臺灣博碩士論文加值系統

The discrete multitone transmission (DMT) has been recognized as one of the most practical technology for high data rate transmission. The IDFT matrix and the DFT matrix
with the size proportional to number of bands are parts of a transmitter and a receiver, respectively. As we know that, IDFT and DFT have only Nlog_2(N) complex multiplications and additions. For many baseband transmission systems such as DSL, the channels and signals are real. Complex conjugation is done at the DMT transmitter to get real signals. In this thesis, we consider DCT based transceivers which involve only real signals and operations. Like DFT systems, DCT based systems also have Nlog_2(N) complexity. Both DFT and DCT based transceivers belong to the class of block transmission scheme where data are sent on a block by block basis and there is no overlapping between blocks. Block transmission suffers from poor frequency response due to short filter length, no longer than block size. Thus, in the presence of narrowband interference (NBI), the performance of both DFT and DCT based systems degrades significantly as many tones will be affected by the NBI.

To improve the frequency response of transceivers, the perfect reconstruction (PR) filter bank transceivers (FBTs), having the advantage of high stopband attenuation, are often proposed as mitigative solutions in past. To reduce the complexity, the cosine modulated filter bank transceiver (CMFBT) have been studied for DSL applications. However, for frequency selective channels, the CMFBT suffers from severe ISI and ICI. In this thesis, we consider the CMFBT that can mitigate the ISI and ICI effects. For both known and unknown channels, the CMFBT problem can be formulated as a SIR optimization problem. The
optimal solution can easily obtained as the eigenvector of a corresponding Rayleigh-Ritz ratio. To further enhance the SIR performance of the CMFBT, an iterative procedure is
proposed for the SIR maximization. Two frequency constraint methods are introduced to preserve the frequency response of the CMFBT over iteration. Simulations show that the
CMFBT with both high SIR and good frequency response can be designed using the proposed methods.

We also verify the performance of our systems in DSL applications. In order to apply the DCT based transceivers and the CMFBTs, we need to design the time domain equalizer
(TEQ) that gives a symmetric target impulse response. A new TEQ design method is proposed for this purpose. For the commonly used 8 DSL loops, the simulation results show that in the absence of NBI, all systems have roughly the same performance whereas in the presence of NBI, the proposed CMFBT has significantly better performance.

In the second part of the thesis, we consider the general solutions of PR block transceivers for unknown multipath channels. We show that under mild assumptions, the a mutually-orthogonal usercode-receiver (AMOUR) is the only solution to such a problem. In other words, there is no transceiver other than the AMOUR that can achieve PR for
unknown multipath channels.

1 Introduction 1

2 FBT System 5
2.1 Introduction to FBT . . 6
2.2 DFT and Cosine Modulated Filter Bank Systems . . 7
2.2.1 Discrete Multitone System . . 10

3 DCT Based Block Transceivers 17
3.1 Diagonalization Properties of DCT Matrices . . 18
3.2 DCT Based Block Transceivers . . 20

4 Cosine Modulated Filter Bank Transceivers 29
4.1 Cosine Modulated Filter Bank Transceivers with DCT-IIE . . 30
4.2 PR Condition . . 31
4.3 SIR Optimization for Known Channel . . 33
4.4 SIR Optimization for Unknown Channel . . 36
4.5 Iterative Algorithm . . 37
4.6 SIR Optimization with Frequency Constraint. . 41
4.6.1 Frequency Constraint with Prefiltering . . 41
4.6.2 Frequency Constraint with Eigenfilter Approach .. 50

5 CMFBT for DSL 57
5.1 MMSE TEQ with Symmetric Target Impulse Response . . 59
5.1.1 MMSE TEQ 1 . . 60
5.1.2 MMSE TEQ 2 . . 61
5.1.3 Special cases of MMSE TEQ 2 . . 63
5.2 Simulation and Performance Analysis . . 65

6 On the Solutions of PR Block Transceivers for Unknown Channels 79
6.1 PR Block Transceiver . . 81
6.2 Other Solutions . . 87

7 Conclusion 97

Bibliography 99

[1] L. J. Cimini, "Analysis and Simulation of a Digital Mobile Channel Using Orthogonal Frequency Division Multiple Access," IEEE Trans. Commun., Vol. 30,
July 1985, pp. 665-675.
[2] A. N. Akansu, et. al., "Orthogonal Transmultiplexers in Communication: A Review," IEEE Trans. Signal Processing, Vol. 46, April 1998, pp. 979-995.
[3] P. P. Vaidyanathan,Multirate Systems and Filter Banks. Englewood Cliffs, NJ:Prentice Hall, 1993.
[4] V. Sanchez, P. Garcia, A.M. Peinado, J.C. Segura, A.J. Rubio, "Diagonalizing Properties of the Discrete Cosine Transforms," IEEE Trans. Signal Processing,
Vol. 43, NO. 11, Nov 1995, pp. 2631-2641
[5] S. A. Martucci, "Symmetric Convolution and the Discrete Sine and Cosine Transforms," IEEE Trans.Signal Processing, Vol. 42, NO. 5, May 1994, pp. 1038-1051
[6] N. Al-Dhahir and H. Minn, "A New Multicarrier Transceiver Based on the Discrete Cosine Transform," in IEEE WCNC, March 2005. Vol. 42, NO. 5, May 1994, pp.
1038-1051
[7] A. D. Rizos, J. G. Proakis, and T. Q. Nguyen, "Comparison of DFT and Cosine Modulated Filter Banks in Multicarrier Modulation," IEEE Global Telecommunications Conference, pp.687-691, 1994.
[8] S. Govardhanagiri, T. Karp, P. Heller and T. Nguyen, "Performance Analysis of Multicarrier Modulations Systems using Cosine Modulated Filter Banks," Proc.
ICASSP, pp. 1405-1408, March 1999.
[9] B. Farhang-Boroujeny, L. Lin, "Analysis of Post-Combiner Equalizers in Cosine-Modulated Filterbank-Based Transmultiplexer Systems," IEEE Trans. Signal Pro-
cessing, Vol. 51, NO.12, Dec, 2003.
[10] N. J. Fliege and G. Rosel, "Equalizer and Crosstalk Compensation Filters for DFT Polyphase Transmultiplexer Filter Banks," Proc. ISCAS, 1994.
[11] P. P. Vaidyanathan and T. Nguyen, "Eigenfilters: A New Approach to Least Square FIR Filter Design and Applicaitions Including Nyquist Filters," Circuits
and systems, IEEE Transactions on, Vol. 34, Issue: 1, Jan 1987, pp. 481-486
[12] D. Falconer and F. Magee, "Adaptive Channel Memory Truncation for Maximum Likelihood Sequences Estimation," Bell Syst. Tech. J., Vol. 52, No. 9, pp. 1541-1562, Nov, 1973.
[13] R. K. Martin, C. R. Johnson, Jr., Ming Ding and B. L. Evans, "Infinite Length Results for Channel Shortening Equalizers," IEEE Signal Processing Workshop on Signal Processing Advances in Wireless Communication, Rome, Italy, June, 2003.
[14] N. Al-Dhahir and J. M. Cio , "Efficiently Computed Reduced-Parameter Input-Aided MMSE Equalizers for ML Detection: A unified Approach," IEEE Trans. Information Theory, Vol. 42, No. 3, May 1996.
[15] P. J. W. Melsa, R. C. Younce and C. E. Rohrs, "Impulse Response Shortening for Discrete Multitone Transceiver," IEEE Trans. Communications Vol. 44, No. 12, Dec 1996.
[16] G. Arslan, M. Ding, B. Lu, M. Milosevic, Z. Shen, and B. L. Evans. MATLAB DMTTEQ Toolbox Version 3.1. Univ. Texas, Austin, TX. Available:http://www.ece.utexas.edu/ bevans/projects/adsl/dmtteq/dmtteq.html
[17] G. Arslan, B. L. Evans and S. Kiaei, "Equalization for Discrete Multitone Trainsceivers to Maximize Bit Rate," IEEE Trans. Signal Processing, Vol. 49, NO. 12, Dec 2001.
[18] A. Scaglion, G. B. Giannakis, "Design of User Codes in QS-CDMA Systems for MUI Elimination in Unknown Multipaths," Communications Letters, IEEE Vol.3, No. 2, Feb 1999.
[19] G. B. Giannakis, Z. Wang, A. Scaglionw, S. Barbarossa, "AMOUR-Generalized Multicarrier Transceivers for Blind CDMA Regardless of Multipath," IEEE Trans. Communications, Vol. 48, No. 12, Dec 2000.
[20] S. M. Phoong, Y. B. Chang, C. Y. Chen, "DFT Modulated Filter Bank Transceivers for Multipath Fading Channels," IEEE Trans. Signal Processing, VOL. 53, NO. 1, Jan 2005.