I haven't finished coding today, so I'll update it later, probably around one or two in the morning. The main reason is that this chapter is really a bit laborious. It took me two or three hours to write it, but it's still a big difference. I just need to refresh it after the update.

This chapter will do.

Abstract: In order to reduce the secure transmission delay of multiple heterogeneous network data, a secure transmission technology for multiple heterogeneous network data based on machine learning is designed. By selecting the data source and defining the importance of data attributes, preprocessing the multiple heterogeneous network data,

And establish a multi-path parallel transmission architecture. On this basis, machine learning methods are used to estimate effective bandwidth and parameter filtering processing. Finally, bandwidth scheduling and channel security protocol systems are established, thereby completing secure transmission of multiple heterogeneous network data based on machine learning.

.Experimental results show that the secure transmission of multi-heterogeneous network data based on machine learning effectively reduces data transmission delays, reduces data transmission interruptions and data packet loss rates, and meets the design needs of data transmission technology.

Keywords: machine learning; multiple heterogeneous networks; secure data transmission; network data preprocessing; parallel transmission architecture

2k

1 Introduction

At present, communication technology is developing rapidly, and various networks have obvious characteristics. After years of reform and innovation, the transmission rate of wireless access technology is gradually approaching the limit. In this context, in order to meet various business needs, multi-network writing is needed

.However, the traditional writing mechanism cannot be used simultaneously and efficiently in the use of network transmission resources, cannot effectively ensure efficient transmission services, and will increase energy consumption in transmission, resulting in interference problems during the transmission process. Therefore, many

Scholars have carried out research on multiple network data transmission methods. In the literature [1], Shi Lingling and Li Jingzhao studied the secure data transmission mechanism in heterogeneous networks. The mechanism mainly uses an optimization-based AES-GCM authentication encryption algorithm and a

The secure data transmission mechanism combined with SHA's digital signature algorithm performs data transmission; in literature [2], Zhou Jing and Chen Chen studied a data security model based on heterogeneous networks, which encrypts the data in advance, and then

A secure transmission channel is established to transmit data. The above two methods can achieve certain results, but there are still certain shortcomings. In view of the above shortcomings, this paper applies machine learning methods to the secure transmission of data in multiple heterogeneous networks.

To solve the existing problems. The experimental results show that the multi-heterogeneous network data secure transmission technology studied this time effectively solves the existing problems and has certain practical application significance.

2 Multivariate heterogeneous network data preprocessing

In the secure transmission of data on multiple heterogeneous networks, a lot of data is useless. For this reason, it is necessary to select relevant data sources from the multiple network data for transmission, thereby improving the accuracy and efficiency of data transmission. In the process of selecting effective data sources

In, importance is used to measure the relationship between data attributes [3-4] to capture highly correlated data. The calculation expression is as follows: (1) In formula (1), T represents the comprehensive sum of all data sources.

The number of tables, (i, j) represents the correlation between the sample source classes. Based on the judgment of the importance of the data source, the collection of data tables with the highest degree of correlation can be selected and the irrelevant tables can be reduced. After the selection of the above important data sources is completed,

Analyze data attributes. Since a data source is composed of a set of data attributes, these attribute characteristics can reflect the basic information of the data to be transmitted. It is mainly measured by the correlation of data tuples, and the number of occurrences of tuple data is analyzed, that is

It is defined by tuple data density, and the data tuple density diagram is shown in Figure 1. In Figure 1, ε represents the radius of the specified neighborhood. According to this idea, the weight is assigned to each tuple data in the above data set.

[5-7], its expression is as follows: (2) In formula (2), w(C) represents the attribute weight, w(tk) represents the number of core tuples, δ represents outliers, and w(tb) represents

Number of tuples of edges.

3Multipath parallel transmission architecture

After the above preprocessing is completed, a multipath parallel transmission architecture is established. The main contents are as follows: traffic segmentation in advance. Communication flow segmentation is used by the sender to segment large data blocks into data units of different sizes or the same size [8].

The size is determined by the granularity of communication flow segmentation, which is mainly divided into the following categories: First, in packet-level business segmentation, the packet is the smallest component unit of the data flow. Therefore, the segmentation method has the smallest granularity, and the grouping probabilities are independent of each other and can be sent

to the sending end; second, traffic segmentation at the flow level [9], which encapsulates a specific destination address in the packet header, and then aggregates packets with the same destination address into data flows. These different data flows are independent of each other and passed through a unique

Flow identifiers are used to distinguish them. The use of flow-level segmentation technology can effectively solve the impact of data distortion on multipath transmission [10]. Third, traffic segmentation at the sub-flow level, the data flow with the same destination header is divided into multiple sub-flows.

, the packets in all sub-flows have the same destination address, which solves the load imbalance problem in the flow segmentation algorithm to a certain extent. The multi-path parallel transmission architecture is shown in Figure 2. In addition, in bandwidth aggregation

In the architecture, the scheduling algorithm is the core of determining the service transmission mode and the scheduling order of service subflows [11], ensuring that the service subflows arrive at the receiving end in an orderly manner. Next, we will discuss data scheduling.

4 Bandwidth scheduling plan formulation

For multiplex

�

�

The data that constructs the network

�

lose

�

�

�when�

�

The bandwidth of � path reaches �

�

Determine

�

�

hour

�

�

�The bandwidth of the network�

�

�Continuously increasing

�

�

��

�

� lose �

�

�Ability�

�

�Relatively stable

�

�

�

To improve throughput

�

�

�Allocating too much bandwidth�

�

�Reduce spectrum utilization

�

�

�from�

�

��

�

Causes waste of spectrum resources

�

�

�

In the current increasingly tight spectrum resources,

�

� case

�

�

�For multipath parallelism�

�

Schedule and manage the bandwidth of each channel in the transmission

�
To be continued...