Heterogeneous wireless network adaptive vertical handover method, terminal device and computer readable storage medium

Abstract

This disclosure provides a method for adaptive vertical handover in heterogeneous wireless networks, a terminal device, and a computer-readable storage medium. The method includes: acquiring a time series of the received signal strength of a first network; obtaining a predicted value of the received signal strength based on the time series; obtaining a prediction error based on the measured value of the received signal strength and the predicted value of the received signal strength, and obtaining a link fluctuation index based on the prediction error at multiple times; calculating a threshold shift amount based on the link fluctuation index, and synchronously shifting the reference handover threshold and reference handover back threshold of the first network by the threshold shift amount to generate a dynamic handover threshold and a dynamic handover back threshold; and when the terminal is currently connected to the first network, switching from the first network to a second network in response to the predicted value of the received signal strength being less than the dynamic handover threshold; and when the terminal is currently connected to the second network, switching from the second network to the first network in response to the measured value of the current received signal strength of the first network being greater than or equal to the dynamic handover back threshold.

Description

Technical Field

[0001] This disclosure relates to an adaptive vertical handover method for heterogeneous wireless networks, a terminal device, and a computer-readable storage medium. Background Technology

[0002] With the widespread deployment of various wireless access technologies, terminal devices are often located in areas where heterogeneous networks such as Wi-Fi and cellular networks overlap. To maintain continuous service transmission, terminals need to perform vertical switching between different networks.

[0003] Currently, common vertical handover methods rely on received signal strength for decision-making. These methods determine whether to trigger network handover by using preset handover-out and handover-back thresholds: when the received signal strength of the currently accessed network is lower than the handover-out threshold, the terminal switches to another network; when the received signal strength of the other network is higher than the handover-back threshold, the terminal switches back to the original network. However, the above methods have the following problems: First, since the handover decision is based solely on the received signal strength at the current moment, when the terminal moves rapidly or is blocked, causing the signal to continuously attenuate, the handover trigger often lags behind the deterioration of the link quality, making it difficult to complete the handover in time before the network connection is interrupted.

[0004] Second, in scenarios with strong fluctuations in the wireless environment, short-term jitter in the received signal strength can easily trigger switching conditions, causing the terminal to perform unnecessary round-trip switching between multiple networks, affecting the stability of service transmission.

[0005] Third, fixed handover and handover thresholds are difficult to adapt to time-varying wireless environments: when the link continues to deteriorate, fixed thresholds may not respond in time; when signal fluctuations increase, fixed thresholds may be too sensitive to short-term disturbances, making it difficult to balance the timeliness and stability of handover. Summary of the Invention

[0006] This disclosure provides a method for adaptive vertical handover in heterogeneous wireless networks, a terminal device, and a computer-readable storage medium.

[0007] According to one aspect of this disclosure, an adaptive vertical handover method for heterogeneous wireless networks is provided, comprising: acquiring a time series of the received signal strength of a first network; predicting the received signal strength of the first network at future times based on the time series to obtain a predicted value of the received signal strength; obtaining a prediction error based on the measured value of the received signal strength at the same time and the predicted value of the received signal strength; obtaining a link fluctuation index based on the prediction errors at multiple times, the link fluctuation index being used to evaluate the degree of fluctuation of the received signal strength of the first network; and calculating a threshold shift amount based on the link fluctuation index, the threshold shift amount being calculated based on the link fluctuation index and used to adjust a reference handover threshold and... The reference cut-back threshold is shifted by a certain value; the reference cut-out threshold and reference cut-back threshold of the first network are synchronously shifted by the threshold shift amount to generate dynamic cut-out threshold and dynamic cut-back threshold; and when the terminal is currently connected to the first network, in response to the predicted value of the received signal strength being less than the dynamic cut-out threshold, the terminal's access network is switched from the first network to the second network; when the terminal is currently connected to the second network, in response to the current measured value of the received signal strength of the first network being greater than or equal to the dynamic cut-back threshold, the terminal's access network is switched from the second network to the first network, wherein the first network and the second network are heterogeneous wireless networks.

[0008] According to one aspect of the technical solution of this disclosure, by obtaining the time series of the received signal strength of the first network and predicting the received signal strength at future times based on the time series, a predicted value of the received signal strength is obtained, so that the handover decision can perceive the changing trend of the link quality in advance, rather than relying solely on the measured value of the received signal strength at the current time, thereby solving the handover lag problem caused by making decisions solely based on the received signal strength at the current time. The prediction error is obtained by comparing the received signal strength measurement and the predicted received signal strength at the same time. The link fluctuation index is obtained based on the prediction errors at multiple times. The link fluctuation index is used to evaluate the degree of received signal strength fluctuation in the first network. Then, the threshold shift is calculated based on the link fluctuation index. The reference handout threshold and reference handback threshold of the first network are synchronously shifted by the threshold shift to generate dynamic handout threshold and dynamic handback threshold. This allows the handover determination conditions to be adaptively adjusted according to the degree of link fluctuation: when the link fluctuation increases, the synchronously shifted dynamic handout threshold and dynamic handback threshold make handout trigger earlier and handback trigger more difficult, thereby suppressing ping-pong handover; when the link fluctuation decreases, the dynamic handout threshold and dynamic handback threshold fall back to the reference level, restoring normal handover response capability. This solves the problem that fixed handout threshold and handback threshold are difficult to balance handover timeliness and stability.

[0009] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method, predicting the received signal strength of the first network at a future time based on the time series includes: processing the time series using Kalman filtering to obtain a predicted value of the received signal strength of the first network at a future time.

[0010] In the technical solution of this embodiment, since Kalman filtering can recursively estimate based on the changing trend of the received signal strength, when the access network link quality continues to decline, the predicted value reflects this deterioration trend earlier than the current measurement value, thereby providing a basis for triggering network handover in advance.

[0011] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method of the present disclosure, the time series is processed by Kalman filtering, including: modeling the received signal strength as a state vector containing a received signal strength state component and a received signal strength change trend state component; and performing state prediction and measurement update by Kalman filtering based on the state vector to obtain the predicted value of the received signal strength.

[0012] In this embodiment, the state vector simultaneously includes a received signal strength state component and a received signal strength change trend state component. This allows the Kalman filter to not only estimate the current received signal strength level during the recursive process but also track its rate and direction of change. When the received signal strength exhibits a continuous decreasing trend, the change trend state component drives the predicted value to decrease further, enabling the handover decision to identify link degradation risks earlier and trigger wireless network handover.

[0013] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method, based on the time series, predicting the received signal strength of the first network at a future time includes: processing the time series using at least one of exponential smoothing, autoregressive prediction, or grey prediction to obtain the predicted value of the received signal strength.

[0014] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method of the present disclosure, a link fluctuation index is obtained based on prediction errors at multiple times, including: performing exponential moving average processing on the absolute values ​​of prediction errors at multiple times, and using the processing result as the link fluctuation index.

[0015] In this embodiment, the absolute value of the prediction error is processed by exponential moving average, which smooths out the occasional jumps in a single prediction error, making the link fluctuation index more stably reflect the changing trend of the received signal strength fluctuation. The exponential moving average assigns a higher weight to recent prediction errors, enabling the link fluctuation index to respond quickly when environmental fluctuations increase and gradually decrease when the environment stabilizes. This provides a smoother input for calculating the threshold shift and avoids drastic fluctuations in the handover decision conditions due to individual abnormal errors.

[0016] The heterogeneous wireless network adaptive vertical handover method according to at least one embodiment of the present disclosure obtains a link fluctuation index based on prediction errors at multiple times, including: calculating the variance of prediction errors at multiple times and using the variance as a link fluctuation index.

[0017] In the technical solution of this embodiment, the variance of the prediction error reflects the degree of dispersion of the prediction error from a statistical perspective. When the wireless environment fluctuates more, the variance of the prediction error increases, and the link fluctuation index rises accordingly; when the wireless environment tends to be stable, the variance of the prediction error decreases, and the link fluctuation index decreases accordingly.

[0018] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method of the present disclosure, the prediction error is at least one of absolute error, squared error, moving average error or exponential moving average error.

[0019] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method of the present disclosure, a threshold shift amount is calculated based on the link fluctuation index, specifically including: multiplying the link fluctuation index by the shift gain coefficient as the threshold shift amount.

[0020] In this embodiment, the translation gain coefficient is used to adjust the threshold's sensitivity to link fluctuations: the larger the translation gain coefficient, the greater the threshold shift caused by the same degree of link fluctuation, and the more sensitive the handover judgment condition is to link fluctuations; the smaller the translation gain coefficient, the smoother the response. The mapping method of the translation gain coefficient is simple and easy to adapt to actual needs by adjusting the translation gain coefficient in different terminal or network environments.

[0021] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method of the present disclosure, the threshold shift amount does not exceed the upper limit of the shift.

[0022] In this embodiment, an upper limit is set for the threshold shift amount to prevent excessive shifting of the reference handover threshold and reference handover threshold when the received signal strength fluctuates drastically. This would prevent the dynamic handover threshold from being too high or the dynamic handover threshold from being too high, thus avoiding premature handover or failure to handover back. The upper limit constraint keeps the adjustment of the dynamic threshold within a reasonable range, ensuring that the handover method behaves more controllably in extreme wireless environments.

[0023] The heterogeneous wireless network adaptive vertical handover method according to at least one embodiment of the present disclosure further includes: when the terminal is currently connected to the first network, in response to the predicted value of the received signal strength being less than the dynamic handover threshold for N consecutive time periods, switching from the first network to the second network, where N is an integer greater than 1.

[0024] In this embodiment, the handover is triggered only if the predicted received signal strength is less than the dynamic handover threshold for N consecutive moments, thus avoiding erroneous handovers caused by the predicted value occasionally falling below the dynamic handover threshold at a single moment. In other words, by introducing a multiple continuous confirmation mechanism, the stability of the handover decision is further enhanced. Even when there are brief fluctuations in the predicted received signal strength, network changes will not be triggered immediately, making the terminal's stay in the current access network more reliable.

[0025] According to at least one embodiment of the heterogeneous wireless network adaptive vertical handover method of the present disclosure, obtaining the time series of the received signal strength of the first network includes: periodically collecting the received signal strength of the first network, generating a received signal strength observation sequence, and using the received signal strength observation sequence as the time series.

[0026] In the technical solution of this embodiment, the received signal strength of the first network is acquired by periodic acquisition, and the received signal strength observation sequence is generated as a time series, which provides a data basis for subsequent received signal strength prediction, fluctuation assessment and threshold adjustment.

[0027] According to another aspect of this disclosure, a terminal device is provided, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the heterogeneous wireless network adaptive vertical handover method described in any of the above embodiments.

[0028] According to another aspect of this disclosure, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the heterogeneous wireless network adaptive vertical handover method described in any of the above embodiments.

[0029] Beneficial effects The adaptive vertical handover method for heterogeneous wireless networks disclosed herein obtains the time series of the received signal strength of the first network and predicts the received signal strength at future times, enabling the handover decision to perceive the changing trend of link quality in advance, rather than relying solely on the measured value of the received signal strength at the current time, thereby avoiding handover triggering lagging behind the link deterioration process. By obtaining the prediction error based on the measured value of the received signal strength and the predicted value of the received signal strength at the same time, and constructing a link fluctuation index based on the prediction errors at multiple times, a quantitative assessment of the degree of fluctuation of the received signal strength of the first network is achieved. On this basis, a threshold shift is calculated based on the link fluctuation index, and the baseline handover-out threshold and baseline handover-back threshold are synchronously shifted to generate dynamic handover threshold and dynamic handover-back threshold. When the link fluctuation increases, the dynamic handover threshold and dynamic handover-back threshold are synchronously shifted upward, making handover trigger earlier and handover more difficult to trigger, thus suppressing ping-pong handover. When the link fluctuation decreases, the dynamic threshold falls back to the baseline level to restore normal handover response capability, thereby improving both the timeliness of handover triggering and the stability of network camping in a dynamic wireless environment. Attached Figure Description

[0030] The accompanying drawings illustrate exemplary embodiments of the present disclosure and, together with the description thereof, serve to explain the principles of the present disclosure. These drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this specification.

[0031] Figure 1 This is a flowchart illustrating an adaptive vertical handover method for heterogeneous wireless networks according to one embodiment of the present disclosure.

[0032] Figure 2 yes Figure 1 A flowchart illustrating the sub-steps of step S110.

[0033] Figure 3 yes Figure 1 A flowchart illustrating the sub-steps of step S120.

[0034] Figure 4 yes Figure 1 A flowchart illustrating the sub-steps of step S130.

[0035] Figure 5 yes Figure 1 A flowchart illustrating the sub-steps of step S140.

[0036] Figure 6 yes Figure 1 A flowchart illustrating the sub-steps of step S150.

[0037] Figure 7 This is a schematic diagram of a terminal device bus architecture according to one embodiment of the present disclosure.

[0038] Figure 8 This is a functional block diagram of a terminal device according to one embodiment of the present disclosure. Detailed Implementation

[0039] The present disclosure will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure. Furthermore, it should be noted that, for ease of description, only the parts relevant to the present disclosure are shown in the accompanying drawings.

[0040] It should be noted that, where there is no conflict, the embodiments and features described in this disclosure can be combined with each other. The technical solutions of this disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0041] Unless otherwise stated, the exemplary implementations / embodiments shown are to be understood as providing exemplary features of various details that provide ways in which the technical concepts of this disclosure can be implemented in practice. Therefore, unless otherwise stated, the features of various implementations / embodiments may be additionally combined, separated, interchanged and / or rearranged without departing from the technical concepts of this disclosure.

[0042] In existing technologies, in heterogeneous wireless network environments, mobile terminals often use a fixed dual-threshold method based on Received Signal Strength (RSS) for vertical handover decisions: preset handover-out thresholds and handover-back thresholds are used. When the current RSS is lower than the handover-out threshold, the mobile terminal hands over from the current network; when the RSS is higher than the handover-back threshold, the mobile terminal hands over. However, this method has three problems: it relies solely on the current RSS measurement value, lacking the ability to perceive link change trends; when the signal continues to attenuate, the handover trigger lags behind the link quality deterioration process, easily leading to service interruption; RSS is susceptible to short-term fluctuations due to multipath effects, shadow fading, and measurement noise; the fixed thresholds are sensitive to instantaneous disturbances, easily triggering ping-pong handovers in areas with overlapping signal coverage; and the fixed handover-out and handover-back thresholds cannot be adaptively adjusted according to changes in the wireless environment, making it difficult to simultaneously ensure the timeliness and stability of the handover.

[0043] To address this, the present disclosure proposes the following technical solution. In this solution, the terminal no longer makes handover decisions solely based on the current RSS measurement value. Instead, it obtains the RSS time series of the first network and predicts the RSS for future times based on this time series, thus obtaining an RSS prediction value. This allows the handover decision to anticipate link change trends in advance. Simultaneously, by calculating the prediction error between the RSS measurement value and the RSS prediction value at the same time, and processing the prediction errors at multiple times, a link fluctuation index is obtained. This index is used to quantitatively assess the degree of RSS fluctuation in the first network. Based on this, a threshold shift is calculated according to the link fluctuation index, and the baseline handover-out threshold and baseline handover-back threshold of the first network are synchronously shifted to generate dynamic handover-out threshold and dynamic handover-back threshold. When link fluctuation increases, the dynamic handover-out threshold and dynamic handover-back threshold move upwards synchronously, making handover trigger earlier and handover more difficult to trigger, suppressing ping-pong handover. When link fluctuation decreases, the dynamic handover-out threshold and dynamic handover-back threshold fall back to the baseline level, restoring normal handover response capability. In the handover decision, when the terminal is currently connected to the first network, the predicted RSS value is compared with the dynamic handover threshold. If the predicted RSS value is less than the dynamic handover threshold, a handover is triggered. When the terminal is currently connected to the second network, the current RSS measurement value of the first network is compared with the dynamic handover threshold. If the current RSS measurement value is greater than or equal to the dynamic handover threshold, a handover is triggered.

[0044] The application scenarios disclosed herein include: vertical switching between Wi-Fi and cellular networks for mobile terminals in continuous indoor and outdoor coverage scenarios; switching between different network types when a terminal moves through different hotspot areas in a smart park scenario; network switching when a vehicle moves at high speed through different network coverage areas in a vehicle-mounted mobile communication scenario; and switching between Wi-Fi and cellular networks in mobile office scenarios. In these application scenarios, terminals often face situations such as rapid signal attenuation, short-term drastic signal fluctuations, or signal abrupt changes caused by obstruction. The technical solution disclosed herein can detect link deterioration trends in advance, suppress ping-pong handover, and adaptively adjust handover conditions in these scenarios, thereby improving the continuity and stability of service transmission.

[0045] In some embodiments of this disclosure, the first network may be a primary network that the terminal preferentially accesses, and the second network may be a backup network used to maintain network connectivity when the quality of the primary network deteriorates. For example, the first network may be a Wi-Fi network and the second network may be a cellular network; or, the first network may be a cellular network and the second network may be a Wi-Fi network. The first network and the second network may also be combinations of other wireless access networks of different types or parameters, and this disclosure does not specifically limit them.

[0046] Figure 1This is a flowchart illustrating an adaptive vertical handover method for heterogeneous wireless networks according to one embodiment of the present disclosure.

[0047] like Figure 1 The method M100 shown includes steps S110 to S150. This method can be executed by an electronic device that supports two different types of wireless network access, such as a mobile phone, tablet computer, vehicle-mounted communication terminal, or portable communication device.

[0048] In step S110, the time series of the received signal strength of the first network is obtained.

[0049] Here, "first network" refers to the wireless network that the terminal currently prioritizes accessing or is evaluating, such as a Wi-Fi network. "Time series" refers to the set of received signal strength data arranged in chronological order of sampling time. Received signal strength refers to the power value of the wireless signal received by the terminal, usually measured in dBm, and can be obtained through the operating system's network status interface or the wireless network card driver interface.

[0050] For example, the time series can be a sequence containing received signal strength measurements of multiple consecutive sampling periods. The received signal strength measurement at the current moment is collected once in each sampling period and added to the sequence. The sequence length can be preset to a fixed window length. When a newly collected measurement value is added, the oldest measurement value is removed.

[0051] As an example, the terminal periodically samples the received signal strength of the Wi-Fi network with a sampling period of 1 second. The received signal strength sampled at the current moment is -68dBm, and the received signal strength sampled at the previous three sampling moments were -60dBm, -63dBm, and -66dBm, respectively. Therefore, the time series of the current moment is [-60, -63, -66, -68], in dBm.

[0052] This step continuously collects the received signal strength of the first network and maintains the time series, providing continuous historical observation data for subsequent received signal strength prediction and link fluctuation assessment. This enables the prediction model to make trend inferences based on the trajectory of changes in received signal strength, rather than making handover decisions based solely on isolated single measurements.

[0053] In step S120, based on the time series, the received signal strength of the first network at future times is predicted to obtain the predicted value of the received signal strength.

[0054] Here, "future moment" refers to the next sampling moment or the time point corresponding to the next decision period after the current decision period. The predicted received signal strength refers to the estimated value of the received signal strength at that future moment, calculated using a prediction model based on historical received signal strength data from the time series.

[0055] For example, the prediction model can be a time series prediction method such as Kalman filtering, exponential smoothing, autoregressive prediction, or grey prediction. The prediction model uses the historical variation pattern of received signal strength to extrapolate and estimate the value of received signal strength at future times, so that the handover decision can detect the trend of link quality changes in advance before the received signal strength actually drops below the threshold.

[0056] As an example, the terminal uses Kalman filtering to process the time series. The received signal strength is modeled as a state vector containing both a received signal strength state component and a received signal strength change trend state component. This state vector can be represented as [x, v], where x represents the received signal strength level and v represents the rate of change of the received signal strength. The state prediction step of the Kalman filter predicts the current state vector based on the state estimate and state transition model from the previous time step. The measurement update step corrects the prediction result using the received signal strength measurement value at the current time step, obtaining a smoothed state estimate. Based on the current state estimate, the predicted received signal strength values ​​for future time steps are extrapolated using the state transition model. When the received signal strength shows a continuous decreasing trend, the change trend state component v becomes negative, further reducing the predicted value and enabling the handover decision to identify link degradation risks earlier.

[0057] Step S120 transforms the handover decision from passively responding to the current measurement value to actively predicting the trend of link quality changes by predicting the received signal strength at future times. When the terminal moves rapidly out of the first network coverage area or encounters obstruction causing the received signal strength to continuously attenuate, the predicted received signal strength reflects this deterioration trend earlier than the current received signal strength measurement value, thus providing a basis for solving the handover lag problem caused by relying solely on the received signal strength at the current time for decision-making.

[0058] In step S130, the prediction error is obtained based on the measured value of the received signal strength and the predicted value of the received signal strength at the same time. The link fluctuation index is obtained based on the prediction errors at multiple times. The link fluctuation index is used to evaluate the degree of fluctuation of the received signal strength of the first network.

[0059] Here, "the same moment" refers to a specific sampling moment. At this sampling moment, the terminal acquires both the actual measured value of the received signal strength and the predicted value of the received signal strength for that moment, obtained in the previous prediction step. The prediction error reflects the deviation of the prediction model from the predicted signal strength at that moment, and can be the difference between the measured value and the predicted value of the received signal strength at the same moment. The link fluctuation index is a quantified value obtained by statistically processing the prediction errors at multiple moments, used to assess the degree of fluctuation of the received signal strength over a period of time.

[0060] For example, the prediction error can be represented as the difference between the measured received signal strength and the predicted received signal strength. Methods for statistically processing the prediction errors at multiple times to obtain the link fluctuation index can include: performing an exponential moving average of the absolute values ​​of the prediction errors at multiple times and using the result as the link fluctuation index; or calculating the variance of the prediction errors at multiple times and using the variance as the link fluctuation index. When the wireless environment is stable, the prediction model accurately predicts the received signal strength, the prediction error at each time point is small, and the link fluctuation index is at a low level. When the wireless environment experiences obstruction, enhanced multipath effects, or changes in terminal mobility, the received signal strength fluctuates non-stationarily, the prediction deviation increases, and the link fluctuation index rises accordingly.

[0061] As an example, at a certain sampling moment, the actual received signal strength measurement obtained by the terminal is -70dBm, while the previously predicted received signal strength for that moment was -68dBm. Therefore, the prediction error at that moment is -2dBm. This absolute value of the prediction error, 2dBm, is then subjected to an exponential moving average along with the absolute values ​​of prediction errors from previous moments. The prediction errors from more recent moments are given a higher weight, while the weights of earlier prediction errors decay exponentially. The result is used as the link fluctuation index for the current moment.

[0062] Step S130 calculates the prediction error between the measured and predicted received signal strength values ​​at the same time, and statistically processes the prediction errors at multiple times to obtain the link fluctuation index, thus achieving a quantitative assessment of the degree of received signal strength fluctuation. The link fluctuation index provides a basis for subsequent threshold shifting: when fluctuations increase, the link fluctuation index increases, and the threshold shift amount increases accordingly; when fluctuations decrease, the link fluctuation index decreases, and the threshold shift amount decreases accordingly.

[0063] In step S140, a threshold shift amount is calculated based on the link fluctuation index. The threshold shift amount is a value calculated based on the link fluctuation index and used to shift the baseline cutout threshold and the baseline cutback threshold. The baseline cutout threshold and the baseline cutback threshold of the first network are synchronously shifted by the threshold shift amount to generate a dynamic cutout threshold and a dynamic cutback threshold.

[0064] The reference handover threshold refers to a preset reference value of received signal strength used to determine whether handover from the first network is necessary, and the reference handover return threshold refers to a preset reference value of received signal strength used to determine whether handover back to the first network is necessary. The threshold shift amount refers to the value calculated based on link fluctuation indicators and used to shift the reference handover threshold and the reference handover return threshold. Synchronous shift refers to applying the same magnitude and direction of shift to the reference handover threshold and the reference handover return threshold, causing both thresholds to shift upwards or downwards as a whole, while maintaining the difference between them unchanged. The dynamic handover threshold and the dynamic handover threshold refer to the handover threshold and the handover return threshold actually used for handover decision after being shifted by the threshold shift amount.

[0065] For example, the threshold shift is calculated by multiplying the link fluctuation index by a preset shift gain coefficient. The shift gain coefficient is used to adjust the sensitivity of the threshold shift to the degree of link fluctuation. The dynamic cut-out threshold is equal to the baseline cut-out threshold plus the threshold shift, and the dynamic cut-back threshold is equal to the baseline cut-back threshold plus the threshold shift. An upper limit can be set for the threshold shift. When the calculated threshold shift exceeds the upper limit, the upper limit is taken as the actual threshold shift to avoid the dynamic threshold deviating excessively from the baseline value under extreme fluctuation conditions.

[0066] As an example, the preset baseline handover threshold is -70dBm, and the baseline handover-back threshold is -62dBm. In a certain decision period, the calculated link fluctuation index is 5, and the shift gain coefficient is 0.3, resulting in a threshold shift of 1.5dB. The baseline handover and handover thresholds are simultaneously shifted upwards by 1.5dB, generating a dynamic handover threshold of -68.5dBm and a dynamic handover threshold of -60.5dBm. At this point, the terminal only needs to handover to the second network when the predicted received signal strength of the first network is below -68.5dBm, and only needs to hand back to the first network when the current received signal strength measurement of the first network is above -60.5dBm. As subsequent link fluctuations weaken and the link fluctuation index drops to 0.5, the threshold shift becomes 0.15dB, and the dynamic handover and handover thresholds fall back to -69.85dBm and -61.85dBm respectively, gradually approaching the baseline threshold levels.

[0067] Step S140 synchronously shifts the baseline handover threshold and baseline handover threshold based on link fluctuation indicators, enabling the handover determination conditions to adaptively adjust according to the degree of fluctuation in received signal strength. When link fluctuations increase, the dynamic handover threshold and dynamic handover threshold shift upwards synchronously: on the one hand, the upward shift of the handover threshold raises the handover threshold, allowing the terminal to trigger handover in advance before the received signal strength drops to a very low level, improving the timeliness of the handover response; on the other hand, the synchronous upward shift of the handover threshold raises the handover threshold, ensuring that the terminal only hands back after the received signal strength of the first network recovers to a more stable level, effectively suppressing ping-pong handover caused by a short-term rebound in received signal strength. When link fluctuations weaken, the dynamic threshold gradually falls back to the baseline level, restoring normal handover response capability in a stable environment. The synchronous shift method maintains the buffer space between the baseline handover threshold and the baseline handover threshold unchanged, preserving the stability of the dual-threshold mechanism during dynamic adjustment.

[0068] In step S150, when the terminal is currently connected to the first network, it switches from the first network to the second network in response to the predicted value of the received signal strength being less than the dynamic handover threshold; when the terminal is currently connected to the second network, it switches from the second network to the first network in response to the measured value of the current received signal strength of the first network being greater than or equal to the dynamic handover threshold, wherein the first network and the second network are heterogeneous wireless networks.

[0069] The first network and the second network can be two different types of wireless access networks. For example, the first network can be a Wi-Fi network and the second network can be a cellular network, or the first network can be a cellular network and the second network can be a Wi-Fi network. The first network and the second network can also be two wireless access networks with different parameters. For example, the first network can be a first Wi-Fi network and the second network can be a second Wi-Fi network.

[0070] For example, a terminal can perform a network switching operation through the system network management interface or the connection management module. Specific methods of switching may include: switching the bearer path of the service data flow from the network interface of the currently accessed network to the network interface of the target network; or updating the routing configuration to ensure that subsequent data packets are transmitted through the target network.

[0071] When a terminal is currently connected to the first network, the handover decision uses the predicted value of the received signal strength rather than the current measured value. This allows the terminal to initiate a handover in advance based on the judgment of a continuously deteriorating link trend, even before the received signal strength actually falls below the dynamic handover threshold. Once the terminal has switched to the second network, the handover decision uses the current measured value of the received signal strength of the first network instead of the predicted value. This is because the handover needs to be based on the objective fact that the actual signal quality of the first network has recovered to an accessible level, rather than an estimate of future recovery trends. This avoids premature handover when the signal strength of the first network has only briefly recovered.

[0072] In step S150, the decision to switch out of the first network is based on the comparison between the predicted value of the received signal strength and the dynamic switch-out threshold, and the decision to switch back to the first network is based on the comparison between the current measured value of the received signal strength and the dynamic switch-back threshold. This allows the switch-out decision to respond in advance to the link deterioration trend, and the switch-back decision to be executed only when the received signal strength actually recovers to a level greater than or equal to the dynamic switch-back threshold. This improves the timeliness of handover when the link continues to deteriorate, and at the same time suppresses ping-pong handover when the signal fluctuates for a short time.

[0073] The above technical solutions enable forward-looking perception in handover decisions by predicting future trends in received signal strength, thus addressing the handover lag issue caused by fixed threshold methods relying solely on current received signal strength measurements. Simultaneously, the heterogeneous wireless network adaptive vertical handover method obtains a link fluctuation index by comparing the predicted received signal strength measurements with the actual received signal strength at the same time. This quantitatively assesses the degree of link fluctuation. Based on this, a threshold shift is calculated using the link fluctuation index, and the baseline handover-out and handover-back thresholds are synchronously shifted. This allows the handover decision conditions to adaptively adjust with the degree of link fluctuation. When link fluctuation increases, the dynamic handover-out and dynamic handover-back thresholds are simultaneously raised to enable early handover and suppress ping-pong handover. When link fluctuation decreases, the dynamic threshold falls back to the baseline level, restoring normal responsiveness. This solves the problem that fixed handover-out and handover-back thresholds cannot simultaneously ensure both handover timeliness and stability.

[0074] Figure 2 yes Figure 1 A flowchart illustrating the sub-steps of step S110.

[0075] Regarding step S110, in some embodiments of this disclosure, it includes, for example... Figure 2 Steps S111 to S113 are shown.

[0076] In step S111, the received signal strength of the first network is collected at a preset sampling period to obtain the measured value of the received signal strength at the current moment.

[0077] The sampling period refers to the time interval between two received signal strength acquisition operations, which can be set according to the terminal's movement speed or network environment, such as 1 second or 2 seconds. The acquisition operation obtains the received signal strength of the currently accessed first network by calling the operating system's network status interface or the wireless network card driver interface. The obtained received signal strength is usually expressed in dBm.

[0078] In step S112, the received signal strength measurement value at the current moment is added to the received signal strength observation sequence.

[0079] The received signal strength observation sequence refers to the set of received signal strength measurements arranged chronologically according to the acquisition time. When a newly acquired received signal strength measurement is added to the sequence, if the sequence has already reached the preset sliding window length, the earliest received signal strength measurement in the sequence is removed to maintain a fixed sequence length. The sliding window length determines the amount of historical data involved in the prediction and can be set according to actual needs, for example, a length corresponding to 4 to 10 sampling periods.

[0080] In step S113, the maintained received signal strength observation sequence is output as a time series for use in subsequent prediction steps.

[0081] Step S110, through steps S111 to S113, continuously acquires the received signal strength of the first network using a periodic acquisition method and maintains a fixed-length time series. This time series retains the temporal information of recent changes in the received signal strength and controls the data size through a sliding window mechanism, providing regularly sampled historical observation data for subsequent received signal strength prediction, enabling the prediction model to make trend inferences based on the continuous change trajectory of the received signal strength.

[0082] Figure 3 yes Figure 1 A flowchart illustrating the sub-steps of step S120.

[0083] Regarding step S120, in some embodiments of this disclosure, it may include, for example... Figure 3 Steps S121 to S124 are shown.

[0084] In step S121, the state vector, state transition matrix, observation matrix, state noise covariance matrix, and observation noise covariance matrix of the Kalman filter are initialized.

[0085] In this context, the state vector refers to the set of variables representing the internal state of the system in Kalman filtering. In this disclosure, the received signal strength is modeled as a two-dimensional state vector containing both the received signal strength state component and the received signal strength change trend state component. The state transition matrix represents the linear recursive relationship of the state vector from the previous time step to the current time step, and the observation matrix represents the mapping relationship from the state vector to the observed values. The state noise covariance matrix characterizes the uncertainty in the state transition process, and the observation noise covariance matrix characterizes the noise intensity in the received signal strength measurement process.

[0086] In step S122, the state prediction step of Kalman filtering is performed.

[0087] The state prediction step uses the state estimate and state transition matrix from the previous time step to calculate the prior estimate of the current state, and simultaneously calculates the error covariance matrix of the prior estimate. In the state prediction step, the received signal strength change trend state component in the state vector drives the received signal strength state component to evolve forward: when the received signal strength change trend is negative, the predicted received signal strength level further decreases; when the received signal strength change trend is positive, the predicted received signal strength level further increases.

[0088] In step S123, the measurement update step of Kalman filtering is performed.

[0089] The measurement update step uses the received signal strength measurement acquired at the current moment to correct the prior estimate obtained in the state prediction step. Specifically, it calculates the Kalman gain, corrects the state vector based on the difference between the received signal strength measurement and the observed prediction value estimated in the prior estimate, obtains the posterior estimate of the state at the current moment, and updates the error covariance matrix. The posterior estimate is the smoothed estimated value of the received signal strength at the current moment and the estimated value of the trend of the received signal strength change.

[0090] In step S124, based on the posterior state estimate and state transition matrix at the current time, the predicted value of the received signal strength at the future time is extrapolated.

[0091] Specifically, the state vector estimated posteriorly at the current time is recursively passed forward through the state transition matrix to obtain the state prediction vector at the future time, from which the received signal strength state component is extracted as the received signal strength prediction value.

[0092] Step S120, following steps S121 to S124, utilizes Kalman filtering to recursively estimate and extrapolate the received signal strength time series. The state vector simultaneously contains both the horizontal component and the trend component of the received signal strength, enabling the Kalman filter to track the rate and direction of change in the received signal strength. When the received signal strength continuously decreases, the trend component becomes negative, driving the predicted received signal strength value to be lower than the current measurement value. This allows the handover decision to detect the link deterioration trend even before the current received signal strength measurement value falls below a threshold, providing a data basis for the early handover decision based on the predicted received signal strength value in step S150.

[0093] For example, during the recursive process of Kalman filtering, if the received signal strength measurement value decreases sequentially for multiple consecutive cycles, the trend state component will gradually converge to a relatively stable negative value. At this time, even if the received signal strength measurement value at the current moment has not fallen below the reference handover threshold, the predicted received signal strength value at the future moment obtained by extrapolation is already lower than the reference handover threshold or the dynamic handover threshold, and the terminal triggers wireless network handover in advance accordingly.

[0094] In other embodiments of this disclosure, the prediction method in step S120 can replace Kalman filtering with at least one of exponential smoothing, autoregressive prediction, or grey prediction. When using exponential smoothing, the measured values ​​at each moment in the received signal strength time series are assigned weights with exponential decay, with more recent measurements having higher weights and earlier measurements having lower weights. The predicted received signal strength for future moments is obtained through weighted averaging. When using autoregressive prediction, the received signal strength for future moments is represented as a linear combination of several recent measurements in the time series. The model coefficients can be estimated from the historical received signal strength time series using the least squares method or the Yule-Walker equation. When using grey prediction, the received signal strength time series is accumulated once, a grey differential equation is established, and the model parameters are estimated. The predicted received signal strength for future moments is obtained through cumulative subtraction. All of the above alternative prediction methods can extrapolate and estimate the received signal strength for future moments based on the historical received signal strength time series, enabling handover decisions to anticipate link quality changes in advance. This allows for adaptation to different computing resources or implementation environments while meeting the prediction function.

[0095] Figure 4 yes Figure 1 A flowchart illustrating the sub-steps of step S130.

[0096] Regarding step S130, in some embodiments of this disclosure, it may include, for example... Figure 4 Steps S131 to S133 are shown.

[0097] In step S131, the prediction error between the measured value of the received signal strength and the predicted value of the received signal strength at the same time is calculated.

[0098] Here, "the same moment" refers to a certain sampling moment. At this sampling moment, the terminal not only acquires the received signal strength measurement value through step S111, but also obtains the predicted value of the received signal strength at that moment through step S124 in the previous decision period.

[0099] The prediction error can be calculated by subtracting the predicted value of the received signal strength from the measured value. The sign of the result indicates the direction of deviation of the measured value from the predicted value, and the magnitude of the result indicates the magnitude of the deviation.

[0100] In some embodiments of this disclosure, the prediction error can also be represented by at least one of the following: absolute error, squared error, moving average error, or exponential moving average error, between the measured and predicted received signal strength values ​​at the same time. The absolute error is the absolute value of the difference between the measured and predicted received signal strength values; the squared error is the square of the difference between the measured and predicted received signal strength values; the moving average error is the arithmetic mean of the prediction errors over a number of recent times; and the exponential moving average error is the value obtained by performing an exponential moving average on the prediction error sequence. These different error representations can be applied to different channel characteristics and terminal implementation conditions.

[0101] In step S132, the prediction errors at multiple times are statistically processed to obtain the link fluctuation index.

[0102] In some embodiments of this disclosure, the absolute values ​​of prediction errors at multiple time points are processed using an exponential moving average, and the result is used as a link fluctuation index. The exponential moving average is calculated as follows: Let the absolute value of the prediction error at the current time t be e(t), the exponential moving average at the previous time point be s(t-1), and the smoothing coefficient be α (0 < α < 1), then the exponential moving average at the current time point is s(t) = α·e(t) + (1-α)·s(t-1). The smoothing coefficient α controls the response speed to new data; the larger α is, the more sensitive the link fluctuation index is to recent changes in prediction error; the smaller α is, the smoother the link fluctuation index, but the slower the response. The smoothing coefficient α can be configured according to the terminal's moving speed or the characteristics of the wireless environment.

[0103] In other embodiments of this disclosure, the variance of the prediction error at multiple time points is calculated, and this variance is used as an indicator of link fluctuation. Specifically, the variance of the prediction error is calculated for the most recent M time points within a sliding window. When the wireless environment fluctuates more, the dispersion of the prediction error increases, and the variance increases accordingly; when the wireless environment tends to be stable, the dispersion of the prediction error decreases, and the variance decreases accordingly.

[0104] In step S133, the link fluctuation index is output for use in the subsequent threshold shifting step.

[0105] Step S130, following steps S131 to S133, constructs a link fluctuation index by utilizing the prediction error between the measured and predicted received signal strength values ​​at the same time point, combined with prediction error statistics from multiple time points. The magnitude of the prediction error reflects the adaptability of the prediction model to changes in link status: when the link is stable, the time series changes are highly regular, and the prediction error is small; when the link fluctuates drastically, the time series changes are highly volatile, and the prediction error increases. The link fluctuation index quantifies this fluctuation level into a numerical value, serving as the basis for calculating the threshold shift in step S140.

[0106] For example, in a stable indoor environment near a Wi-Fi access point, the received signal strength changes smoothly over time, the Kalman filter prediction is accurate, and the prediction errors at multiple points in time are small. After exponential moving average processing, the link fluctuation index remains at a low level. However, when the terminal moves to an area with people walking around or objects obstructing the view, the received signal strength exhibits non-stationary fluctuations, the prediction error increases, and the link fluctuation index rises accordingly.

[0107] Figure 5 yes Figure 1 A flowchart illustrating the sub-steps of step S140.

[0108] Regarding step S140, in some embodiments of this disclosure, it may include, for example... Figure 5 Steps S141 to S143 are shown.

[0109] In step S141, the threshold shift amount is calculated based on the link fluctuation index.

[0110] Specifically, the threshold shift amount can be obtained by multiplying the link fluctuation index by a preset shift gain coefficient. The shift gain coefficient is used to adjust the sensitivity of the threshold shift to the degree of link fluctuation and can be configured according to the terminal's moving speed, network environment, or service requirements. When the link fluctuation index increases, the threshold shift amount increases accordingly; when the link fluctuation index decreases, the threshold shift amount decreases accordingly.

[0111] In step S142, an upper limit constraint is applied to the threshold shift amount.

[0112] The threshold shift calculated in step S141 is compared with a preset upper limit value. If the threshold shift exceeds the upper limit value, the upper limit value is taken as the actual threshold shift; otherwise, the original calculated value is maintained. The upper limit value is used to prevent the dynamic threshold from deviating excessively from the reference value under extreme fluctuations, ensuring that the switching behavior is within a reasonable range.

[0113] In step S143, the reference cutout threshold and reference cutback threshold of the first network are synchronously shifted by a threshold shift amount to generate dynamic cutout threshold and dynamic cutback threshold.

[0114] Synchronous translation refers to applying the same magnitude and direction of translation to both the baseline cutout threshold and the baseline cutback threshold. The dynamic cutout threshold equals the baseline cutout threshold plus the threshold translation amount, and the dynamic cutback threshold equals the baseline cutback threshold plus the threshold translation amount. After translation, the difference between the baseline cutout threshold and the baseline cutback threshold remains unchanged.

[0115] For example, the terminal presets a baseline handover threshold of -70dBm and a baseline handover-back threshold of -62dBm, with a difference of 8dB. The shift gain coefficient is set to 0.3, and the shift upper limit is set to 5dB. In a certain decision period, the link fluctuation index is calculated to be 4, the threshold shift is 1.2dB, which does not exceed the shift upper limit. The dynamic handover threshold is -68.8dBm, and the dynamic handover-back threshold is -60.8dBm, with a difference of 8dB. When the link fluctuation index rises to 20, the calculated threshold shift is 6dB, which exceeds the shift upper limit by 5dB. Taking 5dB as the actual threshold shift, the dynamic handover threshold is -65dBm, and the dynamic handover-back threshold is -57dBm. When the environment stabilizes, the link fluctuation index drops to 0.5, the threshold shift is 0.15dB, and the dynamic cut-out threshold and dynamic cut-back threshold are close to -69.85dBm and -61.85dBm, respectively.

[0116] Step S140, through steps S141 to S143, completes the calculation process from link fluctuation index to dynamic threshold. When link fluctuation increases, the link fluctuation index increases, the threshold shift increases, and the dynamic handover threshold and dynamic handover threshold shift upwards simultaneously, making it easier for the terminal to hand over to the first network earlier and harder to hand back to the first network. When link fluctuation decreases, the link fluctuation index decreases, the threshold shift decreases, and the dynamic handover threshold and dynamic handover threshold decrease simultaneously, gradually returning to the baseline level. The upper limit constraint mechanism ensures that the dynamic threshold does not deviate from a reasonable range under extreme environments.

[0117] Figure 6 yes Figure 1 A flowchart illustrating the sub-steps of step S150.

[0118] Regarding step S150, in some embodiments of this disclosure, it may include... Figure 6 Steps S151 to S152 are shown.

[0119] In step S151, when the terminal is currently connected to the first network, the predicted value of the received signal strength obtained in step S120 is compared with the dynamic handover threshold generated in step S140. If the predicted value of the received signal strength is less than the dynamic handover threshold, it is determined that the terminal needs to hand over from the first network, and the process proceeds to step S152 to perform network switching; if the predicted value of the received signal strength is greater than or equal to the dynamic handover threshold, the current connection to the first network remains unchanged.

[0120] The handover decision uses the predicted value of the received signal strength instead of the current measured value of the received signal strength. This allows the terminal to detect the link deterioration trend in advance and trigger the handover when the received signal strength is still in the process of gradually decreasing. This avoids the handover delay and service interruption risks caused by waiting until the received signal strength actually falls below the threshold before handover.

[0121] In some embodiments of this disclosure, to further improve the stability of the handover decision and reduce the probability of erroneous handover caused by the predicted value of the received signal strength being accidentally lower than the dynamic handover threshold in a single instance, the handover decision in step S151 can be set as follows: when the terminal is currently connected to the first network, a handover from the first network to the second network is triggered only if the predicted value of the received signal strength is lower than the dynamic handover threshold for N consecutive sampling times, where N is an integer greater than 1. The value of N can be configured according to the terminal's moving speed or the fluctuation characteristics of the wireless environment, for example, N can be 2 or 3.

[0122] In step S152, when the terminal is currently connected to the second network, it continues to periodically monitor the received signal strength of the first network and compares the measured value of the received signal strength of the first network at the current moment with the dynamic switchback threshold generated in step S140. If the measured value of the current received signal strength of the first network is greater than or equal to the dynamic switchback threshold, it is determined that the terminal can switch back to the first network and a switch from the second network to the first network is performed; otherwise, the current connection to the second network remains unchanged.

[0123] The handover decision uses the current received signal strength measurement of the first network, rather than the predicted received signal strength, because the handover decision needs to be based on the certainty that the signal quality of the first network has actually recovered to an accessible level. If the link fluctuates significantly, the dynamic handover threshold will be synchronously raised to a higher level. Even if the received signal strength of the first network rebounds briefly, as long as it does not exceed the dynamic handover threshold, the terminal will not immediately handover, thus avoiding ping-pong handover caused by premature handover before the first network signal has stabilized.

[0124] For example, in a specific scenario, the terminal initially connects to a Wi-Fi network as the first network and a cellular network as the second network. During the terminal's movement, the received signal strength of the Wi-Fi network successively decreases to -60dBm, -63dBm, -66dBm, and -68dBm, showing a continuous downward trend. Kalman filtering captures this downward trend, with the trend state component being negative, extrapolating to a predicted received signal strength of -71dBm for future moments. Simultaneously, due to the continuous decrease in received signal strength, the prediction error increases, the link fluctuation index rises, and the dynamic handover threshold shifts from the baseline value of -70dBm to -68.5dBm. At this point, the predicted received signal strength of -71dBm is less than the dynamic handover threshold of -68.5dBm, causing the terminal to prematurely trigger a handover from the Wi-Fi network to the cellular network. After switching to the cellular network, the terminal continues to monitor the received signal strength of the Wi-Fi network. When the received Wi-Fi signal strength briefly rises to -65dBm, the dynamic switchback threshold remains high (e.g., -60.5dBm) because the link fluctuations have not yet fully subsided. Since the current received signal strength measurement of -65dBm does not exceed the dynamic switchback threshold, the terminal will not immediately switch back to the Wi-Fi network, avoiding ping-pong handover. As the terminal continues to approach the Wi-Fi access point, the received signal strength steadily increases to around -60dBm and the prediction error decreases. The link fluctuation index falls back, and the dynamic switchback threshold gradually recovers to the baseline switchback threshold of -62dBm. The current received signal strength measurement of -60dBm exceeds the dynamic switchback threshold, and the terminal switches back to the Wi-Fi network.

[0125] Step S150, through steps S151 and S152, distinguishes the comparison parameters used for the handover decision and the handover decision: the handover decision is based on the predicted received signal strength, using trend prediction to respond to link deterioration in advance; the handover decision is based on the current measured received signal strength, requiring the first network signal to actually recover and exceed the dynamic handover threshold before handover is allowed. This asymmetric decision mechanism works in conjunction with the received signal strength prediction mechanism and the dynamic threshold synchronous shift mechanism to simultaneously improve the timeliness and stability of handover in dynamic wireless environments.

[0126] The following example illustrates a typical execution process of the technical solution disclosed herein. In this example, the first network is a Wi-Fi network, the second network is a cellular network, the sampling period is 1 second, the baseline cut-out threshold is -70dBm, and the baseline cut-back threshold is -62dBm.

[0127] The terminal initially connects to the Wi-Fi network and continues to move. At four consecutive sampling times, the received Wi-Fi signal strength collected by the terminal is -60dBm, -63dBm, -66dBm, and -68dBm, forming a time series [-60, -63, -66, -68]. Based on this time series, the Kalman filter estimates that the received signal strength is continuously decreasing, with a negative trend state component, predicting a future received signal strength of -71dBm.

[0128] Meanwhile, due to the non-stationary change in the received signal strength, which shows a continuous decrease, the prediction error between the measured and predicted received signal strength values ​​at the same time increases, leading to an increase in the link fluctuation index after exponential moving average processing. The link fluctuation index output by the link fluctuation assessment module is 5, with a shift gain coefficient of 0.3, and the calculated threshold shift is 1.5dB. The baseline cutout threshold of -70dBm and the baseline cutback threshold of -62dBm are simultaneously shifted upwards by 1.5dB, generating a dynamic cutout threshold of -68.5dBm and a dynamic cutback threshold of -60.5dBm.

[0129] At this point, the predicted received signal strength (RSS) of -71 dBm is less than the dynamic handover threshold of -68.5 dBm, causing the terminal to prematurely trigger the handover from the Wi-Fi network to the cellular network. If the traditional fixed dual-threshold method is used, with a handover threshold of -70 dBm, and the current RSS measurement of -68 dBm is still below the threshold, the terminal will not trigger the handover. Instead, it will wait until the RSS continues to drop below -70 dBm before switching. By this time, the link quality has severely deteriorated, easily causing service interruption. The method disclosed in this paper, through a prediction mechanism and a threshold synchronous upward shift mechanism, allows for premature handover to the cellular network while the RSS measurement is still -68 dBm, reserving time for handover execution and service migration.

[0130] After the terminal switches to the cellular network, it continues to periodically monitor the received signal strength of the Wi-Fi network. When the terminal moves back to the Wi-Fi coverage area, the received signal strength of the Wi-Fi network briefly rises to -65dBm. However, because the link fluctuation has not completely subsided, the link fluctuation index remains at a high level, and the dynamic handover threshold is still around -60.5dBm. At this time, the current received signal strength measurement of the Wi-Fi network, -65dBm, does not exceed the dynamic handover threshold of -60.5dBm, so the terminal will not immediately switch back to the Wi-Fi network, thus avoiding ping-pong handover caused by the brief rise in Wi-Fi signal strength.

[0131] As the terminal continues to approach the Wi-Fi access point and enters a stable coverage area, the received signal strength of the Wi-Fi network steadily increases to around -60dBm, the prediction error decreases, the link fluctuation index drops, and the dynamic handover threshold gradually recovers to near the baseline handover threshold of -62dBm. At this point, the current received signal strength measurement of the Wi-Fi network (-60dBm) is greater than or equal to the dynamic handover threshold, and the terminal switches back to the Wi-Fi network, completing a full vertical handover process.

[0132] In the above scenario example, the fixed dual-threshold method not only suffers from handover lag, but also, if the terminal uses a fixed handover threshold of -62dBm for decision-making after switching to the cellular network, a handover will not occur even if the Wi-Fi signal briefly rises to -65dBm, which is below -62dBm. However, if the signal fluctuates drastically and the instantaneous rise exceeds -62dBm, a handover will be triggered, followed by a rapid drop in signal strength, leading to another handover and a ping-pong handover. The method disclosed in this paper synchronizes the mobile-mode handover threshold with the link fluctuation index, raising the handover threshold during signal fluctuations. Handover is only allowed after the signal has stably recovered to a higher level, thus improving both the timeliness and stability of handover in dynamic wireless environments.

[0133] In the technical solution disclosed herein, the following parameters can be adjusted according to the terminal device type, the network types of the first network and the second network, the terminal mobility speed, or service requirements: The sliding window length of a time series is the number of historical measurements involved in the prediction of received signal strength. A larger window length results in smoother predictions but a slower response to trend changes, while a smaller window length responds faster to trend changes but is more susceptible to fluctuations in individual measurements.

[0134] The state noise covariance matrix and the observation noise covariance matrix of the Kalman filter are used to determine the trust weight allocation between the predicted and measured values. The state noise covariance matrix reflects the uncertainty in the state transition process, while the observation noise covariance matrix reflects the noise intensity of the received signal strength measurement. Together, they determine the trust weight allocation between the predicted and measured values ​​in the Kalman filter.

[0135] The smoothing coefficient of the exponential moving average in the calculation of the link volatility index is larger. The larger the smoothing coefficient, the faster the link volatility index responds to recent changes in prediction error. The smaller the smoothing coefficient, the smoother the link volatility index, but the slower the response.

[0136] The translation gain coefficient of the threshold shift is used to adjust the sensitivity of the dynamic cut-out threshold and dynamic cut-back threshold to the link fluctuation index.

[0137] The upper limit of the threshold shift is used to constrain the deviation range of the dynamic threshold under extreme fluctuation environments.

[0138] The reference cut-out threshold and reference cut-back threshold are preset based on the coverage characteristics, signal propagation characteristics, or system handover strategy of the first and second networks.

[0139] The adjustment of the above parameters does not change the core logic of the technical solution disclosed herein, namely, performing short-term prediction based on the received signal strength time series, assessing the link fluctuation degree based on the prediction error, synchronously shifting the handover threshold and handover back threshold based on the link fluctuation degree, and performing a handover decision based on the current access status.

[0140] Figure 7 This is a schematic diagram of a terminal device bus architecture according to one embodiment of the present disclosure.

[0141] This disclosure also provides a terminal device. The hardware structure of the terminal device can be implemented using a bus architecture. For example... Figure 7 As shown, the bus architecture can include any number of interconnect buses and bridges, depending on the specific application of the hardware and overall design constraints. Bus 1100 connects various circuits including one or more processors 1200, memory 1300, and / or hardware modules. Bus 1100 can also connect various other circuits 1400 such as peripherals, voltage regulators, power management circuits, external antennas, etc. Bus 1100 can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Component (EISA) bus, etc. Buses can be classified as address buses, data buses, control buses, etc. For ease of illustration, only one connection line is used in this figure, but this does not mean that there is only one bus or one type of bus.

[0142] For ease of explanation, certain steps of the above method are described in relation to modules. It should be understood that the corresponding module performing one or more steps of the above method may be one or more hardware modules specifically configured to perform the corresponding step, or implemented by a processor configured to perform the corresponding step, or stored in a computer-readable medium for implementation by a processor, or implemented by some combination thereof.

[0143] Figure 8 This is a functional block diagram of a terminal device according to one embodiment of the present disclosure.

[0144] like Figure 8 As shown, the terminal equipment includes: a signal acquisition module 1010, a signal prediction module 1020, a link fluctuation assessment module 1030, a dynamic threshold adjustment module 1040, and a handover decision and execution module 1050.

[0145] The signal acquisition module 1010 is used to acquire the time series of the received signal strength of the first network. Specifically, the signal acquisition module 1010 acquires the received signal strength measurement values ​​of the first network from the operating system network status interface, the wireless network card driver interface, or the network management service at preset intervals, and arranges the received signal strength measurement values ​​at each sampling time in chronological order to generate an observation sequence of the received signal strength as a time series. The function performed by this module corresponds to step S110.

[0146] The signal prediction module 1020 is used to predict the received signal strength of the first network at future times based on the time series output by the signal acquisition module 1010, thereby obtaining a predicted value of the received signal strength. The signal prediction module 1020 can use Kalman filtering to process the time series, suppressing measurement noise while preserving the short-term trend of the received signal strength, thus obtaining a smoothed estimated value of the received signal strength at the current time and an extrapolated predicted value of the received signal strength at future times. The function performed by this module corresponds to step S120.

[0147] The link fluctuation assessment module 1030 is used to obtain the prediction error based on the received signal strength measurement value at the same time and the received signal strength prediction value given by the signal prediction module 1020, and to obtain the link fluctuation index based on the prediction errors at multiple times. The link fluctuation index is used to assess the degree of fluctuation of the received signal strength of the first network. When the link fluctuation index is small, it indicates that the current wireless environment is relatively stable; when the link fluctuation index is large, it indicates that there may be sudden changes such as obstruction, rapid attenuation, or interference. The function performed by this module corresponds to step S130.

[0148] The dynamic threshold adjustment module 1040 calculates the threshold shift amount based on the link fluctuation index output by the link fluctuation assessment module 1030, and synchronously shifts the baseline handout threshold and baseline handback threshold of the first network by the threshold shift amount to generate dynamic handout threshold and dynamic handback threshold. It should be noted that the dynamic threshold adjustment module 1040 does not adjust the hysteresis interval by changing the interval between the baseline handout threshold and the baseline handback threshold, but rather performs a synchronous overall shift of the two thresholds in the same direction based on the link fluctuation index, so that the handover judgment condition moves synchronously with the degree of link fluctuation. When link fluctuation increases, the dynamic handout threshold and dynamic handback threshold move upward synchronously; when link fluctuation decreases, the dynamic handout threshold and dynamic handback threshold move downward synchronously. The difference between the baseline handout threshold and the baseline handback threshold remains unchanged during the shift process. The function performed by this module corresponds to step S140.

[0149] The handover decision and execution module 1050 is used to execute a vertical handover decision and complete the network handover based on the terminal's current access status, the predicted value of the received signal strength, the measured value of the current received signal strength of the first network, the dynamic handover threshold, and the dynamic handover return threshold. Specifically, when the terminal is currently connected to the first network, the predicted value of the received signal strength is compared with the dynamic handover threshold. If the predicted value of the received signal strength is less than the dynamic handover threshold, a handover from the first network to the second network is triggered. When the terminal is currently connected to the second network, the measured value of the current received signal strength of the first network is compared with the dynamic handover return threshold. If the measured value of the current received signal strength is greater than or equal to the dynamic handover return threshold, a handover from the second network to the first network is triggered. The network handover can be executed through the system network management interface, routing configuration, application-bound network interface, or connection management module. The function performed by this module corresponds to step S150.

[0150] The specific implementation of each module in the aforementioned terminal device can be referred to the implementation process of the corresponding steps in the above-described method implementation method of this disclosure, and will not be repeated here.

[0151] The signal acquisition module 1010, signal prediction module 1020, link fluctuation assessment module 1030, dynamic threshold adjustment module 1040, and handover decision and execution module 1050 described above are the main functional modules for implementing this technical solution. In actual products, these modules can be implemented as independent software modules or integrated into the operating system network management service, communication middleware, or application layer connection management program. The above module division method is only illustrative; in actual implementation, multiple modules can be merged into one functional unit, or one module can be split into multiple sub-units. As long as the terminal can execute all the operations of the above-described method steps of this disclosure, it belongs to the implementation method of the technical solution of this disclosure.

[0152] In some embodiments of this disclosure, an electronic device is provided, which may be a mobile phone, tablet computer, vehicle-mounted communication terminal, or portable communication device, etc. The electronic device includes a processor and a memory. The memory stores a computer program, and when the processor executes the computer program, it implements the heterogeneous wireless network adaptive vertical switching method described in any of the above embodiments. The electronic device supports access to at least two different types of wireless networks, such as Wi-Fi networks and cellular networks, and is capable of vertical switching between the two networks by executing the computer program.

[0153] This disclosure also provides a readable storage medium storing a computer program that, when executed by a processor, is used to implement the methods described above. A "readable storage medium" can be any means capable of containing, storing, communicating, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples of a readable storage medium include: an electrical connection with one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable read-only memory (CDROM), etc.

[0154] In some embodiments of this disclosure, the technical solution of this disclosure can be verified by the following methods.

[0155] First, check whether the terminal product collects and uses the received signal strength time series of the first network. If the terminal not only uses the received signal strength measurement at the current moment during network handover, but also continuously records the received signal strength change sequence over a period of time and uses this sequence for subsequent handover judgment, then it may have adopted an implementation method similar to this disclosure.

[0156] Second, check whether the terminal product predicts or judges the trend of the received signal strength. If the terminal can switch to the second network in advance based on the decreasing trend of the received signal strength when the received signal strength of the first network has not fallen below the fixed handover threshold, it indicates that its handover decision may use received signal strength prediction, trend estimation or similar forward-looking judgment mechanism.

[0157] Third, check whether the terminal product dynamically changes the handover threshold based on link fluctuations. Under the same or similar signal strength conditions, if the terminal uses different handover triggering conditions when the signal is stable and when the signal fluctuates sharply—for example, when the fluctuation is stronger, it hands out of the first network earlier and hands back into the first network later—it indicates that it may have dynamically adjusted the handover and handover conditions based on link fluctuations.

[0158] Fourth, check whether the terminal product's logs, debugging information, or configuration files contain parameter information related to this disclosure. If information such as the predicted received signal strength, filter value, prediction error, residual, fluctuation factor, dynamic handover threshold, and dynamic handover back threshold is found, and these parameters are involved in network handover decisions, then this can be used as evidence that the terminal product has used the technical solution disclosed in this disclosure.

[0159] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment / mode or example is included in at least one embodiment / mode or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0160] Those skilled in the art should understand that the above embodiments are merely for illustrating the present disclosure and are not intended to limit the scope of the disclosure. Those skilled in the art can make other changes or modifications based on the above disclosure, and these changes or modifications still fall within the scope of the present disclosure.

Claims

1. A method for adaptive vertical handover in heterogeneous wireless networks, characterized in that, include: Obtain the time series of the received signal strength of the first network; Based on the time series, the received signal strength of the first network at future times is predicted to obtain the predicted value of the received signal strength; The prediction error is obtained by comparing the received signal strength measurement and the predicted received signal strength at the same time. The link fluctuation index is obtained by comparing the prediction errors at multiple times. The link fluctuation index is used to evaluate the degree of fluctuation of the received signal strength of the first network. Based on the link fluctuation index, a threshold shift amount is calculated. The threshold shift amount is a value calculated based on the link fluctuation index and used to shift the baseline cutout threshold and the baseline cutback threshold. The reference cutout threshold and reference cutback threshold of the first network are synchronously shifted by the threshold shift amount to generate dynamic cutout threshold and dynamic cutback threshold; as well as When a terminal is currently connected to the first network, in response to the predicted received signal strength value being less than the dynamic handover threshold, the terminal's access network is switched from the first network to the second network; when a terminal is currently connected to the second network, in response to the current received signal strength measurement value of the first network being greater than or equal to the dynamic handover threshold, the terminal's access network is switched from the second network to the first network, wherein the first network and the second network are heterogeneous wireless networks.

2. The heterogeneous wireless network adaptive vertical handover method according to claim 1, characterized in that, Based on the time series, predicting the received signal strength of the first network at future times includes: The time series is processed using Kalman filtering to obtain the predicted values ​​of the received signal strength of the first network at future times.

3. The heterogeneous wireless network adaptive vertical handover method according to claim 2, characterized in that, The time series is processed using Kalman filtering, including: The received signal strength is modeled as a state vector containing a received signal strength state component and a received signal strength change trend state component. Based on the state vector, Kalman filtering is performed for state prediction and measurement update to obtain the predicted value of the received signal strength.

4. The heterogeneous wireless network adaptive vertical handover method according to claim 1, characterized in that, Link fluctuation metrics are derived from prediction errors at multiple points in time, including: The absolute values ​​of the prediction errors at the multiple time points are processed by exponential moving average, and the processing result is used as the link fluctuation index.

5. The heterogeneous wireless network adaptive vertical handover method according to claim 1, characterized in that, Based on the link fluctuation index, the threshold shift is calculated, specifically including: The product of the link fluctuation index and the translation gain coefficient is used as the threshold translation amount.

6. The heterogeneous wireless network adaptive vertical handover method according to claim 5, characterized in that, The threshold shift amount does not exceed the upper limit of the shift.

7. The heterogeneous wireless network adaptive vertical handover method according to claim 1, characterized in that, Also includes: When the terminal is currently connected to the first network, in response to the predicted value of the received signal strength being less than the dynamic cut-out threshold for N consecutive time periods, the terminal switches from the first network to the second network, where N is an integer greater than 1.

8. The heterogeneous wireless network adaptive vertical handover method according to claim 1, characterized in that, Obtain the time series of the received signal strength of the first network, including: The received signal strength of the first network is periodically collected to generate a received signal strength observation sequence, which is then used as the time series.

9. A terminal device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the heterogeneous wireless network adaptive vertical handover method as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the heterogeneous wireless network adaptive vertical handover method as described in any one of claims 1 to 8.

Images