Edge-side live video transmission scheduling method in dense wireless networks

Abstract

The present invention discloses a method for scheduling live video transmission at the edge of a dense wireless network. The method determines the terminal priority and transmission rate on the network edge device AP and implements transmission scheduling. Specifically, the method includes: (1) a fuzzy logic decision framework: the framework includes input fuzzification, a fuzzy inference engine and a defuzzification output, (2) active queue management based on traffic classification and fuzzy logic: the traffic is allocated to different queues according to the terminal priority determined by the fuzzy logic. (3) adaptive control of wireless transmission rate driven by live video transmission demand. According to the MCS index determined by the fuzzy logic, the MCS index of the terminal is dynamically adjusted to achieve adaptive control of the wireless transmission rate. The method uses fuzzy logic to make terminal priority and rate decisions, occupies very few computing, storage and other resources, and is a lightweight scheduling method.

Description

Technical Field

[0001] The present invention relates to a video transmission optimization technology in a wireless network environment, and in particular to an edge-side live video transmission scheduling method in a dense wireless network. Background Art

[0002] In recent years, IEEE 802.11 wireless local area network (WLAN) technology has developed rapidly, with higher and higher transmission rates, and is widely used in many scenarios such as home audio and video, corporate meetings, large-scale events, etc., to provide convenient and high-speed wireless access network methods. However, with the explosive growth in the number of terminal devices, most WLAN networks are in a user-intensive state, and the rapid popularization of video services such as video audio and video, real-time conferencing, and online education has made live video services in user-intensive wireless network environments more and more common. Considering that in user-intensive environments, there are many types of services, and non-real-time business traffic with lower priority seriously interferes with the transmission of live video traffic. At the same time, in a bottleneck network environment where many terminals compete for shared bandwidth, the terminal throughput will fluctuate, resulting in a decline in user experience. Therefore, although the transmission bandwidth of wireless networks is getting larger and larger, the user experience of live video transmission services is still poor and needs to be optimized urgently.

[0003] In the live video transmission business scenario, most of the work does not consider the impact of the last hop, namely the edge access point (AP), resulting in a poor user experience of live video in a user-dense environment; and does not consider the changes in channel and network status caused by the movement of the terminal, as well as the impact of low-priority services such as browsing the web and downloading files.

[0004] In order to solve the problem of obvious fluctuations in the quality of service (QoS) caused by the mutual competition of bandwidth between terminals in a user-dense wireless network environment, an optimization mechanism on the AP side is proposed, considering that the AP is closer to the user terminal and has certain computing capabilities. Considering the smoothness requirements of live video, a set of optimization mechanisms based on fuzzy logic is designed, which comprehensively considers the terminal channel status to perform reasonable priority scheduling and adjust the transmission rate, ensuring that the QoS indicators of live video data packets on overloaded APs are stable, thereby ensuring the user experience of live video. Summary of the invention

[0005] To solve the above problems, the present invention discloses a method for scheduling edge-side live video transmission in a dense wireless network; in response to the needs of live video services in a user-dense wireless network environment, terminal priority and transmission rate are decided on the network edge device AP, and transmission scheduling is implemented.

[0006] Specifically, we first classify the services, then consider the smoothness requirements of live video, and design an optimization mechanism based on fuzzy logic. We comprehensively consider the terminal channel status to perform reasonable priority scheduling and adjust the transmission rate to ensure that the QoS indicators of live video data packets on overloaded APs are stable, thereby ensuring the user experience of live video.

[0007] The technical solution of the present invention is a method for scheduling edge-side live video transmission in a dense wireless network, which specifically includes the following steps:

[0008] Step (1) Fuzzy logic decision framework: This framework includes input fuzzification, fuzzy inference engine and defuzzification output. It takes throughput information, AP status information and terminal channel information as input, and its output is terminal priority and MCS index.

[0009] Step (2) Active queue management based on traffic classification and fuzzy logic: Classify traffic into various types of business traffic, that is, classify various types of business traffic into live broadcast business and non-live broadcast business; for live broadcast business traffic, use the HTB algorithm to build a tree, use the hierarchical control idea to manage the traffic data of each node, and set the priority of the node. The leaf node has the function of borrowing bandwidth from the intermediate node, and the traffic is allocated to different queues according to the terminal priority determined by fuzzy logic.

[0010] Step (3) Adaptive control of wireless transmission rate driven by live video transmission demand: Using the existing wireless transmission multi-rate mechanism, design active MCS index control combined with the live video transmission service demand. According to the MCS index determined by fuzzy logic, dynamically adjust the MCS index of the terminal to achieve adaptive control of wireless transmission rate.

[0011] Furthermore, the step (1) specifically includes:

[0012] Step (11) constructing an AP side information collection module to obtain terminal throughput information, terminal channel information, AP status information, etc., and taking terminal throughput, terminal packet loss rate, terminal signal-to-noise ratio, terminal air splitting and AP duty cycle (Th, Ls, Sr, Ns, Au) as five linguistic variables as input;

[0013] Step (12) uses the triangle membership μ function to calculate the membership of each input linguistic variable in its own three fuzzy sets, that is, μ H (Th),μ M (Th), μ L (Th), μ H (Ls),μ M (Ls), μ L (Ls), μH (Sr),μ M (Sr), μ L (Sr), μ H (Ns),μ M (Ns), μ L (Ns), μ H (Au),μ M (Au), μ L (Au);

[0014] Step (13) uses the membership value of the input language variable calculated in step (11) on the fuzzy set as a judgment condition and substitutes it into each rule i in the fuzzy rule base;

[0015] Step (14)

[0016] For each rule i, take the minimum value of all the membership values ​​in the judgment conditions to obtain the ignition intensity FS of this rule i , FS i =min[μ i1 (Th),μ i2 (Ls),μ i3 (Sr),μ i4 (Ns),μ i5 (Au)], i1,i2,i3,i4,i5∈{L,H,M}, respectively represent the fuzzy values ​​corresponding to Th, Ls, Sr, Ns, and Au in rule i;

[0017] Step (15) uses the implication operator to calculate the ignition intensity FS of rule i. i Truncate its output fuzzy set;

[0018] Step (16) integrates the output fuzzy values ​​of all rules obtained in step (15);

[0019] (17) The maximum membership method is applied to the integrated output fuzzy value to obtain the exact value of the terminal priority and MCS index.

[0020] Furthermore, the step (2) specifically includes:

[0021] Step (21) classifies the traffic into live service traffic and non-live service traffic, and marks the terminal IP containing the live service traffic;

[0022] Step (22) establishes a tree according to the hierarchical mechanism of the HTB algorithm, with each node of the tree as a class and each leaf node having a subqueue. The root node, intermediate node, and leaf node are assigned a predefined borrowable bandwidth factor Quantum, rate Rate, and maximum rate Ceil, and a priority Prio is defined on the leaf node;

[0023] Step (23) marks each service flow based on iptables, classifies the marked live service flow into the leaf node of HTB, and constructs the subqueue q 1 ,q 2 ,…,q n , distributed in leaf nodes 1 ,leaf 2 ,…,leaf n superior;

[0024] Step (24) The data packet sent by the AP is composed of the data on each queue, and the high priority queue is sent first. The HTB algorithm starts from the bottom of the class tree and searches for nodes in the CAN_SEND state. If a node with the CAN_SEND state is found in a certain layer, it stops;

[0025] Step (25) If the layer is found to be a leaf node, there are m leaf nodes leaf i In the CAN_SEND state (i∈[1,n]), the m leaf nodes are sorted by priority, and the ones with higher priority are sent first; if there are multiple leaf nodes with the same priority, they are polled in these nodes. After each leaf node sends Quantum bytes in its subqueue, it is the turn of the next node to send;

[0026] In step (26), if the intermediate node is found, go down the tree to find the descendant leaf node. If there is only one leaf node in the MAY_BORROW state, lend the surplus Quantum value of the intermediate node to the leaf node so that it can send a data packet; if the intermediate node has multiple descendant leaf nodes in the MAY_BORROW state, the processing steps are the same as step (25), according to the priority and polling operation, the leaf node with the highest priority sends first;

[0027] Step (27) performs scheduling according to the terminal priority obtained by fuzzy logic decision, and dynamically adjusts the terminal traffic in each leaf node.

[0028] Furthermore, the step (3) specifically includes:

[0029] Step (31) obtains the MCS index of the terminal according to the fuzzy logic decision, converts the terminal to be controlled and its MCSindex and stores it in the specified location of the user state;

[0030] Step (32) calls the iw command to call the modified kernel function through the control channel to read the content of the user mode specified location;

[0031] Step (33) converts the read content into corresponding variables, which are executed by the function modified in the kernel to achieve the purpose of modifying the MCS index and controlling the wireless transmission rate of the terminal.

[0032] Beneficial effects of the present invention: Compared with the prior art, the present invention has the following significant advantages:

[0033] 1. The method proposed in the present invention is based on a fuzzy logic decision framework, and implements live video transmission scheduling on the AP side, while considering terminal priority scheduling and rate control. Since the status of each terminal can be obtained in real time on the AP side, the transmission can be optimized from a global perspective, significantly improving the average latency, throughput, packet loss rate and other aspects of the terminal, thereby optimizing the user experience.

[0034] 2. The present invention uses fuzzy logic to make terminal priority and rate decisions, which occupies very few computing and storage resources and is a lightweight scheduling method. Even if it runs on a resource-constrained AP, it can handle a large amount of live video transmission traffic in dense wireless network scenarios and schedule it in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of the fuzzy logic framework diagram of the present invention;

[0036] Figure 2 It is a triangular membership function diagram of each input variable of the present invention;

[0037] Figure 3 This is a schematic diagram of HTB hierarchical management of the present invention;

[0038] Figure 4 is a flow chart of active queue management of the present invention;

[0039] Figure 5 It is the overall framework diagram of the transmission scheduling of the present invention. DETAILED DESCRIPTION

[0040] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. It should be noted that the words "front", "rear", "left", "right", "upper" and "lower" used in the following description refer to directions in the accompanying drawings, and the words "inner" and "outer" refer to directions toward or away from the geometric center of a specific component, respectively.

[0041] The specific implementation modes of the present invention are described in detail below with reference to the accompanying drawings.

[0042] This embodiment provides a method for scheduling live video transmission at the edge side in a dense wireless network. An optimization mechanism is deployed on the edge AP side to formally describe the user's live video stream and perform service classification; on this basis, fuzzy logic theory is used to design and implement a live video data scheduling method based on user priority scheduling and transmission rate decision-making. Based on the fuzzy logic framework, this method mainly includes two aspects: active queue management based on traffic classification and fuzzy logic, and adaptive control of wireless transmission rate driven by live video transmission demand.

[0043] 1. Fuzzy logic decision framework design

[0044] This embodiment designs a fuzzy logic framework, which includes input fuzzification, fuzzy inference engine and defuzzification output, with throughput information, AP status information and terminal channel information as input, and outputs terminal priority and MCS index. Figure 1 The overall steps of the fuzzy logic decision-making framework are given.

[0045] 1.1 Input Fuzzification

[0046] Throughput, packet loss rate, signal-to-noise ratio, air splitting and duty cycle (Th, Ls, Sr, Ns, Au) are taken as five input linguistic variables. For each input linguistic variable, its value is taken as a set of language names, and all its values ​​constitute a language name set, and each language name set corresponds to a fuzzy set. In this embodiment, the number of fuzzy sets for each linguistic variable value is set to 3, namely high, medium and low, represented by High (H), Medium (M) and Low (L) respectively. The membership function calculates the membership of each linguistic variable in its three fuzzy sets. In this embodiment, the triangular membership function is used, and the triangular membership function is defined as follows:

[0047]

[0048] Among them, parameters a and c determine the "foot" of the triangle, and b determines the "peak" of the triangle. This embodiment describes the membership function of each input variable. Figure 2 A membership function graph of each input variable is provided.

[0049] The specific steps of input fuzzification are as follows:

[0050] (1) Build an AP-side information collection module to obtain terminal throughput information, terminal channel information, AP status information, etc., and use terminal throughput, terminal packet loss rate, terminal signal-to-noise ratio, terminal air splitting, and AP duty cycle (Th, Ls, Sr, Ns, Au) as five linguistic variables as input;

[0051] (2) The triangular membership μ function is used to calculate the membership of each input linguistic variable in its own three fuzzy sets, that is, μ H (Th),μ M (Th), μ L (Th), μ H (Ls),μ M (Ls), μ L (Ls), μ H (Sr),μ M (Sr), μ L (Sr), μ H (Ns),μ M (Ns), μ L (Ns), μ H (Au),μ M (Au), μ L (Au).

[0052] 1.2 Fuzzy Reasoning

[0053] This embodiment uses the fuzzy rule form of IF (satisfies a set of conditions) THEN (outputs a set of results) to form a fuzzy rule base. The premise and conclusion in the "IF-THEN" rule are fuzzy concepts. The combination of the fuzzy sets of each input language variable is used as the judgment condition, and the priority Pr of each terminal running the live video and its transmission rate, i.e., the fuzzy set of MCS index, is used as the output. As shown in the first rule of Table 1: When the throughput is low, the packet loss rate is low, the signal-to-noise ratio is medium, the air split is medium, and the duty cycle is high, the conclusion is low priority and medium transmission rate.

[0054] Table 1 Fuzzy rule examples

[0055]

[0056] The fuzzy rule base consists of multiple sub-rules. For the sake of simplicity, we take the fuzzy rule with two inputs and one output as an example. Assume that there are two fuzzy rules R1 and R2 defined as follows:

[0057] R1:IF x is A 1 and y is B 1 THEN z is C 1

[0058] R2:IF x is A 2 and y is B 2 THEN z is C 2

[0059] According to the fuzzy logic operation rules, the ignition intensity of rules R1 and R2 is defined as α1 and α 2 , the calculation steps are:

[0060]

[0061]

[0062] Wherein μ is the membership function, ∧ is the fuzzy AND operation, and the commonly used fuzzy AND operation is the minimum operation or the algebraic product. This embodiment adopts the minimum operation.

[0063] The specific steps of fuzzy reasoning are:

[0064] (1) Substitute the membership value of the input language variable on the fuzzy set calculated in Section 1.1 as the judgment condition into each rule i in the fuzzy rule base;

[0065] (2) For each rule i, take the minimum value of all the membership values ​​in the judgment conditions to obtain the ignition intensity FS of this rule i , FS i =min[μ i1 (Th),μ i2 (Ls),μ i3 (Sr),μ i4 (Ns),μ i5 (Au)], i1,i2,i3,i4,i5∈{L,H,M}, respectively represent the fuzzy values ​​corresponding to Th, Ls, Sr, Ns, and Au in rule i;

[0066] (3) Using the implication operator, according to the ignition intensity FS of rule i i Truncate its output fuzzy set;

[0067] (4) Integrate the output fuzzy values ​​of all rules obtained in (3).

[0068] 1.3 Defuzzification

[0069] Defuzzification converts the fuzzy quantity obtained by fuzzy reasoning into an accurate numerical value. The defuzzification method adopted in this embodiment is the maximum membership method. After defuzzification, the accurate values ​​of the terminal priority and the MCS index are obtained.

[0070] 2. Active queue management based on traffic classification and fuzzy logic

[0071] Traffic is classified into live service traffic and non-live service traffic, and queues are built for their terminals in AP respectively. The optimization mechanism of this embodiment is implemented based on iptables and tc running on OpenWRT, in which the classified traffic algorithm adopts the HTB algorithm, i.e., the hierarchical token bucket algorithm. The HTB algorithm builds a tree, and each node of the tree is used as a class to divide multiple root nodes, intermediate nodes, and leaf nodes. The root node is a classification in a broad sense, and the data packets of all terminals will flow into this node. The root node is usually defined on the network card to be managed. In this embodiment, it is defined on the wlan0 wireless network card on OpenWRT. The intermediate node is the carrier for dividing the leaf node, which can contain multiple leaf node management rules, and the bandwidth of the leaf node is borrowed from the intermediate node. The leaf node is a specific execution rule, which contains its own data packet queue, bandwidth allocation, and priority. Figure 3 The schematic diagram of HTB hierarchical management is provided. In the HTB algorithm, the hierarchical control concept is used to manage the data of each class. Each class applies three modes: HTB_CAN_SEND, HTB_MAY_BORROW and HTB_CANT_SEND, which correspond to the data of this class being able to be sent, being able to borrow the bandwidth of the parent class to send, and the data of this class being unable to be sent. In the HTB algorithm, when several child class nodes borrow bandwidth from the parent class node at the same time, the nodes with higher priority will be satisfied first. If several nodes have the same priority, the remaining available bandwidth will be evenly divided. Figure 4 The process of active queue management is given.

[0072] The specific steps based on traffic classification and active queue management are as follows:

[0073] (1) Classify traffic into live service traffic and non-live service traffic, and mark the terminal IP containing live service traffic;

[0074] (2) A tree is established based on the hierarchical mechanism of the HTB algorithm. Each node of the tree is regarded as a class, and each leaf node has a subqueue. The root node, intermediate node, and leaf node are assigned a predefined borrowable bandwidth factor Quantum, rate Rate, and maximum rate Ceil, and a priority Prio is defined on the leaf node;

[0075] (3) Based on iptables, mark each service flow, classify the marked live service flow into the leaf node of HTB, and build the subqueue q 1 ,q 2 ,…,q n , distributed in leaf nodes 1 ,leaf 2 ,…,leaf n superior;

[0076] (4) The data packet sent by the AP is composed of the data on each queue, and the high-priority queue is sent first. The HTB algorithm starts from the bottom of the class tree and searches for nodes in the CAN_SEND state. If a node with the CAN_SEND state is found in a certain layer, it stops;

[0077] (5) If the layer is found to be a leaf node, there are m leaf nodes. i In the CAN_SEND state (i∈[1,n]), the m leaf nodes are sorted by priority, and the ones with higher priority are sent first; if there are multiple leaf nodes with the same priority, they are polled in these nodes. After each leaf node sends Quantum bytes in its subqueue, it is the turn of the next node to send;

[0078] (6) If the intermediate node is found, go down the tree to find the descendant leaf node. If there is only one leaf node in the MAY_BORROW state, lend the surplus Quantum value of the intermediate node to the leaf node so that it can send a data packet; if the intermediate node has multiple descendant leaf nodes in the MAY_BORROW state, the processing steps are the same as (5), according to the priority and polling operation, the leaf node with the highest priority sends first;

[0079] (7) Scheduling is performed based on the terminal priority obtained by fuzzy logic decision-making, and the terminal traffic is dynamically adjusted in each leaf node.

[0080] 3. Adaptive control of wireless transmission rate driven by live video transmission demand

[0081] The configuration of 802.11n radio rate is realized through MCS (Modulation and Coding Scheme) index. MCS takes the factors affecting the communication rate as the columns of the table and the MCS index as the rows to form a rate table. Therefore, each MCS index actually corresponds to the physical transmission rate under a set of parameters. The current Linux wireless subsystem has a wireless transmission rate adjustment mechanism based on the Minstrel algorithm, which mainly controls the transmission rate of data frames and maximizes the terminal wireless layer throughput. However, in a user-intensive environment, the available wireless channels of the terminal are determined by the channel contention mechanism, and there is a mismatch between its transmission rate and the upper-layer business requirements. For example, for chat messages with low latency and low bandwidth requirements, dynamic adjustment will appropriately reduce its MCS index and reduce its decoding error probability, but the reduction of MCS index will increase the transmission time, which will aggravate the channel contention in a user-intensive environment and increase the latency. Therefore, this example uses the existing wireless transmission multi-rate mechanism to design MCS index active control combined with the needs of live video transmission services.

[0082] The specific steps are as follows:

[0083] (1) Obtain the MCS index of the terminal based on fuzzy logic decision making, convert the terminal to be controlled and its MCS index and store them in the specified location of the user state;

[0084] (2) Call the iw command to call the modified kernel function through the control channel to read the content of the specified location in the user state;

[0085] (3) The read content is converted into corresponding variables and executed by the function modified in the kernel to achieve the purpose of modifying the MCSindex and controlling the wireless transmission rate of the terminal.

[0086] The above disclosure is only a preferred embodiment of the present invention, which cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for scheduling edge-side live video transmission in a dense wireless network, characterized in that: The method comprises the following steps: Step (1) fuzzy logic decision framework: The framework includes input fuzzification, fuzzy inference engine and defuzzification output, with throughput information, AP status information and terminal channel information as input, and its output is terminal priority and MCS index; the step (1) specifically includes: Step (11) constructing an AP side information collection module to obtain terminal throughput information, terminal channel information, and AP status information, and taking terminal throughput, terminal packet loss rate, terminal signal-to-noise ratio, terminal air splitting, and AP duty cycle (Th, Ls, Sr, Ns, Au) as five linguistic variables as input; Step (12) uses the triangle membership μ function to calculate the membership of each input linguistic variable in its own three fuzzy sets, that is, μ H (Th),μ M (Th), μ L (Th), μ H (Ls),μ M (Ls), μ L (Ls), μ H (Sr),μ M (Sr), μ L (Sr), μ H (Ns),μ M (Ns), μ L (Ns), μ H (Au),μ M (Au), μ L (Au); Step (13) uses the membership value of the input language variable calculated in step (11) on the fuzzy set as a judgment condition and substitutes it into each rule i in the fuzzy rule base; Step (14) For each rule i, take the minimum value of all the membership values ​​in the judgment conditions to obtain the ignition intensity FS of this rule i , FS i =min[μ i1 (Th),μ i2 (Ls),μ i3 (Sr),μ i4 (Ns),μ i5 (Au)], i1,i2,i3,i4,i5∈{L,H,M}, respectively represent the fuzzy values ​​corresponding to Th, Ls, Sr, Ns, and Au in rule i; Step (15) uses the implication operator to calculate the ignition intensity FS of rule i. i Truncate its output fuzzy set; Step (16) integrates the output fuzzy values ​​of all rules obtained in step (15); Step (17) applies the maximum membership method to the integrated output fuzzy value to obtain the precise value of the terminal priority and MCS index; Step (2) Active queue management based on traffic classification and fuzzy logic: traffic is classified into various types of business traffic, that is, various types of business traffic are classified into live business and non-live business; for live business traffic, the HTB algorithm is used to build a tree, and the hierarchical control idea is used to manage the traffic data of each node, and the priority of the node is set, and the leaf node has the function of borrowing bandwidth from the intermediate node, and the traffic is allocated to different queues according to the terminal priority determined by fuzzy logic; the step (2) specifically includes: Step (21) classifies the traffic into live service traffic and non-live service traffic, and marks the terminal IP containing the live service traffic; Step (22) establishes a tree according to the hierarchical mechanism of the HTB algorithm, with each node of the tree as a class and each leaf node having a subqueue; assigns a predefined borrowable bandwidth factor Quantum, a rate Rate, and a maximum rate Ceil to the root node, the intermediate node, and the leaf node, and defines a priority Prio on the leaf node; Step (23) marks each service flow based on iptables, classifies the marked live service flow into the leaf nodes of HTB, and constructs subqueues q1, q2, ..., q n , distributed in leaf nodes leaf1, leaf2,…, leaf n superior; Step (24) The data packet sent by the AP is composed of the data on each queue, among which the high priority queue is sent first; the HTB algorithm starts from the bottom of the class tree and searches for nodes in the CAN_SEND state upwards. If a node with the CAN_SEND state is found in a certain layer, it stops; Step (25) If the layer is found to be a leaf node, there are m leaf nodes leaf i In the CAN_SEND state (i∈[1,n]), the m leaf nodes are sorted by priority, and the ones with higher priority are sent first; if there are multiple leaf nodes with the same priority, they are polled in these nodes. After each leaf node sends Quantum bytes in its subqueue, it is the turn of the next node to send; If the intermediate node is found in step (26), the descendant leaf node is searched down the tree; if there is only one leaf node in the MAY_BORROW state, the surplus Quantum value of the intermediate node is lent to the leaf node so that it can send a data packet; if the intermediate node has multiple descendant leaf nodes in the MAY_BORROW state, the processing steps are the same as step (25), and the leaf node with the highest priority is sent first according to the priority and polling operation; Step (27) schedules the terminal priority obtained by fuzzy logic decision and dynamically adjusts the terminal traffic in each leaf node; Step (3) Adaptive control of wireless transmission rate driven by live video transmission demand: utilizing the existing wireless transmission multi-rate mechanism, designing active MCS index control combined with the live video transmission service demand; dynamically adjusting the MCS index of the terminal according to the MCS index determined by fuzzy logic, and realizing adaptive control of wireless transmission rate; the step (3) specifically includes: Step (31) obtains the MCS index of the terminal according to the fuzzy logic decision, converts the terminal to be controlled and its MCS index and stores them in the specified location of the user state; Step (32) calls the iw command to call the modified kernel function through the control channel to read the content of the user mode specified location; Step (33) converts the read content into corresponding variables, which are executed by the function modified in the kernel to achieve the purpose of modifying MCSindex and controlling the wireless transmission rate of the terminal.

Images