發(fā)布時間：2018-09-17 17:49

【摘要】：物聯(lián)網(wǎng)是人們以互聯(lián)網(wǎng)的原型為基礎所提出的一種新型的網(wǎng)絡概念，它能夠實現(xiàn)系統(tǒng)的用戶一端可以在任何物品之間進行關聯(lián)通信和信息數(shù)據(jù)的最終交換。近年來，隨著物聯(lián)網(wǎng)技術的不斷進步，物聯(lián)網(wǎng)覆蓋范圍變得非常廣泛，并且構成物聯(lián)網(wǎng)不同種類的網(wǎng)絡數(shù)量也在不斷增加。物聯(lián)網(wǎng)網(wǎng)關作為數(shù)據(jù)傳輸與轉換的橋梁，其種類和數(shù)量也開始由傳統(tǒng)網(wǎng)絡中的單種類網(wǎng)關逐步向接入不同子網(wǎng)的多種類物聯(lián)網(wǎng)網(wǎng)關的方向發(fā)展。網(wǎng)關的性能直接影響了物聯(lián)網(wǎng)的性能，因此，對物聯(lián)網(wǎng)網(wǎng)關的選擇優(yōu)化具有理論和實用價值。 本文采用了一種新的方式對物聯(lián)網(wǎng)網(wǎng)關的選擇算法進行優(yōu)化，即將物聯(lián)網(wǎng)中接入不同子網(wǎng)的多類網(wǎng)關作為一個整體去優(yōu)化，這種優(yōu)化方式比現(xiàn)有大部分文獻中單純優(yōu)化接入單種類型子網(wǎng)的某種互聯(lián)網(wǎng)網(wǎng)關的局部優(yōu)化方式會取得更好的優(yōu)化效果。因為單純優(yōu)化負責接入某種子網(wǎng)的某個互聯(lián)網(wǎng)網(wǎng)關只能解決局部網(wǎng)關中的問題，而不能從全局角度分析解決問題，所以本文將從全局整體角度去設計與實現(xiàn)物聯(lián)網(wǎng)網(wǎng)關的選擇優(yōu)化。 首先，針對物聯(lián)網(wǎng)中網(wǎng)絡結構和應用需求的動態(tài)變化特性，提出了一種從物聯(lián)網(wǎng)感知層中優(yōu)化選擇節(jié)點，使其成為接入物聯(lián)網(wǎng)網(wǎng)關的算法。這種算法用改進的粒子群搜索選擇算法，在計算過程中將物聯(lián)網(wǎng)網(wǎng)絡帶寬、節(jié)點剩余能量、節(jié)點空間分布等因素作為考慮對象，從而保證了選擇出的網(wǎng)關能夠滿足物聯(lián)網(wǎng)的動態(tài)特性，并改善了物聯(lián)網(wǎng)的地理空間分布不均衡特性。 其次，針對物聯(lián)網(wǎng)網(wǎng)關中出現(xiàn)的網(wǎng)絡擁塞問題，提出了一種多網(wǎng)關負載均衡策略。當選擇多個網(wǎng)關為源節(jié)點提供接入物聯(lián)網(wǎng)服務時，根據(jù)網(wǎng)絡不同鏈路和不同網(wǎng)關的實際情況將負載合理地分配到不同網(wǎng)關中，從而實現(xiàn)不同網(wǎng)關之間的負載均衡，避免網(wǎng)關擁塞現(xiàn)象的出現(xiàn)。 最后，通過NS2網(wǎng)絡模擬軟件對算法進行仿真，，并且從物聯(lián)網(wǎng)網(wǎng)關傳輸數(shù)據(jù)的實時性、可靠性、通信成本以及能耗四個方面對本論文所設計的算法進行詳細的評價。實驗結果表明上述4個指標均比現(xiàn)有方法有不同程度的改善和提高。
[Abstract]:The Internet of things (IoT) is a new network concept based on the prototype of the Internet. It can realize the final communication and the final exchange of information and data between the users of the system. In recent years, with the continuous progress of the Internet of things technology, the Internet of things coverage has become very wide, and the number of different types of Internet of things network is also increasing. As a bridge of data transmission and conversion, the type and number of Internet of things gateways begin to develop from single type gateways in traditional networks to multi-type Internet of things gateways connected to different subnets. The performance of the gateway directly affects the performance of the Internet of things, so it is of theoretical and practical value to select and optimize the gateway of the Internet of things. In this paper, we adopt a new way to optimize the selection algorithm of Internet of things gateway, that is to say, many kinds of gateway connected to different subnets in the Internet of things are optimized as a whole. This optimization method is more effective than the local optimization method of a single type of Internet gateway in most of the existing literatures, which simply optimizes the access to a single type of subnet. Because simply optimizing an Internet gateway that is responsible for accessing a certain subnet can only solve the problems in the local gateway, but not analyze and solve the problem from a global perspective. Therefore, this paper will design and implement the selection and optimization of Internet of things gateway from the global perspective. Firstly, in view of the dynamic characteristics of the network structure and application requirements in the Internet of things, an algorithm is proposed to optimize the selection of nodes from the perception layer of the Internet of things and make it an access gateway to the Internet of things. In this algorithm, the improved particle swarm optimization algorithm is used to calculate the network bandwidth, the residual energy of nodes, the spatial distribution of nodes, and so on. This ensures that the selected gateway can satisfy the dynamic characteristics of the Internet of things and improve the geographical spatial distribution imbalance of the Internet of things. Secondly, a multi-gateway load balancing strategy is proposed to solve the problem of network congestion in Internet of things gateway. When multiple gateways are selected to provide access to the Internet of things for the source nodes, the load is reasonably distributed among different gateways according to the actual conditions of different links and gateways of the network, so as to achieve load balancing among different gateways. Avoid the appearance of gateway congestion. Finally, the algorithm is simulated by NS2 network simulation software, and the algorithm is evaluated in detail from four aspects: real-time, reliability, communication cost and energy consumption. The experimental results show that the above four indexes are improved and improved to some extent compared with the existing methods.
【學位授予單位】：遼寧大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TP391.44;TN929.5

【參考文獻】

相關期刊論文 前3條

1 郭登峰;徐正山;李坤;;物聯(lián)網(wǎng)技術撐起智能電網(wǎng)[J];華北電業(yè);2010年02期

2 沈蘇彬;毛燕琴;范曲立;宗平;黃維;;物聯(lián)網(wǎng)概念模型與體系結構[J];南京郵電大學學報(自然科學版);2010年04期

3 邵威;李莉;;感知中國——我國物聯(lián)網(wǎng)發(fā)展路徑研究[J];中國科技信息;2009年24期

本文編號：2246665