EM Based MAP Channel Estimation for Turbo/Convolutional Coded SFBC/STBC-OFDM Systems

Abstract

OFDM sistemleri ile biutnle,tirilmi, di* kodlayicilarin kullanildigi veya kullanilmadigi sistemler i,cin genelles,tirilmis,tir Bu qali*mada frekans se,ici kanallarda Turbo/Katlamali [3, 4, 5, 6]. Lu ve dig. [3] tarafindan iletilen sembollerin kanal kodlanmi* Uzay-frekans/Uzay-zaman blok kodlu dikgen kestiriminde direk olarak kullanildigi ve kanalin birka, OFDM frekans boliusi,mliu ,ogullama (SFBC/STBC-OFDM) sissembolil boyunca degi,medigi kabul edilerek en biiyiik sonsal temleri i,in alicida kanal bilgisinin olmadigi durumda, alici (MAP) EM tabanli kestirimci turbo kodlanmi* STBC-OFDM tasarimi problemine yakla*imda bulunulmu,tur. Kanal kesistemlerine uyarlanmi*tir. Ayrica kari*tirilmi* sinyallerin kestirimci olarak "Beklenti-Enbiiyiikleme (EM)" algoritmasi stirimi i,in kullanilan genel EM algoritmasi, konvolusyonel kullanilmi* ve algoritmanin karma*ikliginin azaltilmasi icin kodlayicin kullanildigi verici ,esitlemeli OFDM sistemlerine kanal degi,imlerinin rastgele bir siire, olarak modellenip uygulanmi* ve SAGE siuriimiu ile kar*ila*tirilmi*tir [5]. Ma ve Karhunen-Loeve (KL) dik seri a,ilimi uygulanmi*tir. Onerilen dig. [6] ise daha once onerilen yapiyi [5] degi,tirilerek STBCsistemlerin bilgisayar benzetimleri yapilarak, ba*arimlari kanal OFDM ve SFBC-OFDM sistemleri i,in onermi,lerdir. kestirim hatasini g6steren ortalama karesel hata (MSE) ve Onerilen [3, 4, 5, 6] qali*malarindan farkli olarak bit hata orani (BER) olarak kullanilan pilot araliklarina g6re MAP-EM yapisi kanalin her OFDM semboli. i.in degi*tigi incelenmi,tir.

Study and implementation of an iterative decoder for SFBC-OFDM systems with imperfect channel knowledge

“…Let and be two vectors containing the even and odd components of . The received signal can be expressed as follows: (6) where and are diagonal matrices whose elements are and , respectively. Moreover, and are respectively the even and odd component vectors of the channel coefficients between the th transmit antenna and the receive antenna.…”

“…In this case, the resolution of (11) is very difficult, especially when the size of the block is large. We propose to use the EM algorithm [5][6][7] to solve iteratively the problem (11) with a reasonable complexity. Each iteration of the algorithm consists in two steps:…”

“…The solution of this problem is very difficult to find directly. Thus, we propose to use an iterative decoder based on the Expectation Maximization (EM) algorithm [5][6][7] to approach iteratively the solution. Then, we propose to implement the iterative decoder on an FPGA (Field Programmable Gate Array) platform.…”