Increasing Channel Capacity on MIMO System Employing Adaptive Pattern/Polarization Reconfigurable Antenna 1* 2 2 1 1 Helen K. Pan , Gregory Huff , Tyrone Roach , Yorgos Palaskas , Stefano. Pellerano , Parmoon 1 1 1 1 Seddighrad , Vijay K. Nair , Debabani Choudhury , Boyd Bangerter and Jennifer T. Bernhard2
1
2
Commnication Technology Lab, Intel, 2111 NE 25th Ave., Hillsboro, OR 97124, USA University of Illinois at Urbana-Champaign, 1406 W. Green St., Urbana, IL 61801, USA Email:
[email protected] INTRODUCTION
Multiple Input-Multiple Output (MIMO) is a promising technology for improving the throughput, range and reliability in wireless networks. MIMO uses multiple antennas at the transmitter and/or the receiver to increase the data rate or the range of the wireless link. This is done without using extra bandwidth or extra power, but rather by employing digital signal processing power, which is almost free in today’s deeply scaled CMOS technologies. MIMO spatial multiplexing establishes multiple non-interacting “data-pipes” between transmitter and receiver to achieve higher capacity [1-4]. Achieving good MIMO performance requires uncorrelated paths between transmit and receive antennas. Radiation pattern/polarization reconfigurable antennas bring flexibility to achieve better spatial channel forming (lower correlation between paths) and increase throughput. Many researchers have examined the antenna pattern/polarization diversity for MIMO system performance [5-10]. In this paper, a pattern/polarization reconfigurable single turn square spiral microstrip antenna is designed to operate in the 5 GHz indoor band and tested with Inteldeveloped CMOS MIMO transceivers. Different antenna pattern/polarization configurations can be adaptively selected based on MIMO performance such as channel capacity in LOS and various NLOS environments. PATTERN/POLARIZATION RECONFIGURABLE ANTENNA DESIGN The pattern/polarization reconfigurable antenna is a single antenna incorporating RF switches to adaptively tune to different pattern configurations [11-13]. The pattern reconfigurable antenna has two configurations, Broadside (B) and Endfire (E) as shown in Fig. 1. Hard wire connections are implemented to represent the RF switches’ closed state and the empty slot is used to represent the RF switches’ open state in modeling and measurement.
WLAN Band Pattern Reconfigurable Microstrip Patch Antenna: VSWR P 5.00 4.50 4.00 3.50 3.00 2.50 2.00 1.50 1.00
5.0
5.1
5.2
5.3
5.4
5.5
Frequency [GHz]
(a) (b) (c) Fig1. Pattern/polarization reconfigurable antenna configurations (a) broadside, (b) endfire and (c) VSWR plot for both configurations. Broadside Endfire
In broadside configuration, the polarization is linear and aligned with the φ=450 plane. In endfire configuration, the polarization is vertical (Eθ). The measurement results of the pattern/polarization antenna in VSWR and radiation pattern are shown in Figs. 1 and 2. For a 2X2 MIMO system, there are four different combinations of antenna configurations (BB, BE, EB, EE) at both transmitter (TX) and receiver (RX) sides. We chose 9 out of 16 different antenna system
1-4244-0878-4/07/$20.00 ©2007 IEEE
481
configuration combinations, as listed in Table 1, in all the MIMO performance tests due to the symmetry environments on two sides. 0°
0°
0°
5.0
5.0
5.0
5.0
-2.0
-2.0
-2.0
-2.0
-23.0
-23.0
-30.0
-30.0
-30.0
180°
180°
Eθ Eφ
-90°
-23.0
-30.0
-90°
-23.0 -90°
-9.0 -16.0
90°
-9.0 -16.0
90°
-9.0 -16.0
90°
-9.0 -16.0
90°
180°
-90°
0°
180°
(a) (b) (c) (d) Fig. 2.The measrued radiation pattern of pattern/polariztion reconfigurable antenna at 5.28GHz. (a)broadside, phi=0 (xz) plane, (b) broadside, phi=90 (yz) plane, (c)endfire, phi=0 (xz) plane, (d) endfire, phi=90 (yz) plane TX RX
1 BB BB
2 BB BE
3 BE BB
4 BE BE
5 BE EE
6 EE BE
7 EE EE
8 BB EE
9 EE BB
Table 1. Different antenna configurations for 2X2 MIMO system MIMO TEST SETUP A 2x2 MIMO transceiver chip was built using Intel’s 90 nm CMOS technology operating at the 5 GHz 802.11a WLAN band [14-15]. The transceiver with the pattern/polarization antenna testbed setup is shown in Fig. 3.
Radio-TX
Radio-RX A/D
DUT
A/D VSA
VSG
5GHz
10MHz Low-IF MIMO BB data
C
MIMO 2x2 modem
Matlab code
Figure 3: The test setup used to characterize pattern/polarization antenna with 5 GHz transceiver for MIMO performance. Matlab code generates two random MIMO OFDM signals using 64QAM and SFC (Spatial Frequency coding) and loads to two Vector Signal Generators (VSGs). Each packet is 28.15 us long with 5630 samples. Each time sample has a duration of 50 ns (T=1/20MHz=50 ns). Two VSGs are synchronized and repeated transmitting the same packet. The transmitting signal is at 5.28GHz with 20 MHz bandwidth. The output power level can be tuned individually for measurement flexibility at the transmitter. On the receiver side, the MIMO transceiver is implemented to receive MIMO signals, using the vector signal analyzer (VSA) as an A/D
482
converter. The IF outputs of the MIMO digitized radio signals are then passed through timing and frequency synchronization routines, MMSE channel equalization, and Viterbi decoding. MIMO PERFORMANCE ANALYSIS On-the-air tests were performed using the pattern/polarization reconfigurable antennas with 5 GHz 2x2 MIMO CMOS transceivers as an actual MIMO wireless link to verify the functionality of the MIMO system in a real-life environment. Matlab codes extract the received RF signals from two receiving antenna elements and post process real time MIMO channel measured data. For a 2x2 MIMO system (Ntx=Mrx=2), the received signals can be written as:
R1 h11 h12 T1 n1 (1), R = h ⋅ + ⇔ R = H ⋅T + n 2 21 h22 T2 n 2 where R and T are the received and transmitted signal vectors, respectively, n is the receiver noise vector, and H is a 2x2 channel matrix describing the environment between the transmit and receive antennas. Based on channel transfer matrix H , the Shannon capacity for the MIMO system channel is given by the equation as:
SNR C = log 2 det I + ∗ HH H M
(2),
H
Where H is the Hermitian of the matrix, M is the number of receivers and SNR is the estimated channel signal noise ratio. The normalized Shannon capacity Cnorm of the channel can be calculated based on any arbitrary SNR by using the normalized channel transfer matrix Hnorm=H/E[|Hij|2] in Equation (2). The normalized channel capacity is calculated using the measured estimation of the channel transfer matrix for 200 packets with SNR=15dB.
Normalized 2x2 MIMO channel capacity in bit/s/Hz
10 LOS NLOS1 NLOS2 maximum capacity for 2X2 MIMO with SNR 15dB
9.5
9
8.5
8
7.5
7
1
2
3 4 5 6 7 TX/RX antenna configurations based on Talbe 1
8
9
Figure 4: Normalized channel capacity for pattern/polarization antenna with 2X2 5GHz MIMO transceiver for SNR=15dB. As shown in Fig. 4, three different environments were tested with a 10 foot distance between TX and RX antenna systems. Non-Line-of-sight (NLOS) scenario 1 is using a chair and books between TX and RX antenna paths. NLOS scenario 2 has a chair, metal desk and book between TX and RX antennas. The theoretical maximum channel capacity for a 2X2 MIMO system with an average SNR of 15dB is equal to C2x2Max =2log2(1+SNR)=10.0556 bits/s/Hz. Based on the
483
channel capacity testing results, we observed different environments favor different antenna configurations. Moreover, the observed variation in capacity is about 2 bits/s/Hz for a 2X2 MIMO system with SNR of 15dB. The best channel capacity is achieved at 9.43 bits/s/Hz. Different antenna configurations generated different channel transfer matrices. Adaptively tuning the pattern/polarization antenna configuration can help establish better orthogonal spatial channels, optimize the channel capacity and increase throughput. CONCLUSION The fully-integrated 2X2 5GHz MIMO 90nm CMOS radio transceiver was tested with the radiation pattern/polarization single turn square spiral microstrip reconfigurable antennas and the advantages of the scheme were experimentally demonstrated. As the experimental results demonstrated that different environments between transmitter and receiver antenna MIMO system requires different antenna configurations to achieve lower correlation between paths and increase the throughput. ACKNOWLEDGMENTS The authors would like to thank Ralph E. Bishop, Songnan Yang, Qiang Li, S.-Y. Suh, A. Ravi, M. A. Elmala, G. Banerjee, R. B. Nicholls, S. Ling, S. S. Taylor, K. Soumyanath, A. Crouch, and K. Kahn for their contributions. REFERENCES [1] J. H.Winters, ‘‘On the capacity of radio communication systems with diversity in a Rayleigh fading environment,’’ IEEE J. Sel. Areas Commun., vol. SAC-5, no. 5, pp. 871---878, Jun. 1987. [2] G. G. Raleigh and J. M. Cioffi, ‘‘Spatio-temporal coding for wireless communication,’’ IEEE Trans. Commun., vol. 46, pp. 357---366, Mar. 1998. [3] M. Jankiraman, “Space-time codes and MIMO systems”, Artech House, 2004. [4] J. W. M. Rogers et al, “A Fully Integrated Multi-Band MIMO WLAN Transceiver RFIC”, VLSI Circ. Symposium, pp. 290 – 293, June 2005, Kyoto. [5] C. B. Dietrich et al, ‘‘Spatial, polarization, and pattern diversity for wireless handheld terminals,’’ IEEE Trans. Antennas Propag., vol. 49, no. 9, pp. 1271---1281, Sep. 2001. [6] M. R. Andrews et al, ‘‘Tripling the capacity of wireless communications using electromagnetic polarization,’’ Nature, vol. 409, no. 6818, pp. 316---318, Jan. 2001. [7] P. Kyritsi, et al, ‘‘Effect of antenna polarization on the capacity of a multiple element system in an indoor environment,’’ IEEE J. Sel. Areas Commun., vol. 20, no. 6, pp. 1227---1239, Aug. 2002. [8] R. Nabar et al, “Performance of Multiantenna Signaling Techniques in the Presence of Polarization Diversity”, IEEE Trans. on Signal Processing, Vol. 50, pp. 2553-2562, Oct. 2002. [9] T. Svantesson, ‘‘On capacity and correlation of multi-antenna systems employing multiple polarizations,’’ in IEEE Int. Antennas Propagation Symp. Digest, pp. 202---205, Jun. 2002. [10] Michael A. Jensen and Jon W. Wallace, “MIMO wireless channel modeling and experimental characterization,” Space-time Processing for MIMO Communications, 2005. [11] G. H. Huff, J. Feng, S. Zhang and J. T. Bernhard, “A novel radiation pattern and frequency reconfigurable single turn square spiral microstrip antenna,” IEEE Microwave and Wireless Components Lett., Vol. 13, No. 2, pp. 57-59, February 2003. [12] J. T. Bernhard, G. H. Huff, J. Feng, S. Zhang, and G. Cung, “Reconfigurable portable antenna systems for high-speed wireless communication”, IEEE Topical Conference on Wireless Communication Technology, pp. 82 – 83, Oct. 2003. [13] G.H. Huff and J.T. Bernhard, “Integration of packaged RF MEMS switches with radiation pattern reconfigurable square spiral microstrip antennas”, IEEE Transactions on Antennas and Propagation, Volume 54, Issue 2, Part 1, pp. 464 – 469, Feb. 2006. [14] Y. Palaskas et al, “A 5GHz 108Mb/s 2x2 MIMO Transceiver with Fully Integrated 20.5dBm P1dB Power Amplifiers in 90nm CMOS”, Journal of Solid State Circuits, December 2006. [15] Y. Palaskas et al, “A 5GHz, 20dBm Power Amplifier with Digitally-Assisted AM-PM Correction in a 90nm CMOS Process”, Journal of Solid State Circuits, August 2006.
484
