Faculty Publications

Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736

Publications by NITK Faculty

Browse

Filters

2017 - 20202

Settings

Author: search.filters.author.Shripathi Acharya, U.Subject: search.filters.subject.Probability

Search Results

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    A comprehensive framework for Double Spatial Modulation under imperfect channel state information
    (Elsevier B.V., 2017) G.D., G.S.; Koila, K.; Raghavendra, R.; Shripathi Acharya, U.
    The essential requirement for a 5G wireless communication system is the realization of energy efficient as well as spectrally efficient modulation schemes. Double Spatial Modulation (DSM) is a recently proposed high rate Index Modulation (IM) scheme, designed for use in Multiple Input Multiple Output (MIMO) wireless systems. The aim of this scheme is to increase the spectral efficiency of conventional Spatial Modulation (SM) systems while keeping the energy efficiency intact. In this paper, the impact of imperfect channel knowledge on the performance of DSM system under Rayleigh, Rician and Nakagami-m fading channels has been quantified. Later, a modified low complexity decoder for the DSM scheme has been designed using ordered block minimum mean square error (OB-MMSE) criterion. Its performance under varied fading environments have been quantified via Monte Carlo simulations. Finally, a closed form expression for the pairwise error probability (PEP) for a DSM scheme under conditions of perfect and imperfect channel state information has been derived. This is employed to calculate the upper bound on the average bit error probability (ABEP) over aforementioned fading channels. It is observed that, under perfect and imperfect channel conditions DSM outperforms all the other variants of SM by at least 2dB at an average bit error ratio (ABER) of 10?5. Tightness of the derived upper bound is illustrated by Monte Carlo simulation results. © 2017 Elsevier B.V.
  • No Thumbnail Available
    Item
    Index modulation aided multi carrier power line communication employing rank codes from cyclic codes
    (Elsevier B.V., 2020) Raghavendra, M.A.N.S.; Shripathi Acharya, U.
    In a multi-carrier power line communication (mPLC) with dominant Narrowband and Impulse noise, crisscross errors can be clearly observed. In this work, mPLC employing Rank codes with Index modulation (mPLC-IM) has been considered to provide a reliable high data rate communication over the powerline channel. The rank codes required for this implementation have been derived from cyclic codes over GF(qm) viewed as m×n matrices over GF(q). Encoding has been performed by employing the Galois Field Fourier Transform (GFFT) domain description of cyclic codes. This scheme is able to correct a variety of crisscross errors in mPLC-IM The GFFT approach provides an additional degree of freedom that is offered by choice of free transform component indices. It can be used to design an index key scheme which can enhance the physical layer security of an mPLC system. In the absence of knowledge of the index key, it is observed that the probability of error reaches an error floor of ?10?2, highlighting the need for index key for appropriate decoding. Further, a novel check matrix construction is proposed and used in devising a decoding strategy. It is observed that the proposed decoder is capable of correcting any errors of rank ??m?12?. In mPLC-IM with OFDM, the proposed codes over GF(24) provide an asymptotic gain of approximately 3 dB when compared to the uncoded system. For mPLC-IM with multi-tone Frequency Shift Keying (FSK), the proposed RC over GF(24) provides a 25% improvement in symbol error rate (SER) at lower values of p (probability of occurrence of narrowband noise) when compared to Reed-Solomon (RS) based Constant Weight (CW) CW(13,6,5)2?RS[15,14,2]16 codes. Further, a SER improvement of around 30% is achieved using rank codes (RCs) over GF(28) when compared with CW(9,4,4)2?RS[15,14,2]16. In the presence of dominant background noise, the BER graphs show that the proposed codes are equivalent (slightly superior) in performance as that of Low Rank Parity Check (LRPC)/Gabidulin based designs. In the presence of dominant impulse noise, the proposed system is providing significant gain when compared with the Linearly Pre-coded Orthogonal Frequency Division Multiplexing (LP-OFDM) system and LRPC based scheme. Additionally, simulation results show that, in the absence of an index key, the probability of error reaches the error floor, highlighting the need for index key for appropriate decoding. This can be viewed as the code capable of providing an additional layer of security. © 2019 Elsevier B.V.