Intelligent video code rate adjustment and bandwidth allocation method based on deep learning

Abstract

The invention discloses an intelligent video code rate adjustment and bandwidth allocation method based on deep learning, and the method comprises the steps: constructing a code rate allocation neural network, the input of the code rate allocation neural network is a state vector comprising an available bandwidth normalized value, a user cache video duration normalized value and a user active state, and the output of the code rate allocation neural network is a code rate allocation scheme vector of each user; setting a plurality of user states and available bandwidth scenes to obtain training samples, and training the code rate allocation neural network; and when each decision time slot arrives, obtaining a state vector, inputting the state vector into the trained code rate allocation neural network to obtain a code rate allocation scheme, and determining the bandwidth occupied by each user according to the code rate allocation scheme and the current available bandwidth to complete bandwidth allocation. According to the method, the decoding rate and the bandwidth allocation scheme are quickly solved by utilizing the characteristic of high execution speed of a neural network inference process, and the timeliness and the accuracy are improved.

Description

technical field [0001] The invention belongs to the field of communication technology, and more specifically, relates to an intelligent video code rate adjustment and bandwidth allocation method based on deep learning. Background technique [0002] In recent years, the volume of video streaming business has increased dramatically and has occupied a major part of the entire Internet business. In order to adapt to dynamic network conditions and provide better service quality, most video service providers have deployed an HTTP-based Dynamic Adaptive Streaming (Dynamic Adaptive Streaming over HTTP, DASH) solution. In the DASH scheme, a video file is divided into segments of fixed segment length (in seconds). Each segment of the video can be encoded as a different copy using a different bit rate, and each bit rate corresponds to a specific video frame rate and resolution. Clients can adaptively request bitrates (different frame rates and resolutions) for the next segment based ...

AI Technical Summary

Problems solved by technology

When multiple DASH streams compete for the same bottleneck link, the existing adaptive bit rate adjustment algorithm has the following disadvantages: 1) Each client independently makes a bit rate adjustment algorithm without knowing the existence of other clients. Adaptive decision-making, lack of collaboration between clients; 2) Clients can only detect physical bandwidth at the application layer, lacking accurate network bandwidth information

[0014] However, it has been proven that this problem is NP-hard, and the solution cannot be guaranteed to be completed in polynomial time, and it cannot meet the delay-sensitive user business needs at this stage.

Images