Chapter 2 Synchronization and Cell Search Procedure
2.5 Joint Of Cell Search Procedure and CP Length Detection
2.5 Joint Of Cell Search Procedure and CP Length Detection
The Chapter 2.2 to 2.4 introduces the different cell search procedures in the situation that CP length is previous known. The blind CP length detection can refer [10]. In this Chapter we add blind CP length detection in the cell search procedure as shown in Fig. 2.3.
Fig. 2.3 Joint of cell search procedure and CP length detection
Step 1 and Step 4 are the same as before. In step 2, we have to perform SSS self-symmetry correlation to find the starting position of S-SCH. After finding the starting position we decide to judge the CP length and Cell-ID2. In FDD mode, the P-SCH is just in the next symbol. So we can find starting position of P-SCH by
sss IFFT size length, nˆ represents estimated S-SCH starting position of the symbol, and sss nˆpssis
the estimated starting position of P-SCH of the symbol.
In Step 3, take two PSS symbols according to two type CP length and apply correlation computation in the time domain or in the frequency domain. CP length, physical-layer identity within the physical-layer cell identity group (NID(2)) and integer carrier frequency offset (ICFO) will be acquired.
In the time domain:
u
ypssis PSS symbol in the time domain, mdenotes different CP types, m=0represent using normal CP and m=1represent using extended CP. p denote three known PSS time-domain u sequences, uis 25, 29 and 34 according to threeNID(2)is 0, 1, 2. Theε denote ICFO,I βis ICFO range and⊗denotes circular correlation that can diminish symbol timing error.
In the frequency domain:
)
where Ypss is PSS symbol in the frequency domain, m denotes different CP types,
=0
m represent using normal CP and m=1represent using extended CP. P denote three u known PSS frequency-domain sequences, uis 25, 29 and 34 according to threeNID(2) is 0, 1, 2. TheεIdenote ICFO,βis ICFO range.
Chapter 3
Simulation Results and Summary
3.1 Simulation Result
In order to compare the performance of the proposed timing synchronization for LTE cell search, we adopt the tapped delay-line model of SCM channel model described in [6].
There are three scenarios, the suburban macro, urban macro and urban micro. The channel tap delay-line parameters are shown in Table I. We consider one antenna transmitter and receiver. The 15MHz LTE system was considered, with sampling rate of 23.04 MHz, 1024-point DFT/IDFT size, normal CP and 15 kHz subcarrier spacing. We also considered the mismatch of oscillator up to 10 ppm, with corresponds to ε =1.33.
Table I. Tapped delay-line parameters Scenario
Suburban Macro Urban Macro Urban Micro tap Power
(dB)
Delay (ns)
Power (dB)
Delay (ns)
Power (dB)
Delay (ns)
1 0 0 0 0 0 0
2 -2.6682 140.8 -2.220 360 -1.2661 284
3 -6.2147 62.6 -1.718 252.7 -2.7201 204.7
4 -10.4132 401.5 -5.189 1038.7 -4.2973 662.3
5 -16.4735 1382 -9.051 2730 -6.0140 806.6
6 -22.1898 2828 -12.501 4597.7 -8.430 922.7 Fig. 3.1 shows the root mean square error (MSE) of the OFDM timing estimation measured in samples; the MSE is plotted against the SNR with channel scenario of suburban macro. When CFO is large, Method one performs poorly and the timing synchronization is quite off. MSEs of Method two and Method three decrease when SNR increases. Method
Three performs better even under the other two channel scenarios. Nevertheless we still have to confirm PSS detection and NIDcell detection performance.
-5 0 5 10
0 2 4 6 8 10 12 14 16 18x 104
snr
MSE(samples)
Timing Synchronization in Suburban Macro
three PSS PSS symmetry SSS symmetry
Fig. 3.1 Timing synchronization in terms of the MSE measured in samples against the SNR.
Fig. 3.2 shows the PSS detection for NID(2)in time domain and Fig. 3.3 in frequency domain. We can see the performance is related with the channel scenarios. The error probabilities are similar whether by time or frequency-domain methods. With CFO, method two, which is doing timing synchronization with PSS, cannot catch the correct symbol time.
The worse timing synchronization also leads to worse CFO compensation. Method three, which is doing timing synchronization with SSS, is immune to CFO. Thus ISI and ICI are diminished. Consequently the performance of method three is better than method two.
-5 -4 -3 -2 -1 0 1 2 3 4 5 10-5
10-4 10-3 10-2 10-1
snr
errorprobability
N(2) ID detection with different channel in the time domain
suburban macro (PSS symmetry) urban macro (PSS symmetry) urban micro (PSS symmetry) suburban macro (SSS symmetry) urban macro (SSS symmetry) urban micro (SSS symmetry)
Fig. 3.2 PSS detection performance for NID(2)in time domain.
-5 -4 -3 -2 -1 0 1 2 3 4 5 10-5
10-4 10-3 10-2 10-1
snr
errorprobability
N(2) ID detection with different channel in the frequency domain
suburban macro (PSS symmetry) urban macro (PSS symmetry) urban micro (PSS symmetry) suburban macro (SSS symmetry) urban macro (SSS symmetry) urban micro (SSS symmetry)
Fig. 3.3 PSS detection performance forNID(2)in frequency domain.
Fig. 3.4 shows the detection performance forNIDcell. The S-SCH sequences with different
) 1 (
NID has variation just cyclic shift. If the sequence have some interference of other response, the detection result would be wrong. For this reason the SSS detection is very sensitive to ISI and ICI. The detection performance of method two is also worse than method three.
Consequently, method three is shown to have the advantage of lower complexity than method one and better performance than method one and two. In this figure, we also compare the traditional method of using CP repetition to find symbol boundary. The performance is worst.
-5 -4 -3 -2 -1 0 1 2 3 4 5 10-4
10-3 10-2 10-1 100
snr
errorprobability
Ncell ID detection with different channel
suburban macro [SSS symmetry (10 ppm)]
urban macro [SSS symmetry(10 ppm)]
urban micro [SSS symmetry(10 ppm)]
suburban macro [SSS symmetry(10 ppm)]
urban macro [SSS symmetry(10 ppm)]
urban micro [SSS symmetry(10 ppm)]
suburban macro [CP repetition(10 ppm)]
urban macro [CP repetition(10 ppm)]
urban micro [CP repetition(10 ppm)]
Fig. 3.4 Detection performance for NIDcell
3.2 Summary
A cell search procedure is proposed for 3GPP LTE downlink. The cell search procedure utilizes a new timing synchronization method to increase the reliability of cell search results.
The timing synchronization algorithm can effectively combat the impairments caused by CFO with low computation complexity and help achieving better performance in Cell-ID detection.
We verified the idea by link level simulation with 3GPP LTE channel models.
Bibliography
[1] 3GPP TS 36.211 V8.5.0,”Physical Channels and Modulation (Release 8)”, Dec.2008
[2] Farooq Khan, “LTE for 4G broadband”, Cambidge University Press 2009.
[3] M.Sandell, JJ. v.d. Beek and P.O.Borjesson, “Timing and Frequency Synchronization in OFDM System Using Cyclic Prefix”, Proc. IEEE Int. Symp. Synchronization, Essen, Germany, Dec. 1995
[4] Pei-Yun Tsai Hsiang-Wei Chang,’’ A new cell search scheme in 3GPP long term evolution downlink, OFDMA systems’’, Wireless Communications & Signal Processing, 2009. WCSP 2009
[5] Jung-In Kim Jung-Su Han Hee-Jin Roh Hyung-Jin Choi Sch. of Inf. & Commun. Eng., Sungkyunkwan Univ., Suwon, South Korea ,” SSS detection method for initial cell search in 3GPP LTE FDD/TDD dual mode receiver”, Communications and Information Technology, 2009. ISCIT 2009. 9th International Symposium on.
[6] Baum, D.S. Hansen, J. Salo, J. ETH Zurich, Switzerland, ‚‘‘An interim channel model for beyond-3G systems: extending the 3GPP spatial channel model (SCM)’’, Vehicular Technology Conference, 2005. VTC 2005-Spring. 2005 IEEE 61st.
[7] 3GPP TS 36.300 V8.5.0, ”Evolved Universal Terrestrial Radio Access(E-UTRA) and Evolved Universal Terrestrial Radio Access Network(E-UTRAN)”, Dec.2008
(Release 8), Dec.2008
[8] Konstantinos Manolakis, David Manuel Guti´errez Est´evez, Volker Jungnickel, Wen Xu, Christian Drewes, ” A Closed Concept for Synchronization and Cell Search in 3GPP LTE Systems”, Wireless Communications and Networking Conference, 2009. WCNC 2009.
IEEE, 5-8 April 2009
[9] Popovic, Branislav M, Fredrik, “Primary Synchronization Signal in E-UTRA”, Spread Spectrum Techniques and Applications, 2008. ISSSTA ’08. IEEE 10th International Symposium on
[10] S. H. Chen, W. H. He, H. S. Chen, and Y. Lee, “Mode detection, synchronization and channel estimation for DVB-T OFDM receiver” IEEE Global Communication Conference, vol. 5, Nov. 2003, pp. 2416-2420
About the Author
姓 名：蔡耀賢 Yao-Hsien Tsai 出 生 地：台灣省台南市
出生日期：1985. 12. 12
學 歷：1992. 09 ~ 1998. 06 台南縣新市國民小學
學 歷：1998. 09 ~ 2001. 06 台南縣私立黎明高級中學國中部 學 歷：2001. 09 ~ 2004. 06 台南縣私立黎明高級中學高中部 學 歷：2004. 09 ~ 2008. 06 國立成功大學 電機工程學系 學士 學 歷：2008. 09 ~ 2011. 02 國立交通大學 電子研究所系統組
碩士