AMBIENT INTERNET OF THINGS (A-IoT) DEVICES AND METHOD FOR OPERATING ON THE SAME
A class-based adaptive midamble interval configuration for A-IoT devices addresses energy waste by tailoring synchronization to device needs, enhancing energy efficiency and communication reliability.
Patent Information
- Application Number
- PCT/IB2026/054576
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-16
- Filing Date
- 2026-05-08
- Publication Date
- 2026-11-19
AI Technical Summary
Current A-IoT synchronization strategies waste energy due to unnecessary device awakenings caused by applying a single fixed midamble interval to all devices, regardless of their specific synchronization needs, leading to inefficiencies in energy consumption and system performance.
Implement a class-based adaptive configuration of midamble intervals tailored to different synchronization categories of A-IoT devices, based on their expected clock drift, application criticality, and environmental factors, allowing devices to skip unnecessary synchronization requests.
This approach reduces unnecessary energy consumption by optimizing synchronization frequency according to device class, ensuring efficient and scalable communication while maintaining reliable timing alignment.
Smart Images

Figure IB2026054576_19112026_PF_FP_ABST
Abstract
Description
[0001] -SVT084-PC00-
[0002] AMBIENT INTERNET OF THINGS (A-IoT) DEVICES AND METHOD FOR OPERATING ON THE SAME
[0003] DESCRIPTION
[0004] The present invention relates to an ambient internet of things (A-IoT) device and an operating method for said A-IoT device. More particularly, the present invention relates to an improved A-IoT device and an energy-aware operating method for said improved A-IoT device.
[0005] With the growth of Internet of Things (loT) technologies in a great variety of fields, it is expected that the use of very compact loT devices of ultra-low power will increase explosively. However, because such loT devices need to receive power through a battery or wired power connection to operate, they confront limitations in usable environments and occurrence of maintenance costs.
[0006] Ambient loT (A-IoT) is a new concept standardized by Third Generation Partnership Project (3GPP), an international consortium comprising a number of standards organizations which develop protocols for mobile telecommunications. A-IoT aims to support massive numbers of ultra-low-cost, low-power, and maintenance-free loT devices. These A-IoT devices operate by harvesting energy from their surrounding environment, e.g., light, motion, or radio waves, eliminating or minimizing the need for batteries. A-IoT technology represents an evolution of existing cellular loT systems such as Narrowband Internet of Things (NB-loT) and Long-Term Evolution Machine Type Communication (LTE-M) moving towards a more sustainable and scalable future for the Internet of Things.
[0007] Indeed, 3 GPP initially focused on battery-powered loT technologies, such as NB-IoT and LTE-M designed for low-power wide-area applications. However, being battery -based, they showed limited scalability in dense deployments or hard-to-reach environments.
[0008] Ambient loT refers to the use of loT devices that can function with very low energy requirements by harvesting ambient energy. This may comprise energy from light (solar), radio signals, thermal gradients, or motion (vibration). In some cases, A-IoT devices may not have batteries at all or only use tiny energy storage units.
[0009] To overcome the issue about scalability, 3 GPP initiated studies in Release 18 and further refined the work in technical reports TR 38.769 (Ambient loT Study) and TR 38.848 (Evaluation and Scenarios). These reports investigate system requirements, performance evaluation, and architecture needs to support A-IoT.
[0010] A-IoT enables several practical and large-scale applications due to its self-sufficient, maintenance-free nature such as, but not limited to, smart packaging and inventory for-SVT084-PC00-
[0011] tracking items across supply chains without battery maintenance, smart agriculture for soil and weather sensors powered by ambient energy, smart cities for sensors in roads, buildings, and traffic lights, healthcare for wearables and implants without the need for frequent charging or battery replacement, environmental monitoring for widespread sensors in remote or difficult-to-reach locations.
[0012] According to 3GPP standard, in particular for technical reports from TR 38.769 and TR 38.848, A-IoT devices can communicate in two primary ways: backscatter communication, in which A-IoT devices reflect and modulate environmental signals to transmit data and active transmission, in which devices generate their own transmission using collected energy.
[0013] A-IoT devices are typically very small, simple, and low-cost, designed to operate in large quantities with minimal infrastructure support.
[0014] 3 GPP classifies A-IoT devices into three categories based on their energy usage and communication capabilities:
[0015] Battery-Less (BL) in which an A-IoT device operates only through backscatter, without internal energy storage;
[0016] Battery-Assisted (BA) in which an A-IoT device uses minimal stored energy to improve communication, e.g. stronger backscatter signals;
[0017] Battery and Signal -Assisted (BSA) in which an A-IoT device can actively transmit data using stored and harvested energy.
[0018] Moreover, A-IoT devices may use different communication modes depending on the application and available infrastructure. In particular an A-IoT device may use a Monostatic Backscatter mode in which the same node emits and receives the signal, a Bistatic Backscatter mode in which one node emits the carrier signal, and another node receives the reflected signal or a Direct Active Communication in which the A-IoT device actively sends data directly to a base station or another device using harvested energy.
[0019] The 3 GPP TR 38.848 outlines various network configurations for A-IoT:
[0020] Direct-to-Base Station in which an A-IoT device communicates directly with the cellular network;
[0021] Relay via Intermediate Node in which a relay device (e.g. another user equipment, UE) forwards a communication;
[0022] Assisting Node Topology in which another node helps facilitate a communication in either uplink or downlink;
[0023] UE-to-UE in which there is a direct communication between an A-IoT device and a user equipment.-SVT084-PC00-
[0024] At RAN# 106, a plenary meeting of the Technical Specification Group Radio Access Network (TSG RAN), a new work item was approved to study solutions for A-IoT, and that is divide into four parts.
[0025] Physical channel design relating to modulation aspects is a first topic focusing on selecting simple, energy-efficient modulation schemes suitable for low-power A-IoT devices. Given the limited computational resources of A-IoT devices, the modulation needs to be minimalistic yet robust such as digital modulation schemes like On-Off Keying (OOK) or simplified Binary Phase Shift Keying (BPSK).
[0026] The modulation techniques chosen must allow A-IoT devices with very basic radiofrequency (RF) and baseband capabilities to detect and demodulate signals and compatibility with NR waveforms must be maintained while reducing complexity.
[0027] A second topic is related to physical channel design, particularly on aspects of line coding as an aid to synchronization and data integrity such as Manchester, pulse interval coding, Forward Error Correction (FEC) to transmission errors corrections such as simple convolutional or block codes, cyclic redundancy check (CRC) to checks for errors in received data with lightweight error detection, Repetition schemes to increase reliability by retransmitting data, useful in case of intermittent reception or low Signal to Noise Ratio (SNR).
[0028] Timing acquisition and synchronization is a third topic focusing on allowing A-IoT devices to align their timing with a reader, a device that communicates with small loT devices to send commands and timing information, essential function for decoding signals and avoiding interference.
[0029] Furthermore, timing acquisition is necessary for initial access and energy-efficient wake-up scheduling for an A-IoT device. Synchronization techniques comprise the use of preambles, synchronizations sequences and timing advance and supports Time Sensitive Networking (TSN), where applicable.
[0030] Solutions considered in this topic explore also legacy TA mechanisms, enhanced granularity, or Round-Trip Time (RTT) based timing compensation.
[0031] The last of the four topics focuses on how many devices can share the spectrum efficiently and how control information (like scheduling) is conveyed with minimal overhead.
[0032] In particular on multiplexing methods (time / frequency / code) aspects that allow concurrent access for thousands of devices, multiple access schemes aspects that consider the ultra-low-power nature of devices, possibly enabling asynchronous access, scheduling aspects embedded within data transmissions to reduce control signaling and timing relationships aspects defining how devices coordinate their communication windows and avoid collision.-SVT084-PC00-
[0033] Said topics are designed to enable ultra-low-cost, battery-less or energy-harvesting loT devices to communicate effectively within the 5GNR ecosystem, a global standard defined by 3GPP for the new radio, NR, interface of 5G networks.
[0034] Timing acquisition and synchronization is critically important in A-IoT because it enables A-IoT devices to align their internal clocks with the network or reader, allowing correct decoding of signals and reliable data exchange, especially in ultra-low-power scenarios where said devices operate on limited or harvested energy. The legacy schemes are detached from the real-time contextual needs of the A-IoT devices, leading to wasted power cycles when no need arises to receive or send data. Without efficient synchronization, A-IoT devices may miss transmissions, collide with others, or waste energy by waking up unnecessarily, which severely affects their performance and lifespan.
[0035] As A-IoT deployments often involve thousands of A-IoT devices, synchronization ensures scalable, collision-free communication and coordinated access with minimal signaling overhead. Moreover, synchronization plays a central role in the wake-up scheduling of A-loT devices, allowing said devices to remain in a deep sleep state until a need to receive or send data arises, which is essential for energy efficiency and responsiveness.
[0036] In current proposed A-IoT systems, devices are required to wake up periodically for synchronization, regardless of whether an actual need to transmit or receive data exists, leading to unnecessary energy consumption, especially in systems where low-event-rate devices, such as asset trackers or environmental sensors, are present and may remain idle for extended periods.
[0037] The synchronization methods of the known art are insufficient to meet the challenges of A-loT devices, which require meeting both synchronization and energy-aware. Instead, frequent polling to execute synchronization leads to a substantial waste of energy for A-IoT devices.
[0038] In fact, in current proposals, a single fixed midamble interval is applied to all A-IoT devices forcing them to wake up and synchronize.
[0039] A midamble signal is a signal typically used in wireless communication systems as part of a timing synchronization process. It is a signal embedded between transmitted data packets that helps devices align their internal clocks with the transmitter’s clock.
[0040] The midamble signal is used in loT systems for timing synchronization in which the midamble signal provides a reference point for devices to synchronize with the incoming signal, helping them maintain accurate timing. The midamble signal allows loT devices to save energy while maintaining reliable communication since, being a short and periodic signal, it avoids continuous synchronization of devices.-SVT084-PC00-
[0041] Therefore, the midamble signal transmitted at regular intervals helps devices detect and lock onto the correct time window to decode incoming data. For example, in environments with frequent environmental changes (e.g. motion, temperature variations), midamble signals help loT devices detect and adapt to these changes without being constantly active.
[0042] In Ambient loT, A-IoT devices may use a midamble signal to synchronize with a reader or a base station, minimizing power consumption while ensuring accurate data exchange. However, frequent updates do not always benefit all categories of A-IoT devices, for example a static sensor may remain synchronized for hours, in contrast to a mobile tag in a warehouse may drift significantly in seconds. Applying frequent synchronization (midamble decoding) uniformly indifferently to all types of A-IoT devices wastes power, reduces system efficiency, and causes unnecessary processing.
[0043] Similarly, several technical reports and company contributions, such as Ericsson’s symmetric 2D distribution model for A-IoT timing design, underscore the importance of scalable synchronization but do not propose any solution to solve energy consumption issue due to unnecessary awakenings due to the synchronization mechanisms of such devices. In the most recent discussions, including the 3GPP RAN1#12O and RANl#120bis meetings, multiple companies have addressed improvements to synchronization mechanisms to support energy-efficient, large-scale A-IoT deployments. These contributions have typically emphasized enhancements to signal structure, receiver design, and power-efficient timing acquisition but again no proposals were made to solve the issue of power consumption due to unnecessary device wakeups.
[0044] Moreover, as highlighted in the Chair Notes of the 3GPP RANl#120bis meeting, the format and unit, measured in bits after FEC and repetition, of the midamble interval have been agreed upon, but once again nothing has been decided on the single fixed midamble transmission interval.
[0045] Prior art solutions thus highlight a critical gap in current A-IoT synchronization strategies that offer an opportunity to improve the energy efficiency of A-IoT devices by avoiding or at least reducing unnecessary awakenings.
[0046] Therefore, the highlighted issues underscore the need for a solution suitable for improving the energy efficiency of A-IoT devices.
[0047] The present invention aims to solve these and other problems by providing an A-IoT device and a method operating on the same for handling a midamble signal.
[0048] According to an aspect of the present invention, it is described an Ambient Internet of Things, A-IoT, device (10) configured to operate in an
[0049] A-IoT network system in which said A-IoT device (10) receives an R2D signal from a reader-SVT084-PC00-
[0050] device, the A-IoT device (10) comprising:
[0051] a reader unit (20) configured to read an assist flag comprised in said
[0052] R2D signal;
[0053] a class identifier unit (30) configured to memorize a pre-assigned class
[0054] category comprising a time interval value representing a frequency at which said A-IoT device (10) expects to receive a midamble signal in order to maintain a reliable time alignment with a reference clock within acceptable limits and a time tolerance value representative of a tolerable shift from said time interval before a synchronisation is required; an internal clock (70);
[0055] a decision logic unit (40) functionally connected to said reader unit (20), said class identifier unit (30) and said internal clock (70);
[0056] wherein said decision logic unit (40) is configured to:
[0057] reading (210), by means of the reader unit (20), a value of the assist flag comprised in said R2D signal;
[0058] retrieving (220), via the class identifier unit (30), said time interval value and said time tolerance value from said class category to define a time window within which the A-IoT device (10) is configured for skipping a synchronization request contained in said R2D signal;
[0059] checking (230) a first condition, by means of the decision logic unit (40), in which said assist flag value is set to a value representative of a synchronization request and one between a receiving time of said R2D signal and an expected receiving time value calculated by the internal clock (70) falls outside the time window defined by said class category for said A-IoT device (10).
[0060] According to a form of embodiment of the present invention, when an
[0061] application-layer communication requires an accuracy level higher than a first accuracy level provided by the midamble signal for its execution, a Clock Acquisition Part, CAP, portion of an R-TAS signal, is processed to synchronize the internal clock of said A-IoT device according to a second level of accuracy.
[0062] The invention will be described in detail hereinafter through non-limiting embodiments with reference to the attached Figures, wherein:
[0063] Fig. 1 shows a schematic diagram representing an A-IoT device according to the present invention;
[0064] Fig. 2 shows a flowchart of a method for operating with an A-IoT device according to the present invention.
[0065] References used in the description below and attached figures are for convenience only and-SVT084-PC00-
[0066] do not therefore define the scope of protection or the extent of the embodiment’s forms. To address the inefficiencies caused by using a single fixed midamble interval for all Ambient loT devices, this invention proposes a class-based adaptive configuration of midamble intervals.
[0067] In Ambient loT, the term reader device refers to any device suitable for transmitting R2D signals comprising a midamble signal. The midamble signal (comprising a Start Indicator Part, SIP, portion and a Clock Acquisition Part, CAP, portion) is suitable for guiding A-IoT devices through a synchronization process and potentially receiving data from them. The SIP portion and the CAP portion correspond to two different levels of synchronization accuracy, in particular said CAP portion provides a finer timing accuracy with a higher level than the level of accuracy achievable with the SIP portion.
[0068] The reader ensures that the A-IoT device synchronizes its timing with, for example, the base station used in smart cities and industrial networks to provide wide-area coverage and synchronize many devices. The reader plays a vital role in A-IoT ecosystems by providing precise timing information for the devices to operate efficiently.
[0069] An R2D signal refers to a communication link or control signal transmitted from the reader, like a base station or gateway, to an loT device, providing it with synchronization, control, or data messages. In Ambient loT, the R2D signal ensures that A-IoT devices stay synchronized with the reader and the overall network, receiving necessary instructions for data processing or wake-up commands.
[0070] In an Ambient loT environment, the R2D signal might activate A-IoT devices based on the status of the environment (e.g., if a threshold is exceeded, such as motion detection). The A-loT device will listen to the R2D signal to determine whether to wake up and synchronize, optimizing energy consumption while ensuring effective communication.
[0071] Midamble and R2D Signals Work Together. The midamble signal serves as a timing reference point, helping devices synchronize to said signal, while the R2D signal carries instructions about synchronization process, operational status, or commands for A-IoT devices. Together, they ensure efficient communication, coordination, and minimal power usage in an Ambient loT system.
[0072] Nowadays, a midamble signal contained within an R2D signal is sent periodically according to an identical time interval for all A-IoT devices. When the midamble signal is present, an assist flag of the R2D signal is set to 1. This value forces all awake A-IoT devices to synchronize. However, an A-IoT device may not need to synchronize because it is, for example, a static or low-drift device.
[0073] To address the inefficiencies caused by using a single fixed midamble interval for all A-IoT-SVT084-PC00-
[0074] devices, the present invention proposes a class-based adaptive configuration of midamble intervals.
[0075] A-IoT devices are divided into five synchronization category classes from A to E according to their synchronization needs, which are determined by their typical drift tolerance, communication frequency, and use case environment, e.g., static vs. mobile.
[0076] Each class category is assigned a specific midamble interval length, expressed in bits after Forward Error Correction, FEC, and repetition. These bit-length intervals correspond to how often the A-IoT device should receive a midamble for reliable timing alignment within acceptable bounds.
[0077] In addition, each class category is assigned a specific time tolerance value which is a measure of how far an internal clock frequency of an A-IoT device is phase-shifted compared to an optimal synchronized clock frequency.
[0078] The time interval value representing a frequency at which said A-IoT device expects to receive a midamble signal in order to maintain a reliable time alignment with the reader device within acceptable limits and a time tolerance value representative of a tolerable shift from said time interval before a synchronisation through a midamble signal is required. The classification is based on a theoretical assessment of expected oscillator drift during sleep, application criticality, and practical alignment with the OFDM symbol structure (256 bits per symbol). The OFDM symbol is a time unit used in wireless communication. One OFDM symbol duration is used, for example, in SIP signals to balance signal detectability and energy efficiency. This symbol duration, 66.7 microseconds, was agreed upon during the RANl#120bis meeting, considering its practicality for synchronization in A-IoT applications.
[0079] Below is a summary table of the five classes for A-IoT devices according to this invention:
[0080] Class Interval (bits) Estimated Timing Tolerance
[0081] A 0 (No midamble) ±50 ps
[0082] B 16-32 ±10 ps
[0083] C 4-8 ±1 ps
[0084] D 64 ±20 ps
[0085] E 128 ±5 ps
[0086]
[0087] The values in said table are derived from expected clock drifts per class category over typical sleep durations, timing tolerance range, e.g. ±1 to ±50 ps, system’s symbol structure, e.g.-SVT084-PC00-
[0088] 256 bits per OFDM symbol, practical framing alignment, e.g. multiples of 8 or 16 bits. For example, choosing 64 bits in class D instead of 56 or 128 is intentional since 64 aligns cleanly with OFDM frame sizes, offers appropriate sync frequency for devices with ~±20 ps tolerance, and avoids the unnecessary overhead of more frequent midamble signals. Meanwhile, 56 bits would misalign symbol framing, and 128 would risk timing drift for moderate-precision devices.
[0089] Each class category also comprises a value of a drift threshold and a value of an SNR threshold.
[0090] The term drift threshold means how many timing errors an A-IoT device can handle, because of internal clock does not run at the same rate as a reference clock, e.g. a reader device clock, before synchronization becomes necessary.
[0091] Each class is also associated with a drift tolerance range, which may vary depending on the application requirements, device precision, or energy constraints.
[0092] A Class A A-IoT device is the most tolerant to clock drift and can be assigned, for example, to very simple, passive environmental sensors or tags that only need loose synchronization. In said class category A, an acceptable drift value range is comprised between 0 to 200 ms, a preferred drift value range is between 0 to 150 ms, and a most preferred drift value range is between 50 and 100 ms.
[0093] This means that for class category A A-IoT device a drift threshold is 200 ms, a preferred drift threshold is 150 ms and a most preferred drift threshold is 100 ms.
[0094] A Class B A-IoT device require a bit more timing accuracy and can be assigned, for example, to basic motion sensors or alert tags. In said class category B, an acceptable drift value range is comprised between 0 to 150 ms, a preferred drift value range is between 0 to 100 ms, and a most preferred drift value range is between 30 and 80 ms.
[0095] This means that for class category B A-IoT device a drift threshold is 150 ms, a preferred drift threshold is 100 ms and a most preferred drift threshold is 80 ms.
[0096] A Class C A-IoT device operate in moderately time-sensitive scenarios and can be assigned, for example, to consumer wearables, logistics sensors, or energy usage trackers. In said class category C, an acceptable drift value range is comprised between 0 to 100 ms, a preferred drift value range is between 0 to 70 ms, and a most preferred drift value range is between 20 and 50 ms.
[0097] This means that for class category C A-IoT device a drift threshold is 100 ms, a preferred drift threshold is 70 ms and a most preferred drift threshold is 50 ms.
[0098] A Class D A-IoT device is a high-accuracy systems and can be assigned, for example, to synchronized medical monitors or real-time industrial trackers. In said class category D, an-SVT084-PC00-
[0099] acceptable drift value range is comprised between 0 to 70 ms, a preferred drift value range is between 0 to 40 ms, and a most preferred drift value range is between 10 and 30 ms. This means that for class category D A-IoT device a drift threshold is 70 ms, a preferred drift threshold is 40 ms and a most preferred drift threshold is 30 ms.
[0100] A Class E A-IoT device demand the highest synchronization precision and can be assigned, for example, to safety-critical or coordination-sensitive applications. In said class category E, an acceptable drift value range is comprised between 0 to 50 ms, a preferred drift value range is between 0 to 30 ms, and a most preferred drift value range is between 5 and 20 ms. This means that for class category E A-IoT device a drift threshold is 50 ms, a preferred drift threshold is 30 ms and a most preferred drift threshold is 20 ms.
[0101] A Signal-to-Noise Ratio, SNR, is a number that measures how clear a signal is, namely, show how much stronger is a received signal than a noise, i.e. all the interference that is superimposed on the received signal.
[0102] SNR threshold means a value within which the strength of the received signal is still considered acceptable even if affected by noise.
[0103] In general terms, a broad range covering all A-IoT devices regardless of the type of use is between 0 to 40 dB, a more realistic range for A-IoT devices suitable for working under normal conditions is between 10 to 30 dB, finally a smaller range suitable for devices operating in critical or high-precision environments is between 15 and 25 dB. The narrower the range, the more sensitive the A-IoT device is to poor signal quality and will synchronize sooner.
[0104] According to the present invention, a plurality of ranges with each of the classes is defined. For A-IoT devices that can be included in Class A, such as low-power sensors, a general range is between 0 to 20 dB, a preferred range is between 10 to 20 dB, a most preferred range is between 15 to 20 dB.
[0105] This means that for class A A-IoT device an SNR threshold suitable for marking a boundary between a signal having a poor quality and a signal with acceptable or better quality is 0 dB, preferably 10 dB, most preferably 15 dB.
[0106] For A-IoT devices that can be included in Class B, such as smart tags, a general range is between 5 to 25 dB, a preferred range is between 10 to 25 dB, a most preferred range is between 15 to 22 dB.
[0107] This means that for class B A-IoT device an SNR threshold suitable for marking a boundary between a signal having a poor quality and a signal with acceptable or better quality is 5 dB, preferably 10 dB, most preferably 15 dB.
[0108] For A-IoT devices that can be included in Class C, such as logistics beacons, a general range-SVT084-PC00-
[0109] is between 10 to 30 dB, a preferred range is between 15 to 25 dB, a most preferred range is between 18 to 24 dB.
[0110] This means that for class C A-IoT device an SNR threshold suitable for marking a boundary between a signal having a poor quality and a signal with acceptable or better quality is 10 dB, preferably 15 dB, most preferably 18 dB.
[0111] For A-IoT devices that can be included in Class D, such as medical tracking, a general range is between 15 to 35 dB, a preferred range is between 18 to 28 dB, a most preferred range is between 20 to 25 dB.
[0112] This means that for class D A-IoT device an SNR threshold suitable for marking a boundary between a signal having a poor quality and a signal with acceptable or better quality is 15 dB, preferably 18 dB, most preferably 20 dB.
[0113] Finally, for A-IoT devices that can be included in Class E, such as industrial safety device, a general range is between 20 to 40 dB, a preferred range is between 22 to 30 dB, a most preferred range is between 25 to 28 dB.
[0114] This means that for class E A-IoT device an SNR threshold suitable for marking a boundary between a signal having a poor quality and a signal with acceptable or better quality is 20 dB, preferably 22 dB, most preferably 25 dB.
[0115] Later in the document, it will be explained more in detail how these couple of threshold values are used to determine whether the device needs a synchronisation
[0116] Fig. 1 shows a schematic diagram representing a A-IoT device 10 according to an embodiment of the present invention comprising a reader unit 20, a class identifier unit 30, a decision logic unit 40, a midamble decoder unit 50, a clock alignment unit 60, an internal clock 70, a communication interface unit 80, a sleep controller unit 90 and an environment monitor unit 100.
[0117] The reader unit 20, known in the state-of-the-art for A-IoT systems, is responsible for reading a 1 -bit assist flag embedded in the R2D signal control field sent by the reader device. The assist flag specifies when an A-IoT device needs to decode the midamble signal to synchronize with the reader. Said assist flag, represented on one bit, can be set to a first value, e.g. 1, representing a request for synchronization by the reader device towards the A-loT device 10 or to a second value, e.g. 0, whereby such synchronization is not required. An assigned device class, A to E, with associated timing tolerance and interval information is stored within the class identifier unit 30, enabling a A-IoT device 10, according to the present invention, to adaptively manage a midamble interval configuration, and thus a synchronization frequency for said A-IoT device 10, based on characteristics and application context on which said device 10 is used.-SVT084-PC00-
[0118] The class category can generally be stored in the A-IoT device 10 in two ways: pre-assigning it during a device installation phase or configuring it later from the network.
[0119] The class identifier unit 30 therefore represents something completely new for A-IoT devices, since in the known art A-IoT devices there are no class-based devices, nor does the concept of class division itself that allows A-IoT devices to adapt their synchronization behaviour according to their needs and characteristics.
[0120] Whitin the class identifier unit 30 is stored a class category for said A-IoT device 10. However, it should be noted that, said class category isn’t a fixed classification and can change over the time depending on the specific characteristics owned by the A-IoT device 10 and / or and an application context in which A-IoT device 10 is used.
[0121] The term specific characteristics refers to a set of components inside the device that enable it to execute some tasks rather than others.
[0122] The decision logic unit 40, is the logical unit responsible, when the A-IoT device 10 receives a synchronization request from a reader device, for evaluating its own synchronization status to determine whether it is necessary to follow up on said synchronization request or whether said request can be ignored to avoid unnecessary energy-consuming operations.
[0123] The decision logic unit 40, although itself known in the art for A-IoT devices, according to the present invention is reconfigured to execute the new logic introduced with the class category concept to enable the implementation of a system in which A-IoT devices 10 autonomously decide when to execute a synchronization process regardless of the request made by the reader device.
[0124] As will be explained in more detail during the description of the method according to the present invention, the decision logic unit 40 will base its decision by evaluating a plurality of criteria such a value of the assist flag, its own class category, and other internal metrics to decide on synchronization.
[0125] The midamble decoder unit 50, also known in state-of-the-art for A-IoT systems, is only activated when decoding is deemed necessary.
[0126] The clock alignment unit 60, known in state-of-the-art for A-IoT systems, performs synchronization operations.
[0127] The internal clock 70 of a device is a well-known concept in the state of the art, so it will not be described further.
[0128] Similarly to the internal clock 70, the communication interface unit 80, the sleep controller unit 90 and the environment monitor unit 100 are also well-known A-IoT device components of a state-of-the-art that respectively have the function of proceeds to communication when a synchronization is valid, puts an A-IoT device to sleep when a communication is ended-SVT084-PC00-
[0129] and for measuring ambient drift factors and A-IoT device internal clock stability.
[0130] Fig. 2 shows a flowchart of a method 200 for adaptive synchronization of a A-IoT device 10 in an awakened state according to the present invention. This method 200 allows said A-IoT devices 10 to adapt a synchronization frequency based on its own internal status.
[0131] Method 200 involves an initial step 210 in which the reader unit 20 reads a value of an assist flag comprised in a R2D signal sent by a reader device.
[0132] In a second step 220, not shown in Fig. 2, the class identifier unit 30 reads a pre-assigned class category set for the A-IoT device 10.
[0133] Said class category comprising a first value representing a time interval, a second value representing a time tolerance wherein said first and second value defining a time window within which the A-IoT device 10 is suitable for receiving an R2D signal without requiring synchronization.
[0134] This time window allows the A-IoT device 10 to determine whether the timing of its internal clock 70 is still sufficiently aligned with the timing of the reader device.
[0135] Below are two clarifying examples for two d A-IoT devices 10 to which a different a class category has been assigned.
[0136] In a first example is involved a class-based A-IoT static device 10 to which class A has been pre-assigned and assuming that the said device 10 knows that an R2D signal is sent every 100 ms. For a class A A-IoT device 10, there is only a timing tolerance of ±50 ps, in which said time tolerance is a limit of how far off the internal clock 70 can be without causing problems.
[0137] The class-based A-IoT static device 10 keeps track of time using its internal clock 70 and, when 100 ms passes, it expects the next R2D signal. If the signal arrives in the expected time, comprising also the time tolerance value, said device 10 assumes its timing is still accurate and skips the midamble synchronization.
[0138] In a second example is involved a class-based A-IoT static device 10 to which class D has been pre-assigned and assuming that the said device 10 knows that an R2D signal is sent again every 100 m but said class D A-IoT device 10 has more clock drift due, for example, to motion or temperature. When said A-IoT device 10 checks the signal arrival and sees it came too early or too late than the expected receiving time value calculated by the internal clock 70 falling outside time window defined by the time interval and the time tolerance provided for class D.
[0139] This means that the internal clock 70 has shifted and therefore at least the midamble signal will be decoded to perform the synchronization.
[0140] The examples described above describe what the decision logic unit 40 does in a third step-SVT084-PC00-
[0141] 230 of the method 200 to check whether a first condition for which a synchronisation request from the reader device must be fulfilled.
[0142] In particular the decision logic unit 40 check if the assist flag value of a R2D received signal is set to a value representative of a synchronization request from the reader device and one between a receiving time of said R2D signal and an expected receiving time value calculated by the internal clock 70 falls outside the time window defined for the class category preassigned to the A-IoT device 10.
[0143] In a fourth step 240, not shown in Fig. 2, the environment monitor unit 100 is configured to determine a drift value of the internal clock 70 based on the expected receiving time value calculated considering an internal timing history of last awakenings.
[0144] In general, each time the A-IoT device 10 wake-up checks when the last synchronization signal arrived and when said A-IoT device 10 expected it to arrive in order to store a difference, drift, of timing between said two signals, specifically storing only a certain number of last differences, e.g. the last 3 to 5. This plurality of stored differences represents the internal timing history.
[0145] The wider the internal timing historical depth of last awakenings considered, the greater the accuracy of the calculated drift value. For example, the A-IoT device 10 may store the drift values of the last 3-5 or even 3-10 wake-up events, depending on the application and preassigned class of the A-IoT device 10.
[0146] In a first short history example, a A-IoT device 10 only stores the last 3 times it woke up. If one of those timings was affected by temporary interference, the calculated drift might not be very reliable.
[0147] In a second long history example, another A-IoT device 10 stores the last 10 wake-up times. This helps smooth out any anomalies (e.g., noise, rare signal delays), so the drift value it calculates is more accurate and stable.
[0148] The number of stored values, namely a depth of said internal timing history of last awakenings, depends on the context scenario in which said A-IoT device 10 is used. For a simple temperature sensor, a short history (3-5) may be enough whereas for a critical industrial monitor, a longer history (up to 10 or more) improves reliability and avoids false triggering of re-synchronization.
[0149] According to the present invention, the number of stored values also depends on the class category reflecting the accuracy required by the different categories of A-IoT devices 10 and their uses.
[0150] A Class A A-IoT device 10, like a low-cost temperature tag used for basic ambient monitoring, may only store the last 3 to 4 events. Since it doesn’t need high precision and is-SVT084-PC00-
[0151] built for low energy consumption, a small drift is acceptable.
[0152] A Class B A-IoT device 10, like a smart label for tracking inventory in retail, may store 4 to 6 past awakenings. These devices need more regular timing to stay in sync with check-in systems but are not overly sensitive to small drift.
[0153] A Class C A-IoT device 10, such as a wearable patch that monitors heart rate periodically, may use 5 to 7 events in its history. These devices need more accurate internal timing to ensure consistent health data capture.
[0154] A Class D A-IoT device 10, such as an industrial sensor used for detecting vibration anomalies in machines may use 7 to 9 wake events. It needs to detect and report events with higher timing accuracy to avoid errors or missed alerts.
[0155] A Class E A-IoT device 10 used in time-critical tracking applications, e.g. logistics handoffs or transport synchronization, may keep a history of 8 to 10 or more past awakenings. These devices require precise drift correction and minimal synchronization error due to tight timing demands.
[0156] In a fifth step 250, not shown in Fig. 2, the environment monitor unit 100 is also configured to determine a SNR value representative of a SNR signal quality. The environment monitor unit 100 listens to the incoming R2D signal, measures a signal strength and compare it to a background noise. The environment monitor unit 100 therefore calculates the Signal -to-Noise Ratio (SNR) by dividing the signal strength by the noise level. The higher the SNR, the better the quality. A higher SNR, e.g. 30 dB, means better signal quality, while a lower SNR, e.g. 10 dB, indicates a poor signal.
[0157] In a sixth step 260, the decision logic unit 40 check whether a second condition for which a synchronisation request from the reader device must be fulfilled. In particular, the decision logic unit 40 check if the drift value exceeding the drift threshold and / or the SNR value is lower than the SNR threshold defined for the class category pre-assigned to the A-IoT device 10.
[0158] The synchronization request made by the reader device must be satisfied when at least one of the two criteria of the second condition is verified because the internal clock 70 of the A-loT device 10 is shifted out of alignment with the timing of the reader device.
[0159] In a seventh step 270, when at least one of the first and second conditions has occurred, the A-IoT device 10 proceeds to perform a midamble decoding and internal clock 70 synchronization.
[0160] This step 270, shown as a single step in the diagram in Fig. 2, consists of a plurality of substeps.
[0161] In a first sub-steps 270a, the midamble decoder unit 50 decode the SIP signal portion suitable-SVT084-PC00-
[0162] for synchronizing the internal clock 70 of the A-IoT device 10 according to a first level of accuracy.
[0163] In particular, the clock alignment unit 60 verifies the accuracy of the signal obtained because of decoding the SIP signal portion by comparing it with an internal threshold value, which may be defined according to several factors comprising, for example, ambient signals and the typology of application task that must be executed.
[0164] Internal threshold values, which set the limit for acceptable timing errors, depend on the application. For example, a temperature logger might accept a timing error of ±50-100 ps, while a more precise system, like an industrial controller, might only allow ±5-10 ps. If the comparison shows an insufficient accuracy, the clock alignment unit 60 proceeds with the decoding of the CAP signal portion as described below in a second sub-steps 270a of the seventh step 270. The accuracy evaluation is dynamic and based on specific requirements for several applications
[0165] The need for timing accuracy after decoding the SIP signal depends on the application. Some devices can work with basic, coarse timing, while others need very precise synchronization. For example, a sensor only needs low accuracy because it just reports fill levels occasionally, so it uses SIP for coarse synchronization. An environmental monitor (like for humidity) also needs low to moderate accuracy, as it can tolerate some timing drift. A wearable fitness tracker may need moderate accuracy, syncing periodically, and might use SIP or CAP if extra precision is required. A motion-triggered light controller may need moderate to high accuracy if it’s part of a larger system and could require CAP for better sync. A smart grid power monitor requires high accuracy to align power phases, so it uses both SIP and CAP for precise timing. Finally, a robotic actuator needs very high accuracy, always using CAP to ensure tight coordination between parts.
[0166] When additional timing accuracy is required for starting / resuming a communication that the first level of accuracy achievable with the SIP portion decoding cannot satisfy, in a second sub-step 270b of said of the seventh step 270 the clock alignment unit 60 identifies the CAP signal portion comprised in said midamble signal transmitted by the reader device.
[0167] Subsequently in a third sub-step 270c the midamble decoder unit 50 decode the CAP signal portion suitable for synchronizing the internal clock 70 of the A-IoT device 10 according to a second level of accuracy.
[0168] Once the appropriate portion of the midamble signal has been decoded to ensure the correct level of timing accuracy, in fourth sub-step 270d the clock alignment unit 60 synchronizes the internal clock 70 of said A-IoT device 1010 with a timing of said reader device.
[0169] On the other hand, when neither of the first and second conditions is met, the A-IoT device-SVT084-PC00-
[0170] 10 does not perform any internal clock 70 synchronization skipping the step 270 described above.
[0171] Once the awakened A-IoT device 10 has finished its checks for the purpose of determining whether the synchronization request sent by the reader device should be executed or not, in an eighth step 280 the communication interface unit 80 starts or continues a communication with a networked device, preferably the reader device.
[0172] At the end of the communication step 280, in a ninth step 290 the sleep controller unit 90 generates a status signal suitable for switching the A-IoT device 10 from the active state into a deep-sleep state.
[0173] The present invention overcomes the highlighted issues of the prior art that cannot be solved by the known solutions.
[0174] The method 200 of the present invention, avoiding periodic synchronization of the A-IoT device 10 minimizing the energy required to perform the synchronization process through an adaptive midamble interval configuration management, and thus a synchronization frequency, for the class-based A-IoT device 10 based on characteristics and application context in which the A-IoT device 10 is used.
[0175] The present description has tackled some of the possible variants, but it will be apparent to the man skilled in the art that other embodiments may also be implemented, wherein some elements may be replaced with other technically equivalent elements. The present invention is not therefore limited to the explanatory examples described herein, but may be subject to many modifications, improvements or replacements of equivalents parts and elements without departing from the basic inventive idea, as set out in the following claims.
Claims
-SVT084-PC00-CLAIMS1. Ambient Internet of Things, A-IoT, device (10) configured to operate in anA-IoT network system in which said A-IoT device (10) receives an R2D signal from a reader device, the A-IoT device (10) comprising:a reader unit (20) configured to read an assist flag comprised in saidR2D signal;a class identifier unit (30) configured to memorize a pre-assigned classcategory comprising a time interval value representing a frequency at which said A-IoT device (10) expects to receive a midamble signal in order to maintain a reliable time alignment with a reference clock within acceptable limits and a time tolerance value representative of a tolerable shift from said time interval before a synchronisation is required; an internal clock (70);a decision logic unit (40) functionally connected to said reader unit (20), saidclass identifier unit (30) and said internal clock (70),wherein said decision logic unit (40) is configured to:reading (210), by means of the reader unit (20), a value of the assist flag comprised in said R2D signal;retrieving (220), via the class identifier unit (30), said time interval value and said time tolerance value from said class category to define a time window within which the A-IoT device (10) is configured for skipping a synchronization request contained in said R2D signal;checking (230) a first condition, by means of the decision logic unit (40), in which said assist flag value is set to a value representative of a synchronization request and one between a receiving time of said R2D signal and an expected receiving time value calculated by the internal clock (70) falls outside the time window defined by said class category for said A-IoT device (10).
2. The A-IoT device (10) according to claim 1, wherein said A-IoT device (10) further comprises an environment monitor unit (100) functionally connected to said decision logic unit (40) and said pre-assigned class category further comprises a drift threshold value representing a number of timing errors handled by the A-IoT device (10) due to a different operating speed of the internal clock (70) compared to the reference clock and a SNR threshold value representing a strength of the received midamble signal, said decision logic unit (40) is further configured to:determining (240), by means of the environment monitor unit (100), a drift value based on said expected receiving time value calculated by said internal clock (70)-SVT084-PC00-considering an internal timing history of last awakenings;determining (250), by means of said environment monitor unit (100), a SNR value representative of a SNR signal quality;checking (260) a second condition, by means of the decision logic unit (40), in which said drift value exceeding said drift threshold and / or said SNR value is lower than said SNR threshold;3. The A-IoT device (10) according to claim 1 and 2, wherein said A-IoT device (10) further comprises a midamble decoder unit (50) configured to decode a midamble signal comprised in said R2D signal and a clock alignment unit (60) configured to synchronize the internal clock (70) with a decoded midamble signal, said midamble decoder unit (50) and clock alignment unit (60) functionally connected to said decision logic unit (40) and said decision logic unit (40), when at least one of the first and second conditions is satisfied decision is further configured to:decoding (270a), by means of the midamble decoder unit (50), a Start Indicator Part, SIP, portion of said midamble signal suitable for synchronizing the internal clock (70) according to a first level of accuracy;synchronizing (270d), via said clock alignment unit (60), the internalclock (70) with a timing of said reader device.
4. The A-IoT device (10) according to claim 3, wherein when at least one applicationlayer communication requires an accuracy level higher than the first accuracy level said decision logic unit (40) is further configured to:identifying (270b), by means of the midamble decoder unit (50), a Clock Acquisition Part, CAP, portion, said CAP portion comprised in said midamble signal transmitted by said reader device;decoding (270c), by means of the midamble decoder unit (50), said CAP portion signal suitable for synchronizing the internal clock (70) with a timing of said reader device according to a second level of accuracy;synchronizing (270d), via said clock alignment unit (60), the internalclock (70) with a timing of said reader device.
5. The A-IoT device (10) according to claim3 or 4, wherein said A-IoT device (10) further comprises a communication interface unit (80) and a sleep controller unit (90) functionally connected to said decision logic unit (40), said decision logic unit (40) is further configured to:initiating or resuming (280), via said communication interface unit (80), at least an application-layer communication to a networked device, preferably the reader-SVT084-PC00-device.generating (290), by means of said sleep controller unit (90), a status signal suitable for switching said A-IoT device (10) into a deep-sleep state.
6. The A-IoT device (10) according to any of the preceding claims, wherein an internal threshold value, used by said midamble decoder unit (50) to verifies the level of accuracy of a signal obtained decoding the SIP portion signal required by said at least one communication, is between ±50-100 ps and ±5-10 ps.
7. The A-IoT device (10) according to any of the preceding claims, wherein said SNR threshold for class A A-IoT device (10) is 0 dB, preferably 10 dB, most preferably 15 dB; for class B A-IoT device (10) is 5 dB, preferably 10 dB, most preferably 15 dB; for class C A-IoT device (10) is 10 dB, preferably 15 dB, most preferably 18 dB; for class D A-IoT device (10) is 15 dB, preferably 18 dB, most preferably 20 dB; for class E A-IoT device (10) is 20 dB, preferably 22 dB, most preferably 25 dB.
8. The A-IoT device (10) according to any of the preceding claims, wherein said drift threshold for class A A-IoT device (10) is 200 ms, preferably 150 ms, most preferably 100 ms; for class B A-IoT device (10) is 150 ms, preferably 100 ms, most preferably 80 ms; for class C A-IoT device (10) is 100 ms, preferably 70 ms, most preferably 50 ms; for class D A-IoT device (10) is 70 ms, preferably 40 ms, most preferably 30 ms; for class E A-IoT device (10) is 50 ms, preferably 30 ms, most preferably 20 ms;9. The A-IoT device (10) according to any of the preceding claims, wherein a depth of said internal timing history of last awakenings depends comprises a plurality of value depending on said class category pre-assigned to said A-IoT device (10), in particular A-class category A-IoT device (10) comprises a historical depth of 3 to 4 events, B-class category A-IoT device (10) comprises a historical depth of 4 to 6 events, a C-class category A-IoT device (10) comprises a historical depth of 5 to 7 events, a D-class category A-IoT device (10) comprises a historical depth of 7 to 9 events, a E-class category A-IoT device (10) comprises a historical depth of 8 to 10 events.
