A multi-hardware collaborative self-organizing network video backhaul control system

Abstract

This invention relates to the field of general control or regulation system technology, and discloses a multi-hardware collaborative self-organizing network video backhaul control system, including: a PV contamination prevention unit, a stability regulator, a control priority definition module, and a bitrate allocation unit. The PV contamination prevention unit removes local fault parameters before calculating the global instability index; the control priority definition module monitors user interface events to update priorities instantaneously; the stability regulator generates a single global adjustment factor; and the bitrate allocation unit combines the factor and priority to generate differentiated video stream control commands. This invention transforms open-loop competition into closed-loop control, solving the technical problems of global control failure due to local faults and control commands lagging behind command intentions.

Description

Technical Field

[0001] This invention relates to a self-organizing network video backhaul control system with multi-hardware collaboration, belonging to the general field of control or regulation system technology. Background Technology

[0002] Currently, in ad hoc network video backhaul applications, there is an inherent contradiction in the operational mechanism between the video source's need for a continuously high bitrate to ensure image quality and the drastic bandwidth fluctuations of the ad hoc network link due to high mobility or obstruction. This contradiction is particularly prominent in scenarios where multiple hardware components must backhaul, such as emergency command and tactical reconnaissance, leading to competition for the bottleneck resource of the air interface. Existing technologies, such as congestion control mechanisms in the communications field aimed at optimizing network throughput or ensuring link fairness, operate by having each distributed terminal detect and compete for bandwidth. When the total network capacity momentarily falls below the total demand of concurrent video streams due to topology changes, this disorderly competition can easily lead to systemic deep congestion, causing all video streams, regardless of their importance, to collectively fail, manifesting as frozen images, pixelation, or black screens, affecting the continuity of situational awareness.

[0003] Existing improvement approaches mostly focus on optimizing the network communication layer architecture, addressing the problem by building more robust underlying links. However, this does not address the fundamental flaws in the upper-layer control logic. For example, Chinese invention patent CN115379595A discloses a method and system for a decentralized wireless broadband self-organizing network of multiple unmanned vehicles. It aims to establish a flat, equal-status communication network among unmanned vehicles by using decentralized, self-organizing link-layer networking technology. The core is to ensure the establishment and self-healing of the network, solving basic connectivity issues. It can work normally when network resources are abundant. However, the decentralized and equal-status characteristics naturally revert to the fair competition state mentioned above when facing bandwidth bottlenecks. It does not provide an upper-layer collaborative control and regulation mechanism to proactively manage congestion or achieve differentiated quality of service assurance.

[0004] Therefore, the technical problem to be solved by this invention is how to design a closed-loop regulation system with multi-hardware collaboration, transform open-loop competition into closed-loop control, and introduce priority as the core regulation basis to ensure the continuity of critical tasks when network resources fluctuate. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: a multi-hardware collaborative self-organizing network video backhaul control system, comprising a system status perception module, a control priority definition module, and a video backhaul control logic module;

[0006] The video return control logic module includes a PV contamination defense unit, a stability quantization unit, a stability regulator, and a bitrate allocation unit.

[0007] The system status awareness module is used to monitor the network status parameters of each of the N video streams in the ad hoc network in real time;

[0008] The control priority definition module is used to receive the control priority of N video streams; access and monitor user interface events at the receiving end; when a user interface event indicates that a certain video stream is in user focus state, the control priority of that video stream is automatically and instantaneously increased.

[0009] The PV contamination defense unit is used to establish a health benchmark based on the network state parameters of N video streams, and to identify the video stream whose network state parameters deviate extremely from the health benchmark as a contamination source; and when transmitting the network state parameters to the stability quantization unit, the parameters of the video stream identified as a contamination source are removed.

[0010] The stability quantification unit is used to calculate and generate system instability indicators that characterize the global congestion stability of ad hoc networks based solely on network state parameters that have not been removed.

[0011] The stability regulator is used to calculate and generate a single global regulation factor based on the deviation between the system instability index and the preset system stability setpoint through a control law.

[0012] The bitrate allocation unit is used to determine the target quality bitrate for each of the N video streams based on the control priority defined by the control priority definition module; and to perform a function operation between the target quality bitrate and the global adjustment factor to generate video stream control instructions.

[0013] Preferably, the PV pollution defense unit is used to calculate the statistical median of the network status parameters of N video streams as the first benchmark value of the health benchmark; calculate the absolute deviation of the median of the network status parameters as the dispersion index of the health benchmark; and compare the deviation of the network status parameters of a certain video stream from the first benchmark value with the product of the dispersion index and the preset adjustment value; when the deviation is greater than the product, the video stream is determined to be a pollution source.

[0014] Preferably, the bitrate allocation unit is used to calculate and generate video stream control instructions through the following function operation: ,in, For the first Video stream control commands for the video stream; For the first Target quality bitrate of the video stream; It is a single global regulating factor.

[0015] Preferably, the user focus state includes at least one of the following: the video stream is displayed in full screen, the video stream is pinned to the top of the window, or the video stream has activated audio.

[0016] Preferably, the stability regulator is a PI controller, and the system stability setpoint is always zero.

[0017] Preferably, the stability quantification unit is used to calculate the statistical variance of at least one parameter and the time rate of change of at least one parameter based on the network state parameters that have not been removed, and to perform a weighted operation on the statistical variance and the time rate of change through a weighted algorithm to generate a system instability index.

[0018] Preferably, the stability regulator is used to monitor whether the global regulation factor it calculates reaches the preset saturation limit, and when the global regulation factor reaches the saturation limit, it suspends the cumulative calculation of the integral term in the control regulation law.

[0019] Preferably, the video return control logic module triggers an active optimization mode when the system instability index is lower than a preset stability threshold; the bitrate allocation unit, in active optimization mode, selects the video stream with the lowest control priority from N video streams as a probe according to the control priority; when calculating the video stream control command of the probe, it actively superimposes a preset bitrate increment as the optimization probe measurement; the stability regulator monitors the response of the system instability index to the optimization probe measurement, and when the response causes the system instability index to exceed the stability threshold, the control regulation law automatically adjusts the global regulation factor to offset the optimization probe measurement.

[0020] Preferably, the video return control logic module determines in real time the number of valid streams that are not identified as pollution sources among the N video streams, and the system stability setpoint is dynamically adjusted based on the number of valid streams through a preset nonlinear compensation function.

[0021] Compared with the prior art, the beneficial effects of the present invention are:

[0022] 1. A master-slave cascaded control architecture is provided, which decouples the two control objectives of system stability and priority allocation. The system monitors and adjusts the system instability indicators that characterize network jitter or packet loss rate through the stability regulator in the outer loop, making them approach the zero setpoint. Through the bitrate allocation unit in the inner loop, the adjustment results of the outer loop are executed, and the allocation is differentiated according to the priority set by the user. The bandwidth prediction problem is transformed into a system stability adjustment problem, actively maintaining a stable adjustment state and achieving controllable resource allocation.

[0023] 2. In the process variable calculation stage, a data verification mechanism is set up between the original state perception and the global stability quantification. Using the statistical benchmark of multi-channel video stream parameters, before calculating the global system instability index, extreme abnormal data caused by local terminal failure is identified and eliminated. This enables the main control loop and stability regulator to distinguish between global congestion and local faults, avoiding misleading global control and erroneously suppressing all healthy video streams due to a single pollution source, thus ensuring the system's adjustment stability under partial failure conditions.

[0024] 3. The internal regulation law of the stability regulator includes anti-integral saturation clamping logic, which monitors the regulator's own output. When the output reaches the saturation limit, the regulator actively suspends the accumulation calculation of the integral term, solving the problem of excessive accumulation of the integral term when the regulator is interrupted for a long time. When the network recovers from an extreme interruption, the regulator itself is not stuck in a saturation state, but responds to the recovery of process variables in real time and starts the regulation function, avoiding the delay of the system falling into control failure after the network is restored. Attached Figure Description

[0025] Figure 1 This is a diagram illustrating the closed-loop control system architecture and data flow of the present invention.

[0026] Figure 2 This is a simulation curve of the dynamic adjustment and differentiated protection of the system of the present invention;

[0027] Figure 3 This is a schematic diagram illustrating the analysis of existing technical problems and control defects of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below, but the scope of protection of this invention is not limited by the following specific embodiments.

[0029] This invention provides a multi-hardware collaborative self-organizing network video backhaul control system, including a system status awareness module, a control priority definition module, and a video backhaul control logic module as the control core. The video backhaul control logic module is further decoupled into a PV pollution defense unit, a stability quantization unit, a stability regulator, and a bitrate allocation unit, forming a cascaded control or master-slave control architecture. The system status awareness module is used to monitor all components in the self-organizing network in real time. The monitoring module obtains the network status parameters of each video stream. This monitoring can be based on communication protocols, such as the report block in the RTP / RTCP protocol, to acquire raw parameters such as end-to-end latency, latency jitter, and packet loss rate for each video stream, providing real-time raw process data for subsequent control and adjustment. The control priority definition module provides a basic human-machine interface for receiving control priority information. The video stream has a manually set static control priority; it further integrates and monitors user interface events on the receiving end's display console. When a video stream is detected to be performing user focus operations such as full-screen display, window on top, or audio activation, this module immediately and automatically raises the control priority of that video stream to the preset highest level; in the video return control logic module, the control priority comes from the system status awareness module. The original network state parameters are sent to the PV pollution defense unit; this is used to prevent a terminal from causing a 100% packet loss rate due to physical failure, such as antenna damage, and thus polluting the overall network state assessment; the operating procedure is to perform data identification before calculating global indicators, based on... Similar parameters for path flow, with time delay jitter. For example, a statistical health benchmark is established; in practice, this benchmark is calculated by... The statistical median of the path parameters is defined as follows: As the first benchmark value, the absolute deviation of the median is calculated and defined as follows: As a dispersion index; this cell traverses Road flow, will the first Parameters of path flow When compared with the benchmark, the degree of deviation, i.e. and The deviation is greater than the preset adjustment value. With dispersion index When multiplying by , the first If a path is identified as a pollution source, when passing network state parameters to the next level, all parameters of that path identified as a pollution source are actively removed.

[0030] Preset adjustment value in PV pollution defense unit This sensitivity is used to define the identification of contamination sources. In a benchmark test containing N=20 healthy streams, parameters such as latency jitter (Ji) of all streams were collected, and the ratio of their respective deviations from the statistical median (Mediansys) to the median absolute deviation (MAD) was calculated. The statistical distribution showed that its 99.9th percentile was around 3.5. In another test containing a single faulty stream (whose packet loss rate instantly became 100%), the corresponding ratio of the faulty stream exceeded 50. This effectively eliminates extreme faults while avoiding misjudging statistical fluctuations of healthy streams as contamination sources. The value range is set to 3.0 to 5.0, with a preferred value of 3.5; the network state parameters of the purified but not removed healthy video stream are sent to the stability quantization unit; this is used to abstract the controlled object from the network state into a controllable single process variable, denoted as... The logic is that it doesn't care about the absolute values ​​of network parameters, such as a latency of 500ms, but rather their stability; based on the parameters that haven't been removed, it calculates the statistical variance of at least one parameter, such as the severity of latency jitter. And the rate of change of at least one parameter over time, such as the rate of increase of the packet loss rate. ; Through weighted algorithms, for example: Generate a single scalar system instability index ;in and These are the preset weighting coefficients; The value characterizes the global congestion stability of the system; when the network is stable, The value approaches 0 when the network experiences drastic fluctuations or increased congestion. Greater than 0; weighting coefficients in the stability quantization unit and Used to balance the system's response to time delay jitter variations Changes in packet loss rate The system's response sensitivity was examined in a set of simulation tests under different conditions: high jitter and low packet loss, and low jitter and high packet loss. and Weighted comparison regulator output The impact, test results show, when When the weight of (corresponding packet loss rate change) is less than 0.5, the response to a sudden drop in total network bandwidth is delayed, which can easily lead to deep congestion. When the weight is higher than 0.8, it becomes overly sensitive to occasional non-congestion packet loss, leading to unnecessary rate suppression. and The weight ratio is preferably set at : When the ratio is around 0.3:0.7, it exhibits timely response and stable adjustment under both operating conditions.

[0031] System instability indicators As a process variable, it is fed into a stability controller; forming the main loop of the cascaded control, its function is stability adjustment in control engineering; in specific implementations, this controller is a standard PI controller, a proportional-integral controller; stability setpoint As a control target, it can be constantly set to zero. This represents the control objective of pursuing absolute system stability; in another implementation, to address the scale mismatch problem of different statistical background noise under different network sizes, this given value... The number of effective flows output by the PV contamination defense unit can be determined based on the number of effective flows. Through a preset nonlinear compensation function, such as Make dynamic adjustments so that the controller can... It can tolerate reasonable statistical background noise when the noise level is relatively high; the PI controller has continuous calculation error. By controlling the regulation law (which includes proportional and integral terms), a single global regulation factor is calculated and generated. This factor can be normalized to the [0,1] interval. This means there is no need to suppress it. The representative must maximize suppression; to address the recovery failure issue caused by excessive accumulation of the PI integral term after extreme operating conditions, such as prolonged network outages, this stability regulator integrates anti-integral saturation logic to monitor its own output. ,when When the preset saturation limit, such as 1.0, is reached, the regulator immediately suspends the accumulation calculation of the integral term in its control law to ensure that the network recovers once the network is restored. (Fallback), the regulator can immediately leave the saturation state and instantaneously restore its regulating function; the stability regulator, as a PI controller, controls the proportional gain in the regulating law. and integration time It is used to determine the system's response characteristics to PV and SP deviations. In a simulation test environment, by applying a step network congestion disturbance, the process response curve of the control system is measured, and the pure delay time is extracted from it. With the response time constant T, experimental results show that when Values ​​below 0.8T / At that time, the system adjustment response is too slow, and the PV overshoot is large. Values ​​higher than 1.2T / hour, The output is prone to oscillation, when Value less than When / 0.2, the integral action is too strong, causing oscillations during the recovery phase. Value greater than When the value is 0.5, the speed of eliminating static error is too slow, therefore and The value of is determined within a preferred range after weighing the above response characteristics and balancing the need for stable response with the elimination of steady-state error.

[0032] Global regulation factor output by the main loop (stability regulator) The output of the control priority definition module Road control priority Simultaneously, the data is sent to the rate allocation unit; this unit forms a slave loop in the cascaded control, and its function is to execute the total adjustment of the master loop and allocate data differentially according to priority; the operating procedure is based on the control priority defined by the module, which updates the control priority in real time. ,for Each video stream determines its target quality bitrate. ,Should This represents the bitrate that the stream should achieve when the network is unrestricted; it is a quantification of priority. The target bitrate for each stream... Global suppression with the main loop Perform function calculations to generate the final video stream control commands. The function operation in the specific implementation is as follows: This control architecture ensures that when the global network deteriorates ( When the bitrate increases (e.g., to 0.7), the bitrate of all video streams decreases proportionally to their target quality bitrate (i.e., ...). ), systematically and collaboratively reduce the overall network demand, high-priority flows (their The bitrate allocated to (large) It is still higher than the low-priority flow (its Small); To address the issue of slow passive bandwidth detection after network recovery, this control system also features an active optimization mode; when the stability regulator (main loop) determines that the system is in a stable state ( When the bitrate falls below the stability threshold, this mode is triggered; the bitrate allocation unit (from the loop) automatically selects the video stream with the lowest control priority as the probe, and calculates its... At that time, a preset bitrate increment is actively superimposed as an optimization probe. The consequences of this probing action are monitored in real time by the stability regulator (main loop). If the probing results in... If the stability threshold is exceeded (i.e., the attempt fails), the main loop control law (PI) will automatically respond and increase. This adjustment automatically and instantaneously pushes the optimization probe back through the above function calculation, so that the system automatically returns to stability.

[0033] Example 1: In an emergency command scenario, a control system deployed in the command center is used to coordinate and regulate three concurrently transmitted video streams. These three video streams are: UAV A, UAV B, and UAV C. Drone B And individual soldier C The initial static settings in the control priority definition module will... Set as high priority Set to medium priority Set to low priority, corresponding target quality bitrate They are respectively , , In the initial stage, the ad hoc network links are stable, and the system state awareness module detects that the network state parameters of each flow are stable. Based on this, the stability quantification unit calculates the system instability index. Approaching zero; the stability regulator acts as the main loop, with a stability setpoint. It is always zero, therefore and The deviation is close to zero, and the global adjustment factor output by the PI controller is... Also zero; the rate allocation unit, as a slave ring, is based on... The function operation is for three-way flow computation. The system operates at the target quality bitrates of 8Mbps, 4Mbps, and 2Mbps for each stream. When UAV B maneuvers at high speed to an area obstructed by a mountain, the ad hoc network topology undergoes a sudden change, and the total network bandwidth drops instantaneously. The system state awareness module detects that network state parameters of all three video streams, such as latency jitter and packet loss rate, begin to rise rapidly. The stability quantization unit receives these parameters and calculates the system instability index. Rising to 0.7; the main loop PI controller of the stability regulator detected... and When a deviation occurs, the control law responds immediately, adjusting the global control factor. The rate is increased from 0 to 0.6; the rate allocation unit receives the following data in the next control cycle. ,in accordance with The function operation instantly generates three new video stream control commands, namely... , , The data is then distributed to each terminal; this process transforms the open-loop collapse process into a closed-loop adjustment process, maintaining the continuity of the video stream when the network deteriorates by proactively and differentiatedly reducing the total bitrate requirement.

[0034] The system is currently in Under the adjustment state, individual soldier C is The terminal equipment suffered physical damage, causing network status parameters such as packet loss rate to instantly jump to 100%; The raw parameters are sent to the PV pollution defense unit; procedures are established based on health benchmarks, and calculations are performed. , , The statistical median, i.e. The absolute deviation from the median is ;because and Still in The corresponding controlled stable state, and The value deviates, and the judgment is made through its identification logic. It is a flow that deviates extremely from the healthy baseline; the PV contamination defense unit actively rejects it when passing parameters to the stability quantification unit. The parameters; the stability quantization unit is therefore based on and The parameters are used to calculate the global Calculation results The output of the main loop stability regulator remains around 0.7. Maintain at 0.6; avoid Failure data affects the whole Calculations are performed to prevent the stability regulator from incorrectly adjusting the settings due to misjudgment. Upgraded to version 1.0 to ensure that the control of the two video streams from drones A and B is not affected by... The system malfunctioned and the operation was interrupted; the commander discovered individual soldier C on the display console. The screen is black. Clicking on the video window of Soldier C in the UI interface will display it in full screen; the UI event monitoring function of the control priority definition module will capture the user focus state event of this full-screen display in real time; and immediately and automatically... The control priority is raised to the highest level, with a target quality bitrate. The instantaneous update speed increased from 2Mbps to 8Mbps; this update... Feed to the bitrate allocation unit; Still offline; when soldier C troubleshoots and the device comes back online after a few seconds, network parameters return to normal, the PV pollution defense unit no longer removes it, and the calculation results of the stability quantization unit and stability regulator... and Both remain at 0.7 and 0.6; the bitrate allocation unit then immediately uses this automatically increased value. and the current Calculate new control commands for it, that is The obtained bit rate ensures instantaneous alignment of drone A with the highest priority; by monitoring user interface events, the control priority is automatically updated, so that the adjustment setpoint of the control system is automatically consistent with the user's operation focus.

[0035] Example 2: To verify the effectiveness of the control system of the present invention at the control level, a network simulation test environment was constructed to simulate a network containing... The self-organizing network of concurrent video streams has a total available bandwidth that can be dynamically adjusted during testing. The experiment includes a sample group and two control groups. The sample group fully deploys the publicly disclosed control system, including a system state awareness module, a control priority definition module, and a video backhaul control logic module containing a PV pollution defense unit, a stability quantization unit, a stability regulator, and a bitrate allocation unit. Control group A uses an open-loop competition method, where all five video streams are continuously sent at their target bitrate of 5Mbps, lacking closed-loop adjustment capability. Control group B deploys the same closed-loop adjustment system as the sample group, but its PV pollution defense unit is disabled; that is, the original network state parameters of all five streams are not filtered and are used entirely to calculate system instability indicators. In all tests, the control priority definition modules of both the sample group and control group B set one of the video streams (Stream1) as high priority. The remaining four streams (Stream2-5) are set to low priority. ).

[0036] The experiment consisted of two phases, and the test results are shown in Table 1. In the first phase, a global congestion scenario was simulated, instantaneously reducing the total available bandwidth of the ad hoc network from 30Mbps to 10Mbps. During this phase, all five video streams remained healthy. In control group A, due to open-loop contention, severe conflicts occurred among the five streams, leading to a surge in the global packet loss rate to 45.2%, and the average bitrate of all streams was suppressed to below 1.8Mbps, resulting in a collective failure. In control group B, the control systems of both control groups and the prototype of this invention responded correctly, and the stability regulator output a non-zero global adjustment factor. The bitrate allocation unit performs differentiated allocation accordingly, maintaining the bitrate of high-priority streams at 6.1 Mbps and suppressing the bitrate of low-priority streams to the range of 0.9 Mbps to 1.0 Mbps. The total system demand is controlled below the link capacity, and the global packet loss rate is kept at a stable level below 2.5%. In the second phase, a partial failure condition is simulated. While maintaining the total bandwidth at 10 Mbps, the terminal of one low-priority video stream (Stream5) is artificially made to fail, generating extremely abnormal network state parameters. The state of control group A deteriorates further. Control group B, due to the lack of PV pollution defense, has its system instability indicators polluted. The calculations led to misjudgment in the main loop, causing the stability regulator output to saturate. (Close to 1.0), incorrectly suppressing the bitrate of all video streams, including high-priority streams, to 0.1 Mbps, causing a control failure; the PV contamination defense unit of the present invention comes into play at this time, using the discrimination logic of statistical median and absolute deviation of median to determine that Stream5 is the contamination source and remove its parameters, thus eliminating the uncontaminated streams. This makes the stability regulator Maintaining at the Phase 1 level, the bitrate of the high-priority stream was kept stable at 6.0 Mbps, unaffected by localized faults.

[0037] Table 1: Comparison of System Performance under Different Control Strategies

[0038]

[0039] Experimental data shows that the control system provided by this invention, compared with the open-loop competition method, can maintain the system in a stable adjustment state with a low packet loss rate when the global network is congested, and realize differentiated code rate allocation according to control priority. Furthermore, the data confirms the synergistic effect of the PV pollution defense unit and the stability regulator, which enables the control system to distinguish between global congestion and local faults by eliminating local fault parameters before the main loop calculation.

[0040] Example 3: This example combines Figures 1 to 3 This describes a multi-hardware collaborative self-organizing network video backhaul control system, such as... Figure 1 As shown, the network status data source for N video streams is monitored in real time by the system status perception module to obtain the network status parameters of the N streams. These parameters are sent to the PV pollution defense unit to eliminate local fault parameters and prevent pollution sources from misleading the global picture, thereby outputting healthy network status parameters. The stability quantification unit is used to calculate the global system instability index and send it to the main loop of the stability regulator. Based on the deviation, a single global adjustment factor is generated. Independent user interface events such as full-screen display and audio activation are monitored by the controlled priority definition module, which updates the priority instantaneously and outputs the updated control priority. The bitrate allocation unit combines the global adjustment factor from the main loop with the updated control priority from the priority module to generate differentiated video stream control commands, which are then sent to the N video sources to form a closed-loop control.

[0041] like Figure 2 As shown, the horizontal axis represents the initial stage, global congestion, local failure, user focus switching to the stage of recovery and stable operation; the left vertical axis represents the bit rate in Mbps; and the right vertical axis represents the global adjustment factor. The curve shows that during the global congestion phase, the global adjustment factor... As the rate increases, all bitrates decrease. The bitrate of high-priority drone A is consistently higher than that of medium-priority drone B and low-priority soldier C. During a local fault phase, the global adjustment factor... While the high and medium priority bitrates remained stable, only the low priority bitrate of soldier C dropped to zero. During user focus switching, soldier C's priority was increased, and its bitrate surpassed that of drone B; for example... Figure 3As shown, in ad hoc network video backhaul, link characteristics such as drastic bandwidth fluctuations, high mobility and obstruction, and sudden changes in network topology, combined with terminal competition strategies such as fair competition, individual detection and bandwidth grabbing, and continuous high bitrate requirements, lead to defects in control mechanisms such as open-loop state, distributed mechanism, lack of unified adjustment mechanism, and lack of system coordination, such as inability to differentiate and sacrifice low priority, lack of priority coordination guarantee, and inability of multiple terminals to coordinate actions.

[0042] Example 4: This example describes the system stability setpoint used for calibrating a stability regulator. The procedure for a nonlinear compensation function, which is used to solve for different numbers of effective flows. The corresponding control system has different statistical background noise; calibration is performed on a network simulation test platform, which has pre-loaded the calibrated stability quantization unit parameters. ) and stability regulator parameters ( First step, set the number of valid flows. Under non-congested conditions, the control system operates stably for 600 seconds, and the system instability indicators output by the stability quantification unit are collected. Calculate the statistical mean of the time series. Obtain baseline data points ( The second step is to maintain a non-congested link and repeat the above steps, setting the effective flow count sequentially. , , The corresponding statistical means were measured respectively. , , The third step is to fit the data points (5, 0.041), (10, 0.059), (20, 0.085), and (30, 0.108) to... The function model; A value is given to dynamically adjust the system stability. The number of effective flows, The adjustment constant is to be calibrated; the adjustment constant is determined by least squares regression analysis. The nonlinear compensation function Ultimately, it is pre-set in the stability regulator, and the controller, during real-time operation, determines the effective flow quantity based on the PV contamination defense unit. This function is dynamically called to calculate the system stability setpoint under the current operating conditions. To avoid controllers being in large Under operating conditions, the statistical background noise of health is misjudged as control error.

[0043] Example 5: This example is a more advanced implementation of the control system. The internal operating logic of the stability quantization unit is further defined to handle a low-priority video stream. This causes slight network jitter, while high-priority streams... With medium priority flow All maintain stable specific operating conditions; under these operating conditions, set Individual instability It is 0.4. , If a simple averaging algorithm is used, the global... Calculated as This value may exceed the stability threshold, such as 0.1, triggering the stability regulator to generate... This leads to an unnecessary reduction and The bitrate; to address this situation, the stability quantization unit in this embodiment is limited to executing PV-Priority weighted collaborative logic; this unit calculates the global... Previously, all were obtained from the control priority definition module. Healthy flow refers to the control priority of flows that are not removed by PV contamination defense units. ,Should It can be quantified as its target quality bitrate. The value, let The stability quantization unit no longer performs a simple averaging, but instead uses a priority-weighted aggregation algorithm to calculate the global system instability index. : ,in, It is the first Individual instability indicators of road flow This refers to its corresponding priority weight; applying this weighting algorithm, under the above operating conditions, that is... , The new global computation for: The calculated value of 0.057 is lower than 0.133, which is lower than the stability threshold of 0.1. Therefore, the stability regulator determines that the system is still in a stable state and outputs... Keep it at 0; this mechanism enables the main loop controller to sense... Its control intention Alignment reduces sensitivity to disturbances from low-priority flows and avoids disturbances from high-priority flows. bitrate due to Slight fluctuations should not be suppressed unnecessarily.

[0044] Example 6: This example is used to calibrate the system stability setpoint in the stability regulator. The procedure for the nonlinear compensation function; calibration is performed on the network simulation test platform, with the calibrated stability quantization unit parameters preloaded. ) and stability regulator parameters ( First step, set the number of valid flows. Under non-congested conditions, the control system operates stably for 600 seconds, and the system instability indicators output by the stability quantification unit are collected. Calculate the statistical mean of the time series. Obtain baseline data points ( The second step is to maintain a non-congested link and repeat the above steps, setting the effective flow count sequentially. , , The corresponding statistical means were measured respectively. , , The third step is to fit the data points (5, 0.041), (10, 0.059), (20, 0.085), and (30, 0.108) to... The function model, A value is given to dynamically adjust the system stability. The number of effective flows, The adjustment constant is to be calibrated; the adjustment constant is determined by least squares regression analysis. Nonlinear compensation function Finally, it is pre-set in the stability regulator, and during real-time operation, it determines the effective flow quantity based on the PV contamination defense unit. This function is dynamically called to calculate the system stability setpoint under the current operating conditions. .

[0045] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A self-organizing network video backhaul control system with multi-hardware collaboration, characterized in that, It includes a system status awareness module, a control priority definition module, and a video return control logic module; The video return control logic module includes a PV contamination defense unit, a stability quantization unit, a stability regulator, and a bitrate allocation unit. The system status awareness module is used to monitor the network status parameters of each of the N video streams in the ad hoc network in real time; The control priority definition module is used to receive the control priority of N video streams; Access and monitor user interface events at the receiving end; When a user interface event is detected indicating that a certain video stream is in the user focus state, the control priority of that video stream is automatically and instantly increased. The PV pollution defense unit is used to establish a health benchmark based on the network status parameters of N video streams, and to identify the video stream whose network status parameters deviate extremely from the health benchmark as a pollution source. Furthermore, when transmitting network state parameters to the stability quantization unit, parameters of the video stream identified as a pollution source are removed; The stability quantification unit is used to calculate and generate system instability indicators that characterize the global congestion stability of ad hoc networks based solely on network state parameters that have not been removed. The stability regulator is used to calculate and generate a single global regulation factor based on the deviation between the system instability index and the preset system stability setpoint through a control law. The bitrate allocation unit is used to determine the target quality bitrate for each of the N video streams based on the control priority definition module output; and to perform a function operation between the target quality bitrate and the global adjustment factor to generate video stream control instructions. The PV pollution defense unit is used to calculate the statistical median of the network status parameters of N video streams as the first benchmark value of the health benchmark; calculate the absolute deviation of the median of the network status parameters as the dispersion index of the health benchmark; and compare the deviation of the network status parameters of a certain video stream from the first benchmark value with the product of the dispersion index and the preset adjustment value; when the deviation is greater than the product, the certain video stream is identified as a pollution source.

2. The self-organizing network video backhaul control system with multi-hardware collaboration according to claim 1, characterized in that, The bitrate allocation unit is used to calculate and generate video flow control commands through the following function: ,in, For the first Video stream control commands for the video stream; For the first Target quality bitrate of the video stream; It is a single global regulating factor.

3. The self-organizing network video backhaul control system with multi-hardware collaboration according to claim 1, characterized in that, User focus state includes at least one of the following: the video stream is displayed in full screen, the video stream is on top of the window, or the video stream has audio activated.

4. The self-organizing network video backhaul control system with multi-hardware collaboration according to claim 1, characterized in that, The stability regulator is a PI controller, and the system stability setpoint is always zero.

5. The self-organizing network video backhaul control system with multi-hardware collaboration according to claim 1, characterized in that, The stability quantification unit is used to calculate the statistical variance of at least one parameter and the time rate of change of at least one parameter based on the network state parameters that have not been removed, and to perform a weighted operation on the statistical variance and the time rate of change through a weighted algorithm to generate a system instability index.

6. The self-organizing network video backhaul control system with multi-hardware collaboration according to claim 4, characterized in that, The stability regulator is used to monitor whether the global regulation factor it calculates reaches the preset saturation limit, and when the global regulation factor reaches the saturation limit, it suspends the cumulative calculation of the integral term in the control law.

7. The self-organizing network video backhaul control system with multi-hardware collaboration according to claim 1, characterized in that, The video return control logic module is used to trigger the active optimization mode when the system instability index is lower than the preset stability threshold; the bitrate allocation unit, in the active optimization mode, selects the video stream with the lowest control priority from N video streams as the probe according to the control priority. When calculating the video stream control commands of the probe, a preset bitrate increment is actively superimposed as the probe measurement for optimization. The stability regulator monitors the system's instability index response to the optimization probe, and when the response causes the system's instability index to exceed the stability threshold, the control law automatically adjusts the global regulation factor.

8. The self-organizing network video backhaul control system with multi-hardware collaboration according to claim 1, characterized in that, The video return control logic module determines in real time the number of valid streams that are not identified as pollution sources among the N video streams, and the system stability setpoint is dynamically adjusted based on the number of valid streams through a preset nonlinear compensation function.

Images