版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進(jìn)行舉報(bào)或認(rèn)領(lǐng)
文檔簡介
1、<p> 外文出處:Prasan Kumar Sahoo著, </p><p> Proc of Computing, and Communications Conference, 2005.[C] </p><p> 出版社: IEEE,2005年 </p><p> 附件:1.外文資料翻譯譯文;2.外文原文 </p
2、><p> 基于拓?fù)浣Y(jié)構(gòu)的分布式無線傳感器網(wǎng)絡(luò)的功率控制</p><p><b> 摘要</b></p><p> 無線傳感器網(wǎng)絡(luò)由大量的傳感器節(jié)點(diǎn)電池供電,限制在一定區(qū)域內(nèi)的隨機(jī)部署的幾個(gè)應(yīng)用。由于傳感器能量資源的有限,他們中的每一個(gè)都應(yīng)該減少能源消耗,延長網(wǎng)絡(luò)的生命周期。在這篇文章中,一種分布式算法的基礎(chǔ)上,提出了無線傳感器網(wǎng)絡(luò)的構(gòu)建一種
3、高效率能源樹結(jié)構(gòu),而無需定位信息的節(jié)點(diǎn)。節(jié)點(diǎn)的能量守恒是由傳輸功率控制完成的。除此之外,維護(hù)的網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)由于能源短缺的節(jié)點(diǎn)也提出了協(xié)議。仿真結(jié)果表明,我們的分布式協(xié)議可以達(dá)到類似集中算法的理想水平的能量守恒,可以延長網(wǎng)絡(luò)的生命周期比其他沒有任何功率控制的分布式算法。</p><p> 關(guān)鍵詞:無線網(wǎng)絡(luò)傳感器,分布式算法,功率控制,拓?fù)浣Y(jié)構(gòu)</p><p><b> 1.引言
4、</b></p><p> 近年來在硬件和軟件的無線網(wǎng)絡(luò)技術(shù)的發(fā)展,使小尺寸、低功耗、低成本、多功能傳感器節(jié)點(diǎn)[1]的基礎(chǔ)上,由傳感、數(shù)據(jù)處理及無線通信組件組成。這些低能量節(jié)點(diǎn)的電池,部署在數(shù)百到成千上萬的無線傳感器網(wǎng)絡(luò)。在無線傳感器網(wǎng)絡(luò)系統(tǒng)、音視頻信號處理系統(tǒng),使用更高的發(fā)射功率和轉(zhuǎn)發(fā)數(shù)據(jù)包相似的路徑是種主要消費(fèi)傳感器的能量。除此之外,補(bǔ)充能量的電池更換和充電幾百節(jié)點(diǎn)上的傳感器網(wǎng)絡(luò)應(yīng)用的大部分地區(qū)
5、,特別是在嚴(yán)酷的環(huán)境是非常困難的,有時(shí)不可行。因此,節(jié)能[2],[3],[4]的傳感器節(jié)點(diǎn)是一個(gè)關(guān)鍵問題,如傳感器網(wǎng)絡(luò)的生命周期的完全取決于耐久性的電池。</p><p> 傳感器節(jié)點(diǎn)一般都是自組織建立了無線傳感器網(wǎng)絡(luò),監(jiān)察活動的目標(biāo)和報(bào)告的事件或信息多跳中的基站。有四種主要的報(bào)告模式的傳感器網(wǎng)絡(luò):事件驅(qū)動、隊(duì)列驅(qū)動、期刊、查詢和混合的報(bào)告。在事件驅(qū)動模型, 節(jié)點(diǎn)報(bào)告接收器,同時(shí)報(bào)告遙感一些事件,例如火災(zāi)或水災(zāi)
6、而敲響了警鐘。定期報(bào)告中,節(jié)點(diǎn)模型的數(shù)據(jù)收集和可聚合所需資料,成為集,然后定期的發(fā)送到上游。資料相結(jié)合的方法,稱為數(shù)據(jù)融合[5],[6],[7]和[8],從而降低了數(shù)量的傳輸數(shù)據(jù)。這樣的例子,也可應(yīng)用在這里,如報(bào)告的溫濕度的地方。所以, 集合到一個(gè)單一的類似數(shù)據(jù)包的數(shù)據(jù)融合的遙感數(shù)據(jù)的傳送到接收器的多級跳環(huán)境中,從而保存能源也是在傳感器網(wǎng)絡(luò)中的重要研究問題。</p><p> 在[9]的基礎(chǔ)上, 對傳感器節(jié)點(diǎn)的
7、每個(gè)單元的電源消耗比較進(jìn)行分析,它觀察耗能接收的電源和空閑狀態(tài)幾乎相同,CPU 的功耗是很低。在文獻(xiàn)[10]的基礎(chǔ)上, 在作者建議中的理想的發(fā)射功率評估通過節(jié)點(diǎn)互動與信號衰減節(jié)點(diǎn)的無線傳感器網(wǎng)絡(luò) MAC 協(xié)議的傳輸功率控制。計(jì)算理想的發(fā)射功率的反復(fù)改進(jìn)和存儲當(dāng)前的理想發(fā)射功率,為每個(gè)相鄰的節(jié)點(diǎn)。在[11], 作者介紹了拓?fù)淇刂频臒o線傳感器網(wǎng)絡(luò),于一體的有效子網(wǎng)和短躍點(diǎn)方法來達(dá)到節(jié)能降耗的兩級策略。分析是在非對稱無線鏈接并不罕見,具有不同
8、的最大傳輸范圍,在異構(gòu)無線設(shè)備的網(wǎng)絡(luò)拓?fù)淇刂频膯栴}。詳細(xì)分析了在[12]。因?yàn)楣?jié)點(diǎn)是異構(gòu)的,他們有不同的最大傳輸功率和廣播范圍,需要可調(diào)整的功率控制的分布式天線。作者在[13]中的采取一套主動節(jié)點(diǎn)和節(jié)點(diǎn)的傳播范圍,建議盡量減少總功率消耗的無線傳感器網(wǎng)絡(luò)的最小電源配置方法。</p><p> 在[14],作者提出了一個(gè)分析路由協(xié)議的范圍的可變傳動方案。從他們的分析研究表明,該算法可以提高可變傳動范圍全面的網(wǎng)絡(luò)性能
9、。以LEACH [15]為基礎(chǔ)的算法,這些算法是讓一些節(jié)點(diǎn)使用較高的發(fā)射功率幫助鄰居傳輸數(shù)據(jù)到了 BS。然而, LEACH需要全球的傳感器網(wǎng)絡(luò)的知識,并且假定每個(gè)節(jié)點(diǎn)接近BS。在[16]中,兩個(gè)局部拓?fù)浣Y(jié)構(gòu)的控制算法,并提出了異構(gòu)多跳無線網(wǎng)絡(luò)的非均勻傳輸范圍。雖然這個(gè)協(xié)議保護(hù)網(wǎng)絡(luò)的連接和談?wù)撊绾慰刂频耐負(fù)浣Y(jié)構(gòu),它不談網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)和能耗的密度較大的問題,如無線傳感器網(wǎng)絡(luò)節(jié)點(diǎn)。[17]是跨省電種技術(shù),特設(shè)的無線網(wǎng)絡(luò)無顯著降低能耗的能力或連接
10、的網(wǎng)絡(luò)。這是一個(gè)分布式的隨機(jī)算法,為了節(jié)省功率最大對電池進(jìn)行關(guān)閉。但是它使用固定傳輸功率范圍,該算法適用于低密度等IEEE 802.11無線通信網(wǎng)絡(luò)的節(jié)點(diǎn)。</p><p> 在[18],提出了構(gòu)建集中算法進(jìn)行了靜態(tài)的無線網(wǎng)絡(luò)的拓?fù)浣Y(jié)構(gòu)。根據(jù)這一算法的基礎(chǔ)上,初步每個(gè)節(jié)點(diǎn)有它自己的組成部分。然后,它通過合并交互連通到一個(gè)整體上。畢竟,部件連接環(huán)和優(yōu)化的后處理解除功耗的網(wǎng)絡(luò)。雖然該算法[18]是專為無線網(wǎng)絡(luò)拓?fù)浣Y(jié)
11、構(gòu)的優(yōu)化,它是一個(gè)集中并不能改變發(fā)射功率動態(tài)。分布式算法在無線傳感器網(wǎng)絡(luò)的傳輸功率控制提出了[19]。他們指派一個(gè)任意選擇的傳輸功率級傳感器節(jié)點(diǎn)在可能分裂的網(wǎng)絡(luò)中。同樣,他們提出了全球性的解決方案與不同的傳輸功率算法, 拓?fù)浣Y(jié)構(gòu)創(chuàng)造了一個(gè)連接的網(wǎng)絡(luò)和設(shè)定不同的傳輸范圍為所有的節(jié)點(diǎn)。所以,他們的工作能耗的節(jié)點(diǎn)可能更多,因?yàn)樵跓o線傳感器網(wǎng)絡(luò)中的節(jié)點(diǎn)是相鄰的。</p><p> 在無線傳感器網(wǎng)絡(luò)中、通信是能量消耗的主
12、要因素[20]。然而,傳輸功率調(diào)節(jié)控制網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)可以延長壽命及提高無線傳感器網(wǎng)絡(luò)的能力。另外,而非控制發(fā)射功率水平,總是使用一個(gè)固定的高功率水平網(wǎng)絡(luò)的節(jié)點(diǎn)的節(jié)點(diǎn)將迅速減少死亡網(wǎng)絡(luò)的生存時(shí)間。收集數(shù)據(jù),感覺到最重要的信息可能包含一些要求,提供一種連接網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)是非常必要的無線傳感器網(wǎng)絡(luò)。因此,在我們的工作中,我們提出如何控制發(fā)射功率水平的每個(gè)節(jié)點(diǎn)的網(wǎng)絡(luò)來節(jié)約能源。我們提出一個(gè)分布式算法,調(diào)整傳輸功率級別的節(jié)點(diǎn)動態(tài)和構(gòu)建一棵樹和一個(gè)中間
13、功率電平拓?fù)浣Y(jié)構(gòu)之間的最大和最小,在不同的群體,達(dá)到一種連接網(wǎng)絡(luò)的節(jié)點(diǎn)。本算法在一種無線傳感器網(wǎng)絡(luò)中沒有位置信息,建立連接節(jié)點(diǎn)分布式的拓?fù)浣Y(jié)構(gòu)。</p><p> 接下來的文章是有組織的如下。本協(xié)議的系統(tǒng)模型,提出了第二節(jié)。我們的分布式控制協(xié)議被描述在第3節(jié)。性能分析和仿真結(jié)果在第4節(jié)而結(jié)論在第5節(jié)。</p><p><b> 2.系統(tǒng)模型</b></p&g
14、t;<p> 讓我們考慮一種單一的多跳的無線傳感器網(wǎng)絡(luò),傳感器和部署在某些隨機(jī)地理區(qū)域,這樣小的連通性存在不同組的節(jié)點(diǎn),如圖1。它認(rèn)為水槽內(nèi)通信范圍至少一個(gè)節(jié)點(diǎn)的網(wǎng)絡(luò)。在這個(gè)網(wǎng)絡(luò)的連通性孔由于不可抗拒的自然物質(zhì)間隙小另一批節(jié)點(diǎn)之間的差距部署或由于同一地區(qū)的節(jié)點(diǎn),因?yàn)樗麄儾荒芘c最小傳輸功率(Pmin)連接。然而,所有的節(jié)點(diǎn)要么來自同一或不同的組織使用固定傳輸功率電平交流和形成一個(gè)連接的網(wǎng)絡(luò)沒有任何權(quán)力的控制。這個(gè)固定傳輸功
15、率電平可以被假定為最高(Pmax)或最小值和最大值的權(quán)力之間的水平。根據(jù)我們的實(shí)驗(yàn)用云母塵(21)和RF頻率866,表1兆赫茲,0被視為最低(Pmin)和3為最高(Pmax)傳輸功率電平之間的交流,我們認(rèn)為這種價(jià)值節(jié)點(diǎn)在我們的文中中。之前的下一章節(jié)中,我們定義了一些技術(shù)術(shù)語用于我們的協(xié)議。</p><p> 圖1.傳感器節(jié)點(diǎn)部署和連通性隨機(jī)孔之間的不同的組節(jié)點(diǎn)</p><p><b
16、> 表1</b></p><p> 對于不同的電力能源消費(fèi)水平和相應(yīng)的交流,得到了來自我們距離的實(shí)驗(yàn)結(jié)果</p><p><b> 2.1.定義</b></p><p> ?上游和下游的組: 讓 {G1,G2,G3,...} 作為套組節(jié)點(diǎn)分布在某一地區(qū)。 如果兩個(gè)組Gi和 Gj,i ≠ j,以致控制數(shù)據(jù)包轉(zhuǎn)發(fā)的Gi的任
17、何節(jié)點(diǎn)到 Gj,然后Gi被稱作為上游組于 Gj , Gj 是下游組對應(yīng)Gi。</p><p> 例如在圖 1中的組 G1 包含接收器節(jié)點(diǎn)被認(rèn)為是上游組組 G2、 G3、 G4,相對于控件作為數(shù)據(jù)包是最初廣播從組包含到其他組網(wǎng)絡(luò)的接收器。 相對于 G1 下游組 G2、 G3、G4。同樣,如果控制數(shù)據(jù)包 G2 從廣播到這些組,在這種情況下 G3、 G4 被視為下游組的 G2,G2 可以組 G3、 G4,一個(gè)上游組。
18、</p><p> ? 本地躍點(diǎn)計(jì)數(shù) (LHC): 這是表示的控制數(shù)據(jù)包遍歷本地一的組內(nèi)時(shí)它發(fā)送到另一個(gè)節(jié)點(diǎn)的躍點(diǎn)數(shù)的計(jì)數(shù)器。</p><p> LHC 的控制數(shù)據(jù)包的值初始化為 0 和的數(shù)據(jù)包在同一組內(nèi)的每個(gè)后續(xù)跳躍 1 遞增。一般來說 LHC = LHC + 1。如果節(jié)點(diǎn) A 將數(shù)據(jù)包轉(zhuǎn)發(fā)到 B,并且 B 然后將同一個(gè)數(shù)據(jù)包轉(zhuǎn)發(fā)到價(jià)值的 LHC 在控制數(shù)據(jù)包中的 A = 0,B =
19、 1 和 C = 2。</p><p> ? 組躍點(diǎn)計(jì)數(shù) (GHC): 這是表示的控制數(shù)據(jù)包傳遞到其他傳輸從一個(gè)組時(shí)的躍點(diǎn)數(shù)的計(jì)數(shù)器。 GHC的值是唯一的特定組中的所有節(jié)點(diǎn),它會增加 1,如果包傳遞給另一組。 GHC的值初始化為 0,在一般的GHC = GHC + 1,后續(xù)跳躍的數(shù)據(jù)包從一個(gè)組到其他。</p><p> 數(shù)學(xué)上,讓 G = {g1,g2 … …,gn},作為組中的 n
20、傳感器節(jié)點(diǎn)集,m 傳感器節(jié)點(diǎn)的組處于相同或不同的 m 和 n 的值的另一個(gè)組。 如果GHC值 = p,gi G, 然后 ,GHC值 = q,其中 p ≠ q,作為 G,不同組的節(jié)點(diǎn)。 在我們的協(xié)議,因?yàn)榻邮掌鞴?jié)點(diǎn)啟動建設(shè)階段,即在 G1,GHC的 G1 在圖 2 中,所有節(jié)點(diǎn)的值是 0 和如果數(shù)據(jù)包從 G1 轉(zhuǎn)發(fā)到像 G2 或 G4 的任何其他組,GHC數(shù)據(jù)包中的值是按 1 遞增。 因此,GHC的值 = 1,為 G
21、2 或 G4。</p><p> ? 父網(wǎng)關(guān) ID (PGID): 使連接與一個(gè)上游組的節(jié)點(diǎn)的組中的所有節(jié)點(diǎn)的節(jié)點(diǎn)稱為父網(wǎng)關(guān)和其 ID 稱為為 PGID。 節(jié)點(diǎn)的每個(gè)組中存在只有一個(gè)父網(wǎng)關(guān)。</p><p> 圖 2. 父和子網(wǎng)關(guān)的不同組的節(jié)點(diǎn)</p><p> ? 子網(wǎng)關(guān): 連接到父網(wǎng)關(guān)下游組的節(jié)點(diǎn)稱為子網(wǎng)關(guān)。組中存在至少一個(gè)子網(wǎng)關(guān)。在某些的情況下如果一組都
22、包含唯一的節(jié)點(diǎn)的單個(gè)節(jié)點(diǎn)被視為為該組的父和子網(wǎng)關(guān)。</p><p> 圖 2,組 G1 的A,B節(jié)點(diǎn)分別為D,C節(jié)點(diǎn)的子網(wǎng)關(guān)。</p><p> ? 節(jié)點(diǎn)能級 (NEL):當(dāng)前節(jié)點(diǎn)的能量級別稱為 NEL。 例如在廣播一控制數(shù)據(jù)包,如果一個(gè)節(jié)點(diǎn)的能量級別是X,NEL在控制數(shù)據(jù)包中作為X的單位。</p><p> 父網(wǎng)關(guān)功率級(?PGPL):任何組的父網(wǎng)關(guān)的發(fā)射功
23、率水平,它可以與子網(wǎng)關(guān)的上游組織連接被稱為父網(wǎng)關(guān)功率級(PGPL)。自從接收器始終是父網(wǎng)關(guān)的組里,它PGPL分配為0。然而,對于其它組中的父網(wǎng)關(guān), Pmin < PGPL Pmax,這可能值在1和3之間,按我們的假設(shè)。</p><p> ? 源 ID (SID): 如果 A 和 B 是相同或者不相同組中的兩個(gè)不同傳感器節(jié)點(diǎn),將數(shù)據(jù)包A發(fā)送到 B,A節(jié)點(diǎn)的ID是源節(jié)
24、點(diǎn),A也是B的源節(jié)點(diǎn)。</p><p> 3.分布式的功率控制協(xié)議</p><p> 在本節(jié)中,我們將提出我們基于拓?fù)浣Y(jié)構(gòu)協(xié)議的功率控制,這是一種動態(tài)的拓?fù)浣Y(jié)構(gòu)。 我們假設(shè)在網(wǎng)絡(luò)中的每個(gè)節(jié)點(diǎn)具有一個(gè)唯一的 ID,他們每個(gè)人都知道在拓?fù)浣Y(jié)構(gòu)之前鄰居的 ID。根據(jù)我們協(xié)議的每個(gè)系統(tǒng)模型,由于每個(gè)組的節(jié)點(diǎn)之間存在的連接孔,我們假設(shè)網(wǎng)絡(luò)可能會斷開連接,如果他們使用低傳輸功率級與另一個(gè)節(jié)點(diǎn)的一組
25、之間,并且可能會消耗更多的精力,如果他們使用最大傳輸功率級進(jìn)行通信。此外,在我們假設(shè)傳輸電源網(wǎng)絡(luò)中的所有節(jié)點(diǎn)級別后部署可能是最大或最小值和最大值之間。因此,我們的協(xié)議,在樹拓?fù)錁?gòu)造節(jié)點(diǎn)使用最小傳動功率級的每個(gè)組之間 (Pmin = 0),整個(gè)網(wǎng)絡(luò)的樹拓?fù)溥B接在不同組節(jié)點(diǎn)中形成并使用有效功率級別 (PTx),這里(Pmin=0)<PTx(Pmax=3),這個(gè)分布式協(xié)議的不同階段在下面將描述。</p><p>
26、<b> 3.1.施工階段</b></p><p> 一旦所有的節(jié)點(diǎn)部署在網(wǎng)絡(luò)中, 就通過廣播最小發(fā)射功率的構(gòu)造數(shù)據(jù)包啟動施工階段 (Pmin = 0) 以與鄰居直接連接如圖 4 (a) 所示。圖 3 所示的構(gòu)造數(shù)據(jù)包格式和數(shù)據(jù)包的參數(shù)初始化為:SID = Sink’s ID, PGID = Sink’s ID, NEL 為接收的功率級,
27、 LHC = 0,GHC = 0, PGPL = 0. 接收器節(jié)點(diǎn)通常接收數(shù)據(jù),則其 PGPL 分配給 0,這是不同的網(wǎng)絡(luò)的其他父網(wǎng)關(guān)。在接到構(gòu)造包,在其最小的傳輸接收器的鄰居電源范圍 (Pmin = 0),掃描包的所有參數(shù)。 他們等待隨機(jī)時(shí)間Wi,并與接收器連接。讓Ni作為第i節(jié)點(diǎn)的鄰居代號在網(wǎng)絡(luò)中的N個(gè)節(jié)點(diǎn),自收到構(gòu)建包中, 第i個(gè)節(jié)點(diǎn)的等待時(shí)間可以被看作是:</p&
28、gt;<p><b> (1)</b></p><p> 在αi是一個(gè)小的隨機(jī)數(shù)兼容CSMA-CA機(jī)制(22)。然后,他們每個(gè)人都重播了構(gòu)建包使用相同的最小功率電平Pmin = 0到他們的鄰居提供必要的參數(shù)對應(yīng)的數(shù)據(jù)包,等待時(shí)間的Ti單位在(2)中得到:</p><p><b> (2)</b></p><
29、p> 第i個(gè)節(jié)點(diǎn)的當(dāng)前的能量級為Ei,βi是一個(gè)非常小的隨機(jī)數(shù),比如0.00001 βi 0.0001.</p><p> 圖 3. 構(gòu)造數(shù)據(jù)包的格式</p><p> 為了避免密集網(wǎng)絡(luò)中的節(jié)點(diǎn)之間的數(shù)據(jù)包碰撞,我們建議接收器還等待Ti單位后廣播構(gòu)造包,然后進(jìn)入信息階段 3.2 節(jié)中所述。 它是應(yīng)注意該接收器,必須至少一個(gè)傳感器節(jié)點(diǎn)的
30、最小值或最大傳輸功率范圍內(nèi)。 但是,如果在接收器沒有找到任何Pmin = 0的鄰居,它與它的鄰居鏈接進(jìn)入信息階段,在經(jīng)過Ti輪候時(shí)間(表 2 和表 3)。</p><p><b> 表 2</b></p><p> 接收器與網(wǎng)絡(luò)的任何節(jié)點(diǎn)施工階段算法</p><p> 算法 1: 施工階段</p><p><
31、b> 表 3</b></p><p> 發(fā)件人和接收信息階段算法</p><p> 算法 2: 信息階段</p><p> 在接到構(gòu)造包,節(jié)點(diǎn)掃描中它的所有參數(shù),并考慮作為其源有最少 LHC 節(jié)點(diǎn)。 接收機(jī)節(jié)點(diǎn)等待無線設(shè)備、連接的源和上文所述,然后按照相同的步驟。 直到一個(gè)節(jié)點(diǎn)不會收到任何構(gòu)造包進(jìn)一步與第一個(gè)樹拓?fù)錁?gòu)造一個(gè)的組的節(jié)點(diǎn)之間如根
32、和其他節(jié)點(diǎn)為它最小傳動功率級中其他節(jié)點(diǎn)的接收器與圖 4 (b) 所示,此過程會繼續(xù)進(jìn)行。我們假定有不同組的節(jié)點(diǎn)或一些節(jié)點(diǎn)之間的連接孔是無法構(gòu)建使用 Pmin 的鏈接,在施工階段以有限的時(shí)間間隔后終止。執(zhí)行信息階段后形成樹拓?fù)湎乱唤M。它是構(gòu)造包始終使用最小傳動功率級傳輸,每次 LHC 都加 1,它從一個(gè)節(jié)點(diǎn)跳到另一個(gè)時(shí)應(yīng)注意。在一個(gè)組中一些節(jié)點(diǎn)也可能是其他的鄰居節(jié)點(diǎn)已收到相同的構(gòu)造數(shù)據(jù)包。然后一個(gè)節(jié)點(diǎn)決定它自己的源節(jié)點(diǎn)的如何? 3.3.1
33、 (A) 段所述,我們曾討論在維護(hù)階段的這部分的問題。</p><p> 圖 4. (a) 隨機(jī)分布在面積的傳感器節(jié)點(diǎn) (b) 第一個(gè)樹拓?fù)浣Y(jié)構(gòu)</p><p><b> 3.2.信息階段</b></p><p> 這一階段的目的是在整個(gè)網(wǎng)絡(luò)的不同組節(jié)點(diǎn)中使用最有效電源級構(gòu)建分布式的樹拓?fù)洹?它通過廣播通知數(shù)據(jù)包使用最大傳輸功率級
34、 (Pmax = 3)。 通知數(shù)據(jù)包的格式如圖 5 所示。它是應(yīng)注意每個(gè)組的節(jié)點(diǎn)具有唯一的父網(wǎng)關(guān)。 為例,接收器是其組中的唯一父網(wǎng)關(guān)。 所以,之前于廣播通知數(shù)據(jù)包,一個(gè)節(jié)點(diǎn)將 PGID 的值通知包復(fù)制從該構(gòu)造,可以區(qū)別另一個(gè)構(gòu)造包的數(shù)據(jù)包。此外,GHC構(gòu)造數(shù)據(jù)包中的值是加 1,然后添加到通知數(shù)據(jù)包的相應(yīng)字段。 代以通知數(shù)據(jù)包中的所需值,廣播使用 Pmax = 3。在接到該數(shù)據(jù)包,一個(gè)節(jié)點(diǎn)知道從其標(biāo)題信息,它是通知數(shù)據(jù)包,并可兼容CSMA
35、-CA 機(jī)制 [22] 的隨機(jī)時(shí)間的等待。此外,從每個(gè)發(fā)送方使用下面的公式估計(jì)其物理距離。</p><p><b> (3)</b></p><p> Pr=ξ×(d-γ)×Pt,</p><p> 其中Pt是一個(gè)節(jié)點(diǎn)使用廣播通知數(shù)據(jù)包的發(fā)射功率。在我們的協(xié)議里,每個(gè)節(jié)點(diǎn)使用對應(yīng)于 Pmax = 3的發(fā)射功率(Pt)廣
36、播一個(gè)信息數(shù)據(jù)包,在表1中,Pr是在接收信息包時(shí)節(jié)點(diǎn)的接收功率,接收功率用變量d?γ,γ是路徑損耗 (衰減),它滿足2 γ 4的條件,這里,比例常數(shù) ξ 被假定為符號起見 1 和 γ 的值通常是為 2 的可用空間。 它是應(yīng)注意的發(fā)送方節(jié)點(diǎn)集被視為為按第 2 節(jié)中給出的定義在上游的分組。在接到通知包,可以用式 (3) 估計(jì) d 發(fā)送方和接收方之間的物理距離。 有效的發(fā)射功率 (PTx),依據(jù)它可
37、以上游組中的發(fā)送方可以通信可以有如下估計(jì):</p><p><b> (4)</b></p><p> 將發(fā)件人廣播告知數(shù)據(jù)包的一組。為網(wǎng)絡(luò)中的全數(shù)字節(jié)點(diǎn)考慮 N</p><p><b> (5)</b></p><p> 接收通知數(shù)據(jù)包的節(jié)點(diǎn)集,獲取通知數(shù)據(jù)包后并使用式 (3),讓{dij
38、}作為發(fā)送方Si和接收方Rj的估計(jì)距離,i = 1, 2, …, m; and j = 1, 2, … n。</p><p><b> (6)</b></p><p> 應(yīng)注意一個(gè)組的節(jié)點(diǎn)可能是另一個(gè)節(jié)點(diǎn)的已收到幾個(gè)通知數(shù)據(jù)包。 因此,每個(gè)節(jié)點(diǎn)使用式 (6) 來找到
39、最短的距離,即 {Dij} = min({dij}),具有本身 (接收) 組的所有節(jié)點(diǎn) (發(fā)送) 之間。 計(jì)算的 {Dij} 值后, 一個(gè)節(jié)點(diǎn)再次使用稱為PTx(ij) 的方程 (3),來估計(jì)有效發(fā)射功率之間最接近的發(fā)送(i) 及 自身(j)。 從我們的 4 節(jié)圖 12 中給出的仿真結(jié)果證明,我們發(fā)現(xiàn)使用最大傳輸功率級的可能性是非常小的高密度網(wǎng)絡(luò)。 因此,值得在這里提到的節(jié)點(diǎn)數(shù)數(shù)可能使用最大動力 (Pmax = 3) 作為有效的傳輸功率
40、與上游組節(jié)點(diǎn)通信。 因此,在我們的協(xié)議里,有效功率級別PTx(ij) 可能是 1 或 2。然而,在最壞的情況PTx(ij) = 3 可用作可能的有效傳輸功率級。</p><p> 圖 5. 通知數(shù)據(jù)包的格式</p><p> 隨機(jī)的時(shí)限已經(jīng)屆滿后,已經(jīng)收到通知數(shù)據(jù)包節(jié)點(diǎn)廣播構(gòu)造包使用最小傳動功率級的第 3.1 施工階段中所述。 數(shù)據(jù)包的 PGPL 字段中給出的上游組可連接節(jié)點(diǎn)的有效功率
41、級別。 GHC的值從通知包復(fù)制到各自領(lǐng)域的構(gòu)造包。 節(jié)點(diǎn)將自己的 ID 添加到 PGID 字段聲明本身為父網(wǎng)關(guān)和其他參數(shù),如 SID,LHC、 GHC和 NEL 也根據(jù)定義添加到相應(yīng)構(gòu)造包的字段里。</p><p> Power control based topology construction for the distributed wireless sensor networks</p>
42、<p><b> Abstract</b></p><p> Wireless sensor network consists of large number of sensor nodes with limited battery power, which are randomly deployed over certain area for several applicat
43、ions. Due to limited energy resource of sensors, each of them should minimize the energy consumption to prolong the network lifetime. In this paper, a distributed algorithm for the multi-hop wireless sensor network is pr
44、oposed to construct a novel energy efficient tree topology, without having location information of the nodes. Energy cons</p><p> Keywords: Wireless sensor network; Distributed algorithm; Power control; Top
45、ology construction</p><p> 1. Introduction</p><p> Recent advances in hardware and software for the wireless network technologies have enabled the development of small sized, low-power, low-co
46、st and multi-functional sensor nodes [1], which consist of sensing, data processing and wireless communicating components. These nodes are operated with very low powered batteries and deployed hundreds to thousands in th
47、e wireless sensor network (WSN). In wireless sensor network, signal processing, communication activities using higher transmission power an</p><p> Sensor nodes are generally self organized to build the wir
48、eless sensor network, monitor the activities of the target and report the event or information to the sink or the base station (BS) in a multi-hop fashion. There are four main reporting models of the sensor network: even
49、t driven, query driven, periodical and mixed reporting. In event driven model, nodes report the sink, while sensing some events such as fire or flood alarm. In periodical reporting model, nodes collect the sensed data an
50、d </p><p> In [9], the power consumption comparison of each unit of sensor node is analyzed and it is observed that the energy consumption of the received power and idle state are almost same and the power
51、consumption of CPU is very low. In [10], the authors propose the transmission power control in MAC protocols for wireless sensor network to assess the ideal transmission power by the nodes through node interaction and si
52、gnal attenuation. The proposed algorithm calculates the ideal transmission power by r</p><p> In [14], authors have proposed an analysis of the routing protocol based on the variable transmission range sche
53、me. From their analysis, it is observed that the variable transmission range scheme can improve the overall network performance. The LEACH [15] based algorithm let some nodes to be the cluster leader and uses the higher
54、transmission power to help the neighbor transmitting data to the BS. However, LEACH needs the global knowledge of the sensor network and assumes each node in the radio </p><p> In [18], the authors present
55、a centralized greedy algorithm to construct an optimized topology for a static wireless network. According to this algorithm, initially each node has its own component. Then, it works interactively by merging the connect
56、ed components until there is just one. After all components are connected, a post-processing removes the loop and optimizes the power consumption of the network. Although this algorithm [18] is meant for an optimized top
57、ology of wireless network, it i</p><p> In WSN, communication is the main factor of the energy consumption [20]. However, transmission power adjustment to control the topology can extend the network lifetim
58、e and enhance the capability of the sensor network. Moreover, without controlling the transmission power level and always using a fixed higher power level for all nodes of the network will make the nodes die quickly and
59、minimize the network life time. Since, the collected sensed data may contain some important information as require</p><p> The rest of the paper is organized as follows. System model of our protocol is pres
60、ented in Section 2. Our distributed power control protocol is described in Section 3. Performance analysis and simulation results are presented in Section 4 and conclusion is drawn in Section 5 of the paper.</p>&
61、lt;p> 2. System model</p><p> Let us consider a multi-hop, homogeneous wireless sensor network, in which sensor nodes are randomly and densely deployed over certain geographical area such that small con
62、nectivity holes exist among different group of nodes, as shown in Fig. 1. It is also assumed that the sink is within communication range of at least one node of the network. The connectivity holes in the network may occu
63、r due to small physical gaps among different group of nodes at the time of deployment or due to gap among th</p><p> Fig. 1. Randomly deployed sensor nodes with connectivity holes among different group
64、 of nodes.</p><p><b> Table 1. </b></p><p> Energy consumption for different power levels and corresponding communication distances, obtained from our experimental result</p>
65、<p> 2.1. Definitions</p><p> ? Upstream and Downstream Groups: Let {G1, G2, G3, …} be the set of group of nodes distributed over certain area. If two groups Gi and Gj, for i ≠&
66、#160;j, such that a control packet is forwarded from any node of Gi to Gj, then Gi is known as the upstream group with respect to Gj and Gj is the downstream group with respect to Gi. </p><p> For example,
67、in Fig. 1, group G1 that contains the sink node is considered as the upstream group with respect to the groups G2, G3 and G4, as the control packet is initially broadcast from the group containing the sink to other group
68、s of the network. G2, G3 and G4 are the downstream groups with respect to G1. Similarly, G2 can be an upstream group for the groups G3 and G4, if control packets are broadcast from G2 to those groups and in that case, G3
69、 and G4 are treated as the downstream groups fo</p><p> ? Local Hop Counts (LHC): It is a counter, which represents the number of hops that a control packet traverses locally within a group, when it is forw
70、arded from one node to other. </p><p> The value of LHC of a control packet is initialized to 0 and incremented by 1 for each subsequent hopping of the packet within the same group. In general, LHC =
71、160;LHC + 1. Within a group, if node A forwards a packet to B, and then B forwards the same packet to C, value of LHC in the control packet of A = 0, B = 1 and C = 2. </p>&
72、lt;p> ? Group Hop Counts (GHC): It is a counter, which represents the number of hops that a control packet passes, when it is transmitted from one group to other. The value of GHC is unique for all nodes of a particu
73、lar group and it is incremented by 1, if the packet is transmitted from one group to other. Value of GHC is initialized to 0 and in general GHC = GHC + 1, for the subsequent hopping of the packet from
74、 one group to other. </p><p> Mathematically, let G = {g1, g2, …, gn} be the set of n sensor nodes in a group and be the set of m sensor nodes in another group, for same or differ
75、ent value of m and n. If value of GHC = p, gi G, then , value of GHC = q, where p ≠ q, as G and are different group of nodes. In our protocol, since sink node initiates the constr
76、uction phase, which is in G1, value of GHC for all nodes of G1 in Fig. 2, is 0 and if the packet is forwarded from G1 to any other groups like G2 or G4, value of GHC in </p><p> ? Parent Gateway ID (PGID):
77、The node that leads all nodes of a group to connect with a node of an upstream group is known as the Parent Gateway and its ID is termed as PGID. In each group of nodes, there exists only one Parent Gateway.</p>&
78、lt;p> Fig. 2. Parent and child gateways of different group of nodes.</p><p> ? Child Gateway: The node that connects to the parent gateway of a downstream group is known as the Child Gateway. In a
79、group, there exists at least one child gateway. In certain cases, if a group contains only one node, that single node is treated as both parent and child gateway for that group. </p><p> In Fig. 2, nodes A
80、and B of group G1 are the child gateways of nodes D and C, respectively. </p><p> ? Node Energy Level (NEL): The current energy level of a node is called NEL. For example, at the time of broadcasting a cont
81、rol packet, if energy level of a node is x units, NEL is assigned as x units in the control packet. </p><p> ? Parent Gateway Power Level (PGPL): The transmission power level of the parent gateway of any gr
82、oup with which it can be connected with the child gateway of an upstream group is known as Parent Gateway Power Level (PGPL). Since, sink is always the parent gateway in its group, its PGPL is assigned to 0. However, for
83、 the parent gateway of other groups, Pmin < PGPL Pmax, which may have value between 1 and 3, as per our assumption.</p><p> ? Source ID (SID): If A and B are two different sensor nod
84、es of the same or different groups such that A sends packet to B, A is the source for B and ID of node A is the Source ID (SID).</p><p> 3. The distributed power control protocol</p><p> In th
85、is section we present our power control based topology construction protocol, which constructs the topology dynamically. We assume that each node in the network has a unique ID and each of them knows its one-hop neighbor
86、’s ID prior to the construction of the topology. As per the system model of our protocol, since connectivity holes exist among each group of nodes, we assume that the network may be disconnected, if they use low transmis
87、sion power level between one group of nodes with anothe</p><p> 3.1. Construction phase</p><p> As soon as the nodes are deployed on the network, the sink initiates the construction phase by b
88、roadcasting a construct packet with minimum transmission power (Pmin = 0) to get connected with its immediate neighbors, as shown in Fig. 4(a). The format of the construct packet is shown in Fig. 3 and the para
89、meters of the packet are initialized as: SID = Sink’s ID, PGID = Sink’s ID, NEL = Sink’s power level, LHC = 0, GHC = 0, PGPL = 0. Since, sink node generally rec
90、eives the data, its PGPL is assign</p><p><b> (1)</b></p><p> where αi is a small random number compatible with CSMA-CA mechanism [22]. Then, each of them rebroadcasts the construc
91、t packet using the same minimum power level Pmin = 0 to their neighbors with necessary increments to the parameters of the construct packet and waits for time Ti units, as defined in Eq. (2).</p><p&g
92、t;<b> (2)</b></p><p> where Ei is the current energy level of ith node and βi is a very small random number such that 0.00001 βi 0.0001.</p><p> Fig. 3.
93、;Format of the construct packet.</p><p> View Within Article</p><p> In order to avoid the packet collision among group of nodes in a dense network, we propose that the sink also waits for Ti
94、units after broadcasting the construct packet and then goes to the information phase, as described in Section 3.2. It is to be noted that sink must be within at least one of the sensor node’s minimum or maximum transmiss
95、ion power range. However, if the sink does not find any neighbor with Pmin = 0, it goes to the information phase to construct the link with its neighbors, af</p><p><b> Table 2. </b>&l
96、t;/p><p> Construction phase algorithm for the sink and any node of the network</p><p> ALGORITHM 1: Construction Phase</p><p><b> Table 3. </b></p><p> In
97、formation phase algorithm for both Sender and Receiver</p><p> ALGORITHM 2: Information Phase</p><p> Upon receiving the construct packets, the nodes scan all parameters in it and consider the
98、 node having least LHC as its source. The receiver nodes wait for Wi units, get connected with its source and then follow the same procedure, as described above. This process continues till a node does not receive any co
99、nstruct packet further and the first tree topology is constructed among the nodes of a group, as shown in Fig. 4(b) with sink as the root and other nodes within minimum transmission power le</p><p> Fig. 4.
100、 (a) Randomly distributed sensor nodes over an area. (b) Construction of the first tree topology.</p><p> 3.2. Information phase</p><p> The purpose of this phase is to construct a distri
101、buted tree topology in the whole network, using most effective power level among different group of nodes. It is accomplished by broadcasting the inform packets using maximum transmission power level (Pmax = 3)
102、. The format of the inform packet is shown in Fig. 5. It is to be noted that each group of nodes has a unique parent gateway. For example, the sink is the unique parent gateway in its group. So, prior to broadcasting the
103、 inform packet, a no</p><p><b> (3)</b></p><p> Pr=ξ×(d-γ)×Pt,</p><p> where Pt is the transmission power that a node uses to broadcast the inform packet. I
104、n our protocol, each node uses the transmission power (Pt) corresponding to Pmax = 3 to broadcast an inform packet, as given in Table 1. Pr is the received power by a node during the reception of an inform pack
105、et. The received power varies with d?γ, where γ is the path loss (attenuation) factor that satisfies 2 γ 4. Here, the proportionality constant ξ is assumed to be 1 for notational simplicity and the
溫馨提示
- 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
- 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
- 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
- 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
- 5. 眾賞文庫僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負(fù)責(zé)。
- 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請與我們聯(lián)系,我們立即糾正。
- 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。
最新文檔
- 分布式溫度傳感器外文翻譯
- 無線傳感器網(wǎng)絡(luò)中基于網(wǎng)絡(luò)結(jié)構(gòu)的分布式估計(jì)研究.pdf
- 無線傳感器網(wǎng)絡(luò)分布式檢測.pdf
- 基于功率控制的無線傳感器網(wǎng)絡(luò)拓?fù)淇刂扑惴ǖ难芯?pdf
- 基于發(fā)射功率調(diào)整的無線傳感器網(wǎng)絡(luò)的拓?fù)淇刂?pdf
- 分布式溫度傳感器外文翻譯
- 無線傳感器網(wǎng)絡(luò)分布式檢測的研究.pdf
- 基于無線傳感器網(wǎng)絡(luò)的分布式濾波算法研究.pdf
- 基于無線傳感器網(wǎng)絡(luò)的分布式定位跟蹤系統(tǒng).pdf
- 面向節(jié)能和容錯的異構(gòu)無線傳感器網(wǎng)絡(luò)分布式拓?fù)淇刂扑惴ㄑ芯?pdf
- 異構(gòu)傳感器網(wǎng)絡(luò)分布式拓?fù)淇刂萍夹g(shù)研究.pdf
- 基于無線傳感器網(wǎng)絡(luò)的分布式算法及應(yīng)用.pdf
- 基于分布式無線傳感器網(wǎng)絡(luò)的漁網(wǎng)監(jiān)測系統(tǒng).pdf
- 基于無線傳感器網(wǎng)絡(luò)的分布式定位算法研究.pdf
- 基于功率調(diào)節(jié)的無線傳感器網(wǎng)絡(luò)拓?fù)淇刂萍夹g(shù)研究.pdf
- 無線傳感器網(wǎng)絡(luò)的分布式時(shí)鐘同步算法.pdf
- 分布式無線傳感器網(wǎng)絡(luò)密鑰管理的研究.pdf
- 無線傳感器網(wǎng)絡(luò)基于聚類的分布式算法分析.pdf
- 基于異構(gòu)無線傳感器網(wǎng)絡(luò)的拓?fù)淇刂?pdf
- 無線傳感器網(wǎng)絡(luò)中基于控制信息分布式更新的路由協(xié)議研究.pdf
評論
0/150
提交評論