SMART ANTENNA AOA ESTIMATION EMPLOYING MUSIC ALGORITHM And DIGITAL BEAMFORMING By VARIABLE STEP-SIZE LMS ALGORITHM With NOVEL MAC PROTOCOL For IEEE 82. T.S.JEYALI LASEETHA, R.SUKANESH 2,. &2. Department Of Electronics And Communication Engineering, Anna University, Chennai.Dr. Sivanthi Aditanar College Of Engineering, Tiruchendur 2.Thiagarajar College Of Engineering, Madurai &2 INDIA laseetha@gmail.com Abstract :- Smart antenna using multiple antennas exploring the SDMA technology has been used in wireless communication systems which provides higher spectrum efficiency, improved link quality and throughput through frequency reuse, reduction of intersymbol interference (ISI) and co-channel interference rejection (CCI)[]. Space Division Multiple Access, where multiple user signals are simultaneously transmitted and received over same conventional time and frequency channels by using multiple antennas, dramatically increases spectral efficiency. SDMA is realized with beam forming and null forming techniques which accounts for co-channel interference rejection. Channel equalization is done using adaptive equalizer to mitigate the effects of ISI. This paper discusses SDMA technology with digital beam forming with the novel idea of including the Variable step-size LMS algorithm along with MUSIC algorithm employed in MAC protocol for IEEE 82. and the results of simulation are analyzed for their performance. Key-words :- smart antenna, adaptive beamforming, intersymbol interference, variable step-size lms algorithm, adaptive equalizer, Angle of Arrival estimation(aoa), MAC protocol, MUSIC algorithm.introduction Adaptive antenna beam forming also known as Digital beam forming and Smart antenna technology has been widely used for system identification, sonar technology, image processing, wireless communications, radar and related fields[2],[3]. This technology provides wireless communication systems with improved spectral efficiency, throughput and reduced ISI and CCI. The base station can suppress the cochannel interference in the service cell. Therefore the mobile user need not bother about explicit interference rejection function. The air interface used in this paper is TDD(Time Division Duplex) /TDMA(Time Division Multiple Access), which the frame structure supports for an efficient DBF/SDMA processing. An variable step size Least Mean Square Algorithm used at the base station computes and applies reception array antenna weights upon receiving uplink channel signal from a desired user, to compensate for propagation spatial characteristics[9]-[]. Then the BS uses the reception array weights as the transmission array weights for downlink transmission. Therefore optimum beam forming is made towards the desired user while rejecting the interference from undesired mobile user. In this paper Digital beam forming technique is presented with 6 number of antenna elements at the base station as ISSN: 79-57 87 ISBN: 978-96-474-62-5
uniform linear array. Section 2 deals with system specification with physical layer protocol. In section 3 the adaptive antenna array algorithm is discussed for array weight calculation and updating the weights using variable step size LMS algorithm. Section 4 deals with the MAC protocol with DBF/SDMA technology. The experimental simulation results are shown and discussed in section 5. The paper is concluded in section 6. 2.Digital Beamforming Using Adaptive Antenna Array System Table : System Parameters Specification required average output signal to noise ratio for a given BER with fading. The arrays can theoretically completely cancel N interferers with M antennas (M>N) and achieve an M-N fold diversity gain[]. Significant suppression of N>M interferers is also possible. However this is at the cost of requiring a receiver for each antenna and tracking the antenna weights at the fading rate (79Hz at 2Ghz and 6mph) versus switching every few seconds with the multibeam antenna. In this paper the array weights for these desired beam patterns are generated by variable step size LMS algorithm using MMSE criterion by receiving the uplink training sequence. For downlink System TDD-TDMA/SDMA Carrier frequency 2.4 GHz Symbol rate Mbps Downlink modulation BPSK Base station Number of antenna 6 elements Antenna topology Uniform Linear Array Antenna spacing λ/2 Adaptive antenna algorithm Variable step-size Least mean square Adaptive antenna processing criteria Minimum mean square error Mobile user Number of antenna element An Antenna array consists of M elements arranged in uniform linear array (ULA) fashion is employed in DBF/SDMA signal processing and is executed in baseband. The block diagram of DBF/SDMA signal processing at the Base station is shown in Fig. A base station mitigates the fading of received power of the desired signal by steering the beam in the desired mobile user s direction while there are N number of sources transmitting the signals at the base station. The base station communicates with them all but picks up or selects the desired signal while nullifying interfering signals from others. M antenna elements can provide an increased gain of M plus a diversity gain against the multipath fading which depends on the correlation of the fading the antennas. The antenna gain can be defined as the reduction in required received signal power for a given average output signal to noise ratio. The diversity gain is defined as the reduction in the Fig. Digital beamforming at the Base station transmission, the array weights which are determined by the uplink,are applied because of the reciprocity nature of TDD channel. The DBF/SDMA system reduces communication overhead and improve the user throughput in TDD operation. 3. Variable Step-Size LMS Algorithm For Array Weights Calculation Digital beam forming employing the reference signal structure and variable step-size LMS algorithm is adopted in this system to update the weights of the beam former and execute realtime tracking operation. The advantage of using an variable step-size LMS algorithm includes network compatibility, portability, mobility and ISSN: 79-57 88 ISBN: 978-96-474-62-5
capability of single chip VLSI implementation. The VSLMS algorithm is one of the MMSE algorithms and is proposed for the channel estimator of a maximum likelihood sequence estimator (MLSE) equalizer[]. In the channel model of a L-length tapped delay line (TDL), let T denote the tap coefficient vector, which becomes the estimated channel impulse response, and let d(n) denote the channel output signal. If the training signal at time n is r(n), then x l (n) = r(n-l), x(n) = [ x (n) x 2 (n) x 3 (n)..x L (n)] --------- () and e(n) = d(n) w(n-) H x(n) ---------- (2) The estimated value of w is obtained iteratively with the VLMS algorithm as[] w(n) = w(n-) + µ g (n)e * (n)x(n) --------- (3) where T and H represent matrix transpose and Hermitian transpose respectively and --------- (4) Where is a constant determined by forgetting factor of the RLS algorithm. The optimized array weights are determined using VSLMS algorithm while running Monto-Carlo Simulations for times. The weights are tabulated as shown below. Table.2. The weights for N = 6 ULA Array values weights W W2 -.3929+.842i W3 -.62695-.65586i W4.942-.3432i W5 -.868+.246i W6 -.8363-.6548i W7.88346-.4282i W8.292+.9363i W9 -.86639-.346i W.67864-.7498i W.32279+.988i W2 -.49-.2444i W3.59632-.74992i W4.439+.79558i W5 -.92756+.757i W6.3487-.95363i 4. Mac Protocol For Smart Antennas While using SDMA technology with directional antennas, a new MAC protocol is important[4]. Traditional MAC protocol with omni - directional antennas are not suitable for the support of new features like SDMA technology which supports IEEE 82.. The theoretical aspects of new MAC protocol suitable for smart antennas with SDMA technology is explored herewith for the proper understanding of how the interferers and desired users are being identified and separated. In the traditional method, Request to Send (RTS) and Clear to Send(CTS) packets are sent omni-directionally in order to enable the transmitter and receiver to locate each other and then sending DATA packet and ACK in direct mode. All these four frames contain information about the duration of the pending handshake, informing the neighbors to avoid starting a new transmission during this period. This is managed by a mechanism called Virtual Carrier Sense. Every station maintains a Network Allocation Vector(NAV). If NAV is zero the station can transmit otherwise if can not. NAV is initially zero. If NAV is a positive number there is a countdown until it reaches zero. When a station hears one of the four frames, it updates its NAV with the duration of the pending handshake preventing itself by transmitting until its NAV reaches zero again. With this scheme every station performs a Virtual Carrier Sense in addition to the physical carrier sense to enhance the resistance of the protocol against collisions. In this scheme, the transmitter starts transmitting its RTS in a predefined direction, assume with beam. Short afterwards it turns its transmission beam on the right sending the same RTS with beam2. It continues this procedure again and again until the transmission of RTS covers all the area around the transmitter (until it sends the RTS with beam M). The RTS contains the information about the duration of the intended four way handshake. As this information is spread around by the circular RTS, the neighbors are informed about the intended transmission. The neighbors after executing an algorithm decide if they will defer their transmission in the direction of transmitter or receiver, if this harms the ongoing transmission. The STA ie the mobile user, that is the destination of the RTS waits until the finish of the circular RTS transmission and then send a directional CTS towards the direction of the transmitter of the RTS. Then the ISSN: 79-57 89 ISBN: 978-96-474-62-5
carrier sensing from the transmitter of the RTS in this phase is performed in an omni-directional mode. If the CTS is received during a predefined period (CTS time out) then the transmitter continues with the transmission of the data packet and the reception of ACK, in a particular direction. By using only the directional transmissions of RTS, CTS, Data and ACK we exploit the benefit of increasing the coverage area, compared with the omni-directional mode of transmission analyzed from Fig. 9 wheree only 4 weights are plotted instead of 6 weights to have a clear view. The weights get adapted within iterations but they reach the steady state after 5 iterations. The optimized adapted array weights are shown in Table.2. The SINR is determined as 48.3722 db. Fig:2 A node with M beams Fig 3. Spatial spectrum estimate of the arriving signal 4. Simulation And Results Analysis In this paper, an Uniform Linear Array of 6 elements is considered and the channel is estimated from the autocorrelation matrix of the incoming signal. The signals are assumed to be arriving from three different directions with the angles [4º 7º 9º]. Additive White Gaussian noise of variance (σ 2 ). is adopted with SNR of. The AOA is estimated using MUSIC and Minvariance method[] [3]. MUSIC algorithm performs better among the two. The spatial spectrum estimate is shown in Fig.3. The polar plot in Fig.4 shows the angle of arrival with the maximum signal strength for the signal coming from 4º. After estimating the angle of arrivals, the beamforming was done using Variable Step- of the Step- size LMS algorithm. The variation size (µ g ) is plotted in Fig.6. This speeds up the convergence of LMS algorithm. The signal from 4º is tracked within iterations and the same is shown in Fig.7. The same can be seen in the Fig.8 and Fig.9 also. In Fig.8 the mean square error is plotted to study the convergence performance. The adaptation of weights is Mean square error 8 5 2 polar plot for AOA Estimation 2 24.5 Fig 4. Polar plot for the AOA estimate - -2-3 Fig.5 Plot of Mean-Sqaure-error convergence for all the three arriving signals 9 27 2.5 2.5 Received Signal and interferences -4 2 3 4 5 6 7 8 9 Sample Interval 6 3 desired interferer interferer2 3 33 desired interferer interferer2 ISSN: 79-57 9 ISBN: 978-96-474-62-5
variable stepsize.3.2. adaptation of weights w w2 w3 w4. value of mu weights.9.8.7.6.5.4 2 3 4 5 6 7 8 9 iteration no. Fig.6 Variable step size(µ g ).3 2 3 4 5 6 7 8 9 Iteration no. Fig.9 Weights adaptation 4 3 2 Desired signal Array output AF n.8.6.4 Signals.2 - -2 2 3 4 5 6 7 8 9 No. of Iterations Fig.7 tracking the desired signal -9-6 -3 3 6 9 AOA (deg) Fig. Radiation pattern for the 6 element ULA 5. Conclusion Mean square error 9 8 7 6 5 4 3 2 2 3 4 5 6 7 8 9 Iteration no. Fig.8 Mean Square Error plot A 6-element smart antenna array has been developed for high data rate of Mbps for TDD- TDMA/SDMA system. A brief system design of the adaptive array using digital beamforming with AOA suitable for Novel MAC protocol is presented. Simulation results demonstrates the performance improvement in terms of BER and SINR using adaptive beamforming techniques multipath environment. Uplink Spatial Division Multiple Access (SDMA) is also demonstrated using the adaptive array system. Near-far effect can be avoided by pre-tuning the transmission power of each terminals according to desired BER performance. Better signal separation capability can be obtained using more degrees of freedom in the beamforming weight vectors, which in turn proportional to the number of antenna elements. This will enhance the system capacity and make the system highly suitable for wireless communication applications. In this environment, the variations of the channel are so ISSN: 79-57 9 ISBN: 978-96-474-62-5
fast that by utilizing a variable step-size factor, the system becomes more adaptive and the channel estimation will provide a more accurate estimation of the data and the use of a variable step size allows each path to be independent of every other path. References : [] J.H.Winters, Smart antennas for wireless systems, IEEE Personal Communications, vol.5, pp.23-27, Feb. 998. [2] R.Kohno, Spatial and temporal communication theory using adaptive antenna array, IEEE Personal Communications, vol.5, no., pp.28-35, Feb. 998. [3] Y.Kikuma, Adaptive signal processing with array antenna, Kagakugijutsu-Shuppan, Tokyo, 998. [4] B.Widrow and S.D.Stearns, Adaptive signal processing, Prentice Hall, Upper Saddle River, USA, 985. [5] Y.Ogawa, M.Ohmiya, and K.Itoh, An LMS adaptive array for multipath fading reduction, IEEE Trans. Aerosp. Electron. Syst., vol.aes- 23, no., pp.7-23, Jan. 987. [6] Y.Ogawa, Y.Nagashima, and K.Itoh, An adaptive antenna system for high-speed digital mobile communications, IEICE Trans. Commun., vol.e75-b, no.5, pp.43-42, May 992. [7] T.Ohgane, Spectral efficiency evaluation of adaptive array base station for land mobile cellular systems, Proc. VTC 94, pp.47-474, May 994. [8] D.M.Brady, Adaptive coherent diversity receiver for data transmission through dispersive media, Conf. Rec., IEEE Int. Conf. Commun., San Francisco, June 97. [9] J.Takada, Optimum antenna spacing in MMSE combining, Proc. First International Symposium on Wireless Personal Multimedia Communications (WPMC 98), pp.49-496, Nov. 998. [] J.H.Winters, Optimum combining in digital mobile radio with cochannel interference, IEEE J. Sel. Areas Commun., vol.sac-2, pp.528-539, July 984.\ [] H.Suzuki, Signal transmission characteristics of diversity reception with leastsquare combining, IEICE Trans., vol.j75- B-II, no.8, pp.524-534, Aug. 992. [2] J.Salz and J.H.Winters, Effect of fading correlation on adaptive arrays in digital mobile radio, IEEE Trans. Veh. Technol., vol.43, no.4, pp.49-57, Nov. 994. [3] Y.Akaiwa, Introduction to Digital Mobile Communication, John Wiley & Sons, New York, 997. [4] S.Haykin, Adaptive filter theory, 3rd ed., Prentice Hall, Upper Saddle River, USA, 996. [5] R.W.Harris and D.M.Chabries, A variable step (VS) adaptive filter algorithm, IEEE Trans. Acoust., Speech & Signal Process., vol.assp- 34, pp.39-36, April 986. [6] A.Taguchi and N.Hamada, A variable step size method for the learning identification, IEICE Trans., vol.j7-a, no.8, pp.663-668, Aug. 988. [7] R.H.Kwong and E.W.Johnston, A variable step size LMS algorithm, IEEE Trans. Signal Process., vol.4, pp.633-642, July 992. [8] E.Eweda and O.Macci, Convergence of an adaptive linear estimation algorithm, IEEE Trans. Autom. Control, vol.ac-29, no.2, pp.9-27, Feb. 984. [9] S.Ahn and P.J.Voltz, Convergence of the delayed normalized LMS algorithm with decreasing step size, IEEE Trans. Signal Process., vol.44, pp.38-36, Dec. 996. [2] S.Kozono and T.Tsuruhara, Correlation between two mobile radio base-station antennas, IEICE Trans., vol.j66-b, no.4, pp.558-559, April 983. [2] S.Denno and Y.Saito, Fast channel impulse response estimation scheme for adaptive MLSE equalizer, IEICE Trans., vol.j78-b-ii, no.4, pp.22-23, April 995. [22] J.Nagumo and A.Noda, A learning method for system identification, IEEE Trans. Autom. Control, vol.ac-2, pp.282-287, 967 ISSN: 79-57 92 ISBN: 978-96-474-62-5