INTERFERENCE AWARE RECEIVER MODELING FOR SFBC TRANSMIT DIVERSITY IN 4G DOWNLINK A.Vinotha PG Scholar Department of ECE, Oxford Engineering College Trichy, tamilnadu, India M.Ashok Raj Assist.prof Department of ECE. Oxford Engineering College Trichy, tamilnadu India Abstract: This project investigates the Interference Rejection Combining (IRC) receiver is effective in improving the cell-edge user throughput because it suppresses inter-cell interference. When assuming LTE-Advanced downlink and open-loop transmit diversity employing the Space Frequency Block Code (SFBC) using Alamouti coding, the IRC receiver must detect the SFBC Signals and suppress the interference signal at the same time. Inorder to achieve this the IRC receiver requires highly accurate weight matrix consists of the channel matrix of the serving cell and the statistics of the covariance matrix. These must be extended in the both frequency and spatial domain at the same time and the extended matrices can be estimated using the downlink Reference Signals (RS) from the serving cell. However, some elements cannot be estimated using a RS based estimation scheme. Therefore this paper proposes statistics of those unknown elements are proposes specifically by inserting zero values. The results of simulations show that the IRC receiver suppresses the inter cell interference and improves the throughput when a cell-edge environment is assumed.inaddition to this the usage of MMSE-IRC receiver is expected to be a significant performance booster and it can suppress both data and CRS interferences. Keywords---Interference rejection combining (IRC),Long term evolution(lte)-advanced, Multiple input Multiple output(mimo),space frequency block code(sfbc). I.INTRODUCTION The various commercial services based on the first release (Release 8) of Long Term Evolution (LTE) have been launched in many countries. In Japan, NTT DOCOMO launched a commercial LTE service in December 2010 under the new service brand of Xi. One key feature for LTE is its use of multi antenna techniques, i.e., Multiple- Input Multiple-Output(MIMO) techniques.for MIMO technique mainly two modes are supported, i.e.,openloop and closed loop MIMO multiplexing. LTE-Advanced (LTE-A) is the project name of the evolved version of LTE that is being developed by 3GPP.LTE-Advanced will meet or exceed the requirements of the International Telecommunication Union (ITU) for the Fourth Generation (4G) radio communication standard known as IMT-Advanced.. After finalizing the specifications for the first release (Release 10) of LTE-Advanced, which is an enhancement of LTE, to satisfy the high level requirements for peak and cell edge user throughput, advanced multiple antenna transmission techniques were investigated. An important goal for LTE-Advanced is to improve the cell- edge user throughput while achieving high-speed and high-capacity communications. The IRC receiver investigated to suppresses the interference signals, reception processing with the aid of multiple antenna branches while detecting the desired signal. Since the IRC receiver strictly generates the received weight matrix based on the MMSE criterion, i.e., including the interference signals, the interference signals can be suppressed according to the spatial degrees of freedom.the IRC receiver requires knowledge of the interference signals in addition to the desired signal, i.e., the covariance matrix and the channel matrix of the serving cell. Here, the serving cell is defined as the cell that transmits only the desired data signals. In LTE or LTE- Advanced, the transmission timing and 128
channel matrices of the interfering cells, which are estimated using the downlink reference signals (RSs), are not known at the receiver for every subframe, i.e., 1 ms. This is because the receiver does not have to frequently update the average received signal power levels of the interfering cells.in this the channel matrix of the serving cell can be estimated using downlink RSs. When assuming LTE/LTE-Advanced, the channel matrix of the serving cell can be estimated using the cell-specific RS (CRS) transmitted from the serving cell. In contrast, the statistics of the covariance matrix should be accurately estimated using the received signals at the IRC receiver since it is difficult to estimate the channel matrix of the interfering cells at the receiver, as previously mentioned. The RS-based estimation scheme that estimates the statistics of the covariance matrix, including only the interference and thermal noise components, using the RS sequence of the serving cell was effective. More specifically, this matrix can be estimated by subtracting the replica symbols of the serving cell generated by the estimated channel matrix and the known RS sequence. Here, transmit diversity assuming single-stream transmission and included in both open and closed loop MIMO transmission. As one type of openloop MIMO transmission, the space frequency block code (SFBC) is supported as a single-stream transmission. Transmit diversity is mainly used for a UE device that is located at the cell edge where the IRC receiver is expected to further increase the user throughput. This paper mainly focus on open loop transmit diversity using the SFBC and investigate the IRC reception scheme. When assuming open-loop transmit diversity using the SFBC, two adjacent subcarriers are used to transmit two information signals and those can be treated as a single-stream transmission. Two approaches are mainly considered for the IRC reception scheme, i.e., whether the processes of demodulating data signals and suppressing the interference signals are performed at the same time or separately. It was shown that a receiver using this scheme perfectly suppressed the interference signals when the degrees of freedom at the receiver exceeded the number of interference signals with Alamouti coding. In contrast, assuming an SFBC with two transmitter antenna branches, a scheme was investigated in which the IRC process for suppressing the interference signals was implemented in the spatial domain after the demodulation process using maximal ratio combining (MRC) in the code (frequency) domain. However, we consider that the maximum gains from the IRC receiver cannot be achieved since the demodulation and suppression of the interference signals are separate processes. Therefore, in this paper, we focus on the former scheme for the IRC Receiver weighted matrix generation. In LTE/LTE-Advanced, all elements of the channel matrix for the serving cell, can be estimated using the downlink RS, i.e., CRS. The principle of Alamouti coding and also the CRS is independently transmitted for each transmitter antenna branch without any precoding processing. The covariance matrix or the receiver weight matrix estimation scheme based on the received data signals, including the interfering signals. The past investigation in the traditional sample matrix inversion (SMI) estimated the covariance matrix or the receiver weight matrix based on the averaging operation using the received signals.. Therefore the performance of the IRC receiver might be limited due to the lack of the number of samples for the averaging operation. Our past investigations in showed that the covariance matrix estimation based on the SMI caused degradation in the user throughput compared with the conventional receiver, which could not suppress the interfering signals except in a cell-edge environment. In contrast the covariance matrix estimation using the RS-based scheme, as previously mentioned. This approach improves the user throughput performance for the closed-loop MIMO transmission modes but some elements of the covariance matrix cannot be estimated when assuming open-loop transmit diversity using the SFBC. This is because the CRS is not transmitted using two adjacent subcarriers, i.e., the interfering Alamouti coded signals transmitted using two adjacent subcarriers cannot be properly estimated using the CRS transmitted from the serving cell. This is a problem for the practical IRC receiver employing open-loop transmit diversity using the SFBC since the elements regarding alamouti coding within the covariance matrix are unknown at the receiver. II.INTERFERENCE AWARE RECEIVER FOR OPEN-LOOP TRANSMIT DIVERSITY This paper investigates to achieves the maximum diversity gain when SFBC using Alamouti coding is employed. For simplicity, channel fluctuations in the time and frequency domains are assumed to be sufficiently small over 129
the duration of 1 Resource Block (RB), which is the minimum assignment unit defined as 12 subcarriers 14 OFDM symbols (one subframe). The two-dimensional received signal vector of the m-th space-frequency block coded information at the i-th receiver antenna branch, yi(m), is expressed as, Fig 1: Illustration of UE moving away from its serving cell. Where, For enhanced receiver, the MMSE-IRC receiver can suppress not only the inter-stream interference but also the inter cell interference when the degrees of freedom at the receiver are sufficient, i.e., the number of receiver antennas is higher than that of the number of desired data streams, and MMSE IRC receiver weight matrix is expressed as follow: where the superscript * denotes the complex conjugate. The SFBC is basically employed using two adjacent subcarriers. Terms and are the received signals of the m th SFBC symbol at the i-th receiver antenna branch, and are the information signals within the m-th SFBC symbol of the q-th cell, hij,q is the channel response in the frequency domain between the i-th receiver antenna branch and the j- th transmitter antenna branch. When the sources of intercell interference indicated as q > 0 in equation are not considered, the recovered information signals, i.e., and at the ith receiver antenna branch are detected using MRC as follows: Where and R denote the estimated channel matrix and covariance matrix, respectively. The MMSE-IRC receiver weight matrix consist of covariance matrix including the sources of intercell interference needs to be estimated. The CRS from the interfering cells cannot be suppressed even if the ideal IRC receiver is assumed. Furthermore, the IRC receiver using the data signal based covariance matrix estimation scheme is evaluated for comparison to the proposed scheme. To estimate the covariance matrix using the RS-based covariance matrix estimation scheme as described in Section I, the 4 4 covariance matrix that extends both the code and spatial domains. When the channel matrices of all cells can be ideally obtained, Ryy(m) is expressed as follows: A. System Model In the LTE-Advanced standardization work performed by 3GPP, realistic modelling of linear MIMO receivers was deemed important because advanced linear interference aware receivers can suppress a part of intra-cell and intercell interference improving downlink system performance. The improvement of LTE Advanced downlink performance provided by IRC-type receivers at the UE side. Here, RI+N(m) is the 4 4 covariance matrix that only includes the interference and thermal noise components. When the number of interfering cells is assumed to be one, RI+N(m) is ideally expressed as follows: 130
Where ra1 and rb12 are assumed to be an unknown elements in the estimated covariance matrix. b. Performance degradation of IRC receiver in LTE-Advanced downlink The IRC receiver can suppress not only the inter-stream interference but also the inter-cell interference when the degrees of freedom at the receiver are high, i.e., the number of receiver antennas is higher than that of the desired data streams. IRC algorithm is highly sensitive to the quality of channel and inter-cell interference covariance matrix estimation.the performance of the IRC receiver depends on the channel Conditions channel estimation, and covariance matrix estimation accuracy. In such a case, the performance of the IRC receiver with realistic channel estimation would be worse than that for the simplified MMSE receiver. III. PROPOSED MMSE-IRC ALGORITHM FOR LTE DOWNLINK RECEIVER. MMSE-Interference Rejection Combining (MMSE-IRC)receivers are the mobile terminal interference rejection and suppression technology to mitigate the effects the interference signal and improve the user throughput. The MMSE-IRC receivers considered independently of noise component instead of handling them as equivalent to noise. The well-known linear MIMO receive algorithm is Minimum mean-square error whose expression is given by, where z(k, l) denotes inter-cell interference and z(k, l) combines with n(k, l). It is assumed that z_, n(k, l) and Xs(k, l) are independent of each other. IRC algorithm is given by where Rzz represents the spatial covariance matrix of intercell interference plus noise. Rzz is obtained through average of each pilot position s estimate In order to build an MMSE receiver that can combat inter-cell interference and improve system capacity and BER, the interference covariance matrix has to be estimated, the more accurate the estimate of the covariance matrix the better the receiver performance. In order to build a receiver filter based on the MMSE principle, an UE has to estimate the received interference covariance matrix. In order to suppress interference originating from other than the serving enbs, estimation of the inter-site interference covariance matrix is needed. Two ways of estimating the interference covariance matrix are shown based on received data samples and reference symbols respectively. IV.RESULTS The paper introduced shows the Statistics of Unknown Elements in Estimated Covariance Matrix As the common assumption for rai and rbuv, we propose the Tx correlation value of 1.0. Based on this assumption, both rai and rbuv become zero. The reason for this assumption is that both rai and rbuv depend on the Tx correlation without any a priori information for each correlation value. Here, we consider the physical phenomenon when the proposed rai and rbuv values, i.e., zero, it is assumed that there is no SFBC transmit diversity effect for the interference signal. where is the cross-correlation matrix between Y (k, l) and, and is the auto-correlation matrix of Y (k, l). The MMSE makes a balance between interference and noise. If inter-cell interference appears, an extension t(1) is defined as 131
increase in the number of interfering cells. Therefore, we can say that the impact of the proposed values for the unknown elements, i.e., zero, is small when the number of the interference signals exceeds the degrees of freedom at the receiver. Fig 4.1: Median values of unknown elements in the estimated covariance matrix. When comparing the throughput performance of the one and two interfering cells the performance of the IRC receiver is severely degraded due to increasing the number of interfering cells. To employ the IRC receiver perfectly, it is clear that the values of rai and rbuv, i.e., all of the complex correlation values, must be estimated. However, when these values cannot be estimated, using a common assumption for rai and rbuv is expected to be effective. This is because the actual IRC receiver weight matrix must be generated based on the actual correlation values otherwise, the accuracy of null beam forming severely degrades. As the common assumption for rai and rbuv, we propose the Tx correlation value of 1.0. Based on this assumption, both rai and rbuv become zero. Fig 4.2 User throughput versus average received SNR. Fig 4.2 shows the throughput performance for each IRC receiver when assuming the modeled interference signals. the performance of the ideal IRC receiver with zero padding and the proposed IRC receiver is slightly degraded according to an Fig 4.3.BER performance of MIMO and MMSE equalizer For comparison with IRC receiver the BER performance of the MRC and ZF receiver is also evaluated. V.CONCLUSION Investigation on the IRC receiver for open-loop transmit diversity employing the SFBC required for the extended covariance matrix estimation, some elements in the covariance matrix could not be estimated using the RS-based estimation scheme. Therefore, in this paper investigated the statistics of these unknown elements and proposed appropriate values, specifically inserting zero values, for those elements assuming the LTE/LTE-Advanced downlink. Simulation results showed that the IRC receiver using the proposed scheme, which had two receiver antenna branches, suppressed the inter cell interference and improved the throughput when a cell-edge environment was assumed. Since IRC receiver suffers from inaccurate channel estimation and has poor performance. So that MMSE-IRC receiver is expected to be a significant performance booster and it can suppress both data and CRS interference. REFERENCES 1. Y. Ohwatari, N. Miki, T. Abe, and H. Taoka, Investigation on advanced receiver employing interference rejection combining in asynchronous network for LTE-Advanced downlink, in Proc. IEEE VTC-Spring, May 2012, pp. 1 6. 2. Y. Léost, M. Abdi, R. Richter, and M. Jeschke, Interference rejection combining in LTE networks, 132
Bell Labs Tech. J., vol. 17, no. 1, pp. 25 49, Jun. 2012. 3. Third-Generation Partnership Project (3GPP), TR 36.829 (V11.0.0), Enhanced performance requirement for LTE UE, Sophia-Antipolis, France, Mar. 2012. 4. Y. Ohwatari, N. Miki, T. Asai, T. Abe, and H. Taoka, Performance of advanced receiver employing interference rejection combining to suppress inter-cell interference in LTE-Advanced downlink, in Proc. IEEE VTC-Fall, Sep. 2011, pp. 1 7. LTE system, Proc. IEEE VTC-Spring, Yokohama, Japan, pp.1 5, 2012. 16. M. Li, J. Tan, and W. Zhang, A channel estimation method based on frequency-domain pilots and timedomain processing for OFDM systems, IEEE Tran consumer Electron.,Vol.50,No.4,pp.1049 1057, 2004. 5. Third-Generation Partnership Project (3GPP), TS 36.211 (V10.2.0), Physical channels and modulation, Sophia-Antipolis, France, Jun. 2011. 6. Third-Generation Partnership Project (3GPP), R4-115213, Reference receiver structure for interference mitigation on enhanced performance requirement for LTE UE, NTT DOCOMO, Tokyo, Japan, Oct. 2011 7. Third-Generation Partnership Project (3GPP), TS 36.101 (V10.4.0), User Equipment (UE) radio transmission and reception, Sophia- Antipolis, France, Oct. 2011. 8. J. Zhang, X. Bi, and Y. Wang, Antenna pairing for space frequency block codes in edge-excited distributed antenna systems, in Proc. IEEE PIMRC, Sep. 2010, pp. 117 122 9. J. Lee, R. Arnott, K. Hamabe, and N. Takano, Adaptive modulation switching level control in high speed downlink packet access transmission, in Proc. 3G Mobile Commun. Technol., May 2002, pp. 156 159. 10. J. P. Kermoal, L. Schumacher, K. I. Pedersen, P. E. Mogensen, and F. Frederiksen, A stochastic MIMO radio channel model with experimental validation, IEEE J. Sel. Areas Commun., vol. 20, no. 6, pp. 1211 1226, Aug. 2002. 11. Y. Hara, Weight convergence of adaptive antennas based on SMI algorithm, IEICE Trans. Commun., vol. J84-B, no. 11, pp. 2027 2039, Nov. 2001 12. S. M. Alamouti, A simple transmit diversity technique for wireless communications, IEEE J. Sel. Areas Commun., vol. 16, no. 8, pp. 1451 1458, Oct. 1998 13. K.C. Hung and D.W. Lin, Pilot-based LMMSE channel estimation for OFDM systems with Powerdelay profile approximation, IEEE Trans. Veh. Technol., Vol.59, No.1, pp.150 159, 2010. 14. [9] F.F. Abari, F.K. Sharifabad and O. Edfors, Low complexity channel estimation for LTE in fast fading environments for implementation on multi-standard platforms, Proc. IEEE VTCFall, Ottawa, Canada, pp.1 5, 2010. 15. K. Pietikainen, F.D. Carpio, etc., System-level performance of interference suppression receivers in 133