3.1. Multimedia transmission big data model
As shown in Fig. 1, in order to achieve the reliability evaluation and system optimization design of wireless multipath transmission, first use the communication protocol to build a wireless multipath transmission channel structure model, improve the balanced multipath suppression method, and output the wireless multipath transmission channel model stability.
According to the link structure and node distribution model of wireless multimedia transmission, the channel impulse response of wireless multimedia transmission is as follows:
$$h\left(t\right)=H\sum _{m=1}^{M}\sum _{k=1}^{k\left(m\right)}{\alpha }_{mk}\delta \left(t-{T}_{m}-{\tau }_{mk}\right)$$
1
Respectively used for the wireless multimedia output spreading loss caused by the transmission of the wireless multimedia transmission channel, the code channel attenuation coefficient of the m-th multimedia transmission channel, the channel transmission loss of the k-th path of the m-th wireless multimedia transmission channel, and its attenuation The relationship between is as follows:
$${K}_{\nu }\left(z\right)=1-\prod _{i=1}^{N}\left[\left(1-{P}_{di}\right)\left(1-{P}_{ei}\right)+{P}_{di}{P}_{ei}\right]$$
2
The impulse attenuation index of the wireless multimedia transmission channel constructed by combining the time domain equalization and frequency domain equalization joint modulation method is:
$$\zeta =\frac{{Q}^{+}\left(\theta \right)w}{{w}^{T}{Q}^{+}\left(\theta \right)w}$$
3
If there is multipath attenuation, the spectral gain of the transmitted signal output by the wireless multimedia network will be obtained through signal eigenvalue decomposition:
$${R}_{\left(t\right)}=\frac{\sqrt{WT}}{WT}sin\left[\pi WT\left(1-\frac{\left|\tau \right|}{T}\right)\right]\text{cos}\left(2\pi {f}_{0}\tau \right)$$
4
$$\left|s\left(f\right)\right|=A\sqrt{\frac{1}{2k}}\left\{{\left[c\left({\nu }_{1}\right)+c\left({v}_{2}\right)\right]}^{2}+{\left[s\left({\nu }_{1}\right)+s\left({\nu }_{2}\right)\right]}^{2}\right\}$$
5
$$E={‖x\left(t\right)‖}^{2}=\sum _{j}\sum _{k}{\left|{C}_{\dot{j}}\left(k\right)\right|}^{2}=\sum _{j}{E}_{\dot{j}}$$
6
The statistical characteristics of wireless multimedia scheduling are as follows:
$$g\left(t\right)=\frac{1}{\sqrt{{a}_{0}}}f\left(\frac{t-{\tau }_{0}}{{a}_{0}}\right)$$
7
Assuming that the channel bandwidth of wireless multimedia transmission is strictly limited, the decision feedback modulation method is adopted to obtain the multipath pulse of the output wireless multimedia as follows:
$${S}_{0}\left(t\right)={a}_{0}\delta \left(t\right)$$
8
At least count the data points that meet the data points in the MP domain, obtain the high-density area D, and create the output multipath. The characteristic components of wireless multipath transmission are obtained through channel compensation:
$${p}_{r}\left(t\right)=p{\left(t\right)}^{*}h\left(t\right)+{n}_{p}\left(t\right)$$
9
Through the above analysis, the adaptive equalization control of sampling decision is performed according to the strength of the inter-code interference in the communication transmission.
$${s}_{k}^{+}=\left\{\begin{array}{c}{V}_{max},{x}_{k}^{+}>{V}_{max}\\ {x}_{k}^{+},0\le {x}_{k}^{+}\le {V}_{max}\end{array}\right.$$
10
The negative signal is expressed as:
$${s}_{k}^{-}={x}_{k}^{-}+{\chi }_{k}$$
11
The sampling decision adaptive equalization control is performed according to the IG intensity of the communication transmission. Under the limited code transmission speed, the regularization parameters of the wireless multimedia transmission channel are as follows:
$${B}_{\left(i+1\right)}={\lambda }_{i}{B}_{\left(i\right)}+{\beta }_{i+1}^{-1}{x}_{i+1}{w}_{i+1}^{T}$$
12
$${C}_{(i+1)}^{-1}={\lambda }_{i}{C}_{\left(i\right)}^{-1}+{\beta }_{i+1}^{-1}{w}_{i+1}{w}_{i+1}^{T}$$
13
Combining the fractional interval sampling method, the balanced configuration and adaptability evaluation of multimedia transmission are carried out, and the recursive calculation form of the quantitative control of multimedia transmission is obtained:
$${C}_{\left(i+1\right)}={\lambda }_{i}^{-1}{C}_{\left(i\right)}-{\beta }_{i+1}^{-1}\alpha u{u}^{T}$$
14
$${D}_{\left(i+1\right)}={D}_{\left(i\right)}+{\beta }_{i+1}^{-1}\alpha {z}_{i+1}{u}^{T}$$
15
According to the frequency domain characteristics of the wireless multi-path transmission channel, the communication output errors are as follows:
$$\epsilon \left(k\right)=d\left(k\right)-y\left(k\right)=d\left(k\right)-\sum _{i=1}^{M}{W}_{i}x\left(k-i\right)$$
16
According to the frequency domain characteristics of the wireless multimedia transmission channel, multipath extension is performed. Under the constraint of the minimum mean square error standard, the offset of wireless multimedia transmission can be as follows:
$${K}_{\nu }\left(z\right)=1-\prod _{i=1}^{N}\left[\left(1-{P}_{di}\right)\left(1-{P}_{ei}\right)+{P}_{di}{P}_{ei}\right]$$
17
The best solution of the sparse representation matrix W of the output wireless multimedia transmission signal:
$${W}_{opt}=arg\underset{W}{\text{min}}\lambda {‖\left(X-DW\right)G‖}_{F}^{2}s\cdot t{‖{w}_{i}‖}_{0}\le k\forall i$$
18
3.2. System simulation experiment
As shown in Fig. 2, in order to verify the performance design of the wireless multimedia transmission system, combined with programming simulation experiments, the sampling time per second of wireless multimedia transmission information is 1s and the amplitude is ± 6, v set the baud rate of the serial port It is 9600, the duration of the low-level reset is set to 20 seconds, and the line multimedia transmission simulation is performed according to the above-mentioned set simulation parameters, and the input wireless multimedia signal model is obtained.
As shown in Fig. 3, the input signal waveform of the wireless multimedia signal model is used as input, and the big data fusion scheduling method is adopted to perform adaptive scheduling and transmission of wireless multimedia transmission to obtain the output signal.
As shown in Fig. 4, after analyzing the output signal model, if a method designed for wireless multimedia transmission is used, it can be found that the amplitude of the output signal changes smoothly between − 1 and 1. This can effectively reduce the signal loss, and by using this method, the interference suppression effect can be improved. The error bit rate refers to the ratio of the number of error bits in the received signal to the total number of bits in a specific time period during data communication. The bit error rate is a measure of the accuracy of data transmission.
3.3. Artificial Intelligence Education System Architecture
As shown in Fig. 5, it is designed for the physical architecture of the intelligent education platform. Before class, the tutor will enter the system operation page through the terminal device after connecting the teaching terminal and tutoring terminal equipment to the network. The client processes high CPU tasks, shares the pressure of the application server, communicates with the server through the WebSocket protocol, and forwards the calculated data to the Nginx load balancing server, and then Nginx forwards the request data to the application server. The application server forwards the request to the corresponding service for processing according to the priority of the data request. The types of requests that the application server can handle include multiple types of requests. Next, the data is sent to the cache server, the requested data is parsed, and then the data stored in the database is accessed, and the data is returned to the client layer by layer according to the path requested by the client, and operations related to business processing are performed. Temporarily store frequently accessed data in the cache server to speed up access.
The view presentation layer displays the user interface. The user clicks on different buttons or enters the information prompted by the system interface to interact with the system according to needs. Every step of the user’s operation will trigger the corresponding URL. The system communicates with the system background through WebSocket. The server performs data processing. After reaching the data layer, it reads and writes the data stored in the database. Finally, the data requested by the user will be returned along the data request path and displayed on the front-end page of the system.
3.4. Database design
As shown in Table 1, the table defines the basic information of the teacher, including teacher id, class id, subject, teacher’s grade and other fields. Among them, the role is used to distinguish the types of teachers, including the main teacher and the tutor, and the tutor No lecture function, the corresponding subject attribute value is empty: the teacher's state attribute value includes 0, 1, 2, where 0 means the teacher is disabled for various reasons, 1 means the teacher is in the normal working stage, 2 means the teacher has resigned and needs Delete the teacher’s information. The teacher's status flag is convenient for the administrator to manage the teacher and update the teacher information table in time.
Table 1
Teacher Information Table
| Serial number | Field name | Types of | Attributes | Description |
| 1 | Tea id | bigint | Primary key, non-empty | Teacher id |
| 2 | tea name | varchar | non empty | Name |
| 3 | shcool Id | int | non empty | Campus id |
| 4 | mail | varchar | non empty | Mailbox |
| 5 | phone | varchar | non empty | Contact details |
| 6 | sex | bigint | non empty | Gender |
| 7 | tea avatar | varchar | non empty | Avatar |
| 8 | role | varchar | non empty | Character |
| 9 | status | bigint | non empty | Status |
| 10 | class id | bigint | non empty | Class id |
| 11 | subject | varchar | Available | Subject |
| 12 | grade | varchar | Available | Teacher's grade |
Table 2
student Information Table
| Serial number | Field name | Types of | Attributes | Description |
| 1 | stu id | bigint | Primary key, non-empty | Student id |
| 2 | stu name | varchar | Non empty | Name |
| 3 | Age | varchar | Non empty | Age |
| 4 | Grade | varchar | Non empty | Grade |
| 5 | Sex | bigint | Non empty | Gender |
| 6 | Stu avatar | varchar | Non empty | Avatar |
| 7 | School id | int | Non empty | Campus id |
| 8 | Class id | bigint | Non empty | Class id |
As shown in Table 2, the student information table defines basic student information, including fields such as student id, class id, etc. The avatar field is used to mark avatars before class, so that the lecturer can interact with students in class. Students use clickers to answer questions in class, so they need to bind the student id with the clicker id in advance. To complete the binding operation, you need to find the student id in the student information table.
Table 3
| Serial number | Field name | Types of | Attributes | Description |
| 1 | class id | bigint | Primary key, non-empty | Classroom id |
| 2 | Class | varchar | non empty | Name |
| 3 | shcool Id | int | non empty | Campus id |
| 4 | Count | int | non empty | Total class size |
| 5 | Count full | int | non empty | Number of students supported |
| 6 | status | bigint | non empty | Class status |
| 7 | Class stasus | int | non empty | Class status |
| 8 | Count status | int | non empty | Class size status |
| 9 | Tutor id | bigint | non empty | Tutor id |
| 10 | Last class | long | non empty | Last class time |
| 11 | Next class | long | non empty | Last class start time |
As shown in Table 3, the class information table includes classroom id, name, campus id, total number of class, number of students that can be supported, status of class number, status of class in class, id of tutor, last class start time, last class time, class Fields such as status, where the number of class status is used to query the relevant restrictions and status of the class size, to determine whether the current enrollment of the class is full, which is convenient for new students to enroll and old students to drop out; the class status judges the pre-class attendance, the field value is 0 And 1, where 0 means no absences, 1 means that there are students absent from class; class status is used to indicate the flow of students and teachers in the class, 0 means that the class has recently joined new students, 1 means that students have dropped out, and 2 means that the class has changed to a new one. Of tutors, 3 indicates that the class has no staff changes recently.
Table 4
Courseware Information Table
| Serial number | Field name | Types of | Attributes | Description |
| 1 | lesson id | bigint | Primary key, non-empty | Courseware id |
| 2 | lesson | varchar | non empty | name |
| 3 | shcool Id | int | non empty | Campus id |
| 4 | Class id | bigint | non empty | Class id |
| 5 | class | varchar | non empty | Class name |
| 6 | subject | varchar | non empty | Subject |
| 7 | Count lesson | int | non empty | Course lectures |
| 8 | Tutor id | bigint | non empty | Tutor id |
| 9 | Tutor name | varchar | non empty | Name of tutor |
| 10 | lecture id | bigint | non empty | Lecturer id |
| 11 | Lecture name | varchar | non empty | Name of the lecturer |
| 12 | duration | long | non empty | duration |
| 13 | count | int | non empty | Number of classes |
As shown in Table 4, the courseware information table includes courseware id, course name, campus id, class id, class name, subject, number of lectures, tutor id, tutor name, lecturer id, lecturer name, duration, number of classes Fields, where the number of classes is used to count the number of times the courseware has been used since its release, as one of the evaluation criteria for the popularity of the course.