Multi-device intelligent switching system and method based on polar code

Abstract

The application discloses a multi-device intelligent switching system and method based on a Polar code, which first collects input event streams in real time on each device side, and obtains original state report data through intelligent analysis. Subsequently, the data is encoded by using the Polar code, so that the anti-interference and error correction capability of the data in the wireless transmission process is effectively enhanced. The encoded state data is efficiently sent to a central receiving end through a star flash wireless interface according to a preset priority, and after the Polar code decoding is completed at the central end, the original state of all devices is restored in time. Based on the emergency degree and the activity state of each device, the central end dynamically updates a priority queue, and automatically distinguishes the priority relationship between a current master device and an optimal device, and intelligently decides whether to switch the master control right. In this way, the application can realize efficient and reliable state reporting and intelligent real-time switching of the master device in a multi-device environment.

Description

Technical Field

[0001] This application relates to the field of mobile data communication technology, and more specifically, to a multi-device intelligent handover system and method based on Polar codes. Background Technology

[0002] With the rapid development of modern wireless communication technology and the Internet of Things (IoT), the number of smart terminal devices owned and controlled by users has surged, encompassing smartphones, wearable devices, smart home terminals, and various industrial and office automation equipment. In application environments where multiple devices coexist, information collaboration, state synchronization, and role switching between devices are particularly important. However, existing multi-device collaboration systems still face many challenges, including but not limited to unstable communication links, insufficient data transmission reliability, high response latency, and a lack of intelligent judgment in device state switching. Traditional device switching mechanisms mostly rely on simple polling, fixed priority, or step-by-step decision-making algorithms, which are difficult to meet the high requirements for real-time performance and robustness in high-concurrency, multi-scenario environments. At the same time, with the increasing variety of terminals and the complexity of user applications, the volume and diversity of device status data continue to increase. How to achieve accurate acquisition and rapid response of multi-source device status while ensuring efficient and secure data transmission has become a core technical bottleneck for intelligent multi-device collaboration.

[0003] To address the aforementioned challenges, the industry urgently needs a multi-device intelligent switching technology solution that balances high reliability and high efficiency to improve overall stability, response speed, and human-computer interaction experience. In multi-device integrated application environments, when dynamically selecting a master control role or switching focus devices among multiple terminals, the system must not only be able to perceive the current activity status, key operations, and emergency states of each device in real time, but also make rapid and accurate switching decisions based on changes in device status. Furthermore, due to limitations in wireless channel noise and interference, as well as the actual constraints of terminal device computing power and energy consumption, higher robustness and low latency requirements are placed on the encoding, transmission, and decoding of device status. Especially in scenarios such as the Internet of Things (IoT), smart homes, and smart offices, frequent data status reporting and master device switching needs must ensure communication quality and complete decisions within a very short time; otherwise, it can easily lead to a decline in user experience or even system failure.

[0004] Based on this, this application proposes a multi-device intelligent handover scheme based on Polar codes. Summary of the Invention

[0005] To address the aforementioned technical problems, this application is proposed. Embodiments of this application propose a multi-device intelligent handover system and method based on Polar codes. Employing Polar code technology, it can significantly improve the anti-interference and error correction capabilities of multi-device status information in complex wireless environments, ensuring the integrity and reliability of critical handover data.

[0006] According to one aspect of this application, a multi-device intelligent handover system and method based on Polar codes is provided, comprising: acquiring input event streams of each device; generating original status report data of each device based on the input event streams of each device, the original status report data including device ID, activity type, urgency level, and timestamp; encoding the original status report data of each device using Polar codes to obtain device status report Polar codes for each device; transmitting the device status report Polar codes of each device to a central receiving end via a StarScan wireless interface; decoding the device status report Polar codes of each device using Polar codes to obtain the original status report data of each device; updating a priority queue based on the original status report data of each device to obtain an updated priority queue; and checking the highest priority device extracted from the priority queue, and determining whether to switch to the highest priority device based on a comparison between the priority of the highest priority device and the priority of the current master device.

[0007] In one possible implementation, raw status report data for each device is generated based on the input event stream of each device, including: parsing the input events of each device and identifying the device ID, activity type, activity data, and timestamp; and generating the urgency level of each device based on the activity data of each device.

[0008] In one possible implementation, generating the urgency level of each device based on the activity data of each device includes: performing structured encoding on the activity data of each device to obtain a set of structured encoded vectors of device activity data; inputting the set of structured encoded vectors of device activity data into a device activity context encoder based on a converter architecture to obtain a set of device activity context semantic encoded vectors; and performing feature decoding regression on each device activity context semantic encoded vector in the set of device activity context semantic encoded vectors to obtain the urgency level of each device.

[0009] In one possible implementation, generating the urgency level of each device based on the activity data of each device includes: performing structured encoding on the activity data of each device to obtain a set of structured encoded vectors of device activity data; inputting the set of structured encoded vectors of device activity data into a device activity context encoder based on a converter architecture to obtain a set of semantic encoded vectors of device activity context; performing information projection mapping on the set of semantic encoded vectors of device activity context to obtain a set of entropy projection vectors of device activity context; and performing feature decoding regression on each entropy projection vector of device activity context in the set of entropy projection vectors of device activity context to obtain the urgency level of each device.

[0010] According to another aspect of this application, a multi-device intelligent switching system based on Polar codes is provided, comprising: an input event stream acquisition module for acquiring input event streams from various devices; a raw status report data generation module for generating raw status report data for each device based on the input event streams from the various devices, the raw status report data including device ID, activity type, urgency level, and timestamp; a Polar code encoding module for encoding the raw status report data of each device using Polar codes to obtain device status report Polar codes for each device; and a device status Polar code transmission module for transmitting the device status Polar codes of the various devices. The status report Ploar code is sent to the central receiver via the StarScan wireless interface; the Polar code decoding module is used by the central receiver to decode the device status report Ploar codes of each device to obtain the original status report data of each device; the priority queue update module is used by the central receiver to update the priority queue based on the original status report data of each device to obtain the updated priority queue; the device switching decision module is used by the central receiver to check the highest priority device extracted from the priority queue, and to determine whether to switch to the highest priority device based on the comparison between the priority of the highest priority device and the priority of the current master device.

[0011] Compared with existing technologies, the multi-device intelligent switching system and method based on Polar codes provided in this application first collects input event streams in real time at each device side and obtains raw status report data through intelligent parsing. Then, the data is encoded using Polar codes, effectively enhancing the anti-interference and error correction capabilities of the data during wireless transmission. The encoded status data is efficiently transmitted to the central receiving end via the StarScan wireless interface according to a preset priority. After the central end completes the Polar code decoding, it promptly restores the original status of all devices. Based on the urgency and activity status of each device, the central end dynamically updates the priority queue and automatically determines the priority relationship between the current master device and the highest priority device, intelligently deciding whether to perform a master control switch. In this way, efficient and reliable status reporting and intelligent real-time switching of master devices are achieved in a multi-device environment. Attached Figure Description

[0012] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0013] Figure 1 The illustration shows a schematic flowchart of a multi-device intelligent handover method based on Polar codes according to an embodiment of this application.

[0014] Figure 2 The figure shows a schematic flowchart of step S2 in the multi-device intelligent handover method based on Polar codes according to an embodiment of this application.

[0015] Figure 3 The figure shows a schematic flowchart of step S22 in the multi-device intelligent handover method based on Polar codes according to an embodiment of this application.

[0016] Figure 4 The figure shows a schematic flowchart of step S3 in the multi-device intelligent handover method based on Polar codes according to an embodiment of this application.

[0017] Figure 5 The figure shows a schematic flowchart of step S6 in the multi-device intelligent handover method based on Polar codes according to an embodiment of this application.

[0018] Figure 6 The figure shows a schematic block diagram of a multi-device intelligent handover system based on Polar codes according to an embodiment of this application. Detailed Implementation

[0019] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0020] More specifically, Figure 1 The illustration shows a schematic flowchart of a multi-device intelligent handover method based on Polar codes according to an embodiment of this application. Figure 1 As shown, the multi-device intelligent handover system and method based on Polar codes includes:

[0021] S1, Obtain the input event stream from each device;

[0022] S2, Based on the input event stream of each device, generate the original status report data of each device, the original status report data including device ID, activity type, urgency level and timestamp;

[0023] S3, Polar code encoding is performed on the original status report data of each device to obtain the device status report Polar code of each device;

[0024] S4, the device status report Ploar code of each device is sent to the central receiving end through the Star Flash wireless interface;

[0025] S5, the central receiving end decodes the Polar code of the device status report of each device to obtain the original status report data of each device.

[0026] S6, the central receiving end updates the priority queue based on the original status report data of each device to obtain the updated priority queue;

[0027] S7, the central receiving end checks the extraction of the highest priority device from the priority queue, and determines whether to switch to the highest priority device based on the comparison between the priority of the highest priority device and the priority of the current master device.

[0028] Specifically, in step S1, the input event streams of each device are acquired. It should be understood that the state information of a single device often cannot directly reflect the true needs and environmental dynamics of the entire user scenario. For example, in scenarios such as smart offices and smart homes, users may operate multiple terminals simultaneously, and the state changes of these devices are highly dynamic and urgent. Therefore, acquiring the input event streams of each device is crucial. These streams not only cover user behaviors such as key inputs, touch actions, voice commands, and sensor triggers, but also include various autonomous detection data of the device itself regarding the environment (such as changes in ambient temperature, device power consumption detection, location information, and network status changes). Only by collecting and processing these input event streams in real time can efficient and rigorous modeling and judgment of device activity, current running tasks, and potential switching needs be achieved.

[0029] Specifically, input event acquisition modules are set up in various devices to acquire input event streams from different devices. Taking practical IoT applications as an example, the input event acquisition module can work in conjunction with hardware device drivers and input management programs at the operating system level. For instance, in a smartwatch, the input event acquisition module can continuously monitor user touch screens, key inputs, and motion and physiological sensor data based on the operating system's open APIs. In smartphones or voice assistants, this module can collect data streams triggered by screen operations, voice interactions, and sensors such as gyroscopes through system calls. In smart home gateway devices, the sources of input event streams may include various heterogeneous data sources such as remote control signals, door magnetic sensors, and ambient light detection. These input events are aggregated in real time through local processes or lightweight daemons, forming structured or semi-structured event streams.

[0030] In one specific embodiment, a user carries a smartphone, a smart bracelet, and a portable voice assistant. Each device has an embedded input event acquisition module to monitor user actions and device status events in real time. For the smartphone, the acquired input event stream can include the following types: events where the user opens a meeting schedule by touching the screen; the system records the event type (touch), related data (meeting schedule ID), device local timestamp, and current GPS location information; and automatically triggering a low battery warning event when the device battery is below 10%. For the smart bracelet, the input event stream can cover events such as acceleration notifications while running, abnormal heart rate alerts, and significant changes in movement direction. The portable voice assistant is responsible for continuously listening for keyword recognition requests in the environment; valid events include detecting calls to the assistant or detecting sudden high-decibel noise in an emergency. The event acquisition module in each device is implemented by a system-timed task or event-driven mechanism, encapsulating the acquired events into a data structure with low latency, marking metadata such as device type, event type, urgency level, and occurrence time.

[0031] During the event aggregation process, a unified reporting interface or message queue mechanism is designed to enable various devices to efficiently and with low bandwidth consumption aggregate the input event streams into the subsequent reporting and encoding processes. For example, the device side uses lightweight IoT protocols such as MQTT and CoAP to transmit the initially filtered and integrated event data to the status report generation module in the form of message packets. This module further transforms the raw events into standardized status report units according to predetermined protocol specifications, including but not limited to the device ID to which the event belongs, the event type enumeration, the urgency level value (which can be obtained from event characteristics and activity type determination), and a reliable local timestamp.

[0032] Specifically, in step S2, based on the input event streams of each device, raw status report data for each device is generated. This raw status report data includes device ID, activity type, urgency level, and timestamp. It should be understood that in a multi-device intelligent switching environment, there are numerous devices, and their states change frequently. The state and behavior of each device at any given moment can affect the resource allocation, task execution, or control allocation of the entire system. Therefore, it is necessary to reflect the dynamic operation of devices in a unified, efficient, and standardized data format in a timely and accurate manner. Based on this, the raw input event streams collected from the hardware layer of each device are aggregated and organized into structured, parsable raw status report data.

[0033] In one embodiment, such as Figure 2 As shown, based on the input event streams of each device, raw status report data for each device is generated, including:

[0034] S21, the input events of each device are parsed to identify the device ID, activity type, activity data, and timestamp. It should be understood that the device ID is the foundation for distinguishing states, managing resources, and mapping scheduling among multiple devices. Without the device ID, the central system cannot accurately trace the unique relationship between a single state data point and a physical device instance, which will directly lead to chaos in scheduling and management. The activity type, as an important attribute describing the semantics of device behavior, helps the system understand the current operating scenario and business actions of the device. For example, a fall event on a smart bracelet, a high-temperature alarm trigger, and a voice call from a smart speaker are all different types of activities. Only by knowing the activity type can the central system take targeted scheduling strategies. The timestamp is used to lock the order of events and is also a key basis for ensuring the synchronization and sorting of data streams across multiple devices. In a multi-threaded or distributed environment, correct control of the timing is crucial, greatly improving the logical consistency of data processing and scheduling efficiency.

[0035] S22, Based on the activity data of each device, generate the urgency level of each device. It should be understood that urgency level is the core quantitative indicator for prioritizing multiple devices. In a multi-device collaborative environment, the priority of some events and behaviors dynamically changes according to their urgency. For example, in a medical scenario, abnormal physiological monitoring has a higher priority than ordinary notifications. Quantifying urgency levels enables efficient queue management and control switching based on the principle of prioritizing urgency.

[0036] In one embodiment, such as Figure 3 As shown, based on the activity data of each device, the urgency level of each device is generated, including: S221, performing structured encoding on the activity data of each device to obtain a set of structured encoded vectors of device activity data; S222, inputting the set of structured encoded vectors of device activity data into a device activity context encoder based on a converter architecture to obtain a set of semantic encoded vectors of device activity context; S223, performing feature decoding regression on each device activity context semantic encoded vector in the set of semantic encoded vectors of device activity context to obtain the urgency level of each device.

[0037] Specifically, the first step is to perform structured encoding on the activity data of each device to obtain a set of structured encoded vectors for device activity data. Different structured encoding methods are used for different types of data. For continuous numerical data, such as heart rate, acceleration, temperature, brightness, and battery percentage, normalization is first performed to map their values ​​to a uniform numerical range. To capture time-series features, a sliding window technique can be used to calculate statistical features within each window, or a filter can be applied to remove noise. For continuous data, a discretization method can be further used to divide it into several intervals, and one-hot encoding can be performed on each interval to reduce the complexity of subsequent models.

[0038] For text data such as voice recognition commands, chat messages, and system notifications, word segmentation is first performed, and then word embedding techniques are used to convert them into structured encoded vectors of device activity data. These word embedding techniques include Word2Vec, GloVe, and FastText. For discrete events such as button presses, touch events, and network connection status changes, one-hot encoding can be directly performed, mapping each event type to a specified position in a high-dimensional sparse vector as 1 and other positions as 0. For example, pressing "up" is mapped to [0,1,0,0], and pressing "down" is mapped to [0,0,1,0]. If the device input includes images or video streams (such as smart cameras), pre-trained convolutional neural networks can be used to extract high-level feature vectors for object recognition and behavior analysis, and these recognition results can be converted into structured feature codes.

[0039] After structured encoding, a set of structured encoded vectors of device activity data with multi-dimensional feature representation at the device level is obtained. While it is feasible to perform downstream judgments directly using these feature vectors, it easily leads to system errors or a lack of accurate understanding of the comprehensive background of multiple events because it ignores the rich contextual interactions and global spatiotemporal dependencies between devices and event sequences. To address this issue, the technical solution of this application introduces a device activity context encoder based on a transformer architecture. This encoder takes the set of structured encoded vectors of device activity data as input and uses a transformer self-attention mechanism to globally model the set of structured encoded vectors of device activity data. Its principle is to weightedly associate each event feature with other events, which can fully explore the co-occurrence relationships, temporal dependencies, and potential conflict and cooperation mechanisms between the same or different devices. In specific implementation, the context encoder adopts a standard multi-layer transformer block. Each transformer layer includes a multi-head self-attention layer, which performs head-weighted aggregation on the feature vector sequence of the input events. Each head focuses on different event interaction details, and positional encoding is combined to ensure the time-series characteristics. Subsequently, feedforward neural networks, layer normalization, and residual connections are used to improve the model's training stability and representational ability. The encoder network ultimately outputs a contextual semantic encoding vector for each event. This vector not only contains information about the event itself but also incorporates sequence context and global information into each output, giving it the higher-level semantics required for urgency discrimination.

[0040] The generation of urgency level is a direct output of the aforementioned model. Its mechanism involves encoding the semantic context of device activity into a vector, followed by post-processing through a specially designed feature decoding and regression network. In a specific embodiment of this application, feature decoding and regression are performed using a small MLP (Multilayer Perceptron) with multiple fully connected layers. This module takes the context encoding vector as input, passes through 2-4 layers of nonlinear activation transformation, and outputs a scalar urgency level. The last layer of the network uses an appropriate activation function, such as sigmoid mapping to the 0-1 interval, or ReLU / tanh for positive numbers or bounded intervals, to standardize the urgency level output. The urgency level can be distinguished into continuous values ​​(fine-grained ranking) and discrete levels (such as high / medium / low urgency, etc.) according to the needs of the actual scenario. This regression network and the transformer context model are connected end-to-end, jointly participating in backpropagation gradient optimization to achieve fully automatic learning from the original event flow to the urgency level.

[0041] In the specific training process, the original input events of each device are first structured and encoded into multi-dimensional vector strings according to a unified template, forming an input batch sequence. Each batch is fed into a device activity context encoder based on a converter architecture, which processes the data through multiple layers of self-attention and feedforward networks to output a complete contextual semantic representation. Subsequently, feature decoding regression is used to perform non-linear mapping on each output and output an urgency score. The loss function measures the consistency between this score and the label, and backpropagation optimizes the network parameters throughout the process. The entire model is optimized for several rounds on the training set, and the hyperparameters are adjusted based on the performance on the validation set. After training, the urgency level can also be output end-to-end for new device event streams that have not encountered a new environment, enabling real-time inference.

[0042] For the urgency level of each device obtained by feature decoding regression of the activity context semantic encoding vector of each device, since Polar code encoding needs to be performed, it needs to be mapped to information bits together with device ID, activity type and timestamp based on mapping rules. Therefore, in order to avoid information mapping imbalance caused by encoding and decoding of converter architecture, it is preferable to perform information projection mapping on the set of device activity context semantic encoding vectors.

[0043] Based on this, in another embodiment, generating the urgency level of each device based on the activity data of each device includes: performing structured encoding on the activity data of each device to obtain a set of structured encoded vectors of device activity data; inputting the set of structured encoded vectors of device activity data into a device activity context encoder based on a converter architecture to obtain a set of semantic encoded vectors of device activity context; performing information projection mapping on the set of semantic encoded vectors of device activity context to obtain a set of entropy projection vectors of device activity context; and performing feature decoding regression on each entropy projection vector of device activity context in the set of entropy projection vectors of device activity context to obtain the urgency level of each device.

[0044] In one sub-implementation of this embodiment, performing an information projection mapping on the set of device activity context semantic encoding vectors to obtain a set of device activity context semantic encoding entropy projection vectors includes: for each feature value v in the set of device activity context semantic encoding vectors i Calculate its information entropy representation v ti =p(v i log2p(v i ), where p(v i ) is the eigenvalue v i The probabilistic form, such as that obtained by activation with the sigmoid activation function, v ti Each feature value in the set of entropy projection vectors is used to semantically encode the device activity context.

[0045] However, such information projection mapping will have the problem of dynamic equilibrium at the end, that is, the final state representation of the eigenvalue at the end of the sequence probabilistic end will decay. Therefore, it is necessary to perform sequence distribution to stable equilibrium so that information projection mapping can be performed within a stable framework of the entire information domain, so as to avoid the end state decay disturbance in the information projection process.

[0046] Based on this, in another sub-implementation of this embodiment, the set of device activity context semantic encoding vectors is subjected to information projection mapping to obtain a set of device activity context semantic encoding entropy projection vectors, including: first, probabilistic activation is performed on each feature value of the set of device activity context semantic encoding vectors to obtain probabilistic feature values.

[0047] Then, for the probabilized eigenvalues ​​p(v) i The final-state decayed radiant intensity is measured using the hyperbolic representation of an exponential function, and the final-state decayed radiant intensity value is expressed as:

[0048]

[0049] Where e represents the natural constant, σ i This represents the final state decay radiation intensity metric.

[0050] Next, the final-state decay radiation intensity metric is subjected to stable frame normalization calibration in the full information domain to obtain the stable frame normalization calibration factor, expressed as:

[0051]

[0052] Where, ρ i In this embodiment, v represents the stable frame normalization calibration factor. ti Each feature value in the set of entropy projection vectors is used to semantically encode the initial device activity context.

[0053] In this way, the risk of dynamic decay radiation spillover under the terminal boundary conditions of the final state decay radiation intensity metric within the full information domain stability framework can be reconciled.

[0054] Therefore, downward coupling modulation of the sequence distribution is performed, that is, based on the final state decay radiation intensity metric and the stable frame normalization calibration factor, phase-transform coupling modulation is applied to the probabilistic eigenvalues ​​to obtain optimized probabilistic eigenvalues, expressed as:

[0055]

[0056] Finally, the optimized probabilistic eigenvalue p(v) is used. i ) ′ To calculate the information entropy representation v′ti =p(v i )′log2p(v i )′, to obtain the set of device activity context semantic encoding entropy projection vectors, where v ′ti Each feature value in the set of entropy projection vectors is used to semantically encode the device activity context.

[0057] In this way, by reasonably integrating the terminal feature value final state representation decay and the full information domain stability framework of the probabilistic information representation sequence, the stable coupling balance of the sequence distribution direction is improved, thereby compensating for the terminal dynamic balance defect of the information projection mapping and improving the Polar code encoding effect of the information bit mapping of the urgency of each device obtained by feature decoding regression.

[0058] Specifically, in step S3, the original status report data of each device is encoded using Polar codes to obtain the device status report Polar codes for each device. It should be understood that, due to physical factors such as environmental noise, bandwidth limitations, signal fading, and multipath interference, if the status data reported by the device is broadcast using traditional plaintext or inefficient encoding, it is highly susceptible to bit flipping, distortion, packet loss, or codeword misjudgment during transmission over the link. This can lead to errors in data restoration at the central end, device scheduling failures, or even abnormal switching of control. Especially in scenarios with multiple concurrent tasks and dense device traffic, unreliable transmission will further exacerbate resource contention and scheduling chaos, even threatening security and user experience. Therefore, this application further introduces Polar codes as the core of the transmission encoding, addressing the shortcomings of existing solutions in reliable transmission under high-noise and complex wireless channels. Polar codes offer advantages such as channel capacity performance approaching the Shannon limit, simple and easy-to-implement structure, support for flexible bit allocation, and high-performance soft-decision decoding.

[0059] In this application's technical solution, Polar code encoding of the device's original status report data provides extremely strong error correction and anti-interference capabilities. This ensures that critical business data (such as master control switching and emergency dispatch instructions) can still be determined and recovered with high confidence even in extreme environments, effectively preventing end-to-end scheduling failures caused by the loss of a single bit and improving the system's robustness and security. Furthermore, the Polar code encoding structure inherently supports an adaptive bit allocation strategy that flexibly selects information bits and frozen bits. This allows for prioritizing and increasing the fault tolerance level of important information based on different devices, event urgency, and data packet length, further serving the dynamic priority business needs of multi-device environments.

[0060] In one embodiment, such as Figure 4 As shown, the original status report data of each device is encoded using Polar codes to obtain the device status report Polar codes for each device, including:

[0061] S31, based on the mapping rules, the device ID, activity type, urgency level and timestamp in the original status report data of each device are mapped to information bits to obtain an information bit sequence;

[0062] S32, calculate the CRC checksum of the information bit sequence, and add the CRC checksum to the end of the information bit sequence to obtain the information bit sequence to be encoded;

[0063] S33, the sequence of information bits to be encoded is filled into the Ploar code input vector to obtain an intermediate vector;

[0064] S34, Perform Polar encoding transformation based on Polar code generation matrix on the intermediate vector to obtain the device status report Polar code of each device.

[0065] Specifically, the raw status report data is first mapped into a sequence of information bits that can participate in Polar code encoding. The mapping step requires fixed-length, reduced mapping for each information component to ensure that all status reports generate a fixed-format bit stream. For example, the device ID can be a unique device identifier, fixedly converted to a 16-bit binary code; the activity type is mapped to a 4- to 8-bit binary code through a predefined enumeration table, ensuring comprehensive coverage of all common activity types; the urgency level is normalized and then encoded with fixed-length multi-bit codes to reflect the priority range (e.g., using 3 bits to represent 8 levels of urgency); the timestamp is usually converted to a 32- or 64-bit binary stream using UNIX time with system seconds or higher precision to ensure accurate timing reproduction. These bit fields are concatenated in a set order to form a preliminary information bit sequence. For logical rigor and communication security, redundant check bits are introduced on top of this. A common practice is to calculate the CRC (Cyclic Redundancy Check) check code for the entire information bit sequence or a specific part thereof and add the CRC code to the end. For example, standard schemes such as CRC-8 and CRC-16 can be used, flexibly selected according to the overall code length and check strength requirements. After the CRC code is involved, the final sequence of information bits to be encoded is obtained.

[0066] After initial information bit encoding, the sequence is embedded into the Polar code input vector. The length of the Polar code input vector is typically a power of 2 (N = 2^n), and longer than the information bit sequence itself. A pre-defined set of information bit indices determines which vector indices are used to place information bits, and the remaining indices are frozen bits with 0 (based on the optimal polarization bit channel allocation principle, weak channels are allocated frozen bits). The specific filling process is as follows: each bit is sequentially placed into the input vector component at the position corresponding to the information bit index, and unoccupied bits are filled with 0. In a specific embodiment, if the Polar code input vector length is 128 and 40 information bits need to be encoded, then firstly, 40 optimal information bit indices are determined according to the channel polarization criterion and the current channel estimation strategy, and these are filled into the encoded bits. The remaining 88 bits are all filled with 0, resulting in an initial intermediate vector of length 128. It should be noted that this process not only affects the error correction performance of the Polar code but also relates to the decoding complexity and latency. The information bit selection strategy should be dynamically and adaptively optimized according to the scenario. In addition, the order of information bit filling must be strictly consistent with the central decoding mapping to ensure consistent decoding and reconstruction.

[0067] That is, in one embodiment, filling the sequence of information bits to be encoded into the Ploar code input vector to obtain an intermediate vector includes: filling the sequence of information bits to be encoded into a preset position of the Ploar code input vector based on the information bit index set to obtain an initial intermediate vector; and filling the non-information index positions in the initial intermediate vector with zero values ​​to obtain the intermediate vector.

[0068] After obtaining the intermediate vector, the core Polar code encoding operation can be performed. Polar code encoding is essentially a linear block code, and its encoding transformation relies on the recursive Kronecker product structure of the polarization matrix. Let the input vector v be N-dimensional (length 2^n), then the Polar code is x = v·G, where G is an N×N Polar code generation matrix. G is typically generated from the fundamental matrix... The polarization generation matrix is ​​derived through recursive multiplication (applying the Kronecker product n times to F), ultimately forming an N-dimensional polarization generation matrix. In practice, efficient algorithms such as bit reversal and fast Fourier transform are used to avoid traversing and storing the entire matrix, greatly improving computational efficiency. The entire Polar encoding process follows these core steps: First, the information and the intermediate vector filled with frozen bits are input. Then, polarization operations at each level are recursively implemented from low to high (which can be understood using a reduction diagram). Bit-level XOR and copy operations are implemented through appropriate software or digital hardware, finally outputting the encoded sequence. After encoding, the original information has been polarized and mapped to highly redundant and dispersed codewords, enabling the realization of keywords with extremely strong anti-interference and error correction characteristics.

[0069] Polar code encoding outputs a fixed-length binary code stream (i.e., device status report Polar code), which can then be sent to downstream transmission or scheduling pipelines. It's worth noting that Polar coding design allows for flexible setting of the code rate (the ratio of information bits to total bits), supports diversity transmission, and soft-decision decoding, among other advanced features. For example, for higher-priority devices or extremely low signal-to-noise ratio scenarios, the system can dynamically adjust the number of information bits and increase channel allocation redundancy to achieve adaptive protection. Simultaneously, with specific soft-output decoders (such as SC, SCL, or BP decoders), even if some data bits are damaged, the true information can be restored to the maximum extent possible, thereby greatly improving end-to-end transmission reliability and stability.

[0070] Specifically, in step S4, the device status report PLoar codes of each device are sent to the central receiving end via the StarSpark wireless interface. It should be understood that in scenarios such as the Internet of Things, smart homes, and industrial control, device terminals need to maintain close data interaction with the central system continuously in a low-latency and high-reliability wireless channel environment; otherwise, it will severely restrict the ability of devices to switch control, synchronize status, and respond promptly. Traditional wireless interfaces such as Bluetooth Classic and traditional Wi-Fi, while possessing data transmission capabilities, can no longer meet the urgent needs of the new generation of multi-device intelligent architectures in terms of comprehensive indicators such as real-time performance, concurrency, multi-priority adaptation, and low power consumption. With the breakthrough in StarSpark technology, the capabilities of wireless interfaces in bandwidth, anti-interference, multi-connectivity, latency control, and high-concurrency scheduling have been greatly improved. Therefore, in the technical solution of this application, the device status report PLoar codes of each device are sent to the central receiving end via the StarSpark wireless interface. The StarSpark interface's protocol stack design combines the advantages of Bluetooth and Wi-Fi, supporting high-speed, high-bandwidth data synchronization while maintaining ultra-low latency and high reliability at the millisecond level. This feature is particularly important for real-time switching of master control among multiple devices, significantly reducing link congestion, improving scheduling accuracy, and greatly reducing service inconsistencies or loss caused by bandwidth resource contention.

[0071] In one embodiment, sending the device status report Polar codes of each device to the central receiving end via the StarScan wireless interface includes: determining the transmission priority of the device status report Polar codes of each device based on the urgency level in the original status report data of each device. It should be understood that during the process of delivering the device status report data obtained by encoding the device status reports with Polar codes to the central receiving end via the StarScan wireless interface according to priority, the device needs to actively perceive its own status changes and task urgency characteristics, and dynamically adjust the priority and scheduling strategy of data reporting. Specifically, this relies on calling an urgency assessment module on the device or middleware before generating the device status report Polar codes. Based on the urgency value recorded in the current status report, a transmission priority label is assigned to the report packets to be sent. For example, the system can define high, medium, and low priorities or perform more granular classification. If the local urgency level is determined to be the highest, its corresponding Polar code data packet obtains the highest level of wireless transmission scheduling rights. The principle for setting the urgency level can be based on specific needs of the event type, such as dangerous action detection, health abnormalities, strong user operation commands, etc. Once an urgency is identified, the transmission process is automatically accelerated.

[0072] The StarScan protocol stack itself supports a multi-priority data scheduling mechanism. At the underlying physical and MAC layers, the StarScan interface allows terminals to map high-priority service packets to priority channels for different service scenarios, achieving preemptive access and advance scheduling to ensure the real-time performance and stability of critical task data. At the system application layer, the protocol stack distinguishes different levels of data packets using Quality of Service (QoS) parameters and a multi-queue mechanism. After a Polar code packet arrives at the local transmission queue, the protocol stack allocates channel resources based on its priority label. For example, Polar code data for high-urgency events (such as medical emergency alarms) is directly inserted into the high-priority queue, using hardware or firmware mechanisms such as fast wake-up and frequency modulation to activate the network card and immediately initiate channel contention, competing for the nearest idle time slot for wireless transmission. Medium and low-priority data may be delayed in scheduling, waiting during service conflicts until allowed to enter the channel. The StarScan interface also supports concurrent data reporting from multiple devices and automatic frequency adjustment to avoid data packet loss and congestion caused by multiple terminals competing for channel space. It coordinates load balancing among multiple terminals through mechanisms such as TDMA (Time Division Multiple Access) and FDMA (Frequency Division Multiple Access). For the device side, it is only necessary to call the StarScan stack interface to submit the Polar code data with priority tags to the local wireless stack using a standard API (such as asynchronous message packet delivery). Subsequent data encapsulation, channel allocation, verification and retransmission will be automatically completed by the StarScan protocol, which greatly simplifies the terminal load and ensures reporting efficiency.

[0073] Specifically, in step S5, the central receiving end decodes the Polar codes of the device status reports of each device to obtain the original status report data of each device. It should be understood that each device only possesses its own input events and partial local computing power, and needs to remotely transmit structured, priority-sensitive status data to the central end (central server, router, edge nodes, etc.). The central end then uniformly completes the global status aggregation, priority judgment, and master control transfer decision among multiple devices. During wireless channel transmission, any changes in the physical environment, such as bandwidth limitations, channel fading, external interference, or sudden noise, can lead to problems such as bit loss, misalignment, flipping, and occasional crashes in the transmitted bit stream. If the device status report data is transmitted in plaintext or uses a simple verification method, the central end is highly susceptible to receiving distorted or even unrecoverable invalid information. More seriously, master device switching is often strongly correlated with critical events such as security scheduling and anomaly alarms. If the central end cannot reliably restore the original status data uploaded by each device, it will directly affect the credibility and security of the entire system's services. Therefore, the central receiving end decodes the Polar code of the device status report of each device to obtain the original status report data of each device.

[0074] Specifically, Polar codes, as a forward error correction coding scheme that theoretically approaches the channel capacity limit, are designed to achieve the conversion from noisy input to noiseless, reliable output through mathematical coding fault tolerance under the premise of interference and impairment in the communication channel. After Polar code encoding of the raw state report data at the device end, the state data is mapped to a long bit sequence with high redundancy and polarization dispersion. Each bit undertakes a part of the information fault tolerance and repair responsibility, greatly enhancing fault resistance. The Polar code decoding process at the central end is used to reverse this polarization mapping, reconstructing the original structured information from the received potentially damaged codewords, ensuring the accuracy of the basic data for decision-making.

[0075] In one embodiment, the Polar code decoding process mainly consists of five stages: reception, soft-decision input, polarization path tracking, information bit reconstruction, and verification correction. First, the central receiver receives Polar codewords from multiple devices via a StarScan wireless interface. For each codeword, the system samples it bit by bit to obtain corresponding soft information such as received signal strength and channel characteristic parameters. Unlike traditional hard-decision systems (where a 1 or 0 is received and the decision is made directly), modern Polar code systems typically employ a soft-decision mechanism, quantizing the receiver output of each bit as a probability / log-likelihood ratio (LLR) to capture ambiguous statistical properties during channel transmission. This soft-information mechanism provides significant error tolerance for subsequent decoding processes.

[0076] Polar code decoding algorithms can be mainly divided into Successive Cancellation (SC) and Successive Cancellation List (SCL) types. Taking the most basic SC decoding algorithm as an example: after receiving a complete codeword of length N = 2^n, a set of LLR input vectors of length N is first initialized, where each LLR represents the probability strength of whether the current bit is 0 or 1. The key to Polar code decoding lies in utilizing its hierarchical polarization structure to recursively perform path segmentation and aggregation from high to low order. The specific process of SC decoding is as follows: starting from the first bit (frozen or information bit), based on the recursive relationship of the decoded bits and polarization matrix, at each step, based on the previous results and LLR scores, the probability of the next position is updated, and the optimal path is given according to the preset zero-freezing rule or blind judgment rule. For frozen bits in Polar codes (usually filled with 0), the receiver automatically forces them to 0 during decoding; for non-frozen bits (that is, actual information bits, including device ID, activity type, urgency, and timestamp, etc.), the corresponding values ​​are obtained by progressively recursively judging. While SC decoding is simple in principle and computationally inexpensive, its robustness to some high-error-rate channels is limited. To further improve decoding performance, especially in multi-interference scenarios, the SCL decoding method is commonly used. Its core strategy is path diversity: in the same recursive stage, it not only retains one decoding path, but also concurrently retains L optimal candidate paths (L is the list size, typically 4, 8, 16, etc.), gradually filtering to the final most likely information vector. The SCL method greatly reduces the risk of single-point flipping destroying the full codeword decoding result and is the standard Polar decoding algorithm in 5G and high-security communications. For SCL decoding, a CRC-assisted mechanism is often combined, that is, the path that passes the CRC check is selected first among all candidate results, thereby improving decoding reliability and accelerating the decision selection.

[0077] After Polar code decoding, the information bit sequence of the original state structured data can be obtained. Since the data packet has been mapped sequentially to fields such as device ID, activity type, urgency level, and timestamp before encoding, the structured report can be restored by simply separating these intervals according to the protocol after decoding. For example, the first 16 bits correspond to the device ID, the next 4 to 8 bits are the activity type, 3 or 8 bits are the urgency level, 32 or 64 bits are the timestamp, and the CRC check bits extracted again according to the length are used for the final data packet integrity confirmation. By recalculating the CRC after receiving and decoding and comparing it with the decoded CRC bits, the probability of codeword misinterpretation can be greatly reduced. Specifically, if the comparison fails, it means that the decoding result is unreliable, and the central end can trigger retransmission, soft decision fault tolerance, or handle the data with a safety fallback to prevent scheduling logic errors due to data distortion.

[0078] Specifically, in step S6, the central receiving end updates the priority queue based on the original status report data of each device to obtain the updated priority queue. It should be understood that in a multi-device intelligent collaborative system, the system needs to rapidly and intelligently allocate master control or resources based on changes in the internal and external environment, end-user needs, and responses to emergencies, ensuring that the most suitable device is promoted to the current master device at the most appropriate time. The priority queue is the core mechanism for solving this multi-scenario dynamic decision-making and scheduling problem. For the central receiving end, by continuously analyzing and updating the priority queue, on the one hand, devices with the most urgent interaction needs or those undertaking critical services are given priority to obtain master control; on the other hand, it can respond promptly to changes in user behavior and emergencies in the environment, improving the overall system's agility, real-time performance, and service experience. In a multi-terminal environment, simple polling or static allocation is insufficient to reflect the personalization and dynamism under multiple events, multiple services, and multiple scenarios. Priority queue management effectively overcomes this shortcoming, achieving optimal utilization of scheduling resources and intelligent guarantee of service timeliness.

[0079] In one embodiment, such as Figure 5 As shown, the central receiving end updates the priority queue based on the original status report data of each device to obtain the updated priority queue, including: S61, extracting the urgency level of each device from the original status report data of each device; S62, generating the priority of each device based on the urgency level of each device; S63, updating the current priority queue based on the priority of each device to obtain the updated priority queue.

[0080] In one specific embodiment, urgency information is first extracted from the raw status report data of each device. The raw status report data is decoded using Polar codes to recover key information including the device's unique ID, activity type, urgency level, and timestamp. During each data report or periodic polling, the central receiving end iterates through all connected devices participating in scheduling, decoding and parsing their latest one or more status report data. Urgency is generally expressed as a floating-point value (0-1), an integer (1-10), or an enumerated type like "low / medium / high," depending on the front-end feature regression model and the system's requirements for priority granularity. For each device, its urgency level is combined with comprehensive indicators such as the important tasks currently pending on the device and its current scenario to calculate a priority score. In a more specific embodiment, a weighted scoring formula can be used, incorporating urgency, event freshness (e.g., seconds or milliseconds since the latest timestamp), device activity, and historical priority in the calculation, thus balancing real-time performance and the continuity of device behavior. A comprehensive evaluation formula such as "Priority = Urgency × Coefficient 1 + Activity × Coefficient 2 - Expiration × Coefficient 3" can be set to quickly increase the priority of relevant devices for highly urgent new events, while the priority of old events or passive devices will decay over time.

[0081] The level of urgency itself is the most direct indicator reflecting the status anomalies and urgency of tasks among multiple devices. For example, in a smart home environment, when a smoke alarm detects a high concentration of smoke, it will submit a status report with extremely high urgency. The system should be immediately prioritized to switch control to the alarm to achieve a rapid response and issue end-to-end alerts. In a health monitoring scenario, when an ECG device detects abnormal waveforms, its urgency level is higher than that of a regular step counter or low-priority office equipment.

[0082] Once the central receiving end obtains the latest urgency data for all devices, it can begin maintaining and updating the priority queue. The queue is implemented using a dynamic stack or a self-balancing priority queue (such as a max-heap or a weighted balanced binary tree) to ensure efficient real-time insertion, lookup, and eviction capabilities. For multiple devices reporting within the same time slice, the priority scores of all devices are aggregated, and the priority queue is automatically reconstructed according to decreasing order. During task execution, the priority queue is polled or automatically triggered, starting with the head device (i.e., the highest priority) to process its requests or grant it control. After a high-priority device switches over, if its urgency level stabilizes (e.g., an abnormal alarm is cleared), the central end will detect its priority decrease and automatically adjust the queue order after the next status report. Throughout the process, the priority queue update rhythm can use a fixed time window or be driven in real-time by new status push events, thus balancing real-time response and resource consumption. Some advanced implementations can also introduce logic such as expiration markers, minimum hold time, or priority inheritance to avoid frequent queue jitter under high-frequency rendering, improving stability and anti-interference capabilities.

[0083] Beyond basic priority score sorting, other examples in this application incorporate multi-dimensional strategies into priority queue maintenance. For instance, if preset master control whitelists, blacklists, or special states (such as silent or hibernating) exist among multiple devices, secondary filtering and selection are performed based on the priority score sorting results, and exception policies are triggered for special events. For example, if a user has indeed set the master device to be disabled for switching, even if a new event is reported with extremely high priority, master control may not be switched; only event warnings and interface reminders will be issued to balance user experience and security protection. In higher-level scenarios, priority queues can also be used in conjunction with machine learning modules to automatically analyze device behavior history, user preferences, and changes in scene patterns, dynamically and adaptively adjusting the scoring and scheduling frequency of the same type of emergency events, further enhancing the system's user-friendliness and intelligent experience.

[0084] Specifically, in step S7, the central receiving end checks the highest priority device extracted from the priority queue and determines whether to switch to the highest priority device based on a comparison between the priority of the highest priority device and the priority of the current master device. It should be understood that in real-world scenarios where multiple devices collaborate, different devices may undertake different tasks, provide different services, or be operated by different users. The reasonable allocation and dynamic transfer of system control are fundamental to improving terminal collaboration capabilities, response efficiency, and intelligent perception. If the master device remains fixed or lingers on a particular device for an extended period, it can easily lead to unfair resource allocation, sluggish user experience, and even failure to respond promptly to emergencies. Therefore, an intelligent switching mechanism based on global information and capable of real-time judgment of device activity and priority is provided.

[0085] Specifically, during continuous operation, whenever the central receiving end completes the decoding of the status report data from each device and successfully updates the priority queue, it needs to immediately extract and determine the highest priority device based on this priority queue. The priority queue data structure is a heap or sorted linked list arranged in descending order of device urgency (and auxiliary evaluation indicators). This queue constantly reflects the task priority, activity level, or probability of emergency events occurring for all devices at the current moment. Checking the highest priority device involves selecting the device node with the highest ranking at that moment as the current potential master control candidate.

[0086] In one embodiment, the central receiver typically extracts the head device of the priority queue as the latest highest-priority device each time it receives a new batch of status report data or periodically polls the priority queue. It maintains the unique ID of the current master device and its corresponding priority value, comparing it numerically, hierarchically, or within a range with the priority of the highest-priority device. The comparison standard can be an absolute priority comparison (e.g., a switchover can be triggered as long as the urgency value is strictly higher), or a switchover threshold or range can be set to prevent frequent fluctuations. For example, it can be stipulated that a switchover will only be triggered when the priority increase (highest-priority device priority - current master device priority) exceeds a set threshold, or when the highest-priority device reaches a specific emergency category, thus achieving a master control rotation mechanism that is neither overly sensitive nor overly secure.

[0087] In a specific embodiment, the comparison and switching logic is implemented through the following process: Each time the central receiving end completes a queue update, it calls the master control decision function to first obtain the ID and priority of the current master device, and then obtain the ID and priority of the head device (highest priority device). If the two IDs are the same or the priority of the highest priority device does not exceed that of the current master device, the existing master control structure is maintained, and no switching is triggered. Conversely, when the priority of the head device is significantly higher than that of the current master device, a switching decision is initiated immediately, including resource reallocation, interface focus migration, and abnormal linkage command issuance. After the switching is completed, the master control device ID is updated synchronously, and all sub-modules are globally notified so that downstream services (such as log recording, event push, automatic linkage, etc.) adapt to the new master device in real time.

[0088] In summary, the multi-device intelligent handover system and method based on Polar codes provided in this application first collects input event streams in real time from each device side and obtains raw status report data through intelligent parsing. Then, the data is encoded using Polar codes, effectively enhancing the anti-interference and error correction capabilities of the data during wireless transmission. The encoded status data is efficiently transmitted to the central receiving end via the StarScan wireless interface according to a preset priority. After the central end completes the Polar code decoding, it promptly restores the original status of all devices. Based on the urgency and activity status of each device, the central end dynamically updates the priority queue and automatically determines the priority relationship between the current master device and the highest priority device, intelligently deciding whether to perform a master control handover. This achieves efficient and reliable status reporting and intelligent real-time handover of master devices in a multi-device environment.

[0089] This application also provides a multi-device intelligent switching system 600 based on Polar codes. Figure 6 The illustration shows a schematic block diagram of a multi-device intelligent handover system based on Polar codes according to an embodiment of this application. Figure 6 As shown, the multi-device intelligent handover system based on Polar codes includes:

[0090] The input event stream acquisition module 601 is used to acquire the input event stream of each device;

[0091] The raw status report data generation module 602 is used to generate raw status report data for each device based on the input event stream of each device. The raw status report data includes device ID, activity type, urgency level and timestamp.

[0092] Polar code encoding module 603 is used to encode the original status report data of each device into Polar code to obtain the device status report Polar code of each device.

[0093] The device status Polar code transmission module 604 is used to send the device status report Polar code of each device to the central receiving end through the Star Flash wireless interface.

[0094] Polar code decoding module 605 is used by the central receiving end to perform Polar code decoding on the device status report Polar codes of each device to obtain the original status report data of each device.

[0095] The priority queue update module 606 is used by the central receiving end to update the priority queue based on the original status report data of each device to obtain the updated priority queue.

[0096] The device switching decision module 607 is used by the central receiving end to check the highest priority device extracted from the priority queue, and to determine whether to switch to the highest priority device based on the comparison between the priority of the highest priority device and the priority of the current master device.

[0097] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0098] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0099] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for intelligent handover of multiple devices based on Polar codes, characterized in that, include: Acquire the input event stream from each device; Based on the input event stream of each device, the original status report data of each device is generated. The original status report data includes device ID, activity type, urgency level and timestamp. The original status report data of each device is encoded using Polar code to obtain the device status report Polar code for each device. The device status report Ploar code of each device is sent to the central receiving end through the Star Flash wireless interface; The central receiving end decodes the Polar code of the device status report of each device to obtain the original status report data of each device. The central receiving end updates the priority queue based on the original status report data of each device to obtain the updated priority queue; The central receiving end checks the highest priority device extracted from the priority queue and determines whether to switch to the highest priority device based on the comparison between the priority of the highest priority device and the priority of the current master device.

2. The multi-device intelligent handover method based on Polar codes according to claim 1, characterized in that, Based on the input event streams of each device, the original status report data of each device is generated, including: The input events of each device are parsed, and the device ID, activity type, activity data, and timestamp are identified from them. Based on the activity data of each device, the urgency level of each device is generated.

3. The multi-device intelligent handover method based on Polar codes according to claim 2, characterized in that, Based on the activity data of each device, the urgency level of each device is generated, including: The activity data of each device is structured and encoded to obtain a set of structured encoded vectors for device activity data; The set of structured encoded vectors of the device activity data is input into the device activity context encoder based on the converter architecture to obtain the set of semantic encoded vectors of the device activity context. Feature decoding regression is performed on each device activity context semantic encoding vector in the set of device activity context semantic encoding vectors to obtain the urgency level of each device.

4. The multi-device intelligent handover method based on Polar codes according to claim 2, characterized in that, Based on the activity data of each device, the urgency level of each device is generated, including: The activity data of each device is structured and encoded to obtain a set of structured encoded vectors for device activity data; The set of structured encoded vectors of the device activity data is input into the device activity context encoder based on the converter architecture to obtain the set of semantic encoded vectors of the device activity context. An information projection mapping is performed on the set of device activity context semantic encoding vectors to obtain a set of device activity context semantic encoding entropy projection vectors; The urgency level of each device is obtained by performing feature decoding regression on each device activity context semantic encoding entropy projection vector in the set of device activity context semantic encoding entropy projection vectors.

5. The multi-device intelligent handover method based on Polar codes according to claim 4, characterized in that, An information projection mapping is performed on the set of device activity context semantic encoding vectors to obtain a set of device activity context semantic encoding entropy projection vectors, including: Each feature value in the set of device activity context semantic encoding vectors is probabilistically activated to obtain probabilistically encoded feature values; For the probabilized eigenvalues, the final state decayed radiation intensity is measured using the hypercurvature representation of an exponential function to obtain the final state decayed radiation intensity metric value. The stable frame normalization calibration factor is obtained by performing a full-information domain normalization calibration on the final state decay radiation intensity metric. Based on the final state decayed radiation intensity metric and the stable frame normalization calibration factor, the probabilistic eigenvalues ​​are subjected to phase-conversion coupling modulation to obtain optimized probabilistic eigenvalues. The information entropy representation is calculated using the optimized probabilistic feature values ​​to obtain a set of device activity context semantic encoding entropy projection vectors.

6. The multi-device intelligent handover method based on Polar codes according to claim 1, characterized in that, The original status report data of each device is encoded using Polar code to obtain the device status report Polar code for each device, including: Based on the mapping rules, the device ID, activity type, urgency level and timestamp in the original status report data of each device are mapped to information bits to obtain the information bit sequence; Calculate the CRC checksum of the information bit sequence and add the CRC checksum to the end of the information bit sequence to obtain the information bit sequence to be encoded; The sequence of information bits to be encoded is filled into the Ploar code input vector to obtain an intermediate vector; The intermediate vector is subjected to Polar encoding transformation based on the Polar code generation matrix to obtain the device status report Polar code for each device.

7. The multi-device intelligent handover method based on Polar codes according to claim 6, characterized in that, The process of filling the sequence of information bits to be encoded into the Ploar code input vector to obtain an intermediate vector includes: The sequence of information bits to be encoded is filled into the preset positions of the Ploar code input vector based on the information bit index set to obtain the initial intermediate vector; The positions of the non-information indexes in the initial intermediate vector are filled with zero values ​​to obtain the intermediate vector.

8. The multi-device intelligent handover method based on Polar codes according to claim 7, characterized in that, Sending the device status report Ploar codes of each device to the central receiving end via the StarFlash wireless interface includes: determining the sending priority of the device status report Ploar codes of each device based on the urgency of the original status report data of each device.

9. The multi-device intelligent handover method based on Polar codes according to claim 1, characterized in that, The central receiving end updates the priority queue based on the original status report data of each device to obtain the updated priority queue, including: Extract the urgency level of each device from the original status report data of each device; Based on the urgency level of each device, a priority is generated for each device; The current priority queue is updated based on the priority of each device to obtain the updated priority queue.

10. A multi-device intelligent handover system based on Polar codes, characterized in that, include: The input event stream acquisition module is used to acquire the input event streams of each device; The raw status report data generation module is used to generate raw status report data for each device based on the input event stream of each device. The raw status report data includes device ID, activity type, urgency level and timestamp. The Polar code encoding module is used to encode the original status report data of each device into Polar codes to obtain the device status report Polar codes of each device. The device status Polar code transmission module is used to send the device status report Polar code of each device to the central receiving end through the Star Flash wireless interface. The Polar code decoding module is used by the central receiving end to decode the Polar codes of the device status reports of each device to obtain the original status report data of each device. The priority queue update module is used by the central receiving end to update the priority queue based on the original status report data of each device to obtain the updated priority queue. The device switching decision module is used by the central receiving end to check the highest priority device extracted from the priority queue, and to determine whether to switch to the highest priority device based on the comparison between the priority of the highest priority device and the priority of the current master device.

Images