《電子技術(shù)應(yīng)用》
您所在的位置:首頁(yè) > 通信與網(wǎng)絡(luò) > 設(shè)計(jì)應(yīng)用 > 基于基擴(kuò)展模型的高移動(dòng)性信道估計(jì)方法
基于基擴(kuò)展模型的高移動(dòng)性信道估計(jì)方法
2017年電子技術(shù)應(yīng)用第6期
黃錦錦,趙宜升,陳忠輝,賴鑫琳,董志翔
福州大學(xué) 物理與信息工程學(xué)院,福建 福州350108
摘要: 針對(duì)鐵路長(zhǎng)期演進(jìn)(LTE-R)通信系統(tǒng),開(kāi)展高移動(dòng)性信道估計(jì)研究。通過(guò)引入基擴(kuò)展模型,將LTE-R系統(tǒng)的信道沖激響應(yīng)擬合為若干基函數(shù)與系數(shù)乘積和的形式。通過(guò)對(duì)基函數(shù)系數(shù)的估計(jì),實(shí)現(xiàn)對(duì)快速時(shí)變信道進(jìn)行近似。通過(guò)仿真對(duì)多項(xiàng)式基擴(kuò)展模型、復(fù)指數(shù)基擴(kuò)展模型、泛化復(fù)指數(shù)基擴(kuò)展模型和優(yōu)化泛化復(fù)指數(shù)基擴(kuò)展模型進(jìn)行性能對(duì)比。仿真結(jié)果表明,優(yōu)化泛化復(fù)指數(shù)基擴(kuò)展模型具有最低的歸一化均方誤差。此外,對(duì)于優(yōu)化泛化復(fù)指數(shù)基擴(kuò)展模型,分別探討了不同移動(dòng)速度、不同基函數(shù)個(gè)數(shù)和不同調(diào)制方式下的估計(jì)性能。仿真結(jié)果顯示,在較高移動(dòng)速度、較少基函數(shù)個(gè)數(shù)及較高階調(diào)制方式下,優(yōu)化泛化復(fù)指數(shù)基擴(kuò)展模型仍然具有較低的歸一化均方誤差。
中圖分類號(hào): TN914
文獻(xiàn)標(biāo)識(shí)碼: A
DOI:10.16157/j.issn.0258-7998.2017.06.029
中文引用格式: 黃錦錦,趙宜升,陳忠輝,等. 基于基擴(kuò)展模型的高移動(dòng)性信道估計(jì)方法[J].電子技術(shù)應(yīng)用,2017,43(6):113-117.
英文引用格式: Huang Jinjin,Zhao Yisheng,Chen Zhonghui,et al. High mobility channel estimation method based on basis expansion model[J].Application of Electronic Technique,2017,43(6):113-117.
High mobility channel estimation method based on basis expansion model
Huang Jinjin,Zhao Yisheng,Chen Zhonghui,Lai Xinlin,Dong Zhixiang
College of Physics and Information Engineering,F(xiàn)uzhou University,F(xiàn)uzhou 350108,China
Abstract: In this paper, the problem of high mobility channel estimation is investigated in the Long-Term Evolution for Railway(LTE-R) communication system. By employing a Basis Expansion Model(BEM), the channel impulse response of the LTE-R system is fitted as the summation of several basis functions multiplied by the corresponding coefficients. The fast time-varying channel can be obtained approximately by estimating the coefficients of the basis functions. The performances of a polynomial BEM, a Complex Exponential BEM(CE-BEM), a Generalized CE-BEM(GCE-BEM), and an Optimization GCE-BEM(OGCE-BEM) are compared by simulation. Simulation results show that the OGCE-BEM has the lowest Normalized Mean Square Error(NMSE). In addition, the estimation performances of the OGCE-BEM are discussed under different moving speeds, different numbers of basis functions, and different modulation modes, respectively. Simulation results reveal that the OGCE-BEM still has a lower NMSE under a higher moving speed, a smaller number of basis functions, and a higher order modulation.
Key words : channel estimation;high mobility;basis expansion model

0 引言

    鐵路長(zhǎng)期演進(jìn)(Long -Term Evolution for Railway,LTE-R)系統(tǒng)是極具前景的高速鐵路通信系統(tǒng)。根據(jù)國(guó)際鐵路聯(lián)盟的規(guī)劃,鐵路移動(dòng)通信系統(tǒng)將從傳統(tǒng)的鐵路全球移動(dòng)通信系統(tǒng)(Global System for Mobile Communications-Railway,GSM-R)直接過(guò)渡到LTE-R系統(tǒng)[1]。對(duì)于LTE-R系統(tǒng),列車移動(dòng)速度通常超過(guò)300 km/h,會(huì)產(chǎn)生嚴(yán)重的多普勒頻移。同時(shí),無(wú)線信道狀態(tài)呈現(xiàn)動(dòng)態(tài)變化特點(diǎn)。如何保證在高移動(dòng)性場(chǎng)景下,仍然能夠?yàn)橛脩籼峁┛煽康臒o(wú)線通信服務(wù),信道估計(jì)是關(guān)鍵。

    信道估計(jì)問(wèn)題已經(jīng)引起了廣泛關(guān)注。根據(jù)是否需要引入導(dǎo)頻信息,信道估計(jì)可以分為盲信道估計(jì)、導(dǎo)頻輔助信道估計(jì)和半盲信道估計(jì)。然而,盲信道估計(jì)[2-3]雖然省去了導(dǎo)頻信息的傳遞,提高了頻帶利用率,但其算法收斂速度較慢,且需要大量的數(shù)據(jù)存儲(chǔ)和復(fù)雜的數(shù)學(xué)運(yùn)算,因此局限應(yīng)用于慢時(shí)變衰落信道,對(duì)數(shù)據(jù)實(shí)時(shí)處理要求不高的地方,不適用于高移動(dòng)性信道估計(jì);導(dǎo)頻輔助信道估計(jì)[4-5]具有較低的算法復(fù)雜度,便于系統(tǒng)的實(shí)現(xiàn),且能實(shí)時(shí)跟蹤C(jī)SI,適合進(jìn)行快速時(shí)變信道估計(jì);半盲信道估計(jì)算法[6-7]雖然在復(fù)雜度和導(dǎo)頻數(shù)量上進(jìn)行了折衷,但復(fù)雜度仍然較高,也不適合對(duì)快速動(dòng)態(tài)變化信道進(jìn)行估計(jì)。基于基擴(kuò)展模型(Basis Expansion Model,BEM)的導(dǎo)頻輔助信道估計(jì)方法[8-9]通過(guò)若干基函數(shù)與系數(shù)乘積和的方式,可以對(duì)快速時(shí)變信道進(jìn)行近似,已經(jīng)引起了廣泛關(guān)注。因此,本文將采用BEM對(duì)高移動(dòng)性信道估計(jì)問(wèn)題進(jìn)行研究。

    本文針對(duì)LTE-R通信系統(tǒng),開(kāi)展高移動(dòng)性場(chǎng)景的信道估計(jì)研究。首先,建立LTE-R系統(tǒng)的信道模型。然后,根據(jù)BEM將LTE-R信道沖激響應(yīng)表示為若干基函數(shù)與系數(shù)乘積和的形式。接下來(lái),通過(guò)對(duì)基函數(shù)系數(shù)進(jìn)行估計(jì)的方式,實(shí)現(xiàn)對(duì)LTE-R信道的擬合。最后,通過(guò)仿真對(duì)4種形式的BEM進(jìn)行性能評(píng)估。

1 LTE-R信道模型

    高速鐵路LTE-R通信系統(tǒng)結(jié)構(gòu)如圖1所示。設(shè)計(jì)分布式基站可以解決高速列車通信問(wèn)題,分布式基站由室內(nèi)基帶處理單元(Building Baseband Unit,BBU)和射頻拉遠(yuǎn)單元(Radio Remote Unit,RRU)組成[10]。BBU位于基站的室內(nèi),RRU被部署在鐵路沿線附近,多個(gè)RRU分別通過(guò)光纖將信號(hào)傳輸?shù)紹BU。分布式基站的設(shè)計(jì)可以擴(kuò)大小區(qū)信號(hào)的覆蓋范圍,在一定程度上可以減少用戶越區(qū)切換次數(shù)。BBU和RRU分別用于處理基帶信號(hào)和射頻信號(hào),由于通過(guò)光纖將基帶信號(hào)從BBU傳輸?shù)絉RU,從而避免了射頻信號(hào)的長(zhǎng)距離傳輸,可以顯著降低傳輸損耗。此外,由于無(wú)線信號(hào)穿過(guò)列車車廂會(huì)造成嚴(yán)重的穿透損耗,為了保證RRU和列車之間的可靠通信,在列車的頂部安裝一個(gè)車載臺(tái)(Vehicular Station,VS)。VS通過(guò)無(wú)線方式與RRU建立連接。同時(shí),在每節(jié)車廂里都安裝一個(gè)中繼器(Repeater,R),中繼器通過(guò)有線方式與VS建立連接。車廂里的不同用戶設(shè)備(User Equipments,UE)可以通過(guò)中繼器連接到網(wǎng)絡(luò)。

tx6-t1.gif

    對(duì)于高速鐵路LTE-R通信系統(tǒng),由于RRU部署在鐵路沿線附近,同時(shí)鐵路沿線存在特殊的地理環(huán)境特征,使得RRU和VS之間除了存在一條直接的視距(Line-of-Sight,LOS)路徑以外,還存在若干條間接的非視距(Non-Line-of-Sight,NLOS)路徑。受文獻(xiàn)[11]、[12]啟發(fā),將信道沖激響應(yīng)表示為:

tx6-gs1-3.gif

式中,σ是非視距信號(hào)歸一化平均功率的標(biāo)準(zhǔn)偏差,A為視距信號(hào)的平均幅度。

2 基于基擴(kuò)展模型的信道估計(jì)方法

2.1 基擴(kuò)展模型信道建模

    采用BEM進(jìn)行信道建模的基本原理是通過(guò)若干基函數(shù)與系數(shù)乘積和來(lái)擬合無(wú)線信道。假設(shè)信號(hào)傳播的多徑數(shù)量為L(zhǎng),對(duì)于第l條路徑,若N點(diǎn)采樣的信道增益向量為hl,同時(shí)基函數(shù)向量為bm,那么,通過(guò)基擴(kuò)展模型擬合信道增益hl可以表示為:

tx6-gs4-6.gif

    此外,根據(jù)基函數(shù)形式的不同,基擴(kuò)展模型可以分為多項(xiàng)式基擴(kuò)展模型(Polynomial BEM,P-BEM)、復(fù)指數(shù)基擴(kuò)展模型(Complex Exponential BEM,CE-BEM)、泛化復(fù)指數(shù)基擴(kuò)展模型(Generalized CE-BEM,GCE-BEM)和優(yōu)化泛化復(fù)指數(shù)基擴(kuò)展模(OptimizationGCE-BEM,OGCE-BEM),分別介紹如下:

    對(duì)于P-BEM是以泰勒級(jí)數(shù)為基礎(chǔ)的,其第m個(gè)基函數(shù)第n個(gè)元素可以表示為[13]

    tx6-gs7.gif

式中,0≤n≤N-1,0≤m≤M-1。 

    對(duì)于CE-BEM,是以傅里葉級(jí)數(shù)理論為基礎(chǔ)的,其第m個(gè)基函數(shù)第n個(gè)元素可以表示為[14]

tx6-gs8-9.gif

    對(duì)于OGCE-BEM,其第m個(gè)基函數(shù)第n個(gè)元素可以表示為[16]

tx6-gs10.gif

2.2 基函數(shù)系數(shù)估計(jì)

    時(shí)變多徑信道的輸入輸出響應(yīng)關(guān)系式有:

tx6-gs11-17.gif

式中,H為N×N維時(shí)域信道循環(huán)矩陣,且表達(dá)式為:

tx6-gs18.gif

    所以,頻域Y表示為:

     tx6-gs19-20.gif

3 仿真結(jié)果和分析

tx6-gs21.gif

    圖2對(duì)比了不同BEM的NMSE性能。在仿真中,移動(dòng)速度是v=350 km/h,調(diào)制方式為64QAM,基函數(shù)個(gè)數(shù)為11??梢钥吹?,GCE-BEM的性能優(yōu)于CE-BEM,原因是GCE-BEM對(duì)多普勒頻譜更加密集的采樣能有效減少CE-BEM的頻譜泄露問(wèn)題。同時(shí),OGCE-BEM的估計(jì)性能優(yōu)于GCE-BEM,這是因?yàn)镺GCE-BEM通過(guò)對(duì)GCE-BEM基函數(shù)頻率的修正,減少了高頻基函數(shù)對(duì)模型帶來(lái)的誤差,使信道估計(jì)性能達(dá)到最優(yōu),彌補(bǔ)了GCE-BEM在高頻點(diǎn)的估計(jì)誤差偏大問(wèn)題。另外,P-BEM估計(jì)性能最差,說(shuō)明依據(jù)泰勒級(jí)數(shù)理論基礎(chǔ)的多項(xiàng)式線性組合的基函數(shù)模型對(duì)于LTE-R系統(tǒng)信道擬合誤差偏大。

tx6-t2.gif

    圖3對(duì)比了不同移動(dòng)速度時(shí)OGCE-BEM的NMSE性能。在仿真中,調(diào)制方式為64QAM,基函數(shù)個(gè)數(shù)為11。當(dāng)移動(dòng)速度為v=350 km/h,噪聲功率為0 dBm時(shí),OGCE-BEM的NMSE達(dá)到10-5,說(shuō)明即使在高階調(diào)制情況下,采用OGCE-BEM進(jìn)行高移動(dòng)性信道估計(jì),也能夠獲得較小的NMSE。

tx6-t3.gif

    圖4對(duì)比了不同基函數(shù)個(gè)數(shù)時(shí)OGCE-BEM的NMSE性能。在仿真中,移動(dòng)速度是v=350 km/h,調(diào)制方式為64QAM。當(dāng)基函數(shù)個(gè)數(shù)為3、噪聲功率為10 dBm時(shí),OGCE-BEM的NMSE仍然能夠達(dá)到10-4。當(dāng)基函數(shù)個(gè)數(shù)為15、噪聲功率為10 dBm時(shí),OGCE-BEM的NMSE能夠達(dá)到10-5。通過(guò)增加一定的基函數(shù)個(gè)數(shù),可以使信道估計(jì)的NMSE性能提高。

tx6-t4.gif

    圖5對(duì)比了不同調(diào)制方式時(shí)OGCE-BEM的NMSE性能。在仿真中,移動(dòng)速度為v=350 km/h,基函數(shù)個(gè)數(shù)為11。當(dāng)噪聲功率為0 dBm時(shí),OGCE-BEM的3種調(diào)制方式的NMSE都可以達(dá)到10-6,且調(diào)制方式為QPSK時(shí),OGCE-BEM的NMSE性能最優(yōu),其次是16QAM,64QAM最差。即使當(dāng)噪聲功率達(dá)到10 dBm時(shí),64QAM的NMSE也能達(dá)到10-5

tx6-t5.gif

4 結(jié)論

    本文研究了高移動(dòng)性場(chǎng)景下LTE-R系統(tǒng)的信道估計(jì)問(wèn)題。根據(jù)BEM,將信道沖激響應(yīng)表示為一系列基函數(shù)與系數(shù)乘積和的形式。通過(guò)對(duì)基函數(shù)系數(shù)進(jìn)行估計(jì),實(shí)現(xiàn)對(duì)LTE-R系統(tǒng)的信道估計(jì)。由于本文只針對(duì)導(dǎo)頻位置的信道狀態(tài)進(jìn)行估計(jì),下一步將考慮插值算法,實(shí)現(xiàn)對(duì)數(shù)據(jù)位置的信道狀態(tài)進(jìn)行估計(jì)。同時(shí),根據(jù)得到的估計(jì)誤差,對(duì)基函數(shù)的個(gè)數(shù)進(jìn)行動(dòng)態(tài)調(diào)整。

參考文獻(xiàn)

[1] HE R S,AI B,WANG G P,et al.High-speed railway communications:from GSM-R to LTE-R[J].IEEE Vehicular Technology Magazine,2016,11(3):49-58.

[2] ZHOU S L,GIANNAKIS G B,SCAGLIONE A.Long codes for generalized FH-OFDMA through unknown multipath channels[J].IEEE Transactions on Communications,2001,49(4):721-733.

[3] SONG S,SINGER A C.Blind OFDM channel estimation using FIR constraints: reduced complexity and identifiability[J].IEEE Transactions on Information Theory,2007,53(3):1136-1147.

[4] BARHUMI I,LEUS G,MOONEN M.Optimal training design for MIMO OFDM systems in mobile wireless channels[J].IEEE Transactions on Signal Processing,2003,51(6):1615-1624.

[5] LI Y,SESHADRI N,ARIYAVASITAKUL S.Channel estimation for OFDM systems with transmitter diversity in mobile wireless channels[J].IEEE Journal on Selected Areas in Communications,1999,17(3):461-471.

[6] MEDLES A,SLOCK D T M.Semiblind channel estimation for MIMO spatial multiplexing systems[C].IEEE Vehicular Technology Conference,2001,2:1240-1244.

[7] MEDLES A,SLOCK D T M.Linear precoding for spatial multiplexing MIMO systems:blind channel estimation aspects[C].IEEE International Conference on Communications(ICC),2002,1:401-405.

[8] RABBI M F,HOU S W,KO C C.High mobility orthogonal frequency division multiple access channel estimation using basis expansion model[J].IET Communications,2010,4(3):353-367.

[9] LIU L,TAO C,QIU J H,et al.Position-based modeling for wireless channel on high-speed railway under a viaduct at 2.35 GHz[J].IEEE Journal on Selected Areas in Communications,2012,30(4):834-845.

[10] ZHAO Y S,LI X,LI Y,et al.Resource allocation for high-speed railway downlink MIMO-OFDM system using quantum-behaved particle swarm optimization[C].IEEE International Conference on Communications(ICC),2013:2343-2347.

[11] GUAN K,ZHONG Z D,AI B.Assessment of LTE-R using high speed railway channel model[C].Third International Conference on Communications and Mobile Computing(CMC),2011:461-464.

[12] TKAC A,WIESER V,POLLAK S.Calculation of impulse response in rician and rayleigh channel[C].ELEKTRO,2012:99-102.

[13] HU D,WANG X D,HE L H.A new sparse channel estimation and tracking method for time-varying OFDM systems[J].IEEE Transactions on Vehicular Technology,2013,62(9):4648-4653.

[14] TUGNAIT J K,HE S C,KIM H.Doubly selective channel estimation using exponential basis models and subblock tracking[J].IEEE Transactions on Signal Processing,2010,58(3):1275-1289.

[15] KHAL R N,ZHANG J R,ZAKHAROV Y V.Robustness of joint bayesian frequency offset and channel estimation based on basis expansion models[C].Conference Record of the Forty-Third Asilomar Conference on Signals,Systems and Computers,Pacific Grove,2009:957-961.

[16] 張欽娟,武穆清,郭起霖,等.OFDM系統(tǒng)中優(yōu)化的泛化復(fù)指數(shù)基擴(kuò)展模型[J].西安電子科技大學(xué)學(xué)報(bào),2012(3):131-135.



作者信息:

黃錦錦,趙宜升,陳忠輝,賴鑫琳,董志翔

(福州大學(xué) 物理與信息工程學(xué)院,福建 福州350108)

此內(nèi)容為AET網(wǎng)站原創(chuàng),未經(jīng)授權(quán)禁止轉(zhuǎn)載。