面向分層異構(gòu)網(wǎng)絡(luò)的資源分配:一種穩(wěn)健分層博弈學習方案
發(fā)布時間:2019-05-21 16:06
【摘要】:該文研究了信道狀態(tài)不確定條件下分層異構(gòu)微蜂窩網(wǎng)絡(luò)中的無線資源分配優(yōu)化問題。首先引入信道不確定模型描述無線信道的隨機動態(tài)性,并將該問題建模為考慮信道不確定度的雙層魯棒斯坦伯格博弈;然后給出了該博弈的均衡點分析;最后提出了一種分布式改進型分層Q學習方案以實現(xiàn)宏基站和微基站的均衡策略搜索。理論分析和仿真表明,所提出的分層博弈模型可以有效抑制由于信道狀態(tài)不確定引起的收益下降。所采用的學習方案較傳統(tǒng)Q學習方案收斂速度明顯加快,更加適用于短時快變的通信環(huán)境。
[Abstract]:In this paper, the optimization of wireless resource allocation in layered heterogeneous microcellular networks with uncertain channel state is studied. Firstly, the channel uncertainty model is introduced to describe the stochastic dynamics of wireless channels, and the problem is modeled as a double-layer robust Steinberg game considering channel uncertainty, and then the equilibrium point analysis of the game is given. Finally, a distributed improved hierarchical Q learning scheme is proposed to realize the equilibrium strategy search of macro base station and micro base station. Theoretical analysis and simulation show that the proposed hierarchical game model can effectively suppress the decline of benefits caused by channel state uncertainty. Compared with the traditional Q learning scheme, the convergence speed of the proposed learning scheme is obviously accelerated, and it is more suitable for short-term and rapidly changing communication environment.
【作者單位】: 解放軍理工大學通信工程學院;洛陽理工學院;南京電訊技術(shù)研究所;解放軍信息工程大學信息系統(tǒng)工程學院;
【基金】:國家自然科學基金(61471395,61401508) 江蘇省自然科學基金(BK20161125)~~
【分類號】:TN929.5
[Abstract]:In this paper, the optimization of wireless resource allocation in layered heterogeneous microcellular networks with uncertain channel state is studied. Firstly, the channel uncertainty model is introduced to describe the stochastic dynamics of wireless channels, and the problem is modeled as a double-layer robust Steinberg game considering channel uncertainty, and then the equilibrium point analysis of the game is given. Finally, a distributed improved hierarchical Q learning scheme is proposed to realize the equilibrium strategy search of macro base station and micro base station. Theoretical analysis and simulation show that the proposed hierarchical game model can effectively suppress the decline of benefits caused by channel state uncertainty. Compared with the traditional Q learning scheme, the convergence speed of the proposed learning scheme is obviously accelerated, and it is more suitable for short-term and rapidly changing communication environment.
【作者單位】: 解放軍理工大學通信工程學院;洛陽理工學院;南京電訊技術(shù)研究所;解放軍信息工程大學信息系統(tǒng)工程學院;
【基金】:國家自然科學基金(61471395,61401508) 江蘇省自然科學基金(BK20161125)~~
【分類號】:TN929.5
【相似文獻】
相關(guān)期刊論文 前10條
1 陳慶章;組建異構(gòu)網(wǎng)應著重考慮的問題和產(chǎn)品現(xiàn)狀[J];計算機時代;1994年03期
2 汪蕓,顧冠群,謝俊清,兌繼英,孫昌平;異構(gòu)網(wǎng)絡(luò)集成方法研究[J];計算機研究與發(fā)展;1997年03期
3 馬義忠;楊紅旗;高彥;専秋峰;;基于移動Agent異構(gòu)網(wǎng)絡(luò)管理的分析與設(shè)計[J];微計算機信息;2008年33期
4 黃川;鄭寶玉;;多無線電協(xié)作技術(shù)與異構(gòu)網(wǎng)絡(luò)融合[J];中興通訊技術(shù);2008年03期
5 吳蒙;季麗娜;王X;;無線異構(gòu)網(wǎng)絡(luò)的關(guān)鍵安全技術(shù)[J];中興通訊技術(shù);2008年03期
6 李R,
本文編號:2482215
本文鏈接:http://sikaile.net/kejilunwen/xinxigongchenglunwen/2482215.html
最近更新
教材專著