Computer simulations
5.3 Time-varying channel estimation
5.3.1 Subspace-based method
Here, we test our proposed method in time variant channel to see how they behave.
Each tap of the time-varying FIR channels varies according to Jakes’ model, and the sample rate is 1 MHz.
First, the subspace-based channel estimator is tested and shown in Fig.5.8 in time
Fig.5.8. Testing of the subspace-based channel estimator in different Doppler frequencies for received blocks equaling to 100
In the above, we use received blocks equaling to 100, however in the 100 blocks the channel is already different, hence next we want to test the situation, which data block equal to 50.
Fig.5.9. Testing of the subspace-based channel estimator in different Doppler frequencies for received blocks equaling to 50
Comparing Fig.5.8 and Fig.5.9, we can see that Fig.5.9 has got better performance since the received data blocks is less. As we all know, the estimator is to get a channel from the received data which have the information of channels, which best suits all the channels. However, the channels are all different. Hence, more blocks leads to poorer performance.
Fig.5.10 show the NMSCE for each block in the 50 received data blocks in fd=60Hz and SNR=15dB. As we can see, around the 25th block, we can get best
to all the 50 blocks. Hence we’ve got this result.
Fig.5.10 Channel error for each block
5.3.2 Performance of PD and DD
DD and PD are tested in the following. For PD we do not average all received block to estimate Hi2(ρk), instead we need to test how long the window size should be to get the better estimate, which is because as the window is longer, we can suppress more noise, but then we can’t follow the variation of the channel.
We can see in Fig.5.11 that both PD and DD improve the performance moreover they resolve the error floor problem occurred in subspace-based estimator. Moreover,
in PD, we can see that “window size=1” outperforms, which is because in fd=50 Hz, the channel change fast, we need to trace the variation by using smaller window size.
Fig.5.11 Test of DD and PD in fd=100Hz
Performance of DD in different data constellation is shown in Fig.5.12.We can see that BPSK, QPSK, and 16QAM improve the performance. However, 64QAM worsen the performance in low SNR since it is easy to make wrong decision.
Fig.5.12 Test of DD for different data constellation in fd=50
Chapter 6 Conclusion
For STC transceivers, multichannel estimation algorithms are needed. However, training sequences consume bandwidth and, thereby, incur spectral efficiency (and thus capacity) loss. For this reason, blind channel estimation me thods receive growing attention. We have shown a subspace-based blind channel estimation algorithm for ST OFDM transmissions and develop the theoretical mean square error of the estimator.
To further improve the channel estimation, we can exploit the finite alphabet property to better the channel estimates. We discuss two different methods, DD and PD, and apply them to ST OFDM system. DD, as implied in the name, needs first to get the hard decision data and then use it to update our estimated channel, while PD is to solve the phase ambiguities after we’ve got the channel power response. However, in ST OFDM the channel power response is hard to get since the received data is
channel we can choose a best window size to get the channel power. Computer simulations have shown that the PD and DD method really improve the NMSCE in static channel and time- varying channel and further resolve the error floor problem occurred in time-varying channel.
However, our PD method for Space-Time OFDM is only utilized in BPSK system, we shall try to apply it to other systems such as QPSK, QAM in the future.
Bibliography
[1] Marc Engels, ”Wireless OFDM systems ~How to make them work?” Kluwer academic publishers, 2002.
[2] John Terry and Juha Heuskala ”OFDM wireless LAN: a theoretical and practical guide ” 201 West 103rd St. Indianapolis, Indiana, USA, 2002.
[3] Richard van Nee, Ramjee Presad “OFDM wireless multimedia communications ” Artech House Boston London, 2000
[4] Shengli Zhou, Georgios B. Giannakis “Space time coding with maximum diversity gains over frequency-selective fading channels” IEEE Signal Processing Letters, vol 8, pp.269-272, October 2001
[5] D. Mihai Ionescu, “On Space–Time Code Design” IEEE Trans. on Wireless Communications, vol. 2, pp. 20-28, January 2003
[6] Z. Liu and Georgios. B. Giannakis, “Space-time block coded multiple access through frequency-selective fading channels,” IEEE Trans. Communications, vol.
49, pp. 1033–1044, June 2001.
[7] Z. Liu, G. B. Giannakis, B. Muquet, and S. Zhou, “Space-time coding for
[8] Li, Justin C. Chuang, and N. R. Sollenberger, “Transmitter diversity for OFDM systems and its impact on high-rate data wireless networks ” IEEE J. Select. Areas Communications, vol. 17, pp.1233-1243, July 1999
[9] Ye , Nambirajan Seshadri, and Sirikiat Ariyavisitakul, “Channel estimation for OFDM systems with transmitter diversity in mobile wireless channels” IEEE J.
Select. Areas Communications, vol. 17, pp.461-471, March 1999
[10] Yongzhe Xie, and Costas N. Georghiades, “ Two EM-Type channel estimation algorithms for OFDM with transmitter diversity” IEEE Trans. Communications, vol. 51, pp.106-115, January 2003
[11] S. M. Alamouti, “A simple transmit diversity technique for wireless communications ,” IEEE J. Select. Areas Communications, vol. 16, pp.1451–1458, October 1998.
[12] B. Hassibi and B. Hochwald ” How much training is needed in multiple-antenna wireless links?.” IEEE Trans. Inform. Theory, vol. 49, pp.951-963, April 2003
[13] Y. Li, N. Seshadri, and S. Ariyavisitakul, “Channel estimation for OFDM systems with transmitter diversity in mobile wireless channels,” IEEE J. Select.
Areas Communications, vol. 17, pp. 461–471, March 1999.
[14] H. Bölcskei, R. W. Heath Jr., and A. J. Paulraj, “Blind channel identification and equalization in OFDM-based multi-antenna systems,” IEEE Trans. Signal
Processing, vol. 50, pp. 96–109, January 2002.
[15] Yonghong Zeng, and Tung-Sang Ng, “Semi- Blind channel estimation method for multiuser multiantenna OFDM Systems” IEEE Trans. Signal Processing, vol 2, pp.1419-1429, May 2004
[16] Z. Liu, G. B. Giannakis, S. Barbarossa, and A. Scaglione, “Transmit antenna space-time block coding for generalized OFDM in the presence of unknown multipath,” IEEE J. Select. Areas Communications, vol. 19, pp.1352–1364, July 2001.
[17] Shengli Zhou, and Georgios B. Giannakis, “Subspace-Based (Semi- ) Blind Channel Estimation for Block Precoded Space-Time OFDM ,” IEEE Trans. on signal processing, vol. 50, pp.1215-1228, May 2002
[18] G. H. Golub and C.F. Van Loan, ”Matrix computations” John Hopkins University Press, 3rd edition, 1996
[19] F. Li, H. Liu, and R. J. Vaccaro, “Performance analysis for DOA estimation algorithms: further unification, simplification, and observations,”
IEEE Trans. Aerosp., Electron. Syst., vol. 29, pp. 1170–1184, October 1993.
[20] Shengli Zhou, and Georgios B. Giannakis, “Finite-alphabet based channel
[21] Shengli Zhou, Georgios B. Giannakis and Anna Scaglione, “Long codes for generalized FH-OFDMA through unknown multipath channels ” IEEE Trans. on Communications, vol. 49, pp 721-733, April 2001