Algorithm for solving power distribution in cognitive radio based on reinforcement learning

A cognitive radio and reinforcement learning technology, applied in the field of power allocation strategy, can solve the problem of incomplete channel information and power allocation, and achieve the effect of effectively adjusting the transmission power

Active Publication Date: 2021-02-12
NORTHWESTERN POLYTECHNICAL UNIV
View PDF4 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a power allocation algorithm based on reinforcement learning to solve cognitive radio, so as to solve the problem that power allocation cannot be performed well under the premise of incomplete channel information in the prior art

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Algorithm for solving power distribution in cognitive radio based on reinforcement learning
  • Algorithm for solving power distribution in cognitive radio based on reinforcement learning
  • Algorithm for solving power distribution in cognitive radio based on reinforcement learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0040] 1. Simulation conditions: 1) The number of CUs is K=6, 2) The transmission power of the PU is P PU = 15dB, 3) the discount factor is β = 0.9, 4) the learning rate of the participants is η a = 0.01, 5) The critic's learning rate is η c = 0.001.

[0041] 2. Simulation content: simulate and compare the relationship between the spectral efficiency (Spectralefficiency, SE) performance of CUs and the time index under different learning algorithm scenarios. The results are as follows: figure 2 . figure 2 Among them, the vertical axis is "spectrum utilization rate of cognitive users"; the horizontal axis is "simulation iteration time".

[0042] Depend on figure 2 Simulation results show that by using Q-learning, continuous-valued states and actions must be quantized, and actual values ​​are replaced by finite discrete-valued approximations. Contrary to our AC-RL algorithm, the Q-learning based power allocation algorithm needs to know the immediate CSI of CUs. Figure 2 ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses an algorithm for solving power distribution in cognitive radio based on reinforcement learning, which comprises the following steps: S1, setting initial value parameters of a deep learning algorithm, S2, setting a scene model related to a CR-NOMA system, and setting an initial state set related to states and actions; S3, when a certain calculation moment t is smaller than or equal to the time value Tmax of the maximum limit, solving a state value at the moment t, calculating a corresponding reward function, and calculating a TD error [delta]t; S4, selecting the next action of the user based on the value function, and updating the initial value function to Q (st, at)<-Q(st, at)+[eta]c[delta]t by using the learning rate and the TD error value function; obtaining a corresponding reward according to the selected execution action, obtaining a strategy function [pi](g), and updating the strategy function [pi](g) to [pi](st, at)<-[pi](st, at)-[eta]a[delta]t; [pi](g); and S5, enabling the TD error value to reach the minimum according to the step S3, continuously iteratively updating, and finally obtaining the maximum reward function value, i.e., ending the allocation algorithm. The problem that in the prior art, power distribution cannot be well conducted on the premise that channel information is incomplete is solved.

Description

technical field [0001] The invention belongs to the technical field of communication, and in particular relates to a power allocation strategy, which can be used to solve the power allocation problem in an underlay cognitive radio network. Background technique [0002] Overlay cognitive radio networks can solve the problem of spectrum scarcity, that is, under the constraint that the interference caused by cognitive users cannot degrade the service quality of primary users, cognitive users can use the same spectrum to transmit simultaneously with primary users. On the other hand, Non-orthogonal Multiple Access (NOMA), as a potential technical challenge to improve the spectrum efficiency of future wireless networks, has fundamentally changed the design of conventional access technologies . Power domain non-orthogonal multiple access technology (Power-domain NOMA) is one of the most popular technologies in NOMA technology. Its core idea is to explore the power domain differenc...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): H04B17/382H04W52/34
CPCH04B17/382H04W52/34
Inventor 梁微温书慧杨思远王大伟高昂李立欣
Owner NORTHWESTERN POLYTECHNICAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap