C. Vijaykumar

Bio: C. Vijaykumar is an academic researcher from Dhirubhai Ambani Institute of Information and Communication Technology. The author has contributed to research in topics: WiMAX & Channel state information. The author has an hindex of 2, co-authored 2 publications receiving 5 citations.

QR-RLS based Adaptive channel shortening IIR-TEQ for OFDM Wireless LAN

Abstract: In this paper, QR-RLS based IIR adaptive channel TEQ(time domain equalizer) to make the OFDM systems robust for delay spreads exceeding the CP(cyclic prefix) is presented. IIR based TEQ results in much shorter length equalizer compared to FIR based TEQ, at the same time guarantee for global convergence within the training period. The length of zeros of the proposed QR-IIR TEQ is restricted to CP. Givens rotation based QRD updation of TEQ coefficients results in computational advantage. The performance of the proposed method is then compared with conventional LMS and QR-FIR TEQ in terms of convergence, computational complexity and BER. The simulation results show that proposed QR-IIR method gives better results as compared to the aforementioned methods.

A Joint QR-LS Based Coarse-Fine Channel Estimation and QR-LRL Detection for Mobile WiMAX 802.16m

Abstract: In this paper, We extended our previous work of QR-RLS based MIMO(Multiple input Multiple output) channel estimation to Mobile Wimax 802.16m system. Mobile wimax system provides high data rate, also fulfills user's requirement like VOD(Video on demand)at very high vehicle speed and also provides better cell coverage area. Channel estimation is crucial part to achieve this goals especially in fast fading environment. Generally, Mobile Wimax systems uses Preamble and Pilots for channel estimation purpose. In the proposed method both preamble and pilots are jointly used for robust channel estimation. At First, QR-RLS Estimator uses Preamble for coarse channel estimation at start of every frame. Once the coarse channel is estimated, then pilots (scattered throughout time-frequency grid) are jointly used with the coarse channel component to derive the channel fading rate. This fading rate is then used to finely estimate the channel at pilot as well as data subcarrier. Thus robust estimation results without adding any overhead. Jointly estimated channel is then used with QR-LRL based data detection, where hard decision values are calculated. Simulation results are shown under various slow-fast channel fading conditions. Results are compared with pilot based channel estimation with LS(least square) interpolation, which shows that joint coarse-Fine estimation gives better performance.

Elimination of cyclic prefix of OFDM systems using filter bank based multicarrier systems

Abstract: The challenges associated with the OFDM system are addressed. OFDM transmits redundancy in terms of Cyclic Prefix (CP), which decreases the spectral efficiency of the system. A pulse shaping filter for the OFDM systems is proposed to remove the constraint of CP. Also, transmit and receive filters in OFDM system suffer from poor frequency response because of the use of rectangular window function. The non-rectangular window based prototype filters are proposed for the transmitter and receiver end. The comparison of performance of OFDM system and the suggested filter bank based multicarrier systems is provided in terms of BER and the stop band attenuation of the constituent filters.

Performance of Downlink SISO NR System using MMSE-IRC Receiver

Abstract: In this work, we investigated the performance of single input single output (SISO) downlink channel considering 5G new radio (NR). A number of parameters such as different modulation schemes, channel coding with varying code rates, scalable numerology μ and 3GPP channel models have been considered for evaluation. In addition, the minimum mean square error-interference rejection combining (MMSE-IRC) technique for interference mitigation and bit error rate (BER) performance is analyzed and presented. We also compared the sum-rate performance of LTE and 5G NR. It is observed that MMSE-IRC receiver successfully mitigates the interferences compared to only MMSE based receiver. Simulation results also show performance improvement over various parameters like sum-rate, interference mitigation and BER compared to prior technologies i.e. 4G-LTE, WiMAX, etc.

Mode division multiplexing zero forcing equalisation scheme using LU factorization

Abstract: Optical networks is considered as the main backbone networks that handled the Internet traffic worldwide. Currently, the Internet traffic has had huge annual growth due to the increment in connected devices. At this rate, it is believed that the current technology in optical network will not able to handle this growth in the near future. Till recently, multiplexing techniques in the optical communication rely on modulation techniques where polarization, amplitude and frequency of the signal are used as the main data carrier. In these techniques, light modes are considered as an undesired effect causing modal dispersion. In contrast, mode division multiplexing (MDM) was introduced as a multiplexing approach which relies on the utilization of the light modes for the benefit of increasing the capacity-distance product of the optical network. As per any new technology, it is still facing a lot of problems preventing it from being commercially standardized and used. One of the main MDM issues is the mode coupling, which is an inventible phenomena occurs when the energy of one mode transfers to another mode during their propagation throughout the optical fibre causes inter-symbol interference (ISI), increasing the bit error rate (BER) and reducing the overall system performance. Different equalization schemes have been proposed so far attempting to mitigate the effect of mode coupling on the MDM optical signal. However, they suffer from high computational complexity and rely on training signals in estimating the optical channel which increases the overhead payload. These technique mainly rely on Least Mean Squared (LMS) and Recursive Least Squared (RLS) algorithms. The purpose of this study is to introduce a Zero Forcing LU-based equalization scheme for MDM. Previous research in the radio domain on multiple-input multiple output (MIMO) and orthogonal frequency division multiplexing (OFDM) demonstrated that zero forcing schemes have low computational complexity compared to current schemes as they equalize the signal without training signals, thus reducing the overhead payload. All of the previous points motivate the work of this study to adapt this approach in optical communications. The study adopts the four stages of the Design Research Methodology (DRM). The initial data was collected from the optical simulator, processed and used to derive the transfer function (H) of the system. Then it was used to develop the equalization scheme in MATLAB. The experimentation on Zero Forcing LU based equalization scheme shows O(N) complexity which is lower than RLS which has O(N2) and faster than LMS, in fact, LMS needs an average of 0.0126 seconds to process the signal while ZF LU-based needs 0.0029 seconds only. On the other hand, the proposed equalization reduces the time delay spread of the channel, resulting three times increment in the capacity of the MDM channel and even lower computational complexity. The main contribution of this study is the reduction of the computational complexity of the previous equalization schemes in MDM. Applying this scheme in real MDM systems can produce more cost effective and smaller digital signal processing (DSP) parts for MDM equipment and can accelerate the work on the standardization of MDM for being commercially used as a multiplexing technique for optical communication networks.

A multi-stage iir time domain equalizer for ofdm systems with isi

Abstract: In an orthogonal frequency division multiplexing (OFDM) system, it is known that when the delay spread of the channel is larger than the cyclic prefix (CP) size, intersymbol interference will occur. The time-domain equalizer (TEQ), designed to shorten the channel impulse response (CIR), is a common device to solve this problem. Conventionally, the TEQ is treated as a finite-impulse-response (FIR) filter, and many TEQ design methods have been proposed. However, a wireless channel typically has multi-path responses, exhibiting FIR characteristics. Thus, the corresponding TEQ will have an infinite impulse response (IIR), and the FIR modeling of the TEQ is inefficient, i.e., the required order for the TEQ will be high. The conventional approach will then suffer from the high computation complexity problem, both in the derivation of TEQ and in the operation of channel shortening. In this paper, we propose a new scheme to overcome these problems. In the derivation of the TEQ, we propose to use a multistage structure, replacing a high-order TEQ with a cascade of several low-order TEQs. In the shortening operation, we propose to use an IIR TEQ approximating a high-order FIR TEQ. Since the ideal TEQ exhibits low-order IIR characteristics, the order of the IIR TEQ can be much lower than the FIR TEQ. Simulations show that while the proposed method can reduce computational complexity significantly, its performance is almost as good as existing methods.

EVD-LRL based joint channel estimation and detection for very large MIMO systems

Abstract: This paper introduces a novel joint channel estimation and detection method for very large MIMO system. Conventionally, orthogonal pilot sequences are used to determine correct CSI(channel state information). It falls behind due to pilot contamination and spectral inefficiency in large MIMO systems. Many authors suggested promising approaches of blind and semi blind channel estimation[2][3], which work well with a trade-off for complexity. Author[1] suggested EVD-ILSP based estimation, which results in high spectral efficiency compared to conventional method, However, suffers from high complexity due to ILSP method. Here, the proposed method uses enlarged QR-LRL[5] based ordered Detection jointly with the EVD based estimated channel, which not only results in less complexity but also provides better BER performance compared to conventional EVD-ILSP method. Thus the throughput of large MIMO system is increased with reduced complexity. Simulation results demonstrate remarkable improvement in the performance of proposed method over conventional method.