本文選題：內(nèi)容分發(fā)網(wǎng)絡(luò) + 流媒體分發(fā)　； 參考：《江西理工大學(xué)》2017年碩士論文

【摘要】：隨著互聯(lián)網(wǎng)技術(shù)和多媒體技術(shù)的發(fā)展,流媒體視頻點播已經(jīng)成為互聯(lián)網(wǎng)上最流行的應(yīng)用之一。因此傳統(tǒng)點對點式的流媒體視頻分發(fā)模型已經(jīng)難以滿足日益增長的需求,內(nèi)容分發(fā)網(wǎng)絡(luò)(Content Delivery Networks,CDN)技術(shù)則在這樣的背景下應(yīng)運而生。目前如何基于CDN網(wǎng)絡(luò)結(jié)構(gòu)架構(gòu)新一代流媒體視頻分發(fā)業(yè)務(wù),使其能夠應(yīng)對高并發(fā)大流量的流媒體視頻分發(fā)業(yè)務(wù)也成為了業(yè)界所研究的熱點問題之一。本文首先對CDN網(wǎng)絡(luò)的拓?fù)浣Y(jié)構(gòu)進行了分析,并以CDN網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)作為基礎(chǔ),提出了適用于流媒體視頻分發(fā)業(yè)務(wù)的分發(fā)拓?fù)浣Y(jié)構(gòu)。針對流媒體視頻分發(fā)業(yè)務(wù)的優(yōu)化問題,分別從費用優(yōu)化和延遲優(yōu)化兩個角度對問題進行了研究。(1)對于費用優(yōu)化問題,CDN網(wǎng)絡(luò)中節(jié)點具有聚集現(xiàn)象,在網(wǎng)絡(luò)的演化過程中自然形成具有協(xié)作能力的社區(qū)結(jié)構(gòu),而且同一社區(qū)用戶往往具備共同的流媒體內(nèi)容偏好,如果CDN的每個分發(fā)節(jié)點可以有針對性的依據(jù)內(nèi)容的種類來進行流媒體內(nèi)容分發(fā),將大大減少分發(fā)過程中的存儲費用、傳輸費用和復(fù)制費用。因此提出一種基于差分演化思想的自適應(yīng)調(diào)整差分演化動態(tài)社區(qū)發(fā)現(xiàn)算法(Improved Differential Evolution Dynamic Community Detection Algorithm,IDEDCD)有效挖掘社區(qū)結(jié)果�；趧討B(tài)社區(qū)結(jié)構(gòu)的劃分結(jié)果進而流媒體視頻內(nèi)容分發(fā),起到大大降低分發(fā)費用的作用。(2)對于延遲優(yōu)化問題,根據(jù)求解動態(tài)調(diào)度問題的方法,以最小流媒體視頻分發(fā)時間為優(yōu)化目標(biāo)建立動態(tài)調(diào)度模型,提出一種基于改進PSO算法求解CDN流媒體視頻分發(fā)問題(Improved Particle Swarm Algorithm for Dynamic Scheduling Video Streaming Services on CDN,IPSO)。一方面改進的PSO算法具有更好的延遲優(yōu)化效果;另一方面在負(fù)載較大的情況下,優(yōu)化效果仍然十分穩(wěn)定。
[Abstract]:With the development of Internet technology and multimedia technology, streaming video on demand has become one of the most popular applications on the Internet. Therefore, the traditional point-to-point video distribution model of streaming media has been difficult to meet the increasing demand, and the content Delivery networks (CDNs) technology emerges as the times require in this context. At present, how to construct a new generation of streaming media video distribution service based on CDN network structure, so that it can deal with the high concurrent and large traffic streaming video distribution service has become one of the hot issues in the industry. In this paper, the topology of CDN network is analyzed, and based on the topology structure of CDN network, a distribution topology for streaming media video distribution services is proposed. Aiming at the optimization problem of streaming media video distribution service, this paper studies the problem from two aspects of cost optimization and delay optimization respectively) for the cost optimization problem, the nodes in CDN network have aggregation phenomenon. In the evolution of the network, the community structure with cooperative ability is formed naturally, and the users of the same community often have the same preference for streaming media content. If each distribution node of CDN can distribute streaming media content according to the type of content, the cost of storage, transmission and replication will be greatly reduced. Therefore, a dynamic community discovery algorithm named improved Differential Evolution Dynamic Community Detection algorithm based on the idea of differential evolution is proposed to effectively mine the community results. Based on the partition result of dynamic community structure and streaming media video content distribution, it can greatly reduce the distribution cost.) for the delay optimization problem, according to the method of solving the dynamic scheduling problem, A dynamic scheduling model based on improved PSO algorithm is proposed to solve the video distribution problem of CDN streaming media based on the minimum video distribution time of streaming media. On the one hand, the improved PSO algorithm has better delay optimization effect; on the other hand, the optimization effect is still very stable in the case of heavy load.
【學(xué)位授予單位】：江西理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP393.02

【參考文獻】

中國期刊全文數(shù)據(jù)庫 前10條

1 暏大海;李元香;龔文引;何國良;;一種求解約束優(yōu)化問題的自適應(yīng)差分進化算法[J];電子學(xué)報;2016年10期

2 費騰;張立毅;;現(xiàn)代智能優(yōu)化算法研究[J];信息技術(shù);2015年10期

3 張震;張雷;;基于CCN的CDN視頻內(nèi)容分發(fā)技術(shù)研究[J];軟件;2015年01期

4 劉陽;季新生;劉彩霞;;一種基于邊界節(jié)點識別的復(fù)雜網(wǎng)絡(luò)局部社區(qū)發(fā)現(xiàn)算法[J];電子與信息學(xué)報;2014年12期

5 王莉;程學(xué)旗;;在線社會網(wǎng)絡(luò)的動態(tài)社區(qū)發(fā)現(xiàn)及演化[J];計算機學(xué)報;2015年02期

6 毛佳昕;劉奕群;張敏;馬少平;;基于用戶行為的微博用戶社會影響力分析[J];計算機學(xué)報;2014年04期

7 逯少華;張曉偉;鮑承強;李文寶;;柯西種群分布的自適應(yīng)范圍粒子群優(yōu)化算法[J];計算機應(yīng)用;2014年04期

8 劉景;;基于云計算環(huán)境的校園網(wǎng)網(wǎng)絡(luò)視頻直播設(shè)計[J];計算機應(yīng)用;2014年02期

9 徐志明;李棟;劉挺;李生;王剛;袁樹侖;;微博用戶的相似性度量及其應(yīng)用[J];計算機學(xué)報;2014年01期

10 王莉;程蘇琦;沈華偉;程學(xué)旗;;在線社會網(wǎng)絡(luò)共演化的結(jié)構(gòu)推斷與預(yù)測[J];計算機研究與發(fā)展;2013年12期

中國博士學(xué)位論文全文數(shù)據(jù)庫 前1條

1 李景濤;P2P環(huán)境下的信任模型與副本方案研究[D];復(fù)旦大學(xué);2006年

中國碩士學(xué)位論文全文數(shù)據(jù)庫 前1條

1 張婷娜;復(fù)雜網(wǎng)絡(luò)模塊度的研究[D];西安理工大學(xué);2010年

，

本文編號：1937365