Network background flow generation method and system based on conditional generative adversarial network

Abstract

The invention relates to a network background flow generation method based on a conditional generative adversarial network, which comprises the following steps: a data acquisition step: acquiring network flow data and conditional information, and vectorizing the acquired network flow data and conditional information into real flow; a model generation step: obtaining an initial generation model anda discrimination model according to the real flow, and training the initial generation model by using the discrimination model through a conditional generative adversarial network to obtain a generation model; and a flow generation step of generating simulated background flow by using a random vector through the generation model.

Description

technical field [0001] The invention relates to the fields of network security and network simulation, in particular to a method and system for generating network background traffic based on conditional generative confrontation network (CGAN). Background technique [0002] Network traffic generation techniques are mostly used for network testing. From the early days of the network to the present, with the continuous growth of network scale and complexity, network traffic generation technology is also constantly developing. In terms of network traffic generation methods, it can be divided into three categories: statistical model-based generation methods, traffic feature-based generation methods, and application / session-based generation methods. [0003] 1. Generation method based on statistical model [0004] Based on statistical analysis theory, this method uses statistical models such as Poisson, ON / OFF, FBM / FGN or multifractal to describe the distribution of network traf...

AI Technical Summary

Problems solved by technology

This method only considers the basic characteristics of a single data packet, the advantages are: simple, low computational load; the disadvantage is: low fidelity, does not consider the mutual influence between various data packets, ignores the inter-protocol and the internal of a single protocol traffic characteristics

It is impossible to give a clear conclusion as to whether the proposed traffic characteristics can fully reflect the characteristics of real network traffic, or whether there are hidden features that have not been discovered in real network traffic.

However, inferring from the nature of cognition, for complex and diverse network traffic, it is very likely that there are more hidden features that have not been discovered.

[0014] (2) The extraction of real network features is very cumbersome

The time, place, user, application, and event are varied and unpredictable

Not only is the feature extraction of real network traffic very cumbersome, but it is also difficult to accurately match real network traffic

[0016] (3) The ease of use of the above network generation method is very poor

In particular, the network traffic modeling process consumes user time and effort, and the effect cannot be guaranteed

Images