發(fā)布時(shí)間：2018-02-05 01:10

  本文關(guān)鍵詞： 無(wú)線資源管理 綠色效率 分布搜索 混合模型粒子群算法 多維資源聯(lián)合優(yōu)化　出處：《西安電子科技大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：由于可再生能源的逐漸減少和全球變暖等問(wèn)題日漸嚴(yán)重,節(jié)能減排成了世界性課題。據(jù)統(tǒng)計(jì),當(dāng)前的ICT行業(yè)能源消耗占世界總能源消耗的2%-6%左右,而且這一比例正在迅速提高。通信行業(yè)的高能耗問(wèn)題由來(lái)已久,特別是蜂窩移動(dòng)通信的快速發(fā)展使得這一問(wèn)題更加突出。針對(duì)通信行業(yè)高能耗的問(wèn)題,“綠色通信”歷史性的成了通信行業(yè)的新任務(wù)。綠色通信不是單純的以減少能源的消耗為首要目標(biāo),而是在保證用戶的業(yè)務(wù)要求的基礎(chǔ)上在有限的通信資源下盡量減少能源的消耗,從而減少對(duì)環(huán)境的污染和溫室氣體的排放。對(duì)于LTE系統(tǒng)中的高能耗問(wèn)題,以綠色效率為目標(biāo)的資源管理方法成為了解決這一問(wèn)題重要手段。LTE系統(tǒng)的業(yè)務(wù)種類和QoS需求更加多樣化,但是不能無(wú)限制地滿足各類不同業(yè)務(wù)的需求或者為了滿足苛刻業(yè)務(wù)的需求而浪費(fèi)大量的資源。在滿足不同用戶需求的條件下如何高效合理的分配無(wú)線資源已成為L(zhǎng)TE系統(tǒng)研究的熱點(diǎn)問(wèn)題。無(wú)線資源管理既要保證系統(tǒng)的吞吐量和能量效率、用戶的公平性等性能,同時(shí)對(duì)算法的復(fù)雜度和收斂性也有相應(yīng)的要求。本文從效率的原始內(nèi)涵出發(fā),分析無(wú)線通信效率的內(nèi)涵、外延與優(yōu)化準(zhǔn)則,把問(wèn)題通過(guò)數(shù)學(xué)語(yǔ)言轉(zhuǎn)換成精確描述、公式化定義的嚴(yán)謹(jǐn)理論模型。針對(duì)LTE網(wǎng)絡(luò),本文構(gòu)造基于綠色效率的具有邊際效應(yīng)遞減的效用函數(shù),并以此來(lái)實(shí)現(xiàn)多維資源聯(lián)合優(yōu)化。多維資源聯(lián)合優(yōu)化算法的計(jì)算復(fù)雜度較高,為了保證算法的實(shí)時(shí)性,本文通過(guò)分步搜索、迭代優(yōu)化的方式,在多項(xiàng)式時(shí)間內(nèi)逼近納什議價(jià)解�；玖Ｗ尤核惴ㄊ且环N啟發(fā)式進(jìn)化算法,其基本思想源于對(duì)自然界中鳥(niǎo)群、魚(yú)類等生物群體覓食行為的仿真研究�；玖Ｗ尤核惴▍�(shù)設(shè)置簡(jiǎn)單、收斂速度快但是算法的達(dá)優(yōu)率較低,容易陷入局部最優(yōu),不能較好的獲得目標(biāo)函數(shù)的全局最優(yōu)解。對(duì)于基本粒子群算法中的缺點(diǎn),本文通過(guò)引入混合模型粒子群優(yōu)化算法自適應(yīng)的調(diào)整算法的全局尋優(yōu)能力和局部尋優(yōu)能力,可以在保證分步搜索的收斂速度的前提下避免陷入局部最優(yōu)陷阱,提高尋優(yōu)能力。最后通過(guò)Matlab軟件來(lái)進(jìn)行平臺(tái)的搭建和算法的仿真,并與傳統(tǒng)的資源管理算法進(jìn)行對(duì)比,分析并驗(yàn)證了多維資源聯(lián)合優(yōu)化算法的優(yōu)點(diǎn)。
[Abstract]:The renewable energy gradually reduced and increasingly serious problems such as global warming, energy saving and emission reduction has become a worldwide issue. According to statistics, the current ICT industry energy consumption accounted for 2%-6% of the total energy consumption around the world, and this proportion is increasing rapidly. Long-standing high energy consumption problems in the communications industry, especially the rapid development of mobile communication the problem is more prominent. According to the communications industry, the problem of high energy consumption, "green communication" has become a new historic task in the communications industry. Green communication is not simply to reduce energy consumption is the primary goal, but to ensure that the user's business requirements based on limited communication resources to minimize energy consumption, thereby reducing the environmental pollution and greenhouse gas emissions. For the problem of high energy consumption in the LTE system, resource management method based on green efficiency as the goal of the In order to solve this problem is an important means of.LTE system and QoS needs more diverse types of business, but can not indefinitely meet the different needs of the business or to meet the demanding needs of the business and a great waste of resources. To meet the needs of different users under the condition of how reasonable and efficient allocation of radio resources has become a hot issue in LTE system research on radio resource management. To ensure the throughput and energy efficiency of the system, the fairness performance of the user, but the complexity of algorithm convergence and also have the corresponding requirements. The efficiency of the original meaning of the content analysis of wireless communication efficiency, extension and optimization criterion, the problem through the language of mathematics into precise description, define rigorous theoretical model. In LTE network, this paper has constructed based on the utility function, the marginal effect of diminishing the number of green efficiency, In order to achieve optimal multidimensional resources. Calculation algorithm of joint optimization of multidimensional resources with high complexity, in order to ensure the real-time algorithm, through step-by-step search, iterative optimization, approximation of Nash bargaining solution in polynomial time. Basic particle swarm algorithm is a heuristic evolutionary algorithm, the basic idea stems from the bird group nature, Simulation Research on the foraging behavior of fish and other biological groups. Parameters of basic particle swarm optimization algorithm is simple and fast convergence but the algorithm success rate is low, easy to fall into local optimal solution can obtain the global optimal objective function better. For basic particle swarm algorithm, by introducing the mixed particle model swarm optimization algorithm of adaptive adjustment algorithm's searching ability and local optimization ability and convergence speed can avoid the premise to ensure the search step by step Finally, we use Matlab software to build platform and algorithm simulation, and compare with the traditional resource management algorithm, analyze and verify the advantages of multi-dimensional resource optimization algorithm.

【學(xué)位授予單位】：西安電子科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TN929.5

【共引文獻(xiàn)】

相關(guān)期刊論文 前10條

1 薛振偉;;管理信息系統(tǒng)流程復(fù)雜度計(jì)算模型研究[J];測(cè)試技術(shù)學(xué)報(bào);2011年04期

2 彭英;劉曼玲;季飛;;軟件估算中錨定效應(yīng)的研究[J];電腦與信息技術(shù);2006年05期

3 趙宇;王欣;;如何從企業(yè)級(jí)度量?jī)?chǔ)存庫(kù)中獲得最大收益[J];電腦知識(shí)與技術(shù);2008年S2期

4 朱方洲;周偉良;姜婷;;基于度量能力成熟度模型的組織級(jí)度量過(guò)程改進(jìn)研究[J];電腦知識(shí)與技術(shù);2009年22期

5 張克宏;;基于規(guī)則的面向?qū)ο髥卧獪y(cè)試研究[J];電腦知識(shí)與技術(shù);2010年08期

6 季飛;彭英;;軟件成本估算模型的研究與本地模型的建立[J];大眾科技;2008年10期

7 郭智恒;潘日晶;;基于改進(jìn)的粒子群算法求NURBS曲面間的最小距離[J];福建師范大學(xué)學(xué)報(bào)(自然科學(xué)版);2010年04期

8 吳智;王曉華;趙明;;一種新的XP過(guò)程軟件可靠性預(yù)測(cè)模型[J];貴州大學(xué)學(xué)報(bào)(自然科學(xué)版);2008年06期

9 呂桂志;;基于TRIZ進(jìn)化及復(fù)雜度對(duì)軟件產(chǎn)品設(shè)計(jì)的研究[J];電子技術(shù);2012年07期

10 張位勇;鄒北驥;;一種改進(jìn)的軟件過(guò)程質(zhì)量度量方法[J];電子科技;2014年04期

相關(guān)博士學(xué)位論文 前5條

1 盛津芳;商業(yè)構(gòu)件評(píng)估方法及關(guān)鍵技術(shù)研究[D];中南大學(xué);2007年

2 何玉安;基于本體的制造網(wǎng)格任務(wù)管理關(guān)鍵技術(shù)研究[D];上海大學(xué);2009年

3 劉錚;大規(guī)模軟件可信性度量分析原理及其方法的研究[D];東北大學(xué);2010年

4 周偉良;軟件開(kāi)發(fā)過(guò)程質(zhì)量與產(chǎn)品質(zhì)量度量方法研究[D];合肥工業(yè)大學(xué);2012年

5 文杏梓;基于用戶需求的可信軟件質(zhì)量屬性評(píng)價(jià)方法研究[D];中南大學(xué);2013年

，

本文編號(hào)：1491683