Performance Comparison of Multiple Input Multiple ...

3 downloads 51 Views 829KB Size Report
Kehinde Odeyemi1* and Erastus Ogunti2. 1Department of Electrical and Electronic Engineering, University of Ibadan, Nigeria. 2Department of Electrical and ...
Copyright © 2014 American Scientific Publishers All rights reserved Printed in the United States of America

Journal of Computational Intelligence and Electronic Systems Vol. 3, 1–8, 2014

Performance Comparison of Multiple Input Multiple Output Techniques for High Data Rate Wireless Communication System Kehinde Odeyemi1 ∗ and Erastus Ogunti2 1

2

Department of Electrical and Electronic Engineering, University of Ibadan, Nigeria Department of Electrical and Electronic Engineering, The Federal University of Technology, Akure, Nigeria

Meeting the demands that are expected from future wireless generation networks poses intriguing challenges for today’s wireless system designers. The demand for higher data rate and better quality of service (QoS) in wireless communications was growing fast in the past few years. Obtaining these requirements becomes challenging for wireless communication systems due to the problems of channel multi-path fading, higher power and bandwidth limitations. One of the most promising solutions to this problem is the Multiple Input Multiple Output (MIMO) system. This paper compared the performance of spatial multiplexing MIMO scheme with beamforming for high data rate wireless communication system. The proposed wireless system was equipped with smart antenna array at both the transmitter and receiver and each of this technique was applied accordingly. The results obtained show that spatial multiplexing technique produced better spectral efficiency than beamforming and improved bit error rate (BER) for the system under the same simulation environment. These two techniques show that the proposed system outperforms the conventional MIMO.

1. INTRODUCTION Multiple-Input Multiple-Output (MIMO) systems that utilize multiple antennas at both transmitter and receiver are able to offer substantial benefits to the rate and reliability.1 2 Various techniques have been suggested for MIMO systems to solve the problems encountered by the wireless communication. This falls into three main approaches: spatial multiplexing scheme,3 spatial diversity scheme4 5 and beamforming. Spatial multiplexing is highly spectral efficient and it creates parallel sub-channels over which separate data streams can be transmitted. The spatial diversity technique is predominantly aimed at improving system reliability because it is used to combat channel fading while beamforming provides a significant increase in performance of wireless communication systems by focusing on the signal energy in a particular direction to increase the received SNR and also reduce interference.6 In a conventional wireless communication system, there is only one antenna at both transmitter and receiver. This system which is called the Single-Input Single-Output (SISO) antenna system suffers a bottleneck in terms of capacity due to the Shannon-Nyquist criterion7 8 and ∗

Author to whom correspondence should be addressed.

J. Comput. Intell. Electron. Syst. 2014, Vol. 3, No. 1

future wireless services demand much higher data bitrate transmission with smaller bit error rate. In order to increase the capacity of the SISO systems to meet such demand, the bandwidth and transmission power have to be increased significantly. As the bandwidth and power are scarce or limited resources, techniques which lead to efficient utilization of these resources are quite necessary in the future wireless systems. There are several ways to increase the bit rate of a wireless communication system, these include choosing a shorter symbol duration T which results into a bandwidth expansion and is typically undesired. This is because; a larger fraction of the frequency spectrum will be occupied, since the required system bandwidth is determined by the symbol rate 1/T . Moreover, wireless channels are typically characterized by multipath signal propagation caused by reflections, scattering, and diffraction,9 a shorter symbol duration can therefore cause an increased degree of inter-symbol interference which may lead to a loss in error performance. Another means of increasing the bit rate is the use of a multicarrier approach10 and multiplex data symbols onto multiple narrow sub-bands, which therefore solved problem of Inter-Symbol Interference (ISI) but still lead to increase in the system bandwidth requirement. The bit rate can also

2326-3008/2014/3/001/008

doi:10.1166/jcies.2014.1061

1

RESEARCH ARTICLE

Keywords: MIMO, Spatial Multiplexing, Beamforming, Smart Antenna, BER, Spectral Efficiency.

Performance Comparison of MIMO Techniques for High Data Rate Wireless Communication System

be enhanced by using higher-order modulation schemes; transmitting more than one bit per data symbol,7 which results in a higher bit rate without bandwidth expansion. However, the error performance will again deteriorate by given the same average transmit power per bit. Finally, one can employ sophisticated source-coding techniques,11 in order to compress the information sequence before transmission but an excessive compression rate will cause signal distortions. In this paper, the MIMO technique is conceived as a promising solution for spectrally efficient transmission technique for wireless communication system. This could increase the system capacity with lower bit error rate substantially without increasing the transmission power and bandwidth.2 Thus, the comparison between the performances of spatial multiplexing MIMO scheme and beamforming technique for high data rate wireless communication system was presented. Spectral efficiency and bit error rate were the two performances metric employed for the comparison of the two techniques.

Odeyemi and Ogunti

in such a way that both the transmitter and the receiver have one or more smart antenna arrays with the transmitter has MT antenna arrays with each array having N antenna elements and there are MR antenna arrays at the receiver with each array having K antenna elements. We assumed that the channel state information (CSI) is only known to the receiver and that the channel has the Rayleigh fading distribution. Spatially uncorrelated complex Gaussian noise was assumed to be added to the faded signal at the receiver. The spacing between the antenna arrays was made to be more than 10 while the antenna element spacing of each antenna array is a half wavelength (/2). When considering beamforming technique for the system as in Figure 1(a), the data was at the transmitter was applied directly to the antenna and beamforming vector was then weighted on the data before transmitted. The vectors W and Z are called the transmit beamforming and receive beamforming vectors, respectively. In case of spatial multiplexing, the data was spitted at the transmitter and each of the data was transmitted to each antenna array as shown in Figure 1(b).

2. SYSTEM MODEL

RESEARCH ARTICLE

In this paper, each of the schemes was applied to the proposed wireless system which was equipped with Smart antenna array at both the transmitter and receiver as illustrated in Figures 1(a) and (b). The system was configured (a)

Beamforming Scheme

3. SYSTEM ANALYSIS Generally, Conventional MIMO systems have a channel impulse response with MT transmit antennas and MR receive antennas given as:12 ⎞ ⎛ h12    h1MT h11 ⎟ ⎜ ⎜ h21 h22    h2MT ⎟ ⎟ ⎜ ⎟ ⎜ H =⎜  (1) ⎟   ⎟ ⎜   ⎟ ⎜ ⎠ ⎝ h MR 1 h MR 2    h MR MT MR ×MT

Where hi j  shows the channel impulse response between the jth transmitter to the ith receiver element and is given as: L  n  − n  (2) hi j  = n=1

(b)

Spatial Multiplexing Scheme

Fig. 1. Proposed wireless system with smart antenna array.

2

Where hi j  multipath channel impulse response, L is the number of paths, n shows the amplitude of the nth path and it obeys independent and identical Rayleigh distribution (i.i.d), · represents the impulse function and n represents the delay of the nth arriving path. Relating this to multiple antenna arrays, case as in Figure 1, this makes the channel matrix to be KMR × NMT a matrix. ⎛ 1 M ⎞ h21    h1 T h1 ⎟ ⎜ 1 M ⎜ h h22    h2 T ⎟ ⎟ ⎜ 2 ⎟ ⎜ (3) H =⎜  ⎟    ⎟ ⎜    ⎟ ⎜ ⎠ ⎝ M h1MR h2MR    hMTR KMR ×NMT

J. Comput. Intell. Electron. Syst. 3, 1–8, 2014

Odeyemi and Ogunti

Performance Comparison of MIMO Techniques for High Data Rate Wireless Communication System

Where h11 is channel fading vector from jth the antenna array at the transmitter to ith antenna array at the receiver. ⎛ j 1 1 1 ⎞ hi 1 hj    hj i K i 2 ⎟ ⎜ j 2 j 2 j 2 ⎟ ⎜h ⎜ i 1 hi 2    hi K ⎟ ⎟ ⎜ (4) hji = ⎜  ⎟  ⎟ ⎜    ⎟ ⎜ ⎠ ⎝ j N N N hi 1 hj    hj i K i 2 K×N

3.1. For Spatial Multiplexing Technique At the transmitter, the transmit signal divided into MT parallel signals S1 n, S2 n     SMT n through the splitter (demultiplexer) and is sent to the different antenna array, thus the transmit signal becomes N × 1 vector as:

3.2. For Beamforming Technique When beamforming is applied to the system, the transmit signal Sn at the transmitter is sent through the antenna array to perform beamforming (transmit beamforming) and the signal becomes: Sˆj n = Wj Sn

(13)

Where Wj is the transmit beamforming weight vector and is given as: Wj = aT  j  ∗ aT  j  = 1 e−j2 dt Sin j /  e−j2 N −1dt Sin j / T

(14) (15)

At the receiver, the received signal at each ith antenna array when noise is added is obtained as:

Where j is the angle of departure (AOD), dt is the distance between the antenna element at jth transmitter antenna array,  is the carrier wavelength, N is the number of element at the jth transmitter antenna array and aT  j ) is the transmit array steering response. After beamforming, sn becomes N × 1 column vector sˆj n. At the receiver side, the receive signal at ith array element is denoted as vector Xn and is given as:

ri n = Xj n + i n

(7)

Xj n = HWj Sn

(8)

The receive beamforming is then weighted on Xn and the output signal after beamforming at the ith receive element antenna array is given as:

sn = S1 n S2 n     SMT n T

(5)

Thus, transmit signal after transmission becomes: Xj n = HSj n

(6)

According to Eq. (6), received signal becomes: ri n = HSj n + i n

ri n =

NMT 

hji nSj n + i n

(9)

M

ri n = h11 nS1 n + · · · + hi T nSMT n + i n (10)

⎡ ⎢ ⎢ ⎢ ⎢ +⎢ ⎢ ⎢ ⎣

1

M

h21

   h1 T

h22



M h2 T

 

 

h2MR

R

MT 

ZiH Xj n + gi n

(17)

ZiH H Sˆj n + ZiH gi n

(18)

j=1

ri t =

MT  j=1

j=1

In Matrix form: ⎤ ⎡ 1 ⎡ h1 r1 ⎥ ⎢ 1 ⎢ ⎢ r2 ⎥ ⎢ h ⎥ ⎢ 2 ⎢ ⎥ ⎢ ⎢ ⎢  ⎥ = ⎢  ⎢  ⎥ ⎢  ⎥ ⎢ ⎢ ⎦ ⎣ ⎣ rMR h1M

ri t =



M hMTR

⎤⎡ ⎥⎢ ⎥⎢ ⎥⎢ ⎥⎢ ⎥⎢ ⎥⎢ ⎥⎢ ⎦⎣

S1 S2   S MT



Zi = aR i

⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦

aR  i  = 1 e−j2 dr Sin j /  e−j2 K−1dr Sin i /

(11)

J. Comput. Intell. Electron. Syst. 3, 1–8, 2014

(20)

ri n =

MT 

ZiH H Sˆj n + i n

(21)

j=1

Then, the received signal can be recovered as: r = HS +

(19)

Where i is the AOA (Angle of Arrival), dr is the distance between the antenna element at ith transmitter array,  is the carrier wavelength, K is the number of element at the ith receiver antenna array and aR  i  is the receive array steering response.



⎥ 2 ⎥ ⎥ ⎥  ⎥  ⎥ ⎥ ⎦ MR

Where Zi is the received beamforming weight vector and is given as:

(12)

Where i n spatially uncorrelated complex Gaussian noise with entry is distributed as ∼ CN0 No  and is given as: i n = ZiH gi n (22) 3

RESEARCH ARTICLE

Where i n spatially uncorrelated complex Gaussian noise with entry is distributed as ∼ CN (0 No ). Therefore, according to Eq. (4), the received signal can be further expressed as:

(16)

Performance Comparison of MIMO Techniques for High Data Rate Wireless Communication System

Since, Sˆj n = Wj Sj n we then substitute for Sˆj n in the Eq. (21).

(23)

In matrix form: ⎤ ⎡ r1 Z H h 1 W1 ⎥ ⎢ 1 1 ⎢ ⎢ r2 ⎥ ⎢ H 1 ⎥ ⎢ Z 2 h 2 W2 ⎢ ⎥ ⎢ ⎢  ⎥ = ⎢  ⎢  ⎥ ⎢ ⎥ ⎢ ⎢  ⎦ ⎣ ⎣ rMT Z H h1 W ⎡

Z1H h21 W2



Z2H h22 W2



 



H ZM h 2 W2 R MR ⎡ ⎤ ⎤ ⎡ S1 n 1 ⎢ ⎥ ⎥ ⎢ ⎢ S2 n ⎥ ⎢ 2 ⎥ ⎢ ⎥ ⎥ ⎢ ⎢ ⎥ ⎥ ⎢ ×⎢  ⎥+⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢ ⎥ ⎥ ⎢ ⎣ ⎦ ⎦ ⎣ MT SMT n MR

MR

1

M

Z1H h1 T WMT

RESEARCH ARTICLE

 is an effective Where H defined as: ⎡ H 1 Z 1 h 1 W1 Z1H h21 W2 ⎢ H 1 ⎢ Z h W2 Z2H h22 W2 ⎢ 2 2 ⎢ ˜ H =⎢   ⎢   ⎣ H ZM h 1 W1 R MR

H ZM h 2 W2 R MR



⎥ M Z2H h2 T WMT ⎥ ⎥ ⎥ ⎥ ⎥  ⎦ M

H  ZM h T WM T R MR

(24)

r = H˜ S +

(25)

channel matrix and is    

M

Z1H h1 T WMT



hi  t = j

⎥ M Z2H h2 T WMT ⎥ ⎥ ⎥ (26) ⎥ ⎥  ⎦ M H ZM h T WM T R MR

aR i ij t aT j T  − n 

√ √ Since aR i  = K and aT j  = N Where · is the Euclidean Vector Norm, thus the effeci j can be approximately tive channel fading element H obtained as: H˜ i j = KN · i j (30) Therefore, the corresponding be formed as: ⎡ 1 1 1 2 ⎢ ⎢ 1 1 1 1 ⎢ ⎢  = KN · ⎢  H  ⎢   ⎢ ⎣  MR  1  MR  2

entire channel matrix can 

1 MT



⎥ 2 MT ⎥ ⎥ ⎥ ⎥   ⎥ ⎥ ⎦     MR  MT 

(31)

Since the element of i j and hi j has the same distribution (i.i.d), then the effective channel matrix in Eq. (26) becomes: H˜ = KN · H (32)

rn = H˜ S +

Where the channel fading vector hji is a matrix of K ×N according to Eq. (4), i j t is the multipath fading components, coupling the first element of the jth antenna array at the transmitter to the ith antenna array at the receiver and it obeys independent and identically Rayleigh-distribution (i.i.d). Since the channel is assumed to be flat, Eq. (27) becomes: hji t = aR i i j t aT j T (28)

(33)

3.3. Detection of the Transmitted Signal by Each Technique In the case of spatial multiplexing, to detect the transmit signal Sn, Zero forcing (ZF) and Minimum Mean Square Error (MMSE) detection algorithm were considered and the receiver was designed using the linear matrix G according to certain algorithm. Thus, according to Eq. (12) the receive signal was obtained as: r = HS +

(34)

Then the detected signal is Sˆ = Gr

(35)

Sˆ = GHS + G

(36)

(27)

i=0

4

(29)

Thus, the receive signal is:

This shows that the channel matrix consists of MIMO channel fading and information concerning AOD and AOA. As a result, H˜ is then transformed from a KMR × NMT channel matrix to a MR ×MT channel matrix H. Due to the strong spatial correlation existing in each antenna array, according to the fading of the first element for each antenna array, the entire steering response of the antenna array is:13 L−1 

Then, the effective channel fading element H˜ i j can be roughly obtained as: H˜ i j = aR i H aR i i j aT j T aT j ∗

M

ri t = ZiH h1i nW1 S1 n + · · · + ZiH hi T nWMT SMT n + i n

Odeyemi and Ogunti

For ZF detection algorithm: GZF = H H H−1 H H

(37)

For MMSE detection algorithm:  GMMSE = H H + H

IKMR o

−1 HH

(38)

J. Comput. Intell. Electron. Syst. 3, 1–8, 2014

Odeyemi and Ogunti

Performance Comparison of MIMO Techniques for High Data Rate Wireless Communication System

If the system uses ZF or MMSE detection algorithm, the effective detection SNR of the qth data streams with linear ZF or MMSE equalizer at the receiver is expressed as:2 14 qZF =

o H H H−1 q q

qMMSE =

H H H

q = 1 2     MT



(39)

o − 1 + IKMR /o −1 q q q = 1 2     MT

Pq KNo

Po MT N

(42)

Po MT KNNo

qMMSE

q = 1 2     MT

Po KNo

(51)

Substitute for average SNR in Eq. (49), the effective SNR is obtained as:  Bf =

KN 2 Po max No

(52)

Thus, the system capacity for wireless system is given by:1 2 NMT  C= log2 1 + q  (53)

(44)

The system capacity for the proposed system when spatial multiplexing is applied is obtained as:   NMT  P (54) log2 1 + H −1 o CZF = H Hk k MT KNNo k=1   NMt  CMMSE = log2 Po  · MT KNNo q=1

Po − 1 = MT KNNo H H H + IKMR /o −1 q q

q = 1 2     MT

(45)

In case of Beamforming scheme, the received signal is obtained according to Eq. (33) and the received signal is then detected as: ˆ Sn = 1/2 max KNSn + i n

(47)

Where o is the average SNR. According to the Eq. (41), the effective SNR becomes:  Bf = o K 2 N 2 H H H J. Comput. Intell. Electron. Syst. 3, 1–8, 2014

The system capacity for the proposed system when the beamforming technique is applied is obtained as:   KN 2 Po max CBf = log2 1 + (56) No

(46)

Where max is the maximum positive eigenvalue of the transmitted signal. Thus, the effective detection SNR for the beamforming technique is obtained as:  Bf = o H˜ H H˜ 

    MT KNNo IKMT −1 −1 H (55) × H H + Po q q

(48)

4. SIMULATION RESULTS This paper provides the simulation results of the spatial multiplexing and beamforming technique for the proposed high data rate wireless communication system. The performance metrics in terms of spectral efficiency and BER of Conventional MIMO are given to compare with the two techniques. The transmitter and the receiver are assumed to have 2 smart antenna arrays at both ends and we examine N and K to be equal to 2, 4 and 8 elements in each array. The spacing between antenna arrays is larger than 10, while the spacing between antenna elements is /2. 5

RESEARCH ARTICLE

Po  H H H −1 q q KNNo MT

(50)

q=1

(43)

Where Po is the total transmitted power. By substituting for o in the Eqs. (39) and (40), then the effective SNR qZF =

Pk KNo

Since the total transmitted power is allocated, then PK = Po , the average SNR becomes: o =

Then, the Eq. (41) becomes: o =

o =

(41)

Where Pq is the transmit power at each jth transmit antenna array. Since the CSI is known at the receive only, then, the transmit power is equally allocated across the transmit antenna array and is obtained as, Pq =

In order to accomplish the use of only the best channel, transmitter chose to allocate all its power to the best channel and no power to the remaining links, thus, the average SNR at the receiver is obtained as:

(40)

Where o is the average SNR at each receiver ith antenna array and is obtained as: o =

Since beamforming utilizes only the strongest subchannel and allocate all power to it, effective is then becomes: (49)  Bf = K 2 N 2 o max

16-QAM modulations are used to modulate the symbols at the transmitter. The channel has the Rayleigh fading distribution, and spatially uncorrelated complex Gaussian noise is added to the faded signal at the receiver. In the case of beamforming technique, the angle spread at each of the transmitter antenna array is 30 degrees and 70 degrees at the receiver side. ZF and MMSE detection are adopted at the receiver when spatial multiplexing technique was applied. These detections are further enhanced by linear nulling and successive interference cancellation algorithm called Vertical-Bell-Laboratory Layered Space-Time Architecture (V-BLAST) to achieve better performance. Spectral efficiency is the capacity of the system which shows the amount of maximum information that can be sent by a wireless communication system. Conventionally, this can be increased by the factor of minMR  MT  without using additional transmits power or spectral bandwidth. This paper shows that maximum spectral efficiency is achievable by increasing the number of elements in each antenna array at both ends of the radio link. Figure 2 shows the spectral efficiency performance of the proposed system when spatial multiplexing (SM) and beamforming (BF) techniques are applied. This result indicates that the spatial multiplexing technique has the best performance with average spectral efficiency of 21.73 b/s/Hz and 14.24 b/s/Hz for MMSE and ZF detection respectively than beamforming scheme with the average spectral efficiency of 11.62 b/s/Hz. The result further shows that the Conventional MIMO system with MT = 2 and MR = 2 has an average spectral efficiency of 4.38 b/s/Hz when MMSE detection was used and 3.54 b/s/Hz for ZF detection which is obviously shown

Odeyemi and Ogunti

that the Conventional MIMO system has a poor capacity performance compared to the other scheme. Thus, this proves that spatial multiplexing technique can be used to achieve higher data rates than beamforming technique. Figure 3 shows the average error produce by the proposed system when MT = 2, MR = 2, K = 2 and N = 2. It is clear that MMSE detection performed better than the ZF detection in the entire scheme. The result shows that spatial multiplexing scheme performs better at lower SNR while beamforming scheme perform better at high SNR. Thus, the spatial multiplexing scheme will provide an average error of 0.0052 for MMSE detection and 0.0073 for ZF detection, but the beamforming technique will provide a high average error of 0.0326. It is also clear from the result that the conventional MIMO system has a poorer performance than the two techniques. The performance of the system was further enhanced by V-BLAST algorithm as shown in the result also. V-BLAST improves the performance of MMSE and ZF detection in the entire schemes. The spatial multiplexing outperforms other schemes with MMSE detection having a better average BER of 0.0029 and ZF detection with an average BER of 0.0039 compare with the beamforming. This proves that V-BLAST has improved the system performance in terms of error reduction by 38.9% and 46.6% for MMSE and ZF detection respectively. Figure 4 shows that the spectral efficiency of the system increases linearly with the number of element in each antenna array. With the MR and MT antenna arrays remaining constant and the elements K and N increased from two to four, the simulation result shows that the capacity performance of spatial multiplexing is better and improved

35 100

Conventional MIMO (ZF) Conventional MIMO (MMSE) 30

SM Scheme (MMSE) 10–1

SM Scheme (ZF) Conventional BF

25

BF Scheme 10–2 20

BER

Spectral Efficiency [bps/Hz]

RESEARCH ARTICLE

Performance Comparison of MIMO Techniques for High Data Rate Wireless Communication System

15

10–3 Conventional MIMO (ZF) Conventional MIMO (MMSE) 10–4

10

SM Scheme (ZF) SM Scheme (ZF-V-BLAST) SM Scheme (MMSE)

5

10–5

SM Scheme (MMSE-V-BLAST) Conventional BF

0

0

2

4

6

8

10

12

14

16

18

20

SNR [dB]

BF Scheme 10–6

0

2

4

6

8

10

12

14

16

18

20

SNR [dB] Fig. 2. Spectral efficiency for the proposed wireless MIMO system when MT = 2, MR = 2, K = 2 and N =2.

6

Fig. 3. BER performance when MT = 2, MR = 2, N = 2 and K = 2.

J. Comput. Intell. Electron. Syst. 3, 1–8, 2014

Odeyemi and Ogunti

Performance Comparison of MIMO Techniques for High Data Rate Wireless Communication System

90

100 SM Scheme (MMSE)

80

SM Scheme (ZF) BF Scheme

10–1

10–2

60 50

BER

Spectral Efficiency [bps/Hz]

70

10–3

40 10–4

30

SM Scheme (ZF) SM Scheme (ZF-V-BLAST)

20

10–5

SM Scheme (MMSE) SM Scheme (MMSE-V-BLAST)

10

BF Scheme 0

10–6 0

2

4

6

8

10

12

14

16

18

20

0

2

4

SNR [dB] Fig. 4. Spectral efficiency for the proposed wireless MIMO system when MT = 2, MR = 2, K = 4 and N = 4.

220 SM Scheme (MMSE) 200

SM Scheme (ZF) BF Scheme

Spectral Efficiency [bps/Hz]

180 160

8

10 12 SNR [dB]

14

16

18

20

BER performance when MT = 2, MR = 2, N = 4 and K = 4.

Figure 5. The spatial multiplexing produces an average spectral efficiency of 159.78 b/s/Hz and 72.87 b/s/Hz for MMSE and ZF detection respectively than beamforming technique with average spectral efficiency of 23.72. Thus, this proves that the capacity of MIMO system can be enhanced by increasing the number of antenna element in each array at the transmitter and receiver. Similarly, Figure 6 shows the BER performance of the system; when N and K were increased from 2 to 4 whiles the antenna array MR and MT remain constant. The result proves that the increase in the antenna array element N and K produced better system performance. It was shown that SM scheme has a better BER performance with average error of 0.000991 for MMSE detection and 0.0014 for ZF detection than BF scheme with 0.0219 error. This indicates a significant improvement in BER compare to when N and K equal to 2 and when the system was enhanced with V-BLAST as in Figure 3.

140

5. CONCLUSION

120 100 80 60 40 20

0

2

4

6

8

10

12

14

16

18

20

SNR [dB] Fig. 5. Spectral efficiency for the proposed wireless MIMO system when MT = 2, MR = 2, K = 8 and N = 8.

J. Comput. Intell. Electron. Syst. 3, 1–8, 2014

The performance comparison between beamforming and spatial multiplexing technique was carried out for a high data rate wireless communication system. The MMSE and ZF MIMO detection algorithm was employed at the receiver and was further enhanced by V-BLAST. Spectral efficiency and BER were the two performance metrics used to determine the efficiency of the techniques. The simulation results show that the spatial multiplexing outperforms the beamforming scheme in spectral efficiency and at lower SNR in BER performance. Thus, this makes the proposed system to have better performance than the Conventional MIMO system for both techniques. It was 7

RESEARCH ARTICLE

beamforming technique. SM will provide an average spectral efficiency 30.57 b/s/Hz when MMSE detection is used and 60.47 b/s/Hz for ZF detection, but the beamforming technique produce an average spectral efficiency of 17.78 b/s/Hz. This shows that spatial multiplexing scheme has an increment of 38.74 b/s/Hz for MMSE detection and 16.33 b/s/Hz for ZF detection when the antenna array element was increased. This was further proved by increasing K and N to eight elements as the result was given in

Fig. 6.

6

Performance Comparison of MIMO Techniques for High Data Rate Wireless Communication System

found that the higher the antenna array element the higher the system spectral efficiency and the better the system reliability for both techniques. The results also show that the MMSE detection has a better performance than the ZF detection for the spatial multiplexing technique and even when enhanced by V-BLAST.

References 1. I. E. Telatar, Capacity of multi-antenna gaussian channels. Europ. Trans. Telecom. 10 (1999). 2. G. Fochini and M. Gans, On limits of wireless communications in a fading environment when using (1998). 3. G. G. Raleigh and J. M. Cioffi, Spatio-temporal coding for wireless communication. IEEE Trans. Commun. 46, 357 (1998). 4. V. Tarokh, H. Jafarkhani, and A. R. Calderbank, Space-time block codes from orthogonal designs. IEEE Trans. Inf. Theory 45, 1456 (1999). 5. S. M. Alamouti, A simple transmit diversity technique for wireless communications. IEEE Journal on Selected Areas in Communications 16 (1998).

Odeyemi and Ogunti

6. M. Jan, S. Robert, L. Lutz, H. G. Wolfgang, and A. H. Peter, Multiple-antenna techniques for wireless communications– A comprehensive literature survey. IEEE Communications Surveys and Tutorials 11, 87 (2009). 7. J. G. Proakis, Digital Communications, 4th edn., McGraw-Hill, New York (2001). 8. C. E. Shannon, A mathematical theory of communication–Part I and II. Bell Syst. Tech. J. 27, 379 (1948). 9. R. Steele, Mobile Radio Communications, IEEE Press, New York (1994). 10. L. Hanzo, M. Muenster, B. J. Choi, and T. Kelle, OFDM and MC-CDMA for broadband multi-user communications, WLANs and broadcasting, John Wiley & Sons/IEEE Press, Chichester (2003). 11. T. M. Cover and J. A. Thomas, Elements of Information Theory, John Wiley & Sons, New York (1991). 12. D. Gore, R. W. Health, Jr, and A. Paularj, Transmit selection in spatial multiplexing system. IEEE Communication Letter 6, 491 (2002). 13. M. S. Akbar, Deconstructing multiantenna fading channels. IEEE Transactions on Signal Processing 50 (2002). 14. L. M. Aris, K. Raj Kumar, and C. Giuseppe, Performance of MMSE MIMO receivers: A large N analysis for correlated channels. IEEE Vehicular Technology Conference (2009), pp. 1–5.

RESEARCH ARTICLE

Received: 4 March 2014. Accepted: 4 March 2014.

8

J. Comput. Intell. Electron. Syst. 3, 1–8, 2014