Theses and Dissertations

Permanent URI for this collectionhttp://ir.daiict.ac.in/handle/123456789/1

Browse

Subject: search.filters.subject.Orthogonal Frequency Division Multiplexing

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    Digital implementation of orthogonal frequency division multiplexing
    (Dhirubhai Ambani Institute of Information and Communication Technology, 2017) Bhanushali, Artiben; Pillutla, Laxminarayana; Bhatt, Amit
    "OFDM (Orthogonal Frequency Division Multiplexing) is the most important part of 4G and 5G technology which is adopted by many standards because of its various advantages. In this thesis, firstly prototype of OFDM system is designed according to 802.11a standard in MATLAB to check its performance and then to implement the same in Verilog using Xilinx ISE 14.4 design suite. FFT (Fast Fourier Transform) and IFFT (Inverse Fast Fourier Transform) are the most complex part of the design which are implemented using radix-2 algorithm. For channel estimation, scattered pilot arrangement is used to insert pilots at regular period in OFDM frame. At receiver side, least-square estimation is used to estimate the channel’s impulse response at known pilot tones. The system analysis is done using MATLAB-Verilog co-simulation in which Verilog transmitter and receiver are connected with MATLAB channel using testbench to generate the text files. This files are used as connecting platform between MATLAB and Xilinx. Performance of channel estimation at different number of multipath is observed which shows that the combination of time interpolation followed by frequency interpolation performs better over only frequency interpolation. In later part, BER (bit error rate) vs. SNR (signal to noise ratio) is analyzed for different number of multipath. From graphs, it is clearly denoted that, BER performance improves with respect to increase in SNR. Finally, fixed point and floating point comparison is carried out which depicts that fixed point system implemented in Verilog performs almost similar to the floating point implementation done in MATLAB."
  • Thumbnail Image
    ItemOpen Access
    Model based channel shortening technique for IEEE 802.11a OFDM system
    (Dhirubhai Ambani Institute of Information and Communication Technology, 2007) Patidar, Pravin; Chakka, Vijaykumar
    Channel shortening tehniques for IEEE802.11a OFDM system are studied. Their performance is evaluated for various standard channel models. A model based channel shortening technique is presented, which reduces the computational comlexity, both during initialization and data transmission. The method is based on modeling of channel impulse response by a pole-zero system, and using the denominator coefficients as the equalizer. Further, it is shown that for channel shortening, calculation of zeros of the model is not required. Comparison of proposed technique with MMSE and MSSNR channel shortening techniques is presented.
  • Thumbnail Image
    ItemOpen Access
    Enhancing data efficiency in OFDM (IEEE 802.11a) using equalization (11A)
    (Dhirubhai Ambani Institute of Information and Communication Technology, 2005) Reddy, S. Gangadhar; Chakka, Vijaykumar
    IEEE 802.11a OFDM signal passed through a dispersive channel introduces ISI & ICI. To avoid ISI & ICI, cyclic prefix (guard interval) is introduced between the OFDM symbols. But the cyclic prefix contributes about 20% (0.8us/4us) of the total symbol duration. Transmission of OFDM symbols without cyclic prefix increases the data efficiency, but introduces ICI & ISI. The ISI & ICI can be removed by equalization at the receiver. Spill over in the frequency domain is same as ICI & ISI. This spill over can be observed at the unused carriers also (carriers numbered from 0-6, 33, 59-64 are unused in IEEE 802.11a). Equalization is done in the frequency domain by making use of spill over in unused carriers. A relationship is brought between the length of the channel impulse response and the number of unused carriers for perfect equalization in the absence or very little AWGN-channel noise. To avoid the noise amplification caused by Zero Forcing equalization and to satisfy the relationship derived, a methodology using Time domain equalizer (TEQ) is proposed and simulation results are presented supporting it. Alternatively, computational efficient algorithm for indoor environment is suggested by transmitting data in unused carriers with the combination of used carriers. By this method, ICI & ISI is removed at the transmitter itself and it even doesn’t amplify the noise.