file selector picks the most diverse and informative multi-model perception data for training CNN / DNN models for various autonomous driving perception tasks.

[0126] Figure 12 This is a flowchart 1200 of a wireless communication method. This method can be performed by network entities (e.g., one or more location servers 168; base station 102; network entities 602, 1102, 1460). This method enables network entities to configure a pool of UEs to collect perceived data for training machine learning models based on a set of specifications (e.g., dataset / task specifications, cost specifications, storage, vehicle deployment, logistics costs, etc.), thereby reducing data storage costs, increasing data diversity, and / or reducing redundant data.

[0127] At point 1202, the network entity can receive a first set of sensed data collected by the first set of UEs, such as combined... Figures 6 to 11 As described. For example, as in combination Figure 11 As described in 1110, network entity 1102 can receive sensing data collected by the first set of UEs 1104. The reception of the first set of sensing data can be achieved by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0128] At 1204, the network entity may configure at least one of the following for a second set of UEs based on a first set of sensed data: a set of embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11As described in 1112, network entity 1102 can analyze currently available sensing data / distributions (including sensing data received from the first set 1104 of UEs at 1110) and compare this currently available sensing data / distribution with a specification set (e.g., a set of target distributions / tasks, etc.) to determine specified additional sensing data (e.g., a set of desired distributions to be collected). Then, based on the specified additional sensing data, network entity 1102 can (e.g., by sending configuration to the second set 1106 of UEs) configure for the second set 1106 of UEs a set of embedded centroids and / or a set of planned routes and time schedules for collecting the additional sensing data. The configuration can be provided by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0129] In one example, to configure a set of planned routes and time plans for collecting a second set of sensed data, a network entity may be configured to evaluate the difference between a first set of sensed data and an existing set of sensed data in the sensed dataset; determine the set of planned routes and time plans for collecting the second set of sensed data based on the evaluated difference; and send the set of planned routes and time plans to the second set of UEs.

[0130] In another example, in order to configure a set of planned routes and time plans for collecting a second set of sensing data, a network entity may be configured to obtain information relating to at least one of the following: data specified for the sensing dataset, existing trajectory data, or cost constraints; determine, based on the information, a set of planned routes and time plans for collecting the second set of sensing data; and send the set of planned routes and time plans to the second set of UEs.

[0131] In another example, in order to configure a set of planned routes and time plans for collecting a second set of sensed data, a network entity may be configured to obtain an indication of the attribute distribution; determine the set of planned routes and time plans for collecting the second set of sensed data based on the indication, map data, and traffic information; and send the set of planned routes and time plans to the second set of UEs.

[0132] At 1208, the network entity can receive a second set of sensed data from the second set of the UE based on at least one of a configured set of planned routes and time plans or a configured set of embedded centroids, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11As described in 1116, network entity 1102 can receive sensing data collected based on the configuration sent at 1112 from the second set of UEs 1106. The reception of the second set of sensing data can be achieved by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0133] In one example, in order to receive a second set of sensed data from a second set of UEs based on a configured set of planned routes and time plans, a network entity may be configured to receive from each UE in the second set of UEs an indication of a portion of the configured set of planned routes and time plans traveled by each UE.

[0134] In another example, the network entity may configure a list of sensed data to be filtered for a second set of sensed data for the UE based on a first set of sensed data, wherein the reception of the second set of sensed data is further based on the list of sensed data to be filtered, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As described in 1118, network entity 1102 can also (e.g., based on sensed data received from the first set of UEs 1104) configure a list of sensed data to be filtered for the second set of UEs 1106. The configuration can be, for example, by... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0135] In another example, a network entity can output a first set of perceived data and a second set of perceived data to a perceived dataset, and receive a set of annotated data based on the output of the first and second sets of perceived data, such as combining... Figure 6 As described. For example, as in combination Figure 6 As described in 650, network entity 1102 can also output sensing data collected from a first set 1104 of UEs and / or a second set 1106 of UEs to a sensing dataset, and receive a set of annotation data based on this output. The output of the first set of sensing data and the second set of sensing data and the reception of the set of annotation data can be, for example, by Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0136] In another example, the network entity may configure at least one of the following for a third set of the UE based on a second set of sensed data: a second set with embedded centroids or a second set of planned routes and time plans for collecting the third set of sensed data; and a third set of sensed data received from the UE's third set based on at least one of the configured second set with planned routes and time plans or the configured second set with embedded centroids, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As described in section 1122, network entity 1102 may repeat one or more of the above steps (e.g., in conjunction with the steps described in sections 1110 to 1120) until sufficient sensed data is collected or the specified task is met. Therefore, network entity 1102 may continue to update and modify planned / proposed routes for different UEs (e.g., vehicles) based on time-evolving specifications. For example, network entity 1102 may configure embedded centroids and / or planned routes and time plans for another set of UEs in the UE pool (e.g., which may have UEs overlapping / non-overlapping with the first set of UEs 1104 and / or the second set of UEs 1106) based on sensed data received from the second set of UEs at 1116, and / or for collecting additional sensed data. Similarly, network entity 1102 may receive additional sensed data from the third set of UEs based on the configured set of planned routes and time plans and / or the configured set of embedded centroids. The configuration of at least one of the following: a second set of embedded centroids or a second set of planned routes and time plans for collecting a third set of sense data, and / or the reception of the third set of sense data may be provided by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0137] In another example, the first set of perceived data may be based on the set of centroids of the distribution representing the features of the first set of perceived data.

[0138] In another example, the first set of UEs may at least partially overlap with the second set of UEs.

[0139] In another example, the first set of sensing data or the second set of sensing data may include at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0140] In another example, network entities can annotate a second set of perceived data.

[0141] Figure 13This is a flowchart 1300 of a wireless communication method. This method can be performed by network entities (e.g., one or more location servers 168; base station 102; network entities 602, 1102, 1460). This method enables network entities to configure a pool of UEs to collect perceived data for training machine learning models based on a set of specifications (e.g., dataset / task specifications, cost specifications, storage, vehicle deployment, logistics costs, etc.), thereby reducing data storage costs, increasing data diversity, and / or reducing redundant data.

[0142] At 1302, the network entity can receive a first set of sensed data collected by the first set of UEs, such as combined... Figures 6 to 11 As described. For example, as in combination Figure 11 As described in 1110, network entity 1102 can receive sensing data collected by the first set of UEs 1104. The reception of the first set of sensing data can be achieved by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0143] At 1304, the network entity may configure at least one of the following for a second set of UEs based on a first set of sensed data: a set of embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As described in 1112, network entity 1102 can analyze currently available sensing data / distributions (including sensing data received from the first set 1104 of UEs at 1110) and compare this currently available sensing data / distribution with a specification set (e.g., a set of target distributions / tasks, etc.) to determine specified additional sensing data (e.g., a set of desired distributions to be collected). Then, based on the specified additional sensing data, network entity 1102 can (e.g., by sending configuration to the second set 1106 of UEs) configure for the second set 1106 of UEs a set of embedded centroids and / or a set of planned routes and time schedules for collecting the additional sensing data. The configuration can be provided by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0144] In one example, to configure a set of planned routes and time plans for collecting a second set of sensed data, a network entity may be configured to evaluate the difference between a first set of sensed data and an existing set of sensed data in the sensed dataset; determine the set of planned routes and time plans for collecting the second set of sensed data based on the evaluated difference; and send the set of planned routes and time plans to the second set of UEs.

[0145] In another example, in order to configure a set of planned routes and time plans for collecting a second set of sensing data, a network entity may be configured to obtain information relating to at least one of the following: data specified for the sensing dataset, existing trajectory data, or cost constraints; determine, based on the information, a set of planned routes and time plans for collecting the second set of sensing data; and send the set of planned routes and time plans to the second set of UEs.

[0146] In another example, in order to configure a set of planned routes and time plans for collecting a second set of sensed data, a network entity may be configured to obtain an indication of the attribute distribution; determine the set of planned routes and time plans for collecting the second set of sensed data based on the indication, map data, and traffic information; and send the set of planned routes and time plans to the second set of UEs.

[0147] At 1308, the network entity can receive a second set of sensed data from the second set of the UE based on at least one of a configured set of planned routes and time plans or a configured set of embedded centroids, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As described in 1116, network entity 1102 can receive sensing data collected based on the configuration sent at 1112 from the second set of UEs 1106. The reception of the second set of sensing data can be achieved by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0148] In one example, in order to receive a second set of sensed data from a second set of UEs based on a configured set of planned routes and time plans, a network entity may be configured to receive from each UE in the second set of UEs an indication of a portion of the configured set of planned routes and time plans traveled by each UE.

[0149] In another example, as shown at 1306, the network entity can configure a list of sensed data to be filtered for a second set of UEs based on a first set of sensed data, wherein the reception of the second set of sensed data is further based on the list of sensed data to be filtered, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As described in 1118, network entity 1102 can also (e.g., based on sensed data received from the first set of UEs 1104) configure a list of sensed data to be filtered for the second set of UEs 1106. The configuration can be, for example, by... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0150] In another example, as shown at 1310, a network entity can output a first set of perceived data and a second set of perceived data to a perceived dataset, and receive a set of annotated data based on the output of the first set of perceived data and the second set of perceived data, such as combining... Figure 6 As described. For example, as in combination Figure 6 As described in 650, network entity 1102 can also output sensing data collected from a first set 1104 of UEs and / or a second set 1106 of UEs to a sensing dataset, and receive a set of annotation data based on this output. The output of the first set of sensing data and the second set of sensing data and the reception of the set of annotation data can be, for example, by Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0151] In another example, as shown at 1312, the network entity may configure at least one of the following for a third set of the UE based on a second set of sensed data: a second set with embedded centroids or a second set of planned routes and time plans for collecting the third set of sensed data; and a third set of sensed data received from the third set of the UE based on at least one of the configured second set of planned routes and time plans or the configured second set with embedded centroids, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11As described in section 1122, network entity 1102 may repeat one or more of the above steps (e.g., in conjunction with the steps described in sections 1110 to 1120) until sufficient sensed data is collected or the specified task is met. Therefore, network entity 1102 may continue to update and modify planned / proposed routes for different UEs (e.g., vehicles) based on time-evolving specifications. For example, network entity 1102 may configure embedded centroids and / or planned routes and time plans for another set of UEs in the UE pool (e.g., which may have UEs overlapping / non-overlapping with the first set of UEs 1104 and / or the second set of UEs 1106) based on sensed data received from the second set of UEs at 1116, and / or for collecting additional sensed data. Similarly, network entity 1102 may receive additional sensed data from the third set of UEs based on the configured set of planned routes and time plans and / or the configured set of embedded centroids. The configuration of at least one of the following: a second set of embedded centroids or a second set of planned routes and time plans for collecting a third set of sense data, and / or the reception of the third set of sense data may be provided by, for example... Figure 14 The trajectory planning component 199, network processor 1412, and / or network interface 1480 of the network entity 1460 are executed.

[0152] In another example, the first set of perceived data may be based on the set of centroids of the distribution representing the features of the first set of perceived data.

[0153] In another example, the first set of UEs may at least partially overlap with the second set of UEs.

[0154] In another example, the first set of sensing data or the second set of sensing data may include at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0155] In another example, network entities can annotate a second set of perceived data.

[0156] Figure 14Figure 1400 illustrates an example of a hardware implementation for network entity 1460. In one example, network entity 1460 may be within core network 120. Network entity 1460 may include at least one network processor 1412. Network processor 1412 may include on-chip memory 1412'. In some aspects, network entity 1460 may also include an additional memory module 1414. Network entity 1460 communicates with CU 1402 directly (e.g., via a backhaul link) or indirectly (e.g., via RIC) through network interface 1480. On-chip memory 1412' and additional memory module 1414 may each be considered as computer-readable media / memory. Each computer-readable media / memory may be non-transitory. Network processor 1412 is responsible for general processing, including executing software stored on the computer-readable media / memory. The software, when executed by the corresponding processor, causes the processor to perform the various functions described above. The computer-readable media / memory may also be used to store data manipulated by the processor when executing the software.

[0157] As discussed above, the trajectory planning component 199 may be configured to receive a first set of sensed data collected by the first set of user equipment (UEs). The trajectory planning component 199 may also be configured to configure at least one of the following for a second set of UEs based on the first set of sensed data: a set of embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data. The trajectory planning component 199 may also be configured to receive a second set of sensed data from the second set of UEs based on at least one of the configured set of planned routes and time plans or the configured set of embedded centroids. The trajectory planning component 199 may be located within a network processor 1412. The trajectory planning component 199 may be one or more hardware components specifically configured to execute the stated process / algorithm, implemented by one or more processors configured to execute the stated process / algorithm, stored in a computer-readable medium for implementation by one or more processors, or some combination thereof. When multiple processors are implemented, the multiple processors may execute the stated process / algorithm individually or in combination. Network entity 1460 may include various components configured for various functions. In one configuration, network entity 1460 may include components for receiving a first set of sensed data collected by the first set of UEs. Network entity 1460 may also include components for configuring at least one of the following for a second set of UEs based on the first set of sensed data: a set with embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data. Network entity 1460 may also include components for receiving a second set of sensed data from the second set of UEs based on at least one of the configured set with planned routes and time plans or the configured set with embedded centroids.

[0158] In one configuration, the components for configuring a set of planned routes and time plans for collecting a second set of sensed data may include configuring network entity 1460 to evaluate the difference between a first set of sensed data and a set of existing sensed data in the sensed dataset; determine a set of planned routes and time plans for collecting the second set of sensed data based on the evaluated difference; and send the set of planned routes and time plans to the second set of UEs.

[0159] In another configuration, the component for configuring a set of planned routes and time plans for collecting a second set of sensing data may include configuring network entity 1460 to obtain information relating to at least one of: data specified for the sensing dataset, existing trajectory data, or cost constraints; determining a set of planned routes and time plans for collecting the second set of sensing data based on the information; and sending the set of planned routes and time plans to the second set of UEs.

[0160] In another configuration, the components for configuring a set of planned routes and time plans for collecting a second set of sensed data may include configuring network entity 1460 to obtain an indication of attribute distribution; determining a set of planned routes and time plans for collecting the second set of sensed data based on the indication, map data, and traffic information; and sending the set of planned routes and time plans to the second set of UEs.

[0161] In another configuration, the components for receiving a second set of sensed data from a second set of UEs based on a configured set of planned routes and time planning may include configuring network entity 1460 to receive from each UE in the second set of UEs an indication of a portion of the configured set of planned routes and time planning traveled by each UE.

[0162] In another configuration, network entity 1460 may also include a component for configuring a list of sensed data to be filtered for a second set of UEs based on a first set of sensed data, wherein the reception of the second set of sensed data is further based on the list of sensed data to be filtered.

[0163] In another configuration, network entity 1460 may further include components for outputting a first set of sensing data and a second set of sensing data to a sensing dataset, and components for receiving a set of annotation data based on the output of the first set of sensing data and the second set of sensing data.

[0164] In another configuration, network entity 1460 may further include components for configuring at least one of the following for a third set of UEs based on a second set of sensed data: a second set with embedded centroids or a second set of planned routes and time planning for collecting sensed data; and components for receiving a third set of sensed data from the third set of UEs based on at least one of the configured second set with planned routes and time planning or the configured second set with embedded centroids.

[0165] In another configuration, the first set of perceived data may be based on a set of centroids of the distribution representing the features of the first set of perceived data.

[0166] In another configuration, the first set of UEs may at least partially overlap with the second set of UEs.

[0167] In another configuration, the first set of sensing data or the second set of sensing data may include at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0168] In another configuration, network entity 1460 may also include components for annotating a second set of perceived data.

[0169] The component may be a trajectory planning component 199 of network entity 1460 configured to perform the functions described by the component.

[0170] Figure 15 This is a flowchart 1500 of a method for wireless communication (or object detection). This method can be performed by a UE (e.g., UE 104; a set of vehicles 604; vehicles 904, 1004; a first set of UEs 1104; a second set of UEs 1106; device 1704). This method can improve the efficiency of sensing data collection and compilation by enabling the UE to filter / select sensing data to be uploaded based on a standard set.

[0171] At point 1504, the UE may send a first set of sensed data collected by the UE to the network entity, such as combined... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1110, the first set 1104 of the UE can send a set of sensed data collected by the first set 1104 of the UE to the network entity 1102. The transmission of the first set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0172] At 1506, the UE may receive configuration from a network entity based on a first set of sensed data for at least one of the following: a set of embedded centroids or a set of planned routes and time plans for collecting a second set of sensed data, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1112, a second set 1106 of UEs (which may include UEs overlapping with the first set 1104 of UEs) may receive configuration from network entity 1102 based on sensing data collected by the first set 1104 of UEs for at least one of the following: a set of embedded centroids or a set of planned routes and times for collecting additional sensing data. The reception of the configuration may be, for example, by... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0173] At point 1508, the UE can collect a second set of perceived data based on its configuration, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1114, the second set 1106 of the UE can collect sensed data based on configuration. The collection of the second set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0174] In one example, to collect a second set of sensed data based on configuration, the UE may be configured to obtain an embedded set from the second set of sensed data, compare the embedded set with a set of embedded centroids, and identify the sensed data to be uploaded based on the comparison. In some specific implementations, to obtain the embedded set from the second set of sensed data, the UE may be configured to perform distance-based file sampling to obtain the embedded set from the second set of sensed data.

[0175] In another example, the UE may select a third set of sensed data from a second set of sensed data collected based on configuration or standards, and send the selected third set of sensed data to the network entity.

[0176] In another example, the UE may receive a list of sensed data to be filtered from a network entity, filter a second set of sensed data based on the list, and send the filtered second set of sensed data.

[0177] In another example, the UE may collect the first set of perceived data based on a set of centroids of the distribution of features representing the first set of perceived data, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 10 As discussed in 1012, the online file selector 620 of vehicle 1004 may include an embedding distance sampler 1006 configured to perform distance-based file sampling using embeddings (e.g., between newly collected embeddings and embedding centroids) with an embedding centroid assigner 616. For example, as shown in 1014, the goal of the embedding distance sampler 1006 could be to find the most relevant / interested samples or perceptual data to retain; for example, if a sample (e.g., sample 2) is located in the target cluster, that sample might be worth retaining compared to samples farther from the target cluster (e.g., sample 1) and could help increase dataset diversity. The collection of the first set of perceptual data based on the centroid set could be, for example, by... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0178] In another example, the UE may send a second set of perceived data to the network entity, such as combined... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1116, the second set 1106 of the UE can send the sensed data collected based on the configuration received at 1112 to network entity 1102. The transmission of the second set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0179] In another example, the UE may send an indication to a network entity of a portion of a configured set of planned routes and time plans for the UE's travel, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1120, the second set of UEs 1106 can also send an indication to network entity 1102 of a portion of a configured set of planned routes and time plans undertaken by the second set of UEs 1106 (e.g., each UE in the second set of UEs 1106). The indication can be sent by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0180] In another example, the UE may receive at least one of the following from a network entity based on a second set of sensed data: a second set of embedded centroids or a second set of planned routes and time plans for collecting a third set of sensed data; a second set of sensed data collected based on at least one of the configured second set of planned routes and time plans or the configured second set of embedded centroids; and transmit the third set of sensed data to the network entity, such as in combination with... Figures 6 to 11 As described. For example, as in combination Figure 11As discussed in section 1122, network entity 1102 may repeat one or more of the above steps (e.g., in conjunction with the steps described in sections 1110 to 1120) until sufficient sensed data is collected or the specified task is met. Therefore, network entity 1102 may continue to update and modify planned / proposed routes for different UEs (e.g., vehicles) based on time-evolving specifications. For example, network entity 1102 may configure embedded centroids and / or planned routes and time plans for another set of UEs in the UE pool (e.g., which may have UEs overlapping / non-overlapping with the first set of UEs 1104 and / or the second set of UEs 1106) based on sensed data received from the second set of UEs at 1116, and / or for collecting additional sensed data. Similarly, network entity 1102 may receive additional sensed data from the third set of UEs based on the configured set of planned routes and time plans and / or the configured set of embedded centroids. In other words, the UE's third set can receive at least one of the following from network entity 1102 based on the sensed data from the UE's second set 1106: another set embedded with centroids or another set of planned routes and time plans for collecting additional sensed data; collect sensed data based on at least one of the configured set of planned routes and time plans or the set embedded with centroids; and send the collected additional sensed data to network entity 1102. The reception of the second set embedded with centroids or the second set of planned routes and time plans for collecting the third set of sensed data, the collection of the second set of sensed data, and / or the transmission of the third set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0181] In another example, the first set of sensing data or the second set of sensing data includes at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0182] In another example, the UE may evaluate whether each piece of perceived data in the second set of perceived data has informational value based on an active learning loop, and send the perceived data in the second set of perceived data that has been evaluated as having informational value to the network entity.

[0183] Figure 16This is a flowchart 1600 of a method for wireless communication (or object detection). This method can be performed by a UE (e.g., UE 104; a set of vehicles 604; vehicles 904, 1004; a first set of UEs 1104; a second set of UEs 1106; device 1704). This method can improve the efficiency of sensing data collection and compilation by enabling the UE to filter / select sensing data to be uploaded based on a standard set.

[0184] At point 1604, the UE may send a first set of sensed data collected by the UE to the network entity, such as combined... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1110, the first set 1104 of the UE can send a set of sensed data collected by the first set 1104 of the UE to the network entity 1102. The transmission of the first set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0185] At 1606, the UE may receive configuration from a network entity based on a first set of sensed data for at least one of the following: a set of embedded centroids or a set of planned routes and time plans for collecting a second set of sensed data, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1112, a second set 1106 of UEs (which may include UEs overlapping with the first set 1104 of UEs) may receive configuration from network entity 1102 based on sensing data collected by the first set 1104 of UEs for at least one of the following: a set of embedded centroids or a set of planned routes and times for collecting additional sensing data. The reception of the configuration may be, for example, by... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0186] At point 1608, the UE can collect a second set of sensed data based on its configuration, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1114, the second set 1106 of the UE can collect sensed data based on configuration. The collection of the second set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0187] In one example, to collect a second set of sensed data based on configuration, the UE may be configured to obtain an embedded set from the second set of sensed data, compare the embedded set with a set of embedded centroids, and identify the sensed data to be uploaded based on the comparison. In some specific implementations, to obtain the embedded set from the second set of sensed data, the UE may be configured to perform distance-based file sampling to obtain the embedded set from the second set of sensed data.

[0188] In another example, the UE may select a third set of sensed data from a second set of sensed data collected based on configuration or standards, and send the selected third set of sensed data to the network entity.

[0189] In another example, the UE may receive a list of sensed data to be filtered from a network entity, filter a second set of sensed data based on the list, and send the filtered second set of sensed data.

[0190] In another example, as shown at 1602, the UE may collect the first set of perceived data based on a set of centroids of the distribution of features representing the first set of perceived data, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 10 As discussed in 1012, the online file selector 620 of vehicle 1004 may include an embedding distance sampler 1006 configured to perform distance-based file sampling using embeddings (e.g., between newly collected embeddings and embedding centroids) with an embedding centroid assigner 616. For example, as shown in 1014, the goal of the embedding distance sampler 1006 could be to find the most relevant / interested samples or perceptual data to retain; for example, if a sample (e.g., sample 2) is located in the target cluster, that sample might be worth retaining compared to samples farther from the target cluster (e.g., sample 1) and could help increase dataset diversity. The collection of the first set of perceptual data based on the centroid set could be, for example, by... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0191] In another example, as shown at 1610, the UE may send a second set of perceived data to the network entity, such as combined... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1116, the second set 1106 of the UE can send the sensed data collected based on the configuration received at 1112 to network entity 1102. The transmission of the second set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0192] In another example, as shown at 1612, the UE may send an indication to a network entity of a portion of a configured set of planned routes and time plans for the UE's travel, such as combining... Figures 6 to 11 As described. For example, as in combination Figure 11 As discussed in section 1120, the second set of UEs 1106 can also send an indication to network entity 1102 of a portion of a configured set of planned routes and time plans undertaken by the second set of UEs 1106 (e.g., each UE in the second set of UEs 1106). The indication can be sent by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0193] In another example, as shown at 1614, the UE may receive from a network entity at least one of the following based on a second set of sensed data: a second set of embedded centroids or a second set of planned routes and time plans for collecting a third set of sensed data; a second set of sensed data collected based on at least one of the configured second set of planned routes and time plans or the configured second set of embedded centroids; and transmit the third set of sensed data to the network entity, such as in combination with... Figures 6 to 11 As described. For example, as in combination Figure 11As discussed in section 1122, network entity 1102 may repeat one or more of the above steps (e.g., in conjunction with the steps described in sections 1110 to 1120) until sufficient sensed data is collected or the specified task is met. Therefore, network entity 1102 may continue to update and modify planned / proposed routes for different UEs (e.g., vehicles) based on time-evolving specifications. For example, network entity 1102 may configure embedded centroids and / or planned routes and time plans for another set of UEs in the UE pool (e.g., which may have UEs overlapping / non-overlapping with the first set of UEs 1104 and / or the second set of UEs 1106) based on sensed data received from the second set of UEs at 1116, and / or for collecting additional sensed data. Similarly, network entity 1102 may receive additional sensed data from the third set of UEs based on the configured set of planned routes and time plans and / or the configured set of embedded centroids. In other words, the UE's third set can receive at least one of the following from network entity 1102 based on the sensed data from the UE's second set 1106: another set embedded with centroids or another set of planned routes and time plans for collecting additional sensed data; collect sensed data based on at least one of the configured set of planned routes and time plans or the set embedded with centroids; and send the collected additional sensed data to network entity 1102. The reception of the second set embedded with centroids or the second set of planned routes and time plans for collecting the third set of sensed data, the collection of the second set of sensed data, and / or the transmission of the third set of sensed data can be achieved by, for example... Figure 17 The sensing data collection component 198, camera 1732, one or more sensors 1718, transceiver 1722, cellular baseband processor 1724 and / or application processor 1706 of the device 1704 are executed.

[0194] In another example, the first set of sensing data or the second set of sensing data includes at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0195] In another example, the UE may evaluate whether each piece of perceived data in the second set of perceived data has informational value based on an active learning loop, and send the perceived data in the second set of perceived data that has been evaluated as having informational value to the network entity.

[0196] Figure 17Figure 1700 illustrates an example of a specific hardware implementation for device 1704. Device 1704 may be a UE, a component of a UE, or implement UE functionality. In some aspects, device 1704 may include at least one cellular baseband processor 1724 (also referred to as a modem) coupled to one or more transceivers 1722 (e.g., cellular RF transceivers). Cellular baseband processor 1724 may include at least one on-chip memory 1724'. In some aspects, device 1704 may also include one or more Subscriber Identity Module (SIM) cards 1720 and at least one application processor 1706 coupled to a Secure Digital Card (SD) card 1708 and a screen 1710. Application processor 1706 may include on-chip memory 1706'. In some aspects, device 1704 may also include a Bluetooth module 1712, a WLAN module 1714, an ultra-wideband (UWB) module 1738, an SPS module 1716 (e.g., a GNSS module), one or more sensors 1718 (e.g., an atmospheric pressure sensor / altimeter; motion sensors such as an inertial measurement unit (IMU), a gyroscope, and / or an accelerometer; light detection and ranging (LIDAR), radio-assisted detection and ranging (RADAR), sound navigation and ranging (SONAR), a magnetometer, audio, and / or other technologies for positioning), an additional memory module 1726, a power supply 1730, and / or a camera 1732. Bluetooth module 1712, UWB module 1738, WLAN module 1714, and SPS module 1716 may include on-chip transceivers (TRX) (or in some cases, only receivers (RX)). Bluetooth module 1712, WLAN module 1714, and SPS module 1716 may include their own dedicated antennas and / or communicate using antenna 1780. Cellular baseband processor 1724 communicates with UE 104 and / or RU associated with network entity 1702 via transceiver 1722 through one or more antennas 1780. Cellular baseband processor 1724 and application processor 1706 may each include computer-readable media / memory 1724', 1706' respectively. Additional memory module 1726 may also be considered as computer-readable media / memory. Each computer-readable media / memory 1724', 1706', 1726 may be non-transitory. Cellular baseband processor 1724 and application processor 1706 are each responsible for general processing, including the execution of software stored on the computer-readable media / memory. When executed by cellular baseband processor 1724 / application processor 1706, the software causes cellular baseband processor 1724 / application processor 1706 to perform the various functions described above. Cellular baseband processor 1724 and application processor 1706 are configured to perform the various functions described above based at least in part on information stored in memory.In other words, the cellular baseband processor 1724 and application processor 1706 can be configured to perform a first subset of the various functions described above without information stored in memory, and can be configured to perform a second subset of the various functions described above based on information stored in memory. The computer-readable medium / memory can also be used to store data manipulated by the cellular baseband processor 1724 / application processor 1706 during software execution. The cellular baseband processor 1724 / application processor 1706 can be a component of the UE 350 and can include at least one of a memory 360 and / or a TX processor 368, an RX processor 356, and a controller / processor 359. In one configuration, the device 1704 can be at least one processor chip (modem and / or application) and includes only the cellular baseband processor 1724 and / or application processor 1706, while in another configuration, the device 1704 can be the entire UE (e.g., see [link]). Figure 3 The UE 350 includes an additional module of the device 1704.

[0197] As discussed above, the sensing data collection component 198 can be configured to send a first set of sensing data collected by the UE to a network entity. The sensing data collection component 198 can also be configured to receive configuration from the network entity based on the first set of sensing data for at least one of the following: a set of embedded centroids or a set of planned routes and time plans for collecting a second set of sensing data. The sensing data collection component 198 can also be configured to collect a second set of sensing data based on this configuration. The sensing data collection component 198 can be located within a cellular baseband processor 1724, an application processor 1706, or both. The sensing data collection component 198 can be one or more hardware components specifically configured to execute the stated process / algorithm, implemented by one or more processors configured to execute the stated process / algorithm, stored in a computer-readable medium for implementation by one or more processors, or some combination thereof. When multiple processors are implemented, the multiple processors can execute the stated process / algorithm individually or in combination. As shown, the apparatus 1704 can include various components configured for various functions. In one configuration, device 1704 (and particularly cellular baseband processor 1724 and / or application processor 1706) may include components for transmitting a first set of sensed data collected by the UE to a network entity. Device 1704 may also include components for receiving from the network entity a configuration based on the first set of sensed data for at least one of: an embedded centroid set or a set of planned routes and time plans for collecting a second set of sensed data. Device 1704 may also include components for collecting a second set of sensed data based on the configuration.

[0198] In one configuration, components for collecting a second set of sensed data based on the configuration may include: configuring device 1704 to obtain an embedded set from the second set of sensed data, comparing the embedded set with a set of embedded centroids, and identifying sensed data for uploading based on the comparison. In some specific implementations, in order to obtain the embedded set from the second set of sensed data, device 1704 may be configured to perform distance-based file sampling to obtain the embedded set from the second set of sensed data.

[0199] In another configuration, the apparatus 1704 may further include components for selecting a third set of sensing data from a second set of sensing data collected based on a configuration-based or criteria-based set, and components for transmitting the selected third set of sensing data to a network entity.

[0200] In another configuration, the apparatus 1704 may further include components for receiving a list of sensing data to be filtered from a network entity, components for filtering a second set of sensing data based on the list, and components for transmitting the filtered second set of sensing data.

[0201] In another configuration, the apparatus 1704 may also include a component for collecting the first set of sensed data based on a set of centroids representing the distribution of features of the first set of sensed data.

[0202] In another configuration, the device 1704 may also include components for sending a second set of sensing data to network entities.

[0203] In another configuration, the apparatus 1704 may also include a component for sending to a network entity an indication of a portion of a configured set of planned routes and time plans for travel by the UE.

[0204] In another configuration, the apparatus 1704 may further include components for receiving at least one of the following from a network entity based on a second set of sensed data: a second set of route and time planning embedded in a centroid or a third set of sensed data for collecting a second set of sensed data; components for collecting a second set of sensed data based on at least one of the configured second set of route and time planning or the configured second set embedded in a centroid; and components for sending the third set of sensed data to the network entity.

[0205] In another configuration, the first set of sensing data or the second set of sensing data includes at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0206] In another configuration, the apparatus 1704 may further include components for evaluating whether each piece of perceived data in the second set of perceived data has informational value based on an active learning loop, and components for sending the perceived data in the second set of perceived data that has been evaluated as having informational value to the network entity.

[0207] The component may be a sensing data collection component 198 of device 1704 configured to perform the functions described therein. As described above, device 1704 may include a TX processor 368, an RX processor 356, and a controller / processor 359. Therefore, in one configuration, the component may be the TX processor 368, the RX processor 356, and / or the controller / processor 359 configured to perform the functions described therein.

[0208] It should be understood that the specific order or hierarchy of the boxes in the disclosed process / flowcharts is merely an example of the exemplary method. It should be understood that the specific order or hierarchy of the boxes in the process / flowcharts may be rearranged based on design preferences. Furthermore, some boxes may be combined or omitted. The appended method claims present the elements of various boxes in a sample order, but are not limited to the given specific order or hierarchy.

[0209] The foregoing description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects. Therefore, the claims are not limited to the aspects described herein but should be given the full scope consistent with the language of the claims. Unless specifically stated otherwise, references to elements in the singular form do not mean “one and only one” but rather “one or more.” Terms such as “if,” “when,” and “simultaneously” do not imply a direct temporal relationship or reaction. That is, these phrases, such as “when,” do not imply an immediate action in response to or during the occurrence of an action, but simply suggest that if a condition is met, then the action will occur, without requiring a specific or immediate temporal constraint on the occurrence of the action. The word “exemplary” is used herein to mean “serving as an example, instance, or illustration.” Any aspect described herein as “exemplary” is not necessarily to be construed as preferred or superior to other aspects. Unless otherwise specifically stated, the term “some” refers to one or more. Combinations such as "at least one of A, B, or C", "one or more of A, B, or C", "at least one of A, B, and C", "one or more of A, B, and C", and "A, B, C, or any combination thereof" include any combination of A, B, and / or C, which may include multiple A, multiple B, or multiple C. Specifically, combinations such as "at least one of A, B, or C", "one or more of A, B, or C", "at least one of A, B, and C", "one or more of A, B, and C", and "A, B, C, or any combination thereof" can be only A, only B, only C, A and B, A and C, B and C, or A and B and C, where any such combination may contain one or more members of A, B, or C. A set should be interpreted as a collection of elements, where the number of elements is one or more. Therefore, for a set of X, X will include one or more elements. When at least one processor is configured to execute a set of functions, the at least one processor is configured to execute the set of functions individually or in any combination. Therefore, each of the at least one processor can be configured to execute a specific subset of the set of functions, wherein the subset is the complete set, a suitable subset of the set, or an empty subset of the set. A processor may be referred to as a processor circuit. A memory / memory module may be referred to as a memory circuit. If a first device receives data from or sends data to a second device, data can be received / sent directly between the first and second devices, or indirectly between the first and second devices through a set of devices. A device configured to "output" or "provide" data (such as transmission, signaling, or messaging) may, for example, transmit data using a transceiver, or may transmit the data to the device that sent the data.A device configured to "acquire" data (such as, transmit, signal, or message) may, for example, receive the data using a transceiver, or may obtain the data from a device that receives the data. Information stored in memory includes instructions and / or data. All structural and functional equivalents of the elements throughout the various aspects described herein that are known to those skilled in the art or will later be known are expressly incorporated herein by reference and are covered by the claims. Furthermore, nothing disclosed herein is intended to be offered to the public, whether or not such disclosure is explicitly recited in the claims. The words "module," "mechanism," "element," "device," etc., cannot replace the word "component." Therefore, no claim element will be construed as a functional component unless the element is explicitly recited using the phrase "component for..."

[0210] As used in this article, the phrase “based on” should not be interpreted as referring to a closed set of information, one or more conditions, one or more factors, etc. In other words, the phrase “based on A” (where “A” can be information, conditions, factors, etc.) should be interpreted as “based on at least A”, unless specifically stated differently.

[0211] The following aspects are merely illustrative and may be combined with other aspects or teachings described herein without limitation.

[0212] Aspect 1 is a method for wireless communication at a network entity, the method comprising: receiving a first set of sensed data collected by the first set of user equipment (UEs) from a first set of UEs; configuring at least one of the following for a second set of UEs based on the first set of sensed data: a set with embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data; and receiving the second set of sensed data from the second set of UEs based on at least one of the configured set of planned routes and time plans or the configured set with embedded centroids.

[0213] Aspect 2 is the method according to aspect 1, wherein configuring the set of planned routes and time plans for collecting the second set of sensed data includes: evaluating the difference between the first set of sensed data and a set of existing sensed data in the sensed dataset; determining the set of planned routes and time plans for collecting the second set of sensed data based on the evaluated difference; and sending the set of planned routes and time plans to the second set of the UE.

[0214] Aspect 3 is the method according to aspect 1 or aspect 2, wherein configuring the set of planned routes and time plans for the second set of sensing data collection includes: obtaining information related to at least one of: data specified for the sensing dataset, existing trajectory data, or cost constraints; determining the set of planned routes and time plans for the second set of sensing data collection based on the information; and sending the set of planned routes and time plans to the second set of the UE.

[0215] Aspect 4 is a method according to any one of Aspects 1 to 3, wherein configuring the set of planned routes and time plans for the second set of collecting sensing data includes: obtaining an indication of attribute distribution; determining the set of planned routes and time plans for the second set of collecting sensing data based on the indication, map data, and traffic information; and sending the set of planned routes and time plans to the second set of the UE.

[0216] Aspect 5 is a method according to any one of Aspects 1 to 4, the method further comprising: configuring a list of sensed data to be filtered for a second set of the UE based on a first set of sensed data, wherein the reception of the second set of sensed data is further based on the list of sensed data to be filtered.

[0217] Aspect 6 is the method according to any one of Aspects 1 to 5, wherein the second set of UEs receiving sensing data from the second set of UEs based on a configured set of planned routes and time plans further includes: receiving from each UE in the second set of UEs an indication of a portion of the configured set of planned routes and time plans traveled by each UE.

[0218] Aspect 7 is a method according to any one of aspects 1 to 6, the method further comprising: configuring at least one of the following for a third set of the UE based on the second set of the sensed data: a second set with embedded centroids or a second set of planned routes and time plans for collecting the third set of the sensed data; and receiving the third set of the sensed data from the third set of the UE based on at least one of the configured second set of planned routes and time plans or the configured second set with embedded centroids.

[0219] Aspect 8 is a method according to any one of aspects 1 to 7, wherein the first set of perceived data is based on a set of centroids of a distribution representing the characteristics of the first set of perceived data.

[0220] Aspect 9 is the method according to any one of aspects 1 to 8, wherein the first set of UEs and the second set of UEs at least partially overlap.

[0221] Aspect 10 is a method according to any one of aspects 1 to 9, wherein the first set of sensing data or the second set of sensing data includes at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0222] Aspect 11 is a method according to any one of aspects 1 to 10, the method further comprising: outputting a first set of perceptual data and a second set of perceptual data to a perceptual dataset; and receiving a set of annotation data based on the output of the first set of perceptual data and the second set of perceptual data.

[0223] Aspect 12 is the method according to any one of aspects 1 to 11, the method further comprising: annotating the second set of perceived data.

[0224] Aspect 13 is an apparatus for wireless communication at a network entity, the apparatus comprising: at least one memory; and at least one processor coupled to the at least one memory and based at least in part on information stored in the at least one memory, the at least one processor being configured individually or in any combination to implement any one of aspects 1 to 12.

[0225] Aspect 14 is the apparatus according to aspect 13, the apparatus further comprising: at least one camera coupled to the at least one processor.

[0226] Aspect 15 is an apparatus for wireless communication at a network entity, the apparatus comprising: components for implementing any one of aspects 1 to 12.

[0227] Aspect 16 is a computer-readable medium (e.g., a non-transitory computer-readable medium) that stores computer-executable code, wherein the code, when executed by a processor, causes the processor to implement any one of aspects 1 to 12.

[0228] Aspect 17 is a method for wireless communication at a user equipment (UE), the method comprising: transmitting to a network entity a first set of sensed data collected by the UE; receiving from the network entity, based on the first set of sensed data, a configuration for at least one of: a set of embedded centroids or a set of planned routes and time plans for collecting a second set of sensed data; and collecting the second set of sensed data based on the configuration.

[0229] Aspect 18 is the method according to aspect 17, the method further comprising: sending the second set of sensing data to the network entity.

[0230] Aspect 19 is the method according to aspect 17 or aspect 18, the method further comprising: receiving from the network entity at least one of the following based on the second set of sense data: a second set of planned routes and time plans embedded with centroids or a third set of sense data for collecting sense data; collecting the second set of sense data based on at least one of the configured second set of planned routes and time plans or the configured second set of embedded centroids; and sending the third set of sense data to the network entity.

[0231] Aspect 20 is a method according to any one of aspects 17 to 19, wherein collecting the second set of sensed data based on the configuration includes: obtaining an embedded set from the second set of sensed data; comparing the embedded set with the set of embedded centroids; and identifying sensed data for uploading based on the comparison.

[0232] Aspect 21 is a method according to any one of aspects 17 to 20, wherein obtaining the set of embeddings from the second set of perceptual data comprises: performing distance-based file sampling to obtain the set of embeddings from the second set of perceptual data.

[0233] Aspect 22 is a method according to any one of aspects 17 to 21, the method further comprising: selecting a third set of sensing data from a second set of sensing data collected based on the configuration or a standard-based set; and sending the selected third set of sensing data to the network entity.

[0234] Aspect 23 is a method according to any one of aspects 17 to 22, the method further comprising: receiving from the network entity a list of sensing data to be filtered; filtering the second set of sensing data based on the list; and sending the filtered second set of sensing data.

[0235] Aspect 24 is the method according to any one of aspects 17 to 23, the method further comprising: sending to the network entity an indication of a portion of a configured set of planned routes and time plans for travel by the UE.

[0236] Aspect 25 is a method according to any one of aspects 17 to 24, wherein the first set of sensed data or the second set of sensed data includes at least one of the following: a set of images captured by at least one camera or a set of point clouds captured or generated by at least one radar.

[0237] Aspect 26 is a method according to any one of aspects 17 to 25, the method further comprising: evaluating whether each piece of perceived data in the second set of perceived data has informational value based on an active learning loop; and sending the perceived data in the second set of perceived data that has been evaluated as having informational value to the network entity.

[0238] Aspect 27 is a method according to any one of aspects 17 to 26, the method further comprising: collecting the first set of perceived data based on a set of centroids of a distribution representing the features of the first set of perceived data.

[0239] Aspect 28 is an apparatus for wireless communication at a user equipment (UE), the apparatus comprising: at least one memory; and at least one processor coupled to the at least one memory and based at least in part on information stored in the at least one memory, the at least one processor being configured individually or in any combination to implement any one of aspects 17 to 27.

[0240] Aspect 29 is an apparatus according to aspect 28, the apparatus further comprising: at least one camera coupled to the at least one processor.

[0241] Aspect 30 is an apparatus for wireless communication at a user equipment (UE), the apparatus comprising: components for implementing any one of aspects 17 to 27.

[0242] Aspect 31 is a computer-readable medium (e.g., a non-transitory computer-readable medium) that stores computer-executable code, wherein the code, when executed by a processor, causes the processor to implement any one of aspects 17 to 27.

Claims

1. An apparatus for wireless communication at a network entity, the apparatus comprising: At least one memory; and At least one processor, coupled to the at least one memory, wherein the at least one processor is configured individually or in any combination as follows: Receive a first set of perception data collected by the first set of user equipment (UE) from the first set of UE; Based on the first set of sensed data, configure at least one of the following for the second set of UEs: a set of embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data; as well as The second set of sensing data is received from the second set of the UE based on at least one of the configured set of planned routes and time planning or the configured set of embedded centroids.

2. The apparatus of claim 1, wherein, in order to configure the planned routes and time plans for collecting the second set of sensing data, the at least one processor is configured individually or in any combination to: Evaluate the differences between the first set of sensed data and the existing set of sensed data in the sensed dataset; The set of planned routes and time plans for collecting the second set of sensory data are determined based on the assessed differences; and The planned route and time plan are sent to the second set of UEs.

3. The apparatus of claim 1, wherein, in order to configure the planned routes and time plans for collecting the second set of sensing data, the at least one processor is configured individually or in any combination to: Obtain information relating to at least one of the following: data specified for the sensing dataset, existing trajectory data, or cost constraints; Based on the information, determine the planned route and time plan for the second set of data collection; and The planned route and time plan are sent to the second set of UEs.

4. The apparatus of claim 1, wherein, in order to configure the planned routes and time plans for collecting the second set of sensing data, the at least one processor is configured individually or in any combination to: Obtain an indication of the attribute distribution; The second set of planned routes and time plans for collecting sensing data are determined based on the instructions, map data, and traffic information. as well as The planned route and time plan are sent to the second set of UEs.

5. The apparatus of claim 1, wherein the at least one processor is further configured, alone or in any combination, to: The second set of sensed data is configured for the UE to filter a list of sensed data based on the first set of sensed data, wherein the reception of the second set of sensed data is further based on the list of sensed data to be filtered.

6. The apparatus of claim 1, wherein, in order to receive the second set of sensed data from the second set of the UE based on a configured set of planned routes and time plans, the at least one processor is further configured individually or in any combination to: Each UE in the second set of UEs receives an instruction for a portion of the configured set of planned routes and time plans for each UE's travel.

7. The apparatus of claim 1, wherein the at least one processor is further configured, alone or in any combination, to: Based on the second set of sensed data, configure at least one of the following for the third set of the UE: a second set with embedded centroids or a second set of planned routes and time plans for collecting the third set of sensed data; and The third set of sensing data is received from the third set of the UE based on at least one of the configured second set of planned routes and time planning or the configured second set of embedded centroids.

8. The apparatus of claim 1, wherein the first set of UEs and the second set of UEs at least partially overlap.

9. The apparatus of claim 1, wherein the first set of sensed data or the second set of sensed data comprises at least one of the following: A collection of images captured by at least one camera, or A collection of point clouds captured or generated by at least one radar.

10. The apparatus of claim 1, wherein the at least one processor is further configured, alone or in any combination, to: Output the first set of perceived data and the second set of perceived data to the perceived dataset, and The set of annotation data is received based on the output of the first set of perceptual data and the second set of perceptual data.

11. A method for wireless communication at a network entity, the method comprising: Receive a first set of perception data collected by the first set of user equipment (UE) from the first set of UE; Based on the first set of sensed data, configure at least one of the following for the second set of UEs: a set of embedded centroids or a set of planned routes and time plans for collecting the second set of sensed data; as well as The second set of sensing data is received from the second set of the UE based on at least one of the configured set of planned routes and time planning or the configured set of embedded centroids.

12. The method of claim 11, wherein the set of planned routes and time plans for configuring the second set of sensing data includes: Evaluate the differences between the first set of sensed data and the existing set of sensed data in the sensed dataset; The planned routes and time plans for the second set of data collection are determined based on the assessed differences. as well as The planned route and time plan are sent to the second set of UEs.

13. The method of claim 11, wherein the set of planned routes and time plans for configuring the second set of sensing data includes: Obtain information relating to at least one of the following: data specified for the sensing dataset, existing trajectory data, or cost constraints; Based on the information, the route and time plan for the second set of data collection are determined. as well as The planned route and time plan are sent to the second set of UEs.

14. The method of claim 11, wherein the set of planned routes and time plans for configuring the second set of sensing data includes: Obtain an indication of the attribute distribution; The second set of planned routes and time plans for collecting sensing data are determined based on the instructions, map data, and traffic information. as well as The planned route and time plan are sent to the second set of UEs.

15. The method according to claim 11, further comprising: The second set of sensed data is configured for the UE to filter a list of sensed data based on the first set of sensed data, wherein the reception of the second set of sensed data is further based on the list of sensed data to be filtered.

16. An apparatus for wireless communication at a user equipment (UE), the apparatus comprising: At least one memory; and At least one processor, coupled to the at least one memory, wherein the at least one processor is configured individually or in any combination as follows: Send a first set of sensed data collected by the UE to the network entity; Based on the first set of sensing data, the network entity receives configuration for at least one of the following: a set of embedded centroids or a set of planned routes and time plans for collecting the second set of sensing data; as well as The second set of sensing data is collected based on the configuration.

17. The apparatus of claim 16, wherein the at least one processor is further configured, alone or in any combination, to: The second set of sensing data is sent to the network entity.

18. The apparatus of claim 16, wherein, in order to collect the second set of sensed data based on the configuration, the at least one processor is configured individually or in any combination to: Obtain the embedded set from the second set of perceived data; Compare the set of embedded centroids with the set of embedded centroids; and The perception data used for uploading is identified based on the comparison.

19. The apparatus of claim 16, wherein the at least one processor is further configured, alone or in any combination, to: Receive a list of perceived data to be filtered from the network entity; The second set of perceived data is filtered based on the list; and The second set of filtered sensory data sent.

20. The apparatus of claim 16, wherein the at least one processor is further configured, alone or in any combination, to: Send an instruction to the network entity regarding a portion of the configured set of planned routes and time plans for the UE to travel.

21. The apparatus of claim 16, wherein the first set of sensed data or the second set of sensed data comprises at least one of the following: A collection of images captured by at least one camera, or A collection of point clouds captured or generated by at least one radar.

22. The apparatus of claim 16, wherein the at least one processor is further configured, alone or in any combination, to: The second set of perceived data is evaluated based on an active learning loop to determine whether each piece of perceived data in the second set of perceived data has informational value; and Sensing data from the second set of sensing data that are evaluated as having informational value are sent to the network entity.

23. The apparatus of claim 16, wherein the at least one processor is further configured, alone or in any combination, to: The first set of perceptual data is collected based on the set of centroids of the distribution of features representing the first set of perceptual data.

24. A method for conducting wireless communication at a user equipment (UE), the method comprising: Send a first set of sensed data collected by the UE to the network entity; Based on the first set of sensing data, the network entity receives configuration for at least one of the following: a set of embedded centroids or a set of planned routes and time plans for collecting the second set of sensing data; as well as The second set of sensing data is collected based on the configuration.

25. The method according to claim 24, further comprising: The second set of sensing data is sent to the network entity.

26. The method of claim 24, wherein the second set of sensing data collected based on the configuration comprises: Obtain the embedded set from the second set of perceived data; Compare the set of embedded centroids with the set of embedded centroids; as well as The perception data used for uploading is identified based on the comparison.

27. The method of claim 24, further comprising: Receive a list of perceived data to be filtered from the network entity; The second set of perceived data is filtered based on the list; as well as The second set of filtered sensory data sent.

28. The method according to claim 24, further comprising: Send an instruction to the network entity regarding a portion of the configured set of planned routes and time plans for the UE to travel.

29. The method of claim 24, wherein the first set of sensed data or the second set of sensed data comprises at least one of the following: A collection of images captured by at least one camera, or A collection of point clouds captured or generated by at least one radar.

30. The method according to claim 24, further comprising: The active learning loop is used to evaluate whether each piece of perceptual data in the second set of perceptual data has informational value. as well as Sensing data from the second set of sensing data that are evaluated as having informational value are sent to the network entity.

Images