A low-power narrowband internet of things communication chip method and system

Abstract

The application relates to the technical field of wireless communication, and particularly discloses a low-power-consumption narrowband Internet of Things communication chip method and system. The method encodes a coverage level index implicitly in a physical layer feature field of an initial access response, embeds a terminal identity fingerprint in an uplink preamble structure, maps a retransmission progress into a resolvable power sequence, deposits a link quality trend into a calibration offset in a downlink response, and is supplemented with a double-track buffer and state synchronization mechanism, so that the base station does not need to add any scheduling signaling, can track terminal coverage level changes throughout the process, accurately deduce the repetition intention, adaptively release redundant resources, and smoothly transition parameter configuration. Thus, under the premise of keeping the original signaling overhead unchanged, the technical problem that the base station cannot dynamically identify and adapt to the repetition transmission requirements of terminals with different coverage levels is solved, and terminal-side power consumption control and system-level spectrum efficiency are taken into account.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, specifically to a low-power narrowband Internet of Things (IoT) communication chip method and system. Background Technology

[0002] In narrowband IoT systems, terminals are commonly deployed in weak coverage scenarios such as underground parking garages, pipe shafts, and deep within warehouse shelves. Uplink speeds are limited by transmit power and receive sensitivity, often requiring repeated transmission of the same data packet to improve the base station's demodulation success rate. Existing technologies typically employ two methods to achieve repeated adaptation: one is for the base station to explicitly indicate the number of repetitions required for each uplink transmission via dedicated downlink signaling; the other is for the terminal to automatically select the number of repetitions based on a pre-configured coverage level table. However, once this level is determined during initial access, it remains fixed and cannot respond to slow changes in link quality.

[0003] The above-mentioned solutions have a prominent technical problem: without increasing uplink scheduling signaling overhead, base stations cannot dynamically identify the actual repetition requirements of terminals with different coverage levels, nor can they adjust adaptation strategies in a timely manner according to changes in the wireless environment. Specifically, while explicit signaling can provide precise control, each round of repetition configuration requires independent scheduling authorization, significantly increasing the load on the control channel; while static level binding saves signaling, it causes terminals to continue transmitting at high repetition rates even after coverage improves, increasing power consumption and interference, and then frequently failing due to insufficient repetition after coverage deteriorates. Neither approach can achieve both low power consumption and high reliability. Summary of the Invention

[0004] This invention aims to provide a low-power narrowband IoT communication chip method and system, which solves the technical problem that base stations cannot dynamically identify and adapt to the repeated transmission needs of terminals with different coverage levels, while taking into account both terminal-side power consumption control and system-level spectrum efficiency.

[0005] To achieve the above objectives, the technical solution adopted by this invention is: a low-power narrowband Internet of Things (IoT) communication chip method, comprising:

[0006] The terminal determines the coverage level index based on the physical layer feature field in the initial access response and selects the corresponding initial value for the number of repetitions. Based on the determined coverage level index, the terminal embeds the coverage level fingerprint sequence in the first uplink transmission.

[0007] Based on the initial value of the selected number of repetitions, the terminal dynamically adjusts the power bias based on the current remaining number of repetitions during each retransmission.

[0008] After receiving the uplink signal, the base station identifies the terminal coverage level by parsing the coverage level fingerprint sequence, and inverts the terminal retransmission progress based on the received signal power sequence formed by the power offset. Combining the inverted terminal retransmission progress and the demodulation soft information distribution characteristics, the base station determines whether the terminal should terminate the repeated transmission prematurely.

[0009] After receiving the downlink confirmation signal from the base station, the terminal calculates the soft information entropy value and decides whether to trigger the coverage level reassessment process based on the comparison result between the entropy value and the preset threshold.

[0010] In response to the coverage level reassessment process triggered by the terminal, the base station constructs a coverage level drift trend model based on the statistics of the terminal's historical retransmission behavior and injects the calibration offset into the downlink response;

[0011] After receiving the downlink response of the injected calibration offset, the terminal starts a dual-track buffering mechanism when the new coverage level index takes effect to achieve a smooth transition between the old and new parameters.

[0012] Based on the operational results of the dual-track buffer mechanism, the base station and the terminal jointly maintain the coverage level status synchronization flag, and achieve bidirectional verification through redundant bits in the downlink response and the uplink power sequence.

[0013] Preferably, the terminal determines the coverage level index and selects the corresponding initial value for repetition count based on the physical layer feature field in the initial access response, including:

[0014] Read the 3-bit physical layer feature field from the initial access response message, and convert the 3-bit physical layer feature field into a decimal integer as the coverage level index;

[0015] Based on the coverage level index, look up the local fixed mapping table to get the corresponding initial value of the repetition count, and assign the initial value of the repetition count to the repetition counter of the uplink transmission task;

[0016] Based on the initial value of the repetition count, adjust the puncture position and interleaving depth of the convolutional code in the physical layer coding link.

[0017] Preferably, based on a determined coverage level index, the terminal embeds a coverage level fingerprint sequence in its initial uplink transmission, including:

[0018] Using the coverage level index as a seed, a binary fingerprint sequence of length 12 is generated, and the binary fingerprint sequence is mapped to a BPSK symbol stream on the real axis;

[0019] The BPSK symbol stream is superimposed on the subcarrier preceding the demodulation reference signal of the physical uplink shared channel;

[0020] After receiving the first uplink signal, the base station extracts the complex response value at the subcarrier and determines the coverage level index by comparing it with the locally stored fingerprint sequence.

[0021] Preferably, based on the selected initial value of the number of repetitions, the terminal dynamically adjusts the power bias based on the current remaining number of repetitions during each retransmission, including:

[0022] After each transmission, decrement the repetition counter by 1 to obtain the remaining number of repetitions;

[0023] Based on the remaining number of repetitions, look up the local power bias table to obtain the current power increment;

[0024] The power increment is converted into a linear domain multiplier and applied to the amplitude of the symbol to be transmitted;

[0025] The base station measures the average received power of multiple received signals for the same uplink transmission task, and uses power sequence normalization to infer the terminal retransmission progress.

[0026] Preferably, by combining the inverted terminal retransmission progress and demodulation soft information distribution characteristics, the base station determines whether the terminal prematurely terminates the repeated transmission, including:

[0027] Continuously monitor the received power sequence of the same uplink transmission task, and record the candidate termination point when the expected power value is not detected at the corresponding time-frequency position;

[0028] Channel equalization and symbol demapping are performed on the received signal to obtain soft bit information;

[0029] Calculate the standard deviation of each bit position in the soft bit information across multiple receptions, and count the percentage of the standard deviation that is less than a preset threshold.

[0030] The percentage of cases with a standard deviation less than a preset threshold is compared with a preset information threshold. When the preset information threshold is met, unused resources are released.

[0031] Preferably, after receiving a downlink confirmation signal from the base station, the terminal calculates a soft information entropy value and determines whether to trigger a coverage level reassessment process based on a comparison between the entropy value and a preset threshold, including:

[0032] Extract the soft bit information from the downlink channel demodulation output and calculate the information entropy of the soft bit information;

[0033] The calculated information entropy is compared with the reference entropy threshold. When the information entropy is greater than the reference entropy threshold plus the tolerance margin, it is determined that the coverage level needs to be reassessed.

[0034] In the next planned uplink data transmission, a re-evaluation request identifier is embedded by modifying the cyclic shift value of the physical layer data block start position.

[0035] Preferably, in response to the coverage level reassessment process triggered by the terminal, the base station constructs a coverage level drift trend model based on the terminal's historical retransmission behavior statistics and injects the calibration offset into the downlink response, including:

[0036] Maintain the terminal's historical behavior table, recording the actual number of repetitions, theoretical initial value, power sequence interruption point number, and average soft information entropy of the terminal's most recent N uplink transmissions;

[0037] Calculate the repetition error rate, interruption shift rate, and entropy growth slope. Then, sum the repetition error rate, interruption shift rate, and entropy growth slope by weight to obtain the overall drift amount.

[0038] The overall drift is mapped to an integer calibration offset and injected into the physical layer feature field of the downlink response.

[0039] Preferably, after receiving the downlink response of the injected calibration offset, the terminal initiates a dual-track buffering mechanism when the new coverage level index takes effect to achieve a smooth transition between the old and new parameters, including:

[0040] Store the new coverage level index and the current coverage level index together in the dual-track register, and mark them as the main track and the auxiliary track, respectively;

[0041] In subsequent uplink transmissions, the transmission success rate is recorded, and the weights of the primary and secondary tracks are updated based on the success rate.

[0042] When initiating uplink transmission, the initial value of the repetition count and the power offset are weighted and synthesized.

[0043] Based on the continuous transmission success rate, a decision is made as to whether to upgrade the auxiliary rail to the main rail and complete the track switching.

[0044] Preferably, based on the operational results of the dual-track buffer mechanism, the base station and the terminal jointly maintain the coverage level status synchronization flag, and achieve bidirectional verification through redundant bits in the downlink response and the uplink power sequence, including:

[0045] After the initial access is completed, initialize the state synchronization flag to the same value;

[0046] After each successful uplink transmission, each entity increments its status synchronization flag cyclically.

[0047] The base station carries the current state synchronization flag in the downlink response, and the terminal compares its own flag with the base station flag;

[0048] During uplink transmission, the terminal maps the state synchronization flag to the starting bias of the power sequence for the base station to verify.

[0049] On the other hand, the present invention proposes a low-power narrowband IoT communication chip system, comprising:

[0050] The coverage level determination unit is used to determine the coverage level index and select the corresponding initial value of the number of repetitions based on the physical layer feature fields in the initial access response.

[0051] A fingerprint embedding unit is used to embed a coverage level fingerprint sequence in the first uplink transmission based on a determined coverage level index;

[0052] The power adjustment unit is used to dynamically adjust the power bias based on the current remaining number of repetitions during each retransmission.

[0053] The coverage level identification unit is used to identify the terminal's coverage level by parsing the coverage level fingerprint sequence after receiving the uplink signal.

[0054] The transmission progress inversion unit is used to invert the terminal retransmission progress based on the received signal power sequence formed by the power offset.

[0055] The resource release unit is used to determine whether the terminal should terminate repeated transmission prematurely by combining the inverted terminal retransmission progress and the distribution characteristics of demodulation soft information.

[0056] The coverage status assessment unit is used to calculate the soft information entropy value after receiving the downlink confirmation signal from the base station, and decide whether to trigger the coverage level reassessment process based on the comparison result of the entropy value and the preset threshold.

[0057] The drift trend modeling unit is used to build a coverage level drift trend model based on the statistics of the terminal's historical retransmission behavior, and inject the calibration offset into the downlink response;

[0058] The dual-track buffer unit is used to activate the dual-track buffer mechanism when the new coverage level index takes effect after receiving the downlink response of the injected calibration offset, so as to achieve a smooth transition between the old and new parameters.

[0059] The status synchronization unit is used to jointly maintain the coverage level status synchronization flag with the terminal, and achieves bidirectional verification through redundant bits in the downlink response and the uplink power sequence.

[0060] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0061] This invention implicitly encodes the coverage level index into the physical layer feature field of the initial access response, embeds the terminal identity fingerprint into the uplink preamble structure, maps the retransmission progress into a resolvable power sequence, and precipitates the link quality trend as a calibration offset in the downlink response. Supplemented by a dual-track buffering and state synchronization mechanism, this allows the base station to track terminal coverage level changes throughout the entire process, accurately infer its retransmission intent, adaptively release redundant resources, and smoothly transition parameter configurations without adding any new scheduling signaling. Thus, while maintaining the original signaling overhead, it solves the technical problem of the base station's inability to dynamically identify and adapt to the retransmission needs of terminals with different coverage levels, while also considering terminal-side power consumption control and system-level spectral efficiency. Attached Figure Description

[0062] Figure 1 This is a flowchart of the low-power narrowband IoT communication chip method of the present invention;

[0063] Figure 2 This is a block diagram of the low-power narrowband IoT communication chip system of the present invention. Detailed Implementation

[0064] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0065] like Figure 1 As shown, this invention proposes a low-power narrowband IoT communication chip method, aiming to enable base stations to dynamically identify and adapt to the repetitive transmission needs of terminals with different coverage levels without increasing uplink scheduling signaling overhead. The method relies on the synergy of the terminal-side autonomous response capability and the base station-side implicit resolution capability. Through the distinguishable design of the physical layer signal structure, coverage level binding of time-frequency resource mapping relationships, contextual continuity modeling of retransmission behavior, and a reverse feedback closed loop of demodulation decision results, a self-consistent, lightweight, and scalable adaptation path is formed. Specifically, it includes the following steps:

[0066] The terminal determines the coverage level index and selects the corresponding initial value of repetition count based on the physical layer feature field in the initial access response. Specifically, this includes: reading a 3-bit physical layer feature field from the initial access response message and converting the 3-bit physical layer feature field into a decimal integer as the coverage level index; looking up the local fixed mapping table based on the coverage level index to obtain the corresponding initial value of repetition count and assigning the initial value of repetition count to the repetition counter of the uplink transmission task; and adjusting the puncturing position and interleaving depth of the convolutional code in the physical layer coding link based on the initial value of repetition count.

[0067] It achieves zero signaling coupling between coverage level identification and repeated configuration, enabling the terminal to establish adaptation parameters the moment the first access response is completed, avoiding waiting for additional scheduling instructions, significantly shortening the uplink transmission start-up delay, and ensuring from the source that the number of repetitions strictly matches the actual coverage capability.

[0068] Based on the determined coverage level index, the terminal embeds a coverage level fingerprint sequence in the initial uplink transmission; specifically, this includes: generating a binary fingerprint sequence of length 12 using the coverage level index as a seed, mapping the binary fingerprint sequence to a BPSK symbol stream on the real axis; superimposing the BPSK symbol stream onto the subcarrier before the demodulation reference signal of the physical uplink shared channel; after receiving the initial uplink signal, the base station extracts the complex response value at the subcarrier and determines the coverage level index by comparing it with the locally pre-stored fingerprint sequence.

[0069] This allows base stations to unambiguously identify terminal coverage levels based solely on the first uplink reception, without needing to parse higher-layer signaling or wait for retransmissions, thus providing a deterministic basis for subsequent resource scheduling and power sequence analysis.

[0070] Based on the initial value of the selected repetition count, the terminal dynamically adjusts the power bias based on the current remaining repetition count during each retransmission. Specifically, this includes: decrementing the repetition counter by 1 after each transmission to obtain the remaining repetition count; looking up the local power bias table based on the remaining repetition count to obtain the current power increment; converting the power increment into a linear domain multiplier and applying it to the amplitude of the symbol to be transmitted; and measuring the average received power of multiple received signals for the same uplink transmission task at the base station, and using power sequence normalization processing to infer the terminal's retransmission progress.

[0071] This allows the base station to accurately determine the current retransmission sequence of the terminal simply by receiving the relative changes in the power sequence, without relying on explicit retransmission counting signaling, thus achieving implicitly decodable repetition progress.

[0072] After receiving the uplink signal, the base station identifies the terminal's coverage level by parsing the coverage level fingerprint sequence and inverts the terminal's retransmission progress based on the received signal power sequence formed by the power offset. Combining the inverted terminal retransmission progress and the distribution characteristics of demodulated soft information, the base station determines whether the terminal should prematurely terminate repeated transmissions. Specifically, this includes: continuously monitoring the received power sequence of the same uplink transmission task; recording candidate termination points when the expected power value is not detected at the corresponding time-frequency position; performing channel equalization and symbol demapping on the received signal to obtain soft bit information; calculating the standard deviation of each bit position in the soft bit information in multiple receptions and statistically analyzing the proportion of standard deviations less than a preset threshold; comparing the proportion of standard deviations less than the preset threshold with a preset signal threshold; and releasing unused resources when the preset signal threshold is met.

[0073] This enables base stations to autonomously, promptly, and reliably identify repeated premature termination behaviors without relying on terminal feedback or additional signaling, and to dynamically reclaim unused resources, thereby improving system resource utilization.

[0074] After receiving the downlink confirmation signal from the base station, the terminal calculates the soft information entropy value and determines whether to trigger the coverage level reassessment process based on the comparison result between the entropy value and a preset threshold. Specifically, this includes: extracting the soft bit information output from the downlink channel demodulation and calculating the information entropy of the soft bit information; comparing the calculated information entropy with a reference entropy threshold; when the information entropy is greater than the reference entropy threshold plus a tolerance margin, it is determined that the coverage level needs to be reassessed; and in the next planned uplink data transmission, embedding a reassessment request identifier by modifying the cyclic shift value of the physical layer data block start position.

[0075] This allows terminals to initiate coverage level reassessment independently based solely on their own demodulation confidence, avoiding repeated configuration mismatches caused by fixed levels and improving the balance between power consumption and reliability during long-term operation.

[0076] In response to the coverage level reassessment process triggered by the terminal, the base station constructs a coverage level drift trend model based on the terminal's historical retransmission behavior statistics and injects the calibration offset into the downlink response. Specifically, this includes: maintaining a terminal historical behavior table, recording the actual number of repetitions completed in the terminal's most recent N uplink transmissions, theoretical initial values, power sequence interruption point numbers, and the average soft information entropy; calculating the repetition deviation rate, interruption shift rate, and entropy growth slope, and weighted summing the repetition deviation rate, interruption shift rate, and entropy growth slope to obtain the comprehensive drift; mapping the comprehensive drift to an integer calibration offset and injecting it into the physical layer feature field of the downlink response.

[0077] This enables base stations to extract coverage change trends from the historical behavior of terminals, rather than relying solely on a single measurement, thereby improving the robustness and anti-interference capabilities of grade calibration and avoiding misadjustments caused by instantaneous channel fluctuations.

[0078] After receiving the downlink response of the injected calibration offset, the terminal initiates a dual-track buffering mechanism when the new coverage level index takes effect to achieve a smooth transition between the old and new parameters. Specifically, this includes: storing the new coverage level index and the current coverage level index together in the dual-track register, marking them as the primary track and secondary track respectively; recording the transmission success rate in subsequent uplink transmissions and updating the primary and secondary track weights based on the success rate; weighting and synthesizing the initial value of the repetition count and the power offset when initiating uplink transmission; and determining whether to promote the secondary track to the primary track based on the continuous transmission success rate, thus completing the track switching.

[0079] To avoid the risk of parameter mutations caused by coverage level switching, a reliable transition path is established between the old and new configurations of the terminal, ensuring the continuity of the link adaptation process and the reliability of services.

[0080] Based on the operational results of the dual-track buffer mechanism, the base station and the terminal jointly maintain the coverage level status synchronization flag, and achieve bidirectional verification through redundant bits in the downlink response and the uplink power sequence. Specifically, this includes: after initial access is completed, the status synchronization flag is initialized to the same value; after each successful uplink transmission, each terminal increments its status synchronization flag cyclically; the base station carries the current status synchronization flag in the downlink response, and the terminal compares its own flag with the base station flag; during uplink transmission, the terminal maps the status synchronization flag to the starting offset of the power sequence for the base station to perform inversion verification.

[0081] Without increasing signaling overhead, the terminal and base station coverage level configuration status can be aligned with zero drift over a long period of time, eliminating the risk of status discrepancies caused by lost responses or decoding errors.

[0082] On the other hand, this invention proposes a low-power narrowband IoT communication chip system, such as... Figure 2 As shown, it includes:

[0083] The coverage level determination unit is used to determine the coverage level index and select the corresponding initial value of the number of repetitions based on the physical layer feature fields in the initial access response.

[0084] A fingerprint embedding unit is used to embed a coverage level fingerprint sequence in the first uplink transmission based on a determined coverage level index;

[0085] The power adjustment unit is used to dynamically adjust the power bias based on the current remaining number of repetitions during each retransmission.

[0086] The coverage level identification unit is used to identify the terminal's coverage level by parsing the coverage level fingerprint sequence after receiving the uplink signal.

[0087] The transmission progress inversion unit is used to invert the terminal retransmission progress based on the received signal power sequence formed by the power offset.

[0088] The resource release unit is used to determine whether the terminal should terminate repeated transmission prematurely by combining the inverted terminal retransmission progress and the distribution characteristics of demodulation soft information.

[0089] The coverage status assessment unit is used to calculate the soft information entropy value after receiving the downlink confirmation signal from the base station, and decide whether to trigger the coverage level reassessment process based on the comparison result of the entropy value and the preset threshold.

[0090] The drift trend modeling unit is used to build a coverage level drift trend model based on the statistics of the terminal's historical retransmission behavior, and inject the calibration offset into the downlink response;

[0091] The dual-track buffer unit is used to activate the dual-track buffer mechanism when the new coverage level index takes effect after receiving the downlink response of the injected calibration offset, so as to achieve a smooth transition between the old and new parameters.

[0092] The status synchronization unit is used to jointly maintain the coverage level status synchronization flag with the terminal, and achieves bidirectional verification through redundant bits in the downlink response and the uplink power sequence.

[0093] In addition, each unit in the above system is also used to implement other steps of the aforementioned low-power narrowband IoT communication chip system, as follows:

[0094] Step 1: Based on the physical layer characteristic fields in the initial access response, the terminal autonomously determines its coverage level and selects the corresponding initial value for the number of repetitions.

[0095] This step begins with the initial access response message received by the terminal after completing the random access procedure. This message is sent by the base station and, in addition to the usual timing advance, uplink resource allocation, and temporary identifier, also embeds a 3-bit physical layer feature field. This field is not an independently defined signaling field, but rather reuses a reserved cyclic prefix extension bit within the PRACH response window that was not used in the original protocol, achieving information carrying without additional overhead while maintaining frame structure compatibility.

[0096] When decoding the response message, the terminal first extracts the 3-bit field and interprets it as an integer. This integer is the coverage level index. It should be noted that this index does not directly correspond to the link budget difference, but reflects the discrete level divided by the base station based on comprehensive criteria such as the strength of the terminal's last random access preamble sequence, arrival time offset, and multipath widening. Its division boundary has been preset on the base station side and is consistent with the network-wide coverage enhancement configuration table.

[0097] The terminal locates and reads the 3-bit physical layer feature field from the initial access response message, and converts the binary value of the field into a decimal integer. This serves as a coverage level index; the index ranges from 0 to 7, with a total of 8 levels, where 0 represents the strongest coverage and 7 represents the weakest coverage. This conversion process does not introduce any external table lookups or additional processing delays; it is entirely completed synchronously by the baseband circuitry at the end of the response message parsing pipeline.

[0098] Terminal index based on known coverage level Query the local fixed mapping table to obtain the initial value of the repetition count for the corresponding level. The mapping table has 8 items, each of which is a non-negative integer and satisfies a monotonically non-decreasing relation, that is, when... There will always be a time when... .For example, , , This mapping relationship is written to the read-only memory area before the chip leaves the factory and cannot be modified at runtime, ensuring consistent behavior across all terminals on the network.

[0099] The terminal will Assign a value to the repeat counter of the current uplink transmission task The retransmission status register is initialized to "awaiting first transmission"; at this point, the terminal has not yet started constructing uplink data packets, but it is clear that this transmission will require at least [number] retransmissions. The same content is repeatedly transmitted. This counter will automatically decrement with each subsequent successful transmission until it reaches zero. The retransmission status register is used to control whether the physical layer encoder enables repetition mode, whether the modulation symbols are mirrored, and whether the power boost is activated.

[0100] The terminal enters the uplink data preparation stage, based on The values ​​are adjusted to determine the puncture position and interleaving depth of the convolutional codes in the physical layer coding chain, so that the length of the final generated code block is exactly matched. The total number of symbols required for multiple repetitions; specifically, if the number of symbols required for a single transmission is... The total length of the target code block is: ;

[0101] in It is determined by the current modulation order, the number of resource units, and the channel coding rate, and is a known constant; Derived from the above-mentioned search results Therefore Fully indexed by coverage level The decision has been made. This adjustment will be completed once before the encoder starts, and will not change the MAC layer group size, but will only affect the physical layer symbol generation rhythm and buffer arrangement.

[0102] Step 2: The terminal embeds a coverage level fingerprint sequence in its initial uplink transmission. This sequence is uniquely generated by the coverage level index and maintains a numerical isomorphic relationship with the initial value of the repetition count.

[0103] In the previous step, the terminal has determined the coverage level index. and the corresponding initial value of the number of repetitions This configuration is used to configure the code block length. However, relying solely on this configuration, the base station cannot determine the actual code block length currently used by the terminal without parsing higher-layer signaling. Therefore, this step introduces a fingerprint sequence that does not occupy additional signaling fields, but is embedded through the phase shift of a scrambling sequence. Its generation rule is similar to... Forming a one-to-one correspondence, and simultaneously with It exhibits numerically derivable isomorphism, allowing the base station to deduce it through receiver signal processing. This allows us to predict potential retransmissions.

[0104] Terminal indexed by coverage level Use the seed to generate a binary fingerprint sequence of length 12. This sequence is generated by a linear feedback shift register with a fixed initial state, and its feedback polynomial is: The initial state is The binary representation is padded with zeros on the left to 12 bits and then XORed with a fixed mask. Obtained. This process ensures that different The fingerprint sequences generated by the values ​​have a cross-correlation of less than 0.15, and the Hamming distance between any two sequences is not less than 5, which provides sufficient discriminative power.

[0105] The terminal will display the fingerprint sequence. The mapping is a set of complex modulation symbols, and the mapping method is as follows: Time output Time output This forms a BPSK symbol stream on the real axis. This symbol stream does not occupy resources independently; instead, it is superimposed on the preamble of the uplink data block. Specifically, a 12-bit spreading sequence is inserted before the demodulation reference signal DMRS of the Physical Uplink Shared Channel (PUSCH). Each chip of this spreading sequence is multiplied by the corresponding... Then, the data is mapped onto the 0th to 11th available subcarriers in the frequency domain via the same subcarrier, forming an implicit fingerprint carrier group.

[0106] During the initial transmission, the terminal sends the fingerprint carrier group along with the normal data, maintaining the same power as the main data symbol without enhancement or attenuation. Since this carrier group is located before the DMRS and its width represents only a very small proportion of the entire uplink bandwidth (e.g., 12 / 1200), its impact on main channel estimation is negligible. After coarse synchronization, the base station receiver can extract this frequency domain signal before DMRS parsing and perform matched filtering detection. This operation does not change the original frame structure, nor does it add any scheduling instructions; it simply utilizes existing idle subcarrier positions to carry low-rate identity information.

[0107] After receiving the initial uplink signal, the base station extracts the complex response value from subcarriers 0 to 11 in the frequency domain, and then restores it using hard decision. Then reverse map to It compares all 8 pre-stored fingerprint sequences and finds the one with the smallest Hamming distance, recording its corresponding index as . This yields an estimate of the terminal's true coverage level. At this point, the base station can immediately infer the initial number of repetitions currently used by the terminal. ;

[0108] This value is completely consistent with the result obtained from the table lookup above, because the fingerprint sequence and... The derivation is unambiguous due to the strict one-to-one correspondence. The base station can reserve a corresponding number of potential retransmission resources for the terminal in the next downlink subframe after the first reception without waiting for subsequent retransmissions.

[0109] Step 3: During each retransmission, the terminal dynamically adjusts the power offset based on the remaining number of repetitions and the channel quality trend, enabling the base station to infer the terminal's retransmission progress from the received signal power sequence.

[0110] In the previous step, the base station has already obtained the terminal coverage level index through the fingerprint sequence. And thus determine its initial number of repetitions. However, during actual transmission, the channel quality may slowly change due to fluctuations in the wireless environment, a drop in battery voltage, or movement. This necessitates fine-tuning the transmission power of subsequent retransmissions to maintain the demodulation success rate at the receiver. If all retransmissions are sent at the same power, the base station cannot distinguish whether increased retransmissions are due to channel degradation or the terminal is executing scheduled retransmissions. Conversely, if power adjustments lack a pattern, it undermines the base station's basis for judging retransmission progress. Therefore, this step designs a power bias sequence whose variation is uniquely determined by the remaining number of retransmissions and exhibits a monotonically increasing characteristic. This allows the base station to accurately determine the current retransmission stage based solely on the power measurements of multiple consecutive received signals.

[0111] After completing the initial transmission, the terminal will set a repeat counter. Subtract 1 to get the remaining number of repetitions. The initial value is The value is decremented by 1 after each subsequent transmission, until it reaches zero. The terminal follows this pattern. For input, look up the local power bias table to obtain the current power increment to be added. This table contains... Each term is a real number in dB and satisfies a strictly monotonically increasing relationship, that is, when For example, if ,but This table is indexed by coverage level. It is irrelevant; it only relates to the remaining number of attempts, and all terminals use the same table to ensure consistent parsing on the base station side.

[0112] The terminal will Convert to linear field multipliers This multiplier is then applied to the amplitude of the current symbol to be transmitted. This operation occurs after digital baseband shaping filtering and before digital-to-analog conversion; it is pure amplitude scaling and does not change the symbol phase or time structure. Since the content of each repeated transmission is exactly the same, only the amplitude is sequentially increased, the receiver can still use the channel response estimated in the first transmission when performing coherent demodulation. It only needs to adjust the received signals according to their corresponding values ​​before combining. Weighted summation is sufficient.

[0113] The terminal performs a second transmission, with the symbol amplitude being the same as the first. Double; thereafter, update before each transmission. And look up the table to get the new , then calculate the new This ensures that each repetition has a uniquely identifiable power scale. This process continues until... That is, the last time the corresponding The amplitude is restored to the baseline value. This forms a line with a length of power sequence Its form is from Completely decided.

[0114] The base station measures the average received power of multiple received signals for the same uplink transmission task, and records the average received power as the first signal. The received power is ,in Because the terminal's transmission power is based on Sequence scaling, while path loss and antenna gain remain constant in the short time, therefore... ;

[0115] in The received power is the reference power corresponding to the initial transmission power, and is an unknown constant. The linear multiplier is obtained from a lookup table; the square relationship stems from the quadratic relationship between power and amplitude. The base station will be measured... Sequence normalization, i.e., calculating the ratio. This can eliminate Impact, direct gains Then compare with all the possibilities stored locally. The sequence list is matched to find the closest match, and the current sequence is deduced from this. Value and This inversion does not depend on the accuracy of channel estimation; reliable identification is achieved only if the power measurement stability is better than 0.3 dB.

[0116] Step 4: Based on the power sequence matching results and the distribution characteristics of demodulated soft information, the base station jointly determines whether the terminal should terminate the repetition prematurely and dynamically releases unused resources.

[0117] The first three steps have achieved base station identification of terminal coverage level, confirmation of initial repetition count, and resolvability of transmission power for each repetition. However, in actual deployment, the terminal may actively stop subsequent repetitions because it is correctly demodulated on the first transmission or the decoding threshold has been met in the first few repetitions. If the base station still proceeds as planned... Reserving all resources would result in wasted spectrum; releasing resources too early might miss subsequent repetitions that have not yet been transmitted by the terminal. Therefore, this step introduces a dual decision mechanism: on the one hand, it identifies repetition termination behavior based on the power sequence interruption point; on the other hand, it assesses the decoding confidence by combining the soft information entropy value of the demodulated output. The two mechanisms jointly decide the timing of resource release, ensuring that it is neither overly conservative nor overly aggressive.

[0118] The base station continuously monitors the received power sequence of the same uplink transmission task, and when it detects the first... After the first reception, the expected number of receptions is... If the expected power value is not detected at the corresponding time-frequency position, and this deficiency persists for more than one subframe period, it is determined as a power sequence interruption. In this case, the base station records the time corresponding to the interruption. The value is temporarily stored as a candidate termination point. This interrupt detection does not rely on an absolute power threshold. Instead, it compares the deviation between the ratio of two consecutive measured power values ​​and the theoretical ratio. It is triggered when the deviation in two consecutive subframes is greater than a set tolerance (e.g., 0.5dB). This tolerance takes into account the effects of hardware gain drift and temperature changes to ensure no false alarms.

[0119] The base station has received the previous The signals are then subjected to channel equalization and symbol demapping to obtain the soft bit information of each demodulated output, denoted as . ,in This represents the number of bits in the code block; subsequently, for each bit position... Calculate its in Standard deviation in group soft information ,in This is the mean at that position; finally, calculate the total for all positions. Medium less than the preset threshold The percentage is denoted as This value reflects the degree of convergence of soft information. The higher the value, the more consistent the demodulation results are across each iteration, and the stronger the decoding reliability.

[0120] Base station will With preset threshold In comparison, if ,and If the decoding is successful, the terminal has completed valid decoding and no further repetition is needed; the remaining resources can be safely released. Otherwise, if... Even in the event of a power outage, the base station will retain a listening window of up to two subframes to prevent the terminal from missing its transmission opportunity due to brief interference. This dual-condition design balances robustness and efficiency. Ensure demodulation quality meets standards. The lower limit prevents premature release (e.g., if the release is interrupted after only one attempt, it will not be recognized).

[0121] Once the base station confirms termination, it immediately generates a resource release command. This command is not issued as new signaling, but rather by adjusting the format of the subsequent downlink control information (DCI) for the terminal. It fills the previously reserved uplink grant field with all zeros, indicating that the grant is invalid. Simultaneously, in the physical downlink control channel (PDCCH), the search space position corresponding to the terminal's C-RNTI scrambling sequence is set to idle, preventing the terminal from decrypting a valid grant in the next listening opportunity, thus naturally ceasing subsequent actions. This release process is represented by "no new grant received" on the terminal side, seamlessly connecting with its autonomous termination behavior without interactive confirmation and without any additional signaling overhead. At this point, the base station can immediately allocate the released time-frequency resources to other terminals awaiting scheduling, achieving a dynamic improvement in spectrum utilization efficiency.

[0122] Step 5: After completing all repeated transmissions or early termination, the terminal autonomously decides whether to initiate a coverage level reassessment based on the soft information entropy value of the last successfully demodulated signal, and triggers an implicit renegotiation process.

[0123] The first four steps have established a closed-loop system for base station identification of terminal coverage levels, the ability to predict repetition counts, the resolvability of power sequences, and a signaling-free coordination mechanism for resource release. However, the wireless environment is time-varying, especially in deep coverage scenarios. Changes in building obstruction, terminal attitude adjustments, and the activation or deactivation of surrounding interference sources can all cause observable shifts in link quality over time, ranging from several seconds to tens of seconds. If the terminal remains stuck in the initial coverage level index for an extended period... This will lead to a continuous mismatch between the number of repetitions in subsequent transmissions and the actual needs. Using high repetitions for excessive coverage will only increase power consumption, while using low repetitions for insufficient coverage will cause demodulation failures. Therefore, this step introduces a terminal-initiated coverage level reassessment mechanism. Its triggering condition does not depend on base station instructions, but is instead determined by the terminal itself through quantitative analysis of the soft information distribution characteristics of the most recent complete reception feedback. It is only activated when the entropy value increases significantly to avoid frequent oscillations. Furthermore, the entire renegotiation process does not generate new signaling, but instead reuses the physical layer structure of the next uplink data packet to carry the reassessment request identifier.

[0124] After the terminal completes this uplink transmission task, if it receives an ACK confirmation from the base station via the downlink control channel, it will extract the soft bit information from the PDSCH demodulation output corresponding to the ACK. ,in The number of bits in the downlink transport block is given; then the information entropy of this soft information sequence is calculated:

[0125] ;

[0126] In this formula, Indicates the first Posterior probability estimate for each bit being 0 Its complement is the probability of being 1; entropy value A higher value indicates that the soft-decision result is closer to an equal probability distribution, resulting in lower decoding confidence and indirectly reflecting a deterioration in the current channel quality compared to the previous assessment. This calculation is performed in the baseband processor using a fixed-point approximation method, without consuming additional instruction cycles.

[0127] The terminal will calculate the result Reference entropy threshold compared to local storage In comparison, this threshold is the average entropy value measured in a typical weak coverage scenario before the terminal leaves the factory, and is fixed at [value missing]. ;like ,in If a tolerance margin is provided to filter out transient noise disturbances, it indicates a systematic degradation in downlink quality, requiring a reassessment of the coverage level; otherwise, the current status is maintained. The value remains unchanged, and the original mapping relationship continues to be used.

[0128] When a reassessment is deemed necessary, the terminal does not immediately initiate random access. Instead, in the next planned uplink data transmission, it embeds a 2-bit reassessment request identifier at the beginning of the physical layer data block. This identifier is not an independent field, but is implemented by modifying the cyclic shift value in the first resource unit of PUSCH: the original protocol specified that the shift value ranged from 0 to 7, and now it is agreed that when the shift value is 6 or 7, it means that a re-evaluation request is carried, where 6 corresponds to "slight deterioration" and 7 corresponds to "significant deterioration"; this agreement has been written into the chip's solidified logic, requiring no software intervention and not changing the MAC layer scheduling process.

[0129] The terminal completes the uplink transmission the prescribed number of times, and its fingerprint sequence, power offset, code length, and other parameters remain based on the old coverage level. The signal is generated; however, after receiving the signal, the base station finds that the cyclic shift value is 6 or 7 when parsing it, and knows that the terminal is requesting a coverage level update. At this time, the base station does not change the current scheduling policy, but marks the request as pending, and in the subsequent downlink response, fills in the new coverage level index in the same 3-bit physical layer feature field of the initial access response message. ;Should It does not take effect immediately, but is used as input for the next uplink transmission and is only enabled after the terminal receives the response; thus forming a three-step implicit negotiation of "request-response-switching", without adding any signaling overhead or interrupting service continuity.

[0130] Step Six: Based on the historical retransmission behavior statistics of terminals and the matching results of the current power sequence, the base station constructs a coverage level drift trend model and injects a calibration offset into the downlink response.

[0131] In step five, although the base station can detect the terminal's reassessment request, it cannot accurately determine the drift direction and magnitude based solely on a single increase in entropy value. Directly using the latest random access preamble strength to reclassify the level is susceptible to misjudgment due to fast fading; relying entirely on terminal reports violates the original design principle of no-signaling. Therefore, this step proposes a base station-side trend modeling method: based on the terminal's past... Using the next uplink transmission as the observation window, we statistically analyze the degree to which the actual number of repetitions deviates from the initial value, the forward shift trend of the power sequence interruption point, and the cumulative rate of change of the soft information entropy in each demodulation. From these, we extract three dimensions of drift indicators and synthesize them into a weighted calibration offset. This information is then encoded into the physical layer feature field of the downlink response for the terminal to use during the next access.

[0132] The base station maintains a terminal-level historical behavior table, recording the terminal's most recent... The following fields were transmitted on the next uplink: actual number of repetitions. Theoretical initial values Power sequence interruption point number and the corresponding soft information entropy mean ;in Indicates the earliest time, This indicates the most recent record. This table is created when the terminal first connects and is updated with each upstream activity. If the record exceeds its length, it will scroll and overwrite the oldest record.

[0133] Base stations calculate three drift metrics: the first being the repetition error rate. The first indicator reflects whether the terminal system experiences a high or low number of interrupts; the second is the interrupt shift rate. If the value is negative, it indicates that the interruption point is advanced successively, meaning that the number of repetitions required for effective decoding is decreasing, and the coverage is improving; the third is the entropy growth slope. This measures the rate of downlink quality degradation. The three indicators are normalized and then weighted and summed to obtain the overall drift.

[0134] ;

[0135] Among them, weight It is derived from the contribution of various indicators to the actual coverage changes in a large number of field tests and has been embedded in the base station baseband processing unit.

[0136] Base station will Mapped to integer calibration offset The mapping rule is: if ,but ;like ,but ;like ,but ;like This mapping is nonlinear, focusing on enhancing the sensitivity to significant improvements or deteriorations in recognition, while suppressing frequent jumps caused by minor fluctuations.

[0137] When generating the initial access response message, the base station changes the value of the original 3-bit physical layer feature field from a simple coverage level index. Change to a new index after calibration. This operation ensures the result remains within a valid range, preventing it from going out of bounds. After receiving this response, the terminal will no longer read the original data in step one. , but This allows for automatic incorporation of historical trend corrections during subsequent accesses. This process does not alter the frame structure, does not increase bit overhead, and only reuses existing field semantics, achieving seamless integration between base station-side intelligent calibration and terminal-side operation.

[0138] Step 7: When the terminal receives the new coverage level index after calibration, it does not immediately switch the mapping relationship, but instead activates a dual-track buffer mechanism to establish a smooth transition window between the old and new sets of parameters.

[0139] In step six, the base station has already calibrated the offset. The code is encoded into the response field, which the terminal can retrieve upon the next access. However, if the terminal discards the old one immediately after receiving the response... , change to new Determining the repetition count and power bias by looking up a table can lead to abrupt behavioral changes: for example, from (correspond ) mutation (correspond If the current channel has not yet stabilized and improved, the new configuration will cause demodulation failure. Therefore, this step designs a dual-track buffering mechanism: when the terminal receives... Instead of immediately replacing the currently active coverage level, the old and new levels are loaded into the buffer in parallel, each maintaining an independent set of repetition counters and power bias sequences. The weights are dynamically allocated based on the success rates of the two most recent uplink transmissions, so that the parameter switching presents a gradual convergence rather than a step jump.

[0140] The terminal parses Then, compare it with the level currently in use. Both are stored in a dual-rail register, labeled "Main Rail" and "Support Rail" respectively; the main rail retains its original designation. All corresponding mapping parameters, including Power bias table Fingerprint sequence generation seeds, etc.; auxiliary tracks are pre-calculated and cached. The corresponding full set of parameters is available, but it is not enabled for the time being.

[0141] In the subsequent two uplink transmissions, the terminal executed the transmission according to the master rail parameters each time, but simultaneously recorded the actual demodulation success rate of that transmission. ,in These represent the first and second attempts, respectively; then the average success rate is calculated. And update the primary and secondary track weights accordingly: if This indicates that the current main track configuration is still reliable, maintaining the main track weight at 1.0 and the auxiliary track weight at 0; if Then let the weight of the main rail be 0.7 and the weight of the auxiliary rail be 0.3; if If the weights are reversed, the main track is reduced to 0.3 and the auxiliary track is increased to 0.7. This weight is only used for parameter mixing in the next step and does not affect the current sending behavior.

[0142] When the terminal prepares to initiate the third uplink transmission, it no longer uses only the main rail or auxiliary rail parameters, but instead performs a weighted synthesis of key parameters: the initial value of the repetition count is taken as... ,in The primary and secondary track weights are calculated, and the result is rounded to the nearest integer; the power bias table then calculates the weights for each remaining repetition. ,calculate The result is also rounded down. This mixing process is completed in real time before transmission, without adding any delay.

[0143] After the terminal completes the third uplink transmission, it reassesses the success rate and updates the weight; if three consecutive transmissions are successful... If the auxiliary rail is promoted to the main rail, the original main rail is cleared, and the rail switch is completed; if this happens three times in a row... If an anomaly is detected, the system will forcibly revert to the original main track and send a special preamble (identifying the anomaly through a specific sequence combination in PRACH format 1b) to the base station, requesting manual verification; otherwise, the dual-track buffer will continue to be maintained. This mechanism enables the terminal coverage level adaptation process to have memory, anti-interference, and reversibility, avoiding systemic instability caused by a single misjudgment.

[0144] Step 8: The base station and the terminal jointly maintain the coverage level status synchronization flag, update the flag at the end of each uplink transmission, and perform bidirectional verification through redundant bits in the downlink response to ensure zero status drift during long-term operation.

[0145] The first seven steps cover the entire lifecycle, including coverage level identification, dynamic adaptation, trend calibration, and smooth handover. However, a potential risk remains: the terminal and base station may differ in their understanding of the currently effective coverage level due to a lost response, decoding error, or hardware reset. For example, the terminal might believe it has already switched to... However, because a certain response was not received correctly, the base station still proceeded as normal. Power sequence matching can lead to subsequent parsing failures, resulting in resource mismatches and retransmission avalanche. To prevent such state drift, this step introduces a lightweight state synchronization mechanism: each party maintains a 2-bit state synchronization flag. This flag does not carry any business meaning; it only indicates whether the current coverage level configuration is consistent. It achieves bidirectional verification through redundant bits in the downlink response and a specific pattern of the uplink power sequence, ensuring that the states of both parties are strictly aligned at any time.

[0146] After initial access is completed, both the terminal and the base station will set the state synchronization flag. Initialize to 00; thereafter, each time an uplink transmission task successfully completes (i.e., the terminal receives an ACK and demodulation is successful), the terminal executes... This process involves a cyclical increment; similarly, the base station performs the same increment operation each time it completes a valid demodulation and ACK transmission for the terminal. This incrementing action is automatically completed by the hardware state machine without software intervention, ensuring a strictly consistent rhythm.

[0147] When the base station generates a downlink response message, it will... The value is filled into the 2-bit redundancy field reserved in the initial access response field (its purpose was not defined in the original protocol, but is reused here); after receiving this response, the terminal compares it with its local... Values ​​and those carried in the response If the two are equal, it indicates that the state synchronization is normal; if they are not equal, the terminal immediately triggers a state reset procedure: suspending subsequent uplink transmission, waiting for the next downlink subframe, re-listening to the redundant bit, and resetting the local state. Force update to received Then continue with the business.

[0148] To prevent one-way verification from failing (e.g., the terminal receives an error) (Subsequent synchronization error) The terminal, in each uplink transmission, will change the local... The value is mapped to the initial bias of the power sequence: specifically, let the standard power sequence be... The first item is In the actual transmission, the entire sequence is cyclically shifted left. The position, that is, the first This send uses After receiving the sequence, the base station tries all four possible sequences. The values ​​are matched, and the one that minimizes the matching error is selected as the current value. The estimated value; this estimated value is related to the value carried in the downlink response. Cross-validation is performed, and state synchronization is confirmed only when both are consistent.

[0149] If during two consecutive uplink transmissions, the terminal locally... Downlink Response Inconsistent, and the power sequence obtained from inversion If the signal is inconsistent with both of these, a deep synchronization fault is determined, and the terminal enters a low-power listening mode: most radio frequency circuits are turned off, and the receiver is only periodically woken up. The terminal listens to the global synchronization beacon broadcast by the base station at a fixed position in each downlink subframe (this beacon is sent once every 10 seconds, occupies a very narrow bandwidth, and the power consumption is negligible). The beacon contains a summary of the synchronization status of all terminals in the network. The terminal uses this summary to quickly locate the status bit corresponding to its own ID and complete the calibration in one go.

[0150] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for a low-power narrowband Internet of Things (IoT) communication chip, characterized in that, include: The terminal determines the coverage level index based on the physical layer feature field in the initial access response and selects the corresponding initial value for the number of repetitions. Based on the determined coverage level index, the terminal embeds the coverage level fingerprint sequence in the first uplink transmission. Based on the initial value of the selected number of repetitions, the terminal dynamically adjusts the power bias based on the current remaining number of repetitions during each retransmission. After receiving the uplink signal, the base station identifies the terminal coverage level by parsing the coverage level fingerprint sequence, and inverts the terminal retransmission progress based on the received signal power sequence formed by the power offset. Combining the inverted terminal retransmission progress and the demodulation soft information distribution characteristics, the base station determines whether the terminal should terminate the repeated transmission prematurely. After receiving the downlink confirmation signal from the base station, the terminal calculates the soft information entropy value and decides whether to trigger the coverage level reassessment process based on the comparison result between the entropy value and the preset threshold. In response to the coverage level reassessment process triggered by the terminal, the base station constructs a coverage level drift trend model based on the statistics of the terminal's historical retransmission behavior and injects the calibration offset into the downlink response; After receiving the downlink response of the injected calibration offset, the terminal starts a dual-track buffering mechanism when the new coverage level index takes effect to achieve a smooth transition between the old and new parameters. Based on the operational results of the dual-track buffer mechanism, the base station and the terminal jointly maintain the coverage level status synchronization flag, and achieve bidirectional verification through redundant bits in the downlink response and the uplink power sequence.

2. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, The terminal determines the coverage level index and selects the corresponding initial value for repetition count based on the physical layer feature fields in the initial access response, including: Read the 3-bit physical layer feature field from the initial access response message, and convert the 3-bit physical layer feature field into a decimal integer as the coverage level index; Based on the coverage level index, look up the local fixed mapping table to get the corresponding initial value of the repetition count, and assign the initial value of the repetition count to the repetition counter of the uplink transmission task; Based on the initial value of the repetition count, adjust the puncture position and interleaving depth of the convolutional code in the physical layer coding link.

3. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, Based on a determined coverage level index, the terminal embeds a coverage level fingerprint sequence in its initial uplink transmission, including: Using the coverage level index as a seed, a binary fingerprint sequence of length 12 is generated, and the binary fingerprint sequence is mapped to a BPSK symbol stream on the real axis; The BPSK symbol stream is superimposed on the subcarrier preceding the demodulation reference signal of the physical uplink shared channel; After receiving the first uplink signal, the base station extracts the complex response value at the subcarrier and determines the coverage level index by comparing it with the locally stored fingerprint sequence.

4. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, Based on the selected initial value for the number of repetitions, the terminal dynamically adjusts the power bias based on the current remaining number of repetitions during each retransmission, including: After each transmission, decrement the repetition counter by 1 to obtain the remaining number of repetitions; Based on the remaining number of repetitions, look up the local power bias table to obtain the current power increment; The power increment is converted into a linear domain multiplier and applied to the amplitude of the symbol to be transmitted; The base station measures the average received power of multiple received signals for the same uplink transmission task, and uses power sequence normalization to infer the terminal retransmission progress.

5. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, Based on the inverted terminal retransmission progress and demodulation soft information distribution characteristics, the base station determines whether the terminal prematurely terminates the repeated transmission, including: Continuously monitor the received power sequence of the same uplink transmission task, and record the candidate termination point when the expected power value is not detected at the corresponding time-frequency position; Channel equalization and symbol demapping are performed on the received signal to obtain soft bit information; Calculate the standard deviation of each bit position in the soft bit information across multiple receptions, and count the percentage of the standard deviation that is less than a preset threshold. The percentage of cases with a standard deviation less than a preset threshold is compared with a preset information threshold. When the preset information threshold is met, unused resources are released.

6. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, After receiving a downlink confirmation signal from the base station, the terminal calculates a soft information entropy value and determines whether to trigger a coverage level reassessment process based on a comparison between the entropy value and a preset threshold, including: Extract the soft bit information from the downlink channel demodulation output and calculate the information entropy of the soft bit information; The calculated information entropy is compared with the reference entropy threshold. When the information entropy is greater than the reference entropy threshold plus the tolerance margin, it is determined that the coverage level needs to be reassessed. In the next planned uplink data transmission, a re-evaluation request identifier is embedded by modifying the cyclic shift value of the physical layer data block start position.

7. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, In response to the coverage level reassessment process triggered by the terminal, the base station constructs a coverage level drift trend model based on the terminal's historical retransmission behavior statistics and injects the calibration offset into the downlink response, including: Maintain the terminal's historical behavior table, recording the actual number of repetitions, theoretical initial value, power sequence interruption point number, and average soft information entropy of the terminal's most recent N uplink transmissions; Calculate the repetition error rate, interruption shift rate, and entropy growth slope. Then, sum the repetition error rate, interruption shift rate, and entropy growth slope by weight to obtain the overall drift amount. The overall drift is mapped to an integer calibration offset and injected into the physical layer feature field of the downlink response.

8. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, After receiving the downlink response of the injected calibration offset, the terminal initiates a dual-track buffering mechanism when the new coverage level index takes effect to achieve a smooth transition between the old and new parameters, including: Store the new coverage level index and the current coverage level index together in the dual-track register, and mark them as the main track and the auxiliary track, respectively; In subsequent uplink transmissions, the transmission success rate is recorded, and the weights of the primary and secondary tracks are updated based on the success rate. When initiating uplink transmission, the initial value of the repetition count and the power offset are weighted and synthesized. Based on the continuous transmission success rate, a decision is made as to whether to upgrade the auxiliary rail to the main rail and complete the track switching.

9. The low-power narrowband Internet of Things communication chip method of claim 1, wherein, Based on the operational results of the dual-track buffer mechanism, the base station and the terminal jointly maintain the coverage level status synchronization flag, and achieve bidirectional verification through redundant bits in the downlink response and the uplink power sequence, including: After the initial access is completed, initialize the state synchronization flag to the same value; After each successful uplink transmission, each entity increments its status synchronization flag cyclically. The base station carries the current state synchronization flag in the downlink response, and the terminal compares its own flag with the base station flag; During uplink transmission, the terminal maps the state synchronization flag to the starting bias of the power sequence for the base station to verify.

10. A low power narrowband internet of things communication chip system for implementing the method according to any one of claims 1 to 9, characterized in that include: The coverage level determination unit is used to determine the coverage level index and select the corresponding initial value of the number of repetitions based on the physical layer feature fields in the initial access response. A fingerprint embedding unit is used to embed a coverage level fingerprint sequence in the first uplink transmission based on a determined coverage level index; The power adjustment unit is used to dynamically adjust the power bias based on the current remaining number of repetitions during each retransmission. The coverage level identification unit is used to identify the terminal's coverage level by parsing the coverage level fingerprint sequence after receiving the uplink signal. The transmission progress inversion unit is used to invert the terminal retransmission progress based on the received signal power sequence formed by the power offset. The resource release unit is used to determine whether the terminal should terminate repeated transmission prematurely by combining the inverted terminal retransmission progress and the distribution characteristics of demodulation soft information. The coverage status assessment unit is used to calculate the soft information entropy value after receiving the downlink confirmation signal from the base station, and decide whether to trigger the coverage level reassessment process based on the comparison result of the entropy value and the preset threshold. The drift trend modeling unit is used to build a coverage level drift trend model based on the statistics of the terminal's historical retransmission behavior, and inject the calibration offset into the downlink response; The dual-track buffer unit is used to activate the dual-track buffer mechanism when the new coverage level index takes effect after receiving the downlink response of the injected calibration offset, so as to achieve a smooth transition between the old and new parameters. The status synchronization unit is used to jointly maintain the coverage level status synchronization flag with the terminal, and achieves bidirectional verification through redundant bits in the downlink response and the uplink power sequence.

Images