Lo2520082015

554 views
524 views

Published on

IJERA (International journal of Engineering Research and Applications) is International online, ... peer reviewed journal. For more detail or submit your article, please visit www.ijera.com

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
554
On SlideShare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
4
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Lo2520082015

  1. 1. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015 BER Comparison of MIMO Systems using Equalization Techniques in Rayleigh Flat Fading Channel Gurpreet Singh* and Pardeep Sharma** * (Department of Electronics and Communication, Shaheed Bhagat Singh State Technical Campus Moga Road (NH-95), Ferozepur-152004) ** (Department of Electronics and Communication, Shaheed Bhagat Singh State Technical Campus Moga Road (NH-95), Ferozepur-152004)ABSTRACT Multiple Input Multiple Output (MIMO) it is very difficult task of designing with high datatechnology is one of the most promising wireless rate and highly reliable wireless communicationtechnologies that can efficiently boost the data systems due to limited power and frequencytransmission rate, improve system coverage, and bandwidth. MIMO [2] technology constitutes aenhance link reliability. By employing multiple breakthrough in wireless communication systemantennas at transmitter and receiver sides, design. The technology offers a number of benefitsMIMO techniques enable a new dimension– the that help meet the challenges posed by both thespatial dimension – that can be utilized in impairments in the wireless channel as well asdifferent ways to combat the impairments of resource constraints. The benefits of MIMOwireless channels. While using MIMO technology that help achieve significant performancetechniques, there is intersymbol interference gains are array gain, spatial diversity gain, spatialpresent between the symbols. This paper will multiplexing gain and interference reduction.focus on Equalization techniques, for Rayleigh Due to multi-path fading, there is aFlat fading. Equalization is a well known distortion of a signal, where one symbolstechnique for combating intersymbol interference with subsequent symbols, is known asinterference. In this paper, we will discuss Intersymbol interference (ISI). Therefore,different types of equalizer like ZF, MMSE, DFE Equalization ideas to remove intersymboland ML. In this paper, we will compare different interference (ISI) can be used. A linear equalizerequalizers with different modulations techniques usually tries to separate the symbols withoutlike BPSK,QPSK,16-QAM.We will find out enhancing the noise. The equalization methods thatwhich modulation techniques is better than we consider in the design of MIMO receiver are ZF,others and then we will compare ZF,MMSE, ML MMSE, DFE and ML.with the best modulation technique.Furthermore, we will conclude which type of 2. MIMO SYSTEM MODELequalizer will provide us better BER We consider single user MIMOperformance. communication system [2] with 2 antennas at the transmitter and 2 antennas at the receiver. ConsiderKeywords- Decision Feedback Equalization that we have a transmission sequence is {x1,(DFE), Interference Intersymbol (ISI), Multiple x2,...........,xn}. In normal transmission, we send x1 inInput Multiple Outpue (MIMO), Minimum Mean the first time slot, x2 in the second time slot and xn inSquare Error (MMSE) and Zero Forcing (ZF) the nth time slot. Now we have two transmit antennas, we may groups the symbols into groups of1. INTRODUCTION two. In the first time slot, send x1 and x2 from the The use of multiple antennas at the first and second antenna. In the second time slot,transmitter and receiver in wireless systems is send x3 and x4 from the first and second antenna andknown as MIMO. Communication [1] is wireless in next time slot x5 and x6 and so on.channels are impaired predominantly by multi-path Let us consider for 2 x 2 MIMO Systemfading. Multipath is the arrival of the transmittedsignal at an intended receiver through differingangles and/or differing time delays and/or differingfrequency shifts due to scattering of electromagneticwaves in the environment. The received signalpower fluctuates in space and/or frequency and/ortime through the random variations of the signals.This random fluctuation in the signal level is known,as fading. Fading can affect the quality andreliability of wireless communication. Additionally 2008 | P a g e
  2. 2. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015 1 𝑇 (3) ℎ11 𝐻~𝑅 2 𝐻𝑤 𝑅𝑥 12 𝑅 𝑇𝑥 Rx where x ~ y denotes that x and y are identical in Tx 1 distribution, RRx and TTx are the normal correlation 1 ℎ21 ℎ12 distribution matrices at the Rx and transmitter (Tx) respectively, and H W  C N M contains i.i.d Tx Rx complex Gaussian entries with zero mean and unitTransmitter ℎ22 Receiver variance. 2 2 For a system with 𝑀 𝑇 transmit antennas and 𝑀 𝑅 receive antennas, the MIMO channel at a given time instant may be represented as a 𝑀 𝑅 × 𝑀 𝑇 matrix H1,1 H1,2 ⋯ H1,MT (5) H2,1 H2,2 … H2,MT H=Figure:1. 2 x 2 MIMO system model ⋮ ⋮ ⋱ ⋮ HMR ,1 HMR,2 ⋯ HMR ,M TThe received signal on the first receive antenna is r1 = h11 s1 + h12 s2 + n1 (1) 3. FADING Fading is used to describe the rapidThe received signal on the second receive antenna is fluctuations of the amplitudes, phases or multipath delays of a radio signal over a short period of time or r2 = h21 s1 + h22 s2 + n2 (2) travel distance, so that large scale path loss effect may be ignored [5]. Fading, or equivalently small- scale fading, is caused by interference between two where, 𝑦1 𝑎𝑛𝑑 𝑦2 are the received symbol or more versions of the transmitted signal whichon the first and second antenna respectively,ℎ11 is arrive at the receiver at slightly different times.the channel from 1 𝑠𝑡 transmit antenna to 1 𝑠𝑡 receive These signals, called multipath waves, combine atantenna,ℎ12 is the channel from 2 𝑛𝑑 transmit the receiver antenna and the corresponding matchedantenna to 2 𝑛𝑑 receive antenna,ℎ21 is the channel filter and provide an effective combined signal. Thisfrom 1 𝑠𝑡 transmit antenna to 2 𝑛𝑑 receive antenna, resulting signal can vary widely in amplitude andℎ22 is the channel from 2 𝑛𝑑 transmit antenna to 2 𝑛𝑑 phase. The rapid fluctuation of the amplitude of areceive antenna, 𝑠1 𝑎𝑛𝑑 𝑠2 are the transmitted radio signal over a short period of time, equivalentlysymbols and 𝑛1 𝑎𝑛𝑑 𝑛2 is the noise on 1 𝑠𝑡 𝑎𝑛𝑑 2 𝑛𝑑 a short travel distance, is such that the large-scalereceive antennas respectively. path loss effects may be ignored. Multipath in the 𝐸𝑞 𝑛 (1) and 𝐸𝑞 𝑛 (2) can be represented in matrix radio channel creates small-scale fading effects. Theform three most important effects are: y1 h11 h12 s1 n1 (3)  Rapid changes in signal strength over a y2 = h21 h22 s2 + n2 small travel distance or time interval  Random frequency modulation due to Therefore, the received vector can be expressed varying Doppler shifts on different multipathas signals y = Hs + n (4)  Time dispersion caused by multipath propagation delaysAnd the complex baseband representation of signal In built up urban areas, fading occurs[15 ] is given by because the height of mobile antennas are well (2) below the height of surrounding structures, so there 𝑃 𝑦= 𝐻𝑥 + 𝑛 is no single line of sight (LOS) the base station [5]. 𝑀 The signal received by mobile at any point in spacewhere y  C N 1 is the received signal vector, may consist of large number of waves having randomly distributed amplitudes, phases and angles x  C M 1 is the transmitted signal vector with zero of arrival. These multipath components combinemean and unit variance, P is the total transmit vectorially at the receiver antenna, and because the N Mpower, H  C is the channel response matrix signal received by mobile is fade [12]. Due towith possibly correlated fading coefficients. In order relative motion between the mobile and the baseto access the performance of MIMO System in station, each multipath wave experiences ancorrelated channel, we adopted a correlation-based apparent shift in frequency. The shift in receivedchannel model which is expressed as [3] signal frequency due to motion is called Doppler 2009 | P a g e
  3. 3. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015shift, and is directly proportional to the velocity and It is clear from the above equation thatdirection of motion of the mobile with respect to the noise power may increase because of the factordirection of arrival of the received multipath wave. (𝐻. 𝐻 𝐻 )−1 .In general if the number of transmitterIf the signal bandwidth is wider than the coherence and receiver antennas is not same, we may multiplybandwidth then different frequencies undergo by Moore–Penrose generalized inverse, pseudo-independent fading and the result is inter-symbol- inverse of H to achieve a similar zero-forcinginterference (ISI). result.In other words, it inverts the effect of channel4. RAYLEIGH FLAT FADING CHANNEL as [3] The fading effect is usually described x ZF = wZF ystatistically using the Rayleigh distribution [7]. The = 𝑥 + H H H −1 𝑛 (9)amplitude of two quadrature Gaussian signals The error performance is directly proportionfollows the Rayleigh distribution whereas the phase connected to the power of H H H −1 𝑛 that is,follows a uniform distribution. The probability H −1 2 H H 𝑛 2.distribution function (PDF) of a Rayleighdistribution is given by [12] 5.2 MINIMUM MEAN SQUARE ERROR 𝑝 𝑟 (1.16) (MMSE) 𝑟 −𝑟 2 If the mean square error between the 𝑒𝑥𝑝 (0 ≤ 𝑟 ≤ ∞ transmitted symbols and the outputs of the detected = 𝜎2 2𝜎 2 symbols, or equivalently, the received SINR is taken 0 (𝑟 < 0) as the performance criteria, the MMSE detector [7] is the optimal detection that seeks to balancewhere σ is the RMS (amplitude) value of the between cancelation of the interference andreceived signal and 𝜎 2 is the average power. reduction of noise enhancement. Let us denote MMSE detector as WMMSE and5. EQUALIZATION TECHNIQUES detection operation by [3]5.1 ZERO FORCING 𝑥 𝑘 = 𝑠𝑔𝑛 [WMMSE y] (10) An ISI channel may be modeled by an The WMMSE maximizes the SINR and minimizes theequivalent finite-impulse response (FIR) filter [5] mean square error which is given by:plus noise. A zero-forcing equalizer uses an inverse 𝐸 𝑥 𝑘 − WMMSE y 𝑇 𝑥 𝑘 − WMMSE y (11)filter to compensate for the channel responsefunction. In other words, at the output of the To solve for x, We know that we need to find aequalizer [4], it has an overall response function matrix WMMSE . The MMSE linear detector forequal to one for the symbol that is being detected meeting this constraint is given by:and an overall zero response for other symbols. If −1 H (12) WMMSE = (H H H + σ2 n I) Hpossible, this results in the removal of theinterference from all other symbols in the absence of Therefore,the noise. Zero forcing is a linear equalization −1 (13) 1method that does not consider the effects of noise. In 𝑊 𝑀𝑀𝑆𝐸 = 𝐻∗ 𝐻 + 𝐼 𝐻∗fact, the noise may be enhanced in the process of 𝑆𝑁𝑅eliminating the interference. MMSE at a high SNR is given byLet us assume the case that MT = MR and H is a full −1 1 (14)rank square matrix. In this case, the inverse of the 𝑊 𝑀𝑀𝑆𝐸 = 𝐻 ∗ 𝐻 + 𝐼 𝐻∗channel matrix H exists and if we multiply both 𝑆𝑁𝑅sides of equation (4) by H-1, we have ≈ 𝐻 𝐻 𝐻 −1 𝐻 𝐻 𝑦H −1 = 𝑥 + 𝑛H −1 (6) At a high SNR MMSE becomes Zero Forcing.From above equation we can see that symbols are 5.3 MAXIMUM LIKELIHOOD (ML)separated from each other. The Linear detection method and SICTo solve for x, we know that we need to find a detection methods require much lower complexitymatrix WZF which satisfies WZF H =1. The Zero than the optimal ML detection, but theirforcing linear detectors for meeting this constraint is performance is significantly inferior to the MLgiven by detection [8]. Maximum Likelihood between WZF = H H H −1 H H (7) received signal vector and the product of all possibleThe covariance matrix of the effected noise may be transmitted signal vector and product of all possiblecalculated as: transmitted signal vectors with the given channel H, 𝐸 (𝑛H −1 ) 𝐻 . 𝑛H −1 (8) and finds the one with minimum distance. = 𝐻 −1 𝐻 . 𝐸 𝑛 𝐻 . 𝑛 . 𝐻 −1 Let C and NT denote a set of signal = 𝑛(𝐻. 𝐻 𝐻 )−1 constellation symbol points and a number of transmit antennas, respectively. Then, ML detection 2010 | P a g e
  4. 4. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015determines the estimate transmitted signal vector x decorrelator the result is used to cancel theas: interference from the received signal vector xML =argmin y−Hx 2 (15) assuming the decision of detection calculates the x ∈C N T Euclidean distance [6] the first stream is correct. For the ZF-SIC, since the interference is already nulled,where, y − Hx 2 Corresponds to the ML metric. the significance of SIC is to reduce the noiseThe ML method achieves the optimal performance amplification by the nulling vector. The nullingas the maximum a posterior detection when all the vector w1 filters the received vector y as:transmitted vectors are likely. However, its T 𝑥 𝑘 = 𝑠𝑔𝑛 [ w1 y] (20)complexity increases exponentially as modulationorder and/or the number of transmit antennas Assuming 𝑥 𝑘 = x1, by substituting x1 from theincreases. The required number of ML metric received vector y, we obtain a modified receivedcalculation is │C│NT, that is the complexity of vector y1 given by:metric calculation exponentially increases with the y1 = y − xk (H)1 (21)number of antennas. where, (H)1 denotes the first column of H. The ML receiver [8] performs optimum We then repeat this operation until all MT bits arevector decoding and is optimal in the sense of detected. Once the first stream is detected, the firstminimizing the error probability. ML receiver is a row of H is useless and will be eliminated. Thereforemethod that compares the received signals with all after the first cancelation the nulling vector for thepossible transmitted signal vectors which is modified second stream need only Mr -1 dimensions. For theby channel matrix H and estimates transmit symbol MMSE detector the significance of SIC is not onlyvector C according to the Maximum Likelihood to minimize the amplification of noise but also theprinciple, which is shown as: cancelation of the interference from other antennas. 2 (16) 𝐶 = 𝑚𝑖𝑛 𝑎𝑟𝑔 𝑦 − 𝐶 ′ 𝐻 𝐹 In addition, there is another opportunity to improve 𝐶 the performance by optimal ordering the SICwhere, F,is the Frobenius norm. Expanding the cost process. The ordering is based on the norm of thefunction using Frobenius norm given by nulling vector. At each stage of cancelation, instead of randomly selecting the stream to detect, we C = min arg Tr y − C ′H H . y (17) C choose the nulling vector that has the smallest norm − C′H to detect the corresponding data stream. This scheme is proved to be the globally optimum ordering more complex. C = min arg Tr y H . y + H H . C ′H . C ′ . H (18) C 5.4.1 ZERO FORCING WITH SIC − H H . C ′H . y OSIC [11] is basically based on subtraction − yH . C′. H of interference of already detected elements of s Considering y H . y is independent of the from the receiver vector r. This results in a modifiedtransmitted codeword so can be rewritten as [3] receiver vector in which effectively fewer interferers are present. In other words, SIC is based on the C (19) subtraction of interference of already detected = min arg⁡ H H . C ′H . C ′ . H] [Tr[ elements s from the received vector x. This results in C a modified receiver vector in which effectively − 2Real Tr H H . C ′H . y ] fewer interferers are present. When Successive Interference Cancellation (SIC) [11] is applied, the order in which the components of s are detected iswhere, .H is a Hermition operator, although ML important to the overall performance of the system.detection offers optimal error performance, it suffers To determine a good detection order, the covariancefrom complexity issues. matrix of the estimation error 𝑠 − 𝑠 𝑒𝑠𝑡 is used. We know that the covariance matrix is given by5.4 SUCCESSIVE INTERFERENCE 𝑄 = 𝐸 ∈. ∈ 𝐻 = 𝜎 2 𝐻 𝐻 𝐻 −1 (22) 𝑛CANCELLATION 𝐻 𝑄 = 𝐸 (𝑠 − 𝑠 𝑒𝑠𝑡 ) 𝑠 − 𝑠 𝑒𝑠𝑡 (23) When signals are detected successively, the = 𝜎 2 𝐻 𝐻 𝐻 −1 𝑛outputs of previous detectors can be used to aid the ≡ 𝜎2 𝑃 𝑛operations of next ones which leads to the decision Where P =𝐻 + 𝐻 + 𝐻directed detection algorithms including SIC, Parallel Let (𝑠 𝑒𝑠𝑡 ) 𝑝 be the pth entry of𝑠 𝑒𝑠𝑡 , then the ―best‖ isInterference cancelation (PIC), and multistagedetection [8]. ZF SIC with optimal ordering, and the one for which 𝑃𝑝𝑝 (i.e., the p-th diagonal elementMMSE-SIC [10] with equal power allocation of P) is the smallest. Because this is estimate withapproach the capacity of the i.i.d. Rayleigh fading the smallest error variance. From the 𝑒𝑞 𝑛 (23) itchannel. After the first bit is detected by the becomes clear that 𝑃𝑝𝑝 is equal to the squared length 2011 | P a g e
  5. 5. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015of row p of 𝐻 + . Hence, finding the minimum 2) From the estimate of the correspondingsquared length row of 𝐻 + is equivalent. elements of s. In case of MMSE:Summarizing, the decoding algorithm consist of 𝑆 𝑒𝑠𝑡 𝑝 = 𝑊 𝑀 𝑥three parts: Where the weight vector W M equals row M Ordering: determine the Tx stream with the lowest (number of transmitting antennas) of the permutederror variance. W Interference Nulling: estimate the strongest Tx 3) While M-1>0 go back to step 1, but now with: signal by nulling out all the weaker Tx signals. 𝐻 ⟶ 𝐻 𝑀−1 = ℎ1 … … … ℎ 𝑀−1 Interference Cancellation: remodulate the databits, subtract their contribution from the receivedsignal vector and return to the ordering step. Figure.3 SIC MMSE Detector Figure.2 SIC Zero Forcing Detector We use the first Zero-Forcing detector to So here we can see that we get optimal ordering bydetect the data stream 𝑠1 (𝑚) decode it and then using MMSE with OSICsubtract this decoded stream from the receivedvector. Assuming the first stream is successfully 6. SIMULATION RESULTSdecoded, and then the second Zero-Forcing detector 6.1 SIMULATION MODELonly needs to deal with 𝑠3 … … … 𝑠 𝑁 𝑡 as interference, The Matlab script performs the following.Generationsince 𝑠1 has been correctly subtracted off. Thus, the of random binary sequencessecond Zero-Forcing detector projects onto a 1. Modulate the binary sequences using BPSK,subspace which is orthogonal to ℎ3 … … … . ℎ 𝑁 𝑡 .This QPSK and 16 QAM.process is continued until the last Zero-Forcing 2. Group them into pair of two symbols and senddetector does not have to deal with any interference two symbols in one time slot.from the other data streams. We assume subtraction 3. Multiply the symbol with the channel and thenis successful in all preceding stages. This SIC add white Gaussian noise.(Successive Interference Cancellation) Zero-Forcing 4. Perform equalization on the received signal anddetector architecture is illustrated in Figure.2 different equalizers are ZF, MMSE and ML [6].So we can see here with respect to ZF, the ZF with 5. Perform hard decision decoding that isOSIC algorithm introduces extra complexity. demodulating the BPSK, QPSK and 16 QAM. Repeat for multiple value of SNR and plot the5.4.2 MMSE WITH SIC simulation result. In order to do OSIC with MMSE [11], thenthe algorithm resulting as follows 6.2 RESULTSCovariance matrix can be written as In Figure.4 We compare a ZF with different 𝑠 − 𝑠 𝑒𝑠𝑡 𝑠 − 𝑠 𝑒𝑠𝑡 𝐻 = 𝜎 2 𝛼𝐼 + 𝐻 𝐻 𝐻 −1 ≡ 𝜎 2 𝑃 modulation and in this we observed that BPSK and 𝑛 𝑛Covariance matrix of the estimation error 𝑠 − 𝑠 𝑒𝑠𝑡 QPSK have the almost same results and 16-QAMwill be used to determine good ordering for have worst results than BPSK and QPSKdetection. modulation. At BER=0.001, there is approximately 31) Compute W (P is obtained while determining dB difference between the BPSK and 16-QAM W). Find the smallest diagonal entry of P and modulation in Zero Forcing. In Figure.5 we compare suppose this is the p-th entry. Permute the p-th the ZF-OSIC with different modulation like BPSK, column of H to be last column and permute the QPSK and 16-QAM. In this graph we observed that rows of W accordingly. BPSK have an equivalent to the QPSK and 16-QAM have worst results than BPSK and QPSK 2012 | P a g e
  6. 6. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015modulation. At BER=0.001, there is approximately 4dB difference between the BPSK and 16-QAMmodulation in Zero Forcing-OSIC. In Figure.(6) wecompare the MMSE with different modulation likeBPSK, QPSK and 16-QAM. In this graph weobserved that BPSK have better results than QPSKand 16-QAM and 16-QAM modulation have worstresults. At BER=0.001, there is approximately 5 dBdifference between the BPSK and 16-QAMmodulation in MMSE-OSIC. In Fig.(7) we comparethe MMSE-OSIC with different modulation likeBPSK, QPSK and 16-QAM. In this graph weobserved that BPSK have better results than QPSKand 16-QAM and 16-QAM modulation have worstresults. At BER=0.001, there is approximately 7 dBdifference between the BPSK and 16-QAMmodulation in MMSE-OSIC. In Figure. (8) wecompare the ML with different modulation likeBPSK, QPSK and 16-QAM. In this graph we Figure.5 Comparison of ZF-OSIC using BPSKobserved that BPSK have better results than QPSKand 16-QAM and 16-QAM modulation have worstresults. At BER=0.001, there is approximately 3dBdifference between the BPSK and QPSK modulationin MMSE-OSIC. In Figure. (9)There is a comparisonbetween the different detectors like MaximumLikelihood (ML), ZF-OSIC, ZF, MMSE andMMSE-OSIC. Here we observed that MaximumLikelihood (ML) have a best performance than otherdetectors and Zero Forcing (ZF) has a worstperformance. If we compare the ZF and ML,performance curve of the two detectors are close toeach other at low SNR but the gap gets larger whenSNR gets higher. When the SNR gets higher, thepost detection of SNR is mainly affected by channelmatrix H. If we compare the MMSE-OSIC and. ZF-OSIC, at BER=0.001 there is an approximately 3 dBdifference between these two detectors. Figure.6 Comparison of MMSE using BPSK Figure.4 Comparison of ZF using BPSK Figure.7 Comparison of MMSE-OSIC using BPSK 2013 | P a g e
  7. 7. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015 Conference on Communications, vol.2, no.11-14,pp-591-595,June 2001. [3] Choo.Y.S, Kim.J. Yang W.Y., and Kang C.G,―Mimo-Ofdm Wireless communication with matlab‖, IEEE PRESS, John Wiley and sons (Asia) Pte Ltd. [4] G. Arslan, B. L. Evans, and S. Kiaei, ―Equalization for Discrete Multitone Receivers To Maximize Channel Capacity‖, IEEE Transactions on Signal Processing, submitted March 30, 2000 [5] C. E. Proakis, ―Digital Communications,‖ McGraw-Hill International Editions, New York, 4th edition, 2000 [6] M. Janakiraman, "Space-time codes and MIMO systems", Artech House, 2004 Figure.8 Comparison of ML using BPSK [7] G. J. Foschini, ―Layered space–time architecture for wireless communication in a fading environment using multi–element antennas,‖ Bell-Labs Techn. J., pp. 41–59, 1996. [8] C.Windpassinger and RF.H Fischer, ―Low- complexity need-Maximum Likelihood detection and precoding for MIMO systems‖ in ITW 2003, Paris, France, March 31-April-4, 2003. [9] A. Paulraj, R.Nabar and D.Gore, ―Introduction to Space Time Wireless Communications‖, Cambridge University Press, May 2003. [10] M.Varanasi and T.Guess, ―Optimum decision feedback multiuser equalization with successive decoding achieves the total capacity of the Gaussian multiple-access channel,‖ Conference Record of the Thirty- Fig.9 Comparison using different detection First Asilomar Conference on signals, technique using BPSK modulation Systems and computers, vol. 2, pp. 1405- 1409, Nov-2-5 1997.7. RESULTS [11] G. J. Foschini, ―Layered space–time In this paper, we studied MIMO V-BLAST architecture for wireless communication insystem performance under Flat Fading Rayleigh a fading environment using multi–elementchannel. Further this system is compared with antennas,‖ Bell-Labs Techn. J., pp. 41–59,different modulation technique and system gets 1996.better result in BPSK modulation and worst result in16-QAM Fig.(9) shows the simulation results for BIOGRAPHYBPSK modulation with different decoding technique Mr.Gurpreet Singh received M.Techand ML gives the best result and ZF gives the worst degree in Electronics andresult. Communication Engineering degree from Jaypee University of Information and Technology, Solan in 2012 andREFERENCES received B.Tech degree from Lovely[1] R. U. Nabar A. J. Paulraj, D. A. Gore and Institutes of Technology in H. Bolcskei, ―An overview of MIMO Electronics and Communication Engineering from communications—a key to gigabit Lovely Institutes of Technology, Phagwara in 2010 wireless,‖ Proceedings of the IEEE, vol. 92, with distinction. Currently, he is working as a no. 2, pp. 198–218, Feb. 2004. Assistant Professor in Shaheed Bhagat Singh State[2] A.Paulraj and R.J.Heath, ―Characterization Technical Campus, Ferozpur, Punjab. His area of of MIMO Channels for Spatial interest is signal processing, MIMO, Wireless Multiplexing Systems ―IEEE International Mobile Communication Engineering, high speed 2014 | P a g e
  8. 8. Gurpreet Singh, Pardeep Sharma / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.2008-2015digital communications and 4G WirelessCommunications Mr. Pardeep Sharma received M.Techdegree in Electronics and CommunicationEngineering from Sant Longowal Institute ofEngineering & Technology in 2012 and receivedB.Tech degree in Electronics and CommunicationEngineering form Swami Vivekanand Institutes ofEngineering & Technology, Banur in 2010.Currently, he is working as a Assistant Professor inShaheed Bhagat Singh State Technical Campus,Ferozepur, Punjab. His area of interest is DigitalSignal processing, VHDL, Wireless MobileCommunication Engineering, Digital Design andAnalysis 2015 | P a g e

×