【摘要】：移動自組織網(wǎng)絡(luò)通過節(jié)點間相互合作的方式來轉(zhuǎn)發(fā)數(shù)據(jù)包,提升網(wǎng)絡(luò)的性能和服務(wù)質(zhì)量。然而,受終端無線信號覆蓋范圍、設(shè)備儲電能力和節(jié)點移動特性等因素限制,使得節(jié)點獲取數(shù)據(jù)時存在不確定性。將服務(wù)內(nèi)容部署在適當?shù)墓?jié)點上會提高節(jié)點獲取數(shù)據(jù)的成功率。但在現(xiàn)實中,節(jié)點是理性的,在沒有利益的情形下,節(jié)點難以與其他節(jié)點合作部署服務(wù)內(nèi)容。為了提高網(wǎng)絡(luò)性能,需要設(shè)計一種有效的激勵方法,鼓勵自私節(jié)點部署服務(wù)內(nèi)容。如何設(shè)計激勵方法,減少部署服務(wù)內(nèi)容的成本,提升用戶獲取數(shù)據(jù)的效率是本文的研究重點。本文主要從以下兩方面對服務(wù)內(nèi)容部署激勵方法進行研究:1.具有全局信息時,針對節(jié)點自私的情形,為了鼓勵節(jié)點積極地參與到服務(wù)內(nèi)容部署中,本文設(shè)計了一種基于虛擬積分的激勵方法。該方法將節(jié)點分為忙碌和空閑兩種狀態(tài),采用激勵的形式讓節(jié)點從忙碌狀態(tài)轉(zhuǎn)換到空閑狀態(tài),參與到服務(wù)內(nèi)容部署過程中。通過解決UKM(uncapacitated k-median)和UFL(uncapacitated facility location)問題確定網(wǎng)絡(luò)中服務(wù)內(nèi)容的位置和數(shù)量。2.具有局部信息時,利用局部網(wǎng)絡(luò)拓撲結(jié)構(gòu)和節(jié)點需求,在r-跳內(nèi)迭代優(yōu)化解決本地UKM和UFL問題,確定了服務(wù)內(nèi)容數(shù)量及位置。本文根據(jù)節(jié)點具有自私性的特性,考慮到節(jié)點轉(zhuǎn)發(fā)服務(wù)內(nèi)容數(shù)據(jù)時產(chǎn)生的能耗問題,利用節(jié)點能耗給出了一種報酬函數(shù),鼓勵了節(jié)點間的合作,有效減少遷移部署服務(wù)內(nèi)容過程中的部署成本。最后,對具有全局信息的情形在E-R隨機圖網(wǎng)絡(luò)模型進行了實驗驗證;對具有局部信息的情形分別在E-R隨機圖和B-A無標度網(wǎng)絡(luò)兩種網(wǎng)絡(luò)模型中進行實驗驗證。實驗結(jié)果表明,無論是哪種情形,本文提出的激勵方法都可以降低部署服務(wù)內(nèi)容成本,促進節(jié)點之間的相互合作,提升網(wǎng)絡(luò)的性能。
[Abstract]:Mobile ad hoc networks (Manets) forward packets through cooperation between nodes to improve network performance and quality of service (QoS). However, due to the limitation of wireless signal coverage, storage capacity and node mobility, there is uncertainty when the node acquires data. Deploying service content on the appropriate node improves the node's success rate in obtaining data. But in reality, nodes are rational, and it is difficult for nodes to cooperate with other nodes to deploy service content without benefit. In order to improve network performance, we need to design an effective incentive method to encourage selfish nodes to deploy service content. How to design incentive methods, reduce the cost of deploying service content and improve the efficiency of data acquisition by users is the focus of this paper. This paper mainly studies the incentive method of service content deployment from the following two aspects: 1. 1. In order to encourage nodes to participate in the deployment of service content actively, an incentive method based on virtual integration is designed for the selfish nodes with global information. In this method, nodes are divided into busy and idle states, and the nodes are converted from busy state to idle state in the form of excitation, and participate in the process of service content deployment. Determine the location and number of service content in the network by solving the UKM (uncapacitated k-median) and UFL (uncapacitated facility location) problems. When local information is available, the local UKM and UFL problems are optimized by using the local network topology and node requirements, and the number and location of service content are determined. In this paper, according to the selfish nature of nodes, considering the problem of energy consumption when nodes transmit service content data, a reward function is given by using node energy consumption, which encourages the cooperation between nodes. Effectively reduce deployment costs during migration of service content. Finally, the global information is verified in the E-R random graph network model, and the local information is verified in the E-R random graph and B-A scale-free network model. The experimental results show that in either case, the incentive method proposed in this paper can reduce the cost of deploying service content, promote the cooperation between nodes and improve the performance of the network.
【學位授予單位】：安徽工業(yè)大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TN929.5

【參考文獻】

相關(guān)期刊論文 前8條

1 吳W，

本文編號：2225244