merged file 1

Author: iaetsdiaetsd

Post on 02-Jun-2018




0 download

Embed Size (px)


  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 1

    Implementation of HDLC Protocol Using Verilog

    K. Durga Bhavani1, B. Venkanna

    2, K. Gayathri


    1,2,3Dept. of ECE


    3Intell Engg. College-Anatapur

    AbstractA protocol is required to transmit data successfu ll y over any network and also to manage the flow at which

    data is transmi tted. HDLC protocol i s the high-level data li nk control protocol established by I nternational Organi zation

    for Standardization (I SO), which is widely used in digital commun ications. H igh-l evel Data L ink Control (H DLC) is th

    most commonly used Layer2 protocol and is sui table for bit ori ented packet transmission mode. This paper discusses the

    Veril og modeli ng of single-channel HDLC Layer 2 protocol and its implementation using Xil inx.

    Keywords- H igh Level Data l ink Control (H DL C), F rame Check Sequence (FCS), and Cycli c Redundancy Chec



    HDLC protocol is the high-level data link control protocol

    established by International Organization for standardization

    (ISO), which is widely used in digital communication and

    are the bases of many other data link control protocols [2].

    HDLC protocols are commonly performed by ASIC

    (Application Specific Integrated Circuit) devices, software

    programming and etc.

    The objective of this paper is to design and implement a

    single channel controller for the HDLC protocol which is the

    most basic and prevalent Data Link layer synchronous, bit-oriented protocol. The HDLC protocol (High Level Data

    link Control) is also important in that it forms the basis for

    many other Data Link Control protocols, which use the same

    or similar formats, and the same mechanisms as employed in


    HDLC has been so widely implemented because it

    supports both half duplex and full duplex communication

    lines, point to point(peer to peer) and multi-point

    networks[1]. The protocols outlined in HDLC are designed

    to permit synchronous, code-transparent data transmission.

    Other benefits of HDLC are that the control information is

    always in the same position, and specific bit patterns used for

    control differ dramatically from those in representing data,

    which reduces the chance of errors.


    The HDLC Protocol Controller is a high-performance

    module for the bit-oriented packet transmission mode. It issuitable for Frame Relay, X.25, ISDN B-Channel (64 Kbits/s

    and D-Channel (16 Kbits/s) The Data Interface is 8-bit wide

    synchronous and suitable for interfacing to transmit and

    receive FIFOs. Information is packaged into an envelope

    called a FRAME [4]. An HDLC frame is structured as



    8 bits 8 bits 8 /16 bits variable 8 8 bits

    Table 1. HDLC Frame

    A. Flag

    Each Frame begins and ends with the Flag Sequence which

    is a binary sequence 01111110. If a piece of data within the

    frame to be transmitted contains a series of 5 or more 1s, thetransmitting station must insert a 0 to distinguish this set of

    1s in the data from the flags at the beginning and end of the

    frame. This technique of inserting bits is called bit-stuffing[3].

    B. Address

    Address field is of programmable size, a single octet or a

    pair of octets. The field can contain the value programmedinto the transmit address register at the time the Frame is


    C. Control

    HDLC uses the control field to determine how to control the

  • 8/10/2019 Merged File 1


    www. Special Issue-1,

    October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 2

    communications process. This field contains the commands,

    responses and sequences numbers used to maintain the data

    flow accountability of the link, defines the functions of the

    frame and initiates the logic to control the movement of

    traffic between sending and receiving stations.

    D. Information or Data

    This field is not always present in a HDLC frame. It is onlypresent when the Information Transfer Format is being used

    in the control field. The information field contains the

    actually data the sender is transmitting to the receiver.

    E. FCS

    The Frame Check Sequence field is 16 bits. The FCS is

    transmitted least significant octet first which contains thecoefficient of the highest term in the generated check

    polynomials. The FCS field is calculated over all bits of theaddresses, control, and data fields, not including any bits

    inserted for the transparency. This also does not include theflag sequence or the FCS field itself. The end of the datafield is found by locating the closing flag sequence and

    removing the Frame Check Sequence field (receiver section)[5].


    In this design, HDLC procedures contain two modules, i.e.

    encoding-and-sending module (Transmitter) and receiving-

    and-decoding module (receiver). The function diagram isshown as below.

    Fig.1. HDLC Block Design

    Form this diagram we know that, transmitter module

    includes transmit register unit, address unit, FCS generationunit, zero insertion unit, Flag generation unit, control and

    status register unit and transmit frame timer and

    synchronization logic unit. Receiver module includes receiveregister unit, address detect unit, FCS calculator unit, zero

    detection unit, flag detection unit, receive control and staturegister unit and frame timer and synchronization logic unit.

    A. Transmitter Module

    The Transmit Data Interface provides a byte-wide

    interface between the transmission host and the HDLC

    Controller. The Transmit data is loaded into the controller on

    the rising edge of Clock when the write strobe input is

    asserted. The Start and End bytes of a transmitted HDLC

    Frame are indicated by asserting the appropriate signals with

    the same timing as the data byte.The HDLC Controller will, on receipt of the first byte of a

    new packet, issue the appropriate Flag Sequence and

    transmit the Frame data calculating the FCS. When the las

    byte of the Frame is seen the FCS is transmitted along with a

    closing Flag. Extra zeros are inserted into the bit stream to

    avoid transmission of control flag sequence within the Frame


    The Transmit Data is available on TxD pin with

    appropriate to be sampled by Clk. If TxEN is de-asserted

    transmit is stalled, and TxD pin is disabled.

    A transmit control register is provided which can enable

    or disable the channel. In addition it is possible to force thetransmission of the HDLC Abort sequence. This will cause

    the currently transmitted Frame to be discarded. The transmi

    section can be configured to automatically restart after an

    abort, with the next frame, or to remain stalled until the host

    microprocessor clears the abort.

    B. Receiver Module

    The HDLC Controller Receiver accepts a bit stream on

    port RxD. The data is latched on the rising edge of Clock

    under the control of the Enable input RxEN. The Flag

    Detection block searches the bit stream for the Flag

    Sequence in order to determine the Frame boundary. Anystuffed zeros are detected and remove and the FCS is

    calculated and checked. Frame data is placed on the Receive

    Data Interface and made available to the host. In addition

    Flag information is passed over indicating the Start and the

    End byte of the HDLC Frame as well as showing any error

    conditions which may have been detected during receipt of

    the Frame.

    In normal HDLC protocol mode, all Receiver Frames are

    presented to the host on the output register. A status registe

  • 8/10/2019 Merged File 1


    www. Special Issue-1,

    October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 3

    is provided which can be used to monitor status of the

    Receiver Channel, and indicate if the packet currently being

    received includes any errors.


    Fig.2. Simulation Waveform

    Device Utilization Report: Clock Frequency: 78.2 MHz

    Resource Used Avail Utilization

    IOs 60 180 33.33%

    Function Generators 205 1536 13.35%

    CLB Slices 103 768 13.41%

    Dffs or Latches 108 1536 7.03%

    Table 2. Synthesis Report


    We designed HDLC protocol sending and receiving RTL

    level modules in Verilog and had them tested successfully,

    which has the following advantages like easy to program andmodify, suitable for different standards of HDLC procedures,

    match with other chips with different interfaces. So thisproposed method can be more useful for many applicationslike a Communication protocol link for RADAR data



    [1] Implementation of HDLC protocol Using FPGA,[IJESAT] International Journal of Engineering Science

    & Advanced Technology, ISSN: 2250-3676, Volume-2

    Issue-4, 11221131.[2] M.Sridevi, DrP.Sudhakar Reddy / International Journa

    of Engineering Research and Applications (IJERA)

    ISSN: 2248-9622 Vol. 2, Issue 5September- October 2012, pp.2217-2219.

    [3] ISO/IEC 13239, Information technology -Telecommunications and Information exchange betweensystems High-level data link control (HDLC)

    procedures, International Organization for

    Standardization, pp 10-17, July 2002.[4] A.Tannenbaum, Computer Networks, Prentice Hall of

    India, 1993.[5] Mitel Semiconductor, MT8952B HDLC Protoco

    Controller, Mitel Semiconductor Inc., pp 2-14, May

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 4

    p Iterative MMSE-PIC Detection Algorithm for

    MIMO OFDM SystemsGorantla Rohini Devi1, K.V.S.N.Raju2, Buddaraju Revathi3

    1Department of ECE, 2Head of ECE Department, 3Asst. Professor, Department of ECE,

    SRKR Engineering College

    Bhimavaram, AP, India

    Abstract- Wir eless communi cation systems are requir ed to provide high data rates, which is essential for many services such

    as video, high quali ty audio and mobile integrated services. When data transmission is aff ected by fading and in terf erence

    eff ects the inf ormation will be altered. Multiple I nput Multi ple Output (MIMO) technique is used to reduce the multipath

    fading. Orthogonal F requency Division Multiplexing (OFDM ) is one of the promising technologies to mitigate the ISI. The

    combination of M IMO-OFDM systems off ers high spectrum effi ciency and diversity gain against multi path fading channels.

    Di ff erent types of detectors such as ZF , MMSE and PI C, I terati ve PIC. These detectors improved the quality of received

    signal in high in terf erence envir onment. Implementations of these detectors veri fi ed the improvement of the BER v/s SNR

    perf ormance. I terati ve PIC technique give best perf ormance in noise envir onment compared to ZF, MMSE and PIC.

    Keywords: Orthogonal Fr equency Division Mul tiplexing (OFDM), Multiple Input Mu ltiple Output (M IMO), Zero Forcing(ZF ), Mi nimum Mean Square Er ror (MMSE), Parallel I nterf erence Cancellation (PI C), Bit Er ror Rate (BER), Signal to

    Noise Ratio (SNR), I nter Symbol Interference (I SI ), Binary Phase Shi ft Keying (BPSK).


    In wireless communication the signal from a transmitter

    will be transmitted to a receiver along with a number of

    different paths, collectively referred as multipath. These

    paths may causes interference from one another and result in

    the original data being altered. This is known as Multipath

    fading. Furthermore wireless channel suffer from co-channel

    interference (CCI) from other cells that share the same

    frequency channel, leading to distortion of the desired signaland also low system performance. Therefore, wireless system

    must be designed to mitigate fading and interference to

    guarantee a reliable communication.

    High data rate wireless systems with very small symbol

    periods usually face unacceptable Inter Symbol Interference

    (ISI) originated from multi-path propagation and their

    inherent delay spread. Orthogonal Frequency Division

    Multiplexing (OFDM) has emerged as one of the most

    practical techniques for data communication over frequency-

    selective fading channels into flat selective channels. OFDMis one of the promising technologies to mitigate the ISI. On

    the other hand, to increase the spectral efficiency of wireless

    link, Multiple-Input Multiple-Output (MIMO) systems [1]. It

    is an antenna technology that is used both in transmitter and

    receiver equipment for wireless radio communication. MIMO

    exploit the space dimension to improve wireless system

    capacity, range, and reliability. MIMO system can be

    employed to transmit several data streams in parallel at the

    same time and on the same frequency but different transmit


    MIMO systems arise in many modern communication

    channels such as multiple user communication and multiple

    antenna channels. It is well known that the use of multiple

    transmit and receive antennas promises sub performance gains

    when compared to single antenna system. The combination

    MIMO-OFDM system is very natural and beneficial since

    OFDM enables support of more antennas and large bandwidth

    since it simplifies equalization in MIMO systems. In MIMO-OFDM system offers high spectral efficiency and good

    diversity gain against multipath fading channels [2][3].

    In MIMO system depends on the different detection

    techniques used at the MIMO receiver. The better detector

    that minimizes the bit error rate (BER) is the maximum

    likelihood (ML) detector. But the ML detector is practically

    difficult as it has computational complexity is exponential. On

    the other hand, linear detectors, such as zero-forcing (ZF) and

    minimum mean square error (MMSE) receivers, have low

    decoding complexity, but detection performance decrease in

    portion to the number of transmit antennas.

    Therefore, there has been a study on a low complexity

    nonlinear receiver, namely, parallel interference cancellation(PIC) receiver, which parallely decodes data streams through

    nulling and cancelling. PIC algorithm [4] relies on a parallel

    detection of the received block. At each step all symbols are

    detected by subtracted from the received block. PIC detection

    is used to reduce the complexity and prevents error

    propagation. The PIC detection uses the reconstructed signal

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 5

    to improve the detection performance by using iteration

    process. Iterative MMSE-PIC detection algorithm [5][6] best

    detection technique compared all nonlinear receivers. For

    improving the performance of overall system, the output of

    detector is regarded as input of the PIC detection to do again.

    By exchanging information between the MIMO detection and

    decoder, the performance of receiver may greatly beenhanced.

    Where number of iteration increases to improve the bit error

    rate (BER) performance.

    PIC introduces parallely, which enables to reduce the

    interference and therefore increases the reliability of the

    decision process. The channel as a flat fading Rayleigh

    multipath channel and the modulation as BPSK has been

    taken. MIMO-OFDM technology has been investigated as the

    infrastructure for next generation wireless networks.


    Consider a MIMO OFDM system with transmitting and

    receiving antennas. When the MIMO technique of spatial

    multiplexing is applied encoding can be done either jointly

    over the multiple transmitter branches.




    Y1 Y2 YL

    Fig1.Schematic of PIC detection for MIMO OFDM system

    According to the block diagram in Figure1 consists of two

    users, one user source while the other user as destination. The

    two users interchange their information as source to different

    instant of time. In MIMO channel model, L simultaneous

    antennas having same data for transmission, while receiver

    has P antennas.

    The binary data are converted into digitally modulated signal

    by using BPSK modulation technique and after that converted

    from serial to parallel through convertor. The digitally

    modulated symbols are applied to IFFT block. After the

    transformation, the time domain OFDM signal at the output of

    the IFFT. After that, Cyclic Prefix (CP) is added to mitigate

    the ISI effect. This information is sent to parallel to serial

    convertor and again, the information symbols are

    simultaneously transmitted over the MIMO channel and laterAWGN noise added at receiver side.

    At the receiver side, firstly serial to parallel conversion occurs

    and cyclic prefix removed. The received signals samples are

    sent to a fast Fourier transform (FFT) block to demultiplex the

    multi-carrier signals and ZF / MMSE / PIC / Iterative-PIC

    detectors is used for separating the user signals at each

    element of the receiver antenna array. Finally demodulated

    outputs and the resulting data combined to obtain the binary

    output data.

    MIMO Techniques:

    Current MIMO system includes MISO and SIMO system thatuses MIMO technique to improve the performance of wireless

    system can be divided into two kinds. One is spatial

    multiplexing which provides a linear capacity gain in relation

    to the number of transmitting antenna and the other is spatial

    diversity schemes which can reduce the BER and improve the

    reliability of wireless link.

    A. Spatial Multiplexing

    The transmission of multiple data stream over more than one

    antenna is called spatial multiplexing. It yields linear (In the

    minimum number of transmit and receive antenna) capacity

    increases, compared to systems with a single antenna at one or

    both sides of the wireless link, at no additional power or

    bandwidth expenditure. The corresponding gain is available if

    the propagation channel exhibits rich scattering and can be

    realized by the simultaneous transmission of independent data

    stream in the same frequency band. The receiver exploits

    difference in the spatial signature induced by the MIMO

    channel onto the multiplexed data stream to separate the

    different signals, there by realizing a capacity gain.

    B. Diversity Schemes

    In which two or more number of signals sent over different

    paths by using multiple antennas at the transmitting and

    receiving side. The space is chosen, in such a way the

    interference between the signals can be avoided. To improve

    the link reliability we are using diversity schemes. Spatial

    diversity improves the signal quality and achieves higher

    signal to noise ratio at the receiver side. Diversity gain is

    obtained by transmitting the data signal over multiple

    independently fading dimensions in time, frequency, and
























  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 6

    space and by performing proper combing in the receiver.

    Spatial diversity is particularly attractive when compared to

    time or frequency diversity, as it does not incur expenditure in

    transmission time or bandwidth. Diversity provides the

    receiver with several (ideally independent) replicas of the

    transmitted signal and is therefore a powerful means to

    combat fading and interference and there by improve linkreliability.

    Two kinds of spatial diversities are considered, Transmitter

    diversity and Receiver diversity. There are two famous space

    time coding schemes. Space time block code (STBC) and

    Space time trellis code (STTC).



    The co-channel interference is one of the major limitations in

    cellular telephone network. In the case of cellular network

    such as 3G or beyond 3G (4G), the co-channel interference is

    caused by the frequency reuse. Our main idea is to reject the

    co- channel interference in MIMO-OFDM cellular systems.

    To eliminate the inter symbol interference (ISI) different types

    of highly interference channel equalization techniques are

    used. MIMO-OFDM detection method consists of linear and

    nonlinear detection methods. Linear equalizers are ZF [7] and

    MMSE [8] and nonlinear equalizers are PIC and Iterative PIC.

    1. Zero Forcing (ZF) equalizer:

    Zero forcing Equalizer is a linear equalization algorithm used

    in communication systems, it inverse the frequency response

    of the channel. The output of the equalizer has an overall

    response function equal to one of the symbol that is beingdetected and an overall zero response for the other symbols. If

    possible, this results in the removal of the interference from

    all other symbols in the absence of the noise.

    Zero Forcing is a linear method that does not consider the

    effects of noise. In fact, the noise may be enhanced in the

    process of eliminating the interference.

    Consider a 2x2 MIMO system. The received signal on the first

    antenna is given by:


    1 1,1 1 1,2 2 1 1,1 1,2 1


    xh x h x n h h n



    The received signal on the second antenna is given by:


    2 2,1 1 2,2 2 2 2,1 2,2 2


    xh x h x n h h n




    y1 and y2 are the received symbol on the first and second

    antenna, h1,1 is the channel from 1st

    transmit antenna to 1st

    receive antenna, h1,2 is the channel from 1st

    transmit antenna to


    receive antenna, h2,1 is the channel from 2nd


    antenna to 1st

    receive antenna, h2,2 is the channel from 2nd

    transmit antenna to 2nd

    receive antenna, x1, and x2 are thetransmitted symbols and n1 and n2 are the noise on 1

    stand 2


    receive antennas respectively.

    The sampled baseband representation of signal is given by:

    y= Hx+n



    y = Received symbol matrix,

    H = Channel matrix,

    x = Transmitted symbol matrix,

    n = Noise matrix.

    For a system with NT transmit antennas and NR receiver

    antennas, the MIMO channel at a given time instant may berepresented as NT x NR matrix:


    1,1 1, 2 1,

    2 ,1 2 ,2 2 ,

    ,1 ,2



    R R R T



    N N N N

    H H H

    H H HH

    H H H


    To solve for x, we find a matrix W which satisfies WH = I.

    The Zero Forcing (ZF) detector for meeting this constraint is

    given by,

    W = (HHH)




    W= Equalization matrix

    H= Channel matrix

    This matrix is known as the pseudo inverse for a general m x

    n matrix where


    It is clear from the above equation that noise power may

    increase because of the factor (HHH)

    -1. Using the ZF

    equalization approach, the receiver can obtain an estimate ofthe two transmitted symbols and x1 and x2 i.e.




    x= (H









    * *1,1 1, 21,1 2 ,1

    * *2 ,1 2 ,21, 2 2 , 2

    Hh hh h

    H Hh hh h

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 7

    2.Minimum Mean Square Error (MMSE) Equalizer:

    A MMSE estimator is a method in which it minimizes the

    mean square error (MSE), which is a universal measure of

    estimator quality. The most important characteristic of MMSE

    equalizer is that it does not usually eliminate ISI totally butinstead of minimizes the total power of the noise and ISI

    components in the output. If the mean square error between

    the transmitted symbols and the outputs of the detected

    symbols, or equivalently, the received SNR is taken as the

    performance criteria, the MMSE detector [9] is the optimal

    detection that seeks to balance between cancelation of the

    interference and reduction of noise enhancement.

    The received signal on the first receive antenna is,


    1 1,1 1 1,2 2 1 1,1 1,2 1


    xy h x h x n h h n



    The received signal on the second antenna is,


    2 2,1 1 2,2 2 2 2,1 2,2 2


    xy h x h x n h h n




    y1, y2 are the received symbol on the 1st

    and 2nd


    respectively, h1,1 is the channel from 1st

    transmit antenna to 1st

    receive antenna, h1,2 is the channel from 1st

    transmit antenna to


    receive antenna, h2,1 is the channel from 2nd


    antenna to 1st

    receive antenna, h2,2 is the channel from 2nd

    transmit antenna to 2nd

    receive antenna, x1, x2 are the

    transmitted symbols and n1, n2 is the noise on 1


    , 2



    The above equation can be represented in matrix notation as


    1,1 1,21 1 1

    2,1 2,22 2 2

    h h x n

    h h x n


    Equivalently, y = Hx+n

    To solve for x, we know that we need to find a matrix W

    which satisfies WH=I. The Minimum Mean Square Error

    (MMSE) linear detector for meeting this constraint is givenby,





    Using MMSE equalization, the receiver can obtain an estimate

    of the two transmitted symbols x1, x2, i.e.




    x= (H








    3.Parallel Interference Cancellation (PIC):

    Here the users symbols are estimated in a parallel manner.

    This detects all layers simultaneously by subtractinginterference from other layers regenerated by the estimation

    from ZF or MMSE criteria.

    PIC detection is used to reduce the complexity and prevents

    error propagation. The parallel MMSE detector consists of

    two or more stages. The first stage gives a rough estimation of

    substreams and the second stage refines the estimation. The

    output can also be further iterated to improve the performance.

    The first stage will be implemented by using either ZF or

    MMSE detection algorithm. The MMSE detector minimizes

    the mean square error between the actually transmitted

    symbols and the output of the linear detector is








    By using MMSE detector the output of the first stage is

    d = Dec(W.y) (14)

    Where, W is the parameter of Equalization matrix which is

    assumed to be known and Dec(.) is the decision operation. In

    each a vector symbol is nulled.

    This can be written as

    S=I.d (15)

    Where, I is identity matrix and d is rough Estimated symbols

    of MMSE.

    The PIC detection algorithm can be expressed as

    R=y-H.S (16)

    Hence S is the estimated symbols of MMSE Equalizer. The

    estimated symbol using the detection scheme of the

    appropriate column of the channel matrix

    Z= Dec(W.R) (17)


    R is the output of PIC Equalizer

    W is the parameter of MMSE Equalization matrix

    Z is the estimated symbols of PIC Equalizer

    4.Iterative PIC detection:

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 8

    In which, the estimated signal by decoder is used to

    reconstruct the transmitted code signal. The PIC detection

    uses the reconstructed signal to improve the detection

    performance by using iterative process.

    PIC cancellation estimates and subtract out all the interference

    for each user in parallel in order to reduce the time delay. At

    iteration process the output of PIC detector is given it as input.Combing MMSE detection with the PIC cancellation directly

    impacts on the global performance of the systems and also on

    the associated complexity. The complexity directly linked

    with the number of iterations for the detection.

    The Iterative PIC detection scheme based on MIMO system

    algorithm is given by:

    For i = 1: nT

    nT - 1

    c = y - H (: , J). Z


    E = Dec (W. c)(18)


    E is the estimation of transmitted symbols of iterative PIC


    W is the MMSE equalization matrix,

    c is the output of iterative PIC detector,

    nT is the number of transmitting antennas.


    In all simulation results shown by using four equalizers (ZF,

    MMSE, PIC and Iterative PIC) in MIMO OFDM system.Rayleigh fading channel is taken and BPSK modulation

    scheme was used. Channel estimation as well as

    synchronization is assumed to be ideal. We analyze the BER

    performance of data transmission in Matlab software.

    Fig. 2. BER for BPSK modulation with ZF and MMSE

    equalizers in 2x2 MIMO-OFDM system.

    From the plot it is clear that 2x2 MIMO-OFDM system with

    MMSE equalizer for case of pure equalization compared to ZF

    equalizer. Modulation scheme employed here is BPSK.

    Fig. 3. Performance comparison of PIC and Iterative PICequalizers in 2x2 MIMO-OFDM system.

    From the plot it is clear that 2x2 MIMO-OFDM system with

    Iterative PIC equalizer for case of pure equalization compared

    to PIC equalizer. The code BER of proposed scheme is

    produced after iteration. when iteration increases the BER is

    significantly improved. From simulation results the proposed

    scheme Iterative PIC is quite effective compared to PIC.

    Modulation scheme employed here is BPSK.

    Fig. 4. Performance comparison of ZF, PIC and Iterative PIC

    equalizers in 2x2 MIMO-OFDM system.

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 9

    From the plot it is clear that 2x2 MIMO-OFDM system with

    Iterative PIC equalizers for case of pure equalization

    compared to ZF, MMSE, and PIC equalizer. The code BER of

    proposed scheme is produced after iteration. when iteration

    increases the BER is significantly improved. The Zero

    Forcing equalizer removes all ISI and is ideal only when the

    channel is noiseless. From simulation results the proposedscheme Iterative PIC is quite effective compared to ZF and

    PIC. Modulation scheme employed here is BPSK.

    Fig .5. Performance comparison of ZF, MMSE, PIC and

    Iterative PIC equalizers in 2x2 MIMO-OFDM system .

    From the plot it is clear that 2x2 MIMO-OFDM system with

    Iterative PIC equalizers for case of pure equalization

    compared ZF, MMSE, and PIC equalizer. The code BER of

    proposed scheme Iterative PIC is produced after iteration.

    when iteration increases the BER is significantly improved.

    From simulation results the proposed scheme is quite effective

    in all simulation configurations. However, Iterative PIC

    detection scheme is better in the diversity gain and when the

    intefrence comes from the other layers is completely

    cancelled. Modulation scheme employed here is BPSK.


    The combination of MIMO-OFDM systems are used toimprove the spectrum efficiency of wireless link reliability in

    wireless communication systems. Iterative PIC scheme for

    MIMO OFDM systems transmission including the feasibility

    of using the priori information of the transmit sequence of

    MMSE compensation. Performance of Iterative PIC detection

    technique is better compared to ZF, MMSE, PIC using BPSK

    modulation scheme in high interference environment. The

    simulation result shows that the performance of proposed

    scheme is greatly improved compared to other detection

    receivers for MIMO-OFDM systems.


    Any type of modulation techniques such as QPSK or QAM

    will integrate the channel encoding part.


    [1] I. E. Telatar, Capacity of multiple-antenna Gaussian

    channels, Eur. Trans. Telecommun., vol. 10, no. 6, pp.

    585595, Nov/Dec. 1999.

    [2] G. J. Foschini and M. J. Gans, On limits of wireless

    communications in a fading environment when using

    multiple antennas, Wirel. Pers. Commun., vol. 6, no. 3,

    pp. 311335, Mar. 1998.[3] A. Paulraj, R. Nabar, and D. Gore,Introduction to Space

    Time Wireless Communications, 1st ed. Cambridge, U.K.:

    Cambridge Univ. Press, 2003

    [4] Junishi Liu, Zhendong Luo,Yuanan Liu, MMSEPIC


    Transaction Communication., vol.1, oct.2005 .

    [5] Hayashi,H.Sakai , Parallel Interference Canceller with

    Adaptive MMSE Equalization for MIMO-OFDM

    Transmission, France telecom R&D Tokyo.

    [6] Z.Wang, Iterative Detection and Decoding with PIC

    Algorithm for MIMO OFDM System

    ,Int.J.communication, Network and System Science,published august 2009.

    [7] V.JaganNaveen, K.MuraliKrishna, K.RajaRajeswari

    "Performance analysis of equalization techniques forMIMO systems in wireless communication" International

    Journal of Smart Home, Vol.4, No.4, October, 2010

    [8] Dhruv Malik, Deepak Batra "Comparison of various

    detection algorithms in a MIMO wireless communication

    receiver" International Journal of Electronics and

    Compute Science Engineering, Vol.1, No 3, page


    [9] J.P.Coon and M. A. Beach, An investigation od MIMO

    single-carrier frequency-domain MMSE equalizer in

    Proc London comm.Symposium, 2002,pp. 237-240.

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 10

    Computational Performances of OFDM using

    Different Pruned Radix FFT AlgorithmsAlekhya Chundru1, P.Krishna Kanth Varma2

    M.Tech Student, Asst Professor Department Of Eelectronics and Communications,

    SRKR Engineering College,

    Andhra Pradesh, India

    Abstract- The Fast Four ier Transform (FFT) and its inverse (I FFT) are very important algori thms in signal processi

    oftware-defi ned radio, and the most promi sing modulation technique i.e. Orthogonal F requency Division Mul tiplex

    OFDM). Fr om the standard structure of OFDM we can f ind that I FF T/FFT modules play the vital r ole for any OFDM ba

    ransceiver. So when zero valued inputs/outputs outnumber nonzero inputs/outputs, then general I FFT/FFT al gorithm

    OFDM is no longer eff icient in term of execution time. It is possible to reduce the execution time by pruning the FFT. I n

    paper we have implemented a novel and eff icient input zero traced radix FFT prun ing (algori thm based on radix-2 DI F F

    adix-4 DI F FFT, radix-8 DIF F FT ). An intui tive comparison of the computational complexity of orthogonal frequency divis

    mul tiplexing (OFDM ) system has been made in terms of complex calculations requi red using di ff erent radix Fast Fouransform techniques with and without pruni ng. The dif ferent transform techniques are intr oduced such as various types of F

    Four ier transform (FF T) as radix-2 FFT, radix-4 FF T, radix-8 FFT, mixed radix 4/2, mixed radix 8/2 and split r adix 2/4. W

    ntui tive mathematical analysis, it has been shown that with the reduced complexity can be offered with prun ing, OFD

    perf ormance can be greatly impr oved in terms of calculati ons needed.

    ndex terms- OFDM (Orthogonal frequency division multiplexing), Fast Fourier Transform (FFT), Pruni ng Techniqu



    Orthogonal Frequency Divisional Multiplexing (OFDM) is

    modulation scheme that allows digital data to be efficiently

    nd reliably transmitted over a radio channel, even in multi-path

    nvironments [1]. In OFDM system, Discrete Fourier

    Transforms (DFT)/Fast Fourier Trans- forms (FFT) are used

    nstead of modulators. FFT is an efficient tool in the fields of

    ignal processing and linear system analysis. DFT isn't

    eneralized and utilized widely until FFT was proposed. But the

    nherent contradiction between FFT's spectrum resolution and

    omputational time consumption limits its application. To match

    with the order or requirement of a system, the common method

    s to extend the input data sequence x(n) by padding number of

    eros at the end of it and which is responsible for a increased

    alue of computational time. But calculation on undesired

    requency is unnecessary. As the OFDM based cognitive radio

    2] has the capability to nullify individual sub carriers to avoidnterference with the licensed user. So, that there could be a

    arge number of zero valued inputs/outputs compare to non-zero

    erms. So the conventional radix FFT algorithms are no longer

    fficient in terms of complexity, execution time and hardware

    rchitecture. Several researchers have proposed different ways

    to make FFT faster by pruning the conventional radix F


    In this paper we have proposed an input zero traced ra

    DIF FFT pruning algorithm for different radix FFT algorith

    suitable for OFDM based transceiver. The computatio

    complexity of implementing radix-2, radix-4, radix-8, mi

    radix and split radix Fast Fourier Transform with and with

    pruning has been calculated in an OFDM system and compa

    their performance. Result shows IZTFFTP of radix algorith

    are more efficient than without pruning.


    OFDM is a kind of FDM (Frequency Divis

    Multiplexing) technique in which we divide a data stream in

    number of bit streams which are transmitted through s

    channels [3].

    The characteristics of these sub-channels are that they orthogonal to each other. As the data that are transmi

    through a sub-channel at a particular time are only a portion

    the data transmitted through a channel so bit rate in a s

    channel can be kept much low. After splitting the data in

    parallel data streams each stream is then mapped to a tone

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 11

    nique frequency and combined together using the Inverse Fast

    ourier Transform (IFFT) to yield the time domain waveform to

    e transmitted [4]. After IFFT is done, the time domain signals

    re then converted to serial data and cyclic extension is added to

    he signal. Then the signal is transmitted. At the receiving side

    we do the reverse process to get original data from the received

    ne [4,5].In case of deep fade, several symbols in single carrier is

    amaged seriously, but in parallel transmission each of N

    ymbol is slightly affected. So even though the channel is

    requency selective, the sub-channel is flat or slightly frequency

    elective. This is why OFDM provide good protection against

    ading [6].

    In an OFDM system there are N numbers of sub-channels.

    fN is high then it will be very complex to design a system with

    N modulators and demodulators. Fortunately, it can be

    mplemented alternatively using DFT/FFT to reduce the high

    omplexity. A detailed system model for OFDM system is

    hown in Figure 1 [5,6].

    Figure1: OFDM System Model


    Discrete Fourier Transform (DFT) computational

    omplexity is so high that it will cause a long computational

    ime and large power dissipation in implementation. Cooley and

    Tukey provided a lot of ways to reduce the computatio

    complexity. From that, many fast DFT algorithms have b

    developing to reduce the large number of the computatio

    complexity, and these fast DFT algorithms are named

    Fourier transform (FFT) algorithms. Decomposing is

    important role in the FFT algorithms. There are

    decomposed types of the FFT algorithm. One is decimationtime (DIT), and the other is decimation-in-frequency (D

    There is no difference in computational complexity betw

    these two types of FFT algorithm. Different Radix D

    algorithms we used are

    A. Radix-2 DIF FFT Algorithm

    Decomposing the output frequency sequence X[k] into

    even numbered points and odd numbered points is the

    component of the Radix-2 DIF FFT algorithm [6]. We

    divideX[k] into 2r and 2r+1, then we can obtain the follow


    2= () (1)

    2+ 1= () (2)

    = 0,1,2, . . , 2 1

    Because the decomposition of the Equation (1)

    Equation (2) are the same, we only use Equation (1) to exp

    as shown in Equation (3).

    2= ()

    + () (3)

    Finally, by the periodic property of twiddle factors, we

    get the even frequency samples as









    = 0,1,2, . . , 2 1Similarly, the odd frequency samples is

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 12

    2+ 1= /+ 2


    = 0,1,2, . . , 1 (5) From Equation4) and (5), we can find out the same components, x[n] and[n+N/2], so we can combine the two equations as one basic

    utterfly unit shown in Figure 2. The solid line means that x[n]

    dds x[n + N / 2] , and the meaning

    f the dotted line is thatx[n] subtractsx[n +N / 2] .

    Figure 2: The butterfly signal flow graph of radix-2 DIF FFT

    We can use the same way to further decompose N-point

    DFT into even smaller DFT block. So from the radix-2 dif FFT,

    here is a reduction of number of multiplications, which is about

    factor of 2, showing the significance of radix-2 algorithm for

    fficient computation. So this algorithm can compute N-point

    FT inN/2 cycles.

    B.Radix-4 DIF FFT

    In case N-data points expressed as power of 4M

    , we can

    mploy radix-4 algorithm [9] instead of radix-2 algorithm for

    more efficient estimation. The FFT length is 4M, where M is theumber of stages. The radix-4 DIF fast Fourier transform (FFT)

    xpresses the DFT equation as four summations then divides it

    nto four equations, each of which computes every fourth output

    ample. The following equations illustrate radix-4 decimation in


    ()= ()








    + ()


    Equation (7) can thus be expressed as


    ()+ ()(+ 4 ) + (1)

    (+ 2 ) + ()(+ 3 4 )


    So, Equation (8) can then be expressed as four N/ 4 point DF

    The simplified butterfly signal flow graph of radix-4 DIF FFT

    shown in Figure 3.

    Figure 3: The simplified butterfly signal flow graph of radix


    This algorithm results in (3/8)N log compmultiplications and (3/2)N log

    complex additions. So

    number of multiplications is reduced by 25%, but the numbeaddition is increased by 50%.

    C.Radix-8 DIF FFT

    Comparing with the conventional radix-2 FFT algorit

    and radix-4 FFT algorithm, the advantage of developing radi

    FFT algorithm is to further decrease the complexities, especi

    the number of complex multiplications in implementation.

    can split Equation (2.1) and replace index k with eight pa

    including 8r, 8r+1,8r+2, 8r+3, 8r+4, 8r+5, 8r+6, and 8r

    Hence, we can rewrite Equation (6) and obtain the Equation (

    (8+ )=

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 13


    + +2 8

    + + 4 8+ +6 8

    + + 8+ +3 8

    + +5 8 + +7 8



    The butterfly graph can be simplified as shown in Figure 4

    Figure 4: The simplified butterfly signal flow graph of radix-8


    D. Mixed radix DIF FFT There are

    wo kinds of mixed-radix DIF FFT algorithms. The first kind

    efers to a situation arising naturally when a radix-q algorithm,where q = 2

    m> 2, is applied to an input series consisting ofN =

    k q

    sequally spaced points, where1 k < m. In this case, out

    f necessity, k steps of radix-2 algorithm are applied either at the

    eginning or at the end of the transform, while the rest of the

    ransform is carried out bys steps of the radix-q algorithm.

    For example if N = 22m+1

    = 2 4m, the mixed-radix

    lgorithm [7][8] combines one step of the radix-2 algorithm and

    m steps of the radix-4 algorithm. The second kind of mixed-

    adix algorithms in the literature refers to those specialized for a

    ompositeN =N0 N1 N2 ...Nk. Different algorithms may

    e used depending on whether the factors satisfy certain

    estrictions. Only the 2 4m of the first kind of mixed-radix

    lgorithm will be considered here.

    The mixed-radix 4/2 butterfly unit is shown in Figure5.

    Figure 5: The butterfly signal flow graph of mixed-radix-4/


    It uses both the radix-22

    and the radix-2 algorithms can perfo

    fast FFT computations and can process FFTs that are not po

    of four. The mixed-radix 4/2, which calculates four butter

    outputs based on X(0)~X(3). The proposed butterfly unit

    three complex multipliers and eight complex adders.

    E. Split-Radix FFT Algorithms

    Split-radix FFT algorithm assumes two or more para

    radix decompositions in every decomposition stage to fu

    exploit advantage of different fixed-radix FFT algorithm. A

    result, a split-radix FFT algorithm generally has fewer count

    adder and multiplication than the fixed-radix FFT algorith

    while retains applicability to all power-of-2 FFT length.

    More computational complexity of the odd frequency te

    than the even frequency terms, so we can further decompose

    odd terms to reduce complexities. If we use radix-2 DIF F

    algorithm for the even frequency terms and the radix-22


    FFT algorithm for the odd parts, we can obtain the split-ra2/4 algorithm [10,11] as shown in the equation

    in the Equation (10).

    2= + + () (10)

    = 0,1,2, . . , 2 1

    (4+ 1)=

    ()+ ()(+ 4 )+(1)(+ 2 ) + ()(+ 3 4 )(11)

    (4+ 3)=

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 14

    ()+ ()(+ 4 )(+ 2 ) + ()(+ 3 4 )


    Thus the N-point DFT is decomposed into one N/2 -point DFT

    without additional twiddle factors and twoN/4 -point DFTs with

    widdle factors. TheN-point DFT is obtained by successive use

    f these decompositions up to the last stage. Thus we obtain a

    DIF split-radix-2/4 algorithm. The signal flow graph of basic

    utterfly cell of split-radix-2/4 DIF FFT algorithm is shown in

    igure 6

    Figure 6: The butterfly signal flow graph of mixed-radix-2/4


    we have





    (2) = +

    4 + + 3


    (1)= ()+ ()(+ 4 )+(1)(+ 2 ) + ()(+ 3 4 )

    (3)= ()+ ()(+ 4 )(+ 2 ) + ()(+ 3 4 )(13)

    As a result, even and odd frequency samples of each basic

    rocessing block are not produced in the same stage of the

    omplete signal flow graph. This property causes irregularity of

    signal flow graph, because the signal flow graph is an L-sh



    To increase the efficiency of the FFT technique sev

    pruning and different other techniques have been proposed

    many researchers. In this paper, we have implemented a n

    pruning technique i.e. IZTFFTP by simple modification

    some changes and also includes some tricky mathemat

    techniques to reduce the total execution time.

    Zero tracing-as in wide band communication system a la

    portion of frequency channel may be unoccupied by the licen

    user, so no. of zero valued inputs are much greater than the n

    zero valued inputs in a FFT/IFFT operation at the transceiv

    Then this algorithm will give best response in terms of redu

    execution time by reducing the no. of complex computat

    required for twiddle factor calculation. IZTFFTP have a str

    searching condition, which have an array for storing the inpu

    output values after every iteration of butterfly calculation. Iinput searching result whenever it found zero at any inp

    simply omit that calculation by considering useful condi

    based on radix algorithm used.

    A Input Zero Traced Radix-2 DIF FFT Pruning

    In radix-2 since we couple two inputs to obtain two outp

    we therefore have 4 combinations of those two inputs at radi

    butterfly. Now there exist three conditions only based u

    zeros at the input.

    No zero at input: No pruning happens in this case, butte

    calculations are same as conventional radix-2. Any one input zero: Output will be only the copied version

    input available, butterfly calculations are reduced compared

    conventional radix-2.

    All zero input: Output is zero and is obtained fr

    mathematical butterfly calculations is zero.

    B. Input Zero Traced Radix-4 DIF FFT Pruning In radi

    since we couple four inputs to obtain four outputs, we theref

    have 16 combinations of those four inputs at radix-4 butter

    Now therefore for radix-4 pruning there exist five conditi

    only based upon zeros at the input.

    No zero at the input: No pruning takes place, butte

    calculations are same as radix-4

    Any one input zero: Output will be only the copied version

    remaining inputs available, butterfly calculations are redu

    compared to radix-4.

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 15

    Any two inputs are zeros: Output will be only the copied

    version of that remaining two inputs available, butterfly

    calculations are reduced compared to radix-4 pruning with

    one zero at input.

    Any three inputs are zeros: Output will be only the copied

    version of that remaining single input available, butterfly

    calculations are reduced compared to radix-4 pruning withtwo zero at input.

    All zeros input: Output is zero and is obtained from

    mathematical calculations is zero.

    C. Input Zero Traced Radix-8 DIF FFT Pruning In radix-8

    ince we couple eight inputs to obtain eight outputs, we

    herefore have 256 combinations of those eight inputs at radix-8

    utterfly. Now therefore for radix-8 pruning there exist seven

    onditions only based upon zeros at the input. Similarly to

    adix-4 pruning, output is the version of non zero input. The

    more the number of zeros at input leads to less mathematical

    alculations compared to radix-8.

    D. Input Zero Traced Mixed radix DIF FFT Pruning If

    consider mixed radix 4/2, it uses the combination of radi

    pruning and radix-4 pruning. Similarly mixed radix 8/2 uses

    combination of radix-2 pruning and radix8 pruning.

    E. Input Zero Traced Split radix DIF FFT Pruning If

    consider spilt radix 2/4, it uses the combination of radi

    pruning and radix-4 pruning.


    In order to compare the computational complexities amo

    the different radix DIF FFT algorithms on OFDM,

    calculations based on the OFDM block sizes have b

    performed which are given in Table 1 and with prun

    comparison in Table 2.

    The speed improvement factors from without to with prun

    of different radix algorithms are seen in Table 3.




    Radix -2 Radix-4 Radix-8Mixed






    cm cadd cm cadd cm cadd cm cadd cm cadd cm cadd

    2 1 2 - - - - - - - - - -

    4 4 8 3 8 - - - - - 0 8

    8 12 24 - - 7 24 10 24 - - 4 24

    16 32 64 24 64 - - 28 64 22 64 12 64

    32 80 160 - - - - 64 160 60 160 36 160

    64 192 384 144 384 112 384 160 384 152 384 92 384

    Table 2: Comparison of complex additions(cadd) and complex multiplications(cm) of different radix algorithms without prunin




    Radix -2 Radix-4 Radix-8Mixed






    cm cadd cm cadd cm cadd cm cadd cm cadd cm cadd

    2 0 2 - - - - - - - - - -

    4 0 8 3 8 - - - - 0 8

    8 12 24 - - 7 24 8 24 - - 4 24

    16 31 64 24 64 - - 26 64 22 64 12 64

    32 76 160 - - - - 64 160 60 160 36 160

    64 179 384 141 384 112 384 157 384 152 384 90 384

    Table 2: Comparison of complex additions(cadd) and complex multiplications(cm) of different radix algorithms with prunin

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 20

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 16





    - 2















    8 1 - 1 1.25 - 1

    16 1.03 1 - 1.07 1 132 1.05 - - 1 1 1

    64 1.07 1.02 1 1.01 1 1

    Table 3: Speed Improvement Factor without to with pruning

    in terms of Multiplications

    Output shows the significant reduction of computational

    omplexity by reducing the total no. of complex operation

    e. both the multiplications and additions compare to the

    rdinary radix FFT operations. The complex multiplications

    nd additions are compared for different radix and pruned

    lgorithms. The

    omparison of complex multiplications for different radixDIF FFT algorithms is shown in Figure 7 and for different

    nput zero traced radix DIF FFT pruned algorithms are shown

    n Figure 8.

    Figure 7: Comparison of complex multiplications fordifferent radix DIF FFT

    Figure 8: Comparison of complex multiplications for

    different Radix DIF FFT pruned algorithms


    The computational performance of an OFDM system

    epends on FFT as in an OFDM system. FFT works as a

    modulator. If the complexity decreases, then the speed

    OFDM system increases. Results shows input zero tra

    radix DIF FFT pruned algorithms are much efficient than

    Radix DIF FFT algorithms as it takes very less time

    compute where number of zero valued inputs/outputs

    greater than the total number of non zero terms, w

    maintaining a good trade-off between time and spcomplexity, and it is also independent to any input data set


    [1] B. E. E. P. Lawrey, Adaptive Techniques for Mu

    User OFDM, Ph.D. Thesis, James Cook Univers

    Townsville,2001, pp. 33-34.

    [2] J. Mitola, III, "Cognitive Radio: An Integrated Ag

    Architecture for Software Defined Radio," Thesis (Ph

    Dept. of Teleinformatics, Royal Institute of Technol

    (KTH), Stockholm Sweden, May 2000.

    [3] S. Chen, Fast Fourier Transform, Lecture Note, Ra

    Communications Networks and Systems, 2005.[4] OFDM for Mobile Data Communications, T

    International Engineering Consortium WEB ProFor

    Tutorial, 2006.

    [5] Andrea Goldsmith, Wireless Communicatio

    Cambridge university press, 2005, ISB


    [6] J.G. Proakis and D.G. Manolakis, Digital Signal Pr

    essing: Principles, Algorithms and Edition, 2002,


    [7] E. Chu and A. George,Inside the FFT Black Box :Se

    & Parallel Fast FourierTransform Algorithms. C

    Press LLC, 2000.[8] B. G. Jo and M. H. Sunwoo, New Continuous-F

    Mixed-Radix (CFMR) FFT Processor Using Novel

    Place Strategy, Electron Letters, vol. 52, No. 5, M


    [9] Charles Wu, Implementing the Radix-4 Decimatio

    Frequency (DIF) Fast Fourier Transform (FF

    Algorithm Using aTMS320C80 DSP, Digital Sig

    Processing Solutions,January 1998.

    [10] P. Duhamel and H. Hollmann, Split-radix F

    Algorithm, Electron Letters, vol. 20, pp 14-16, J


    [11] [4] H. V. Sorensen, M. T. Heideman and C. S. Bur

    On Computing the Split-radixFFT, IEEE Tra

    Acoust., Speech, Signal Processing, vol. ASSP-34,

    152-156,Feb. 1986.
  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 17

    Chaos CDSK Communication SystemArathi. C

    M.Tech Student, Department of ECE,

    SRKR Engineering College,

    Bhimavaram, India

    Abstract:In recent years chaotic communication systems have emerged as an alternative solution to conventional sp

    spectrum systems. The chaotic carrier used in this kind of modulation-demodulation schemes, have unique properties

    make them suited for secure, and multi-user communications. The security of chaos communication system is superi

    other digital communication system, because it has characteristics such as non-periodic, wide-band, non - predictab

    easy implementation and sensitive initial condition. In this paper, a new approach for communication using chaotic sign


    KeywordsChaos Communication System, CDSK


    Previous digital communication technology continually

    used a linear system. However, as this technology reached basiclimit, people started to improve performance of nonlinearcommunication systems applying chaos communication systemsto nonlinear systems [1]. Chaos communication systems have

    the characteristics such as non - periodic, wide-band, non-

    predictability and easy implementation. Also, chaoscommunication system is decided by initial conditions ofequation, and it has sensitive characteristic according to initialcondition, because chaos signal is changed to different signal

    when initial condition is changed [2]. Chaos signal is expressed

    as randomly and non-linearly generated signal. If initialconditions of chaos signal is not exact, users of chaos system are

    impossible to predict the value of chaos signal because of itssensitive dependence on initial conditions [1][3]. As these

    characteristics, the security of chaos communication system is

    superior to other digital communication system.Due to security and other advantages, chaos

    communication systems are being studied continuously. Look atexisting research, in order to solve disadvantage that bit errorrate (BER) performance of this system is bad, chaos

    communication system is evaluated the BER performanceaccording to chaos maps, and find a chaos map that has the best

    BER performance [4]. In addition, chaos users evaluate the BER

    performance according to chaos modulation system [5][6], andpropose a new chaos map that has the best BER performance.

    In this paper, in AWGN and Rayleigh fading channel,BER performances of chaotic CDSK system is evaluated. At

    existing study, we proposed a novel chaos map in order to

    improve the BER performance [7], and we named a novel chaosmap "Boss map".


    A chaotic dynamical system is an unpredicdeterministic and uncorrelated system that exhibits noisbehavior through its sensitive dependence on its conditions, which generates sequences similar to PN sequ

    The chaotic dynamics have been successfully employ

    various engineering applications such as automatic cosignals processing and watermarking. Since the sigenerated from chaotic dynamic systems are noise-like, sensitive to initial conditions and have spread and flat spe

    in the frequency domain, it is advantageous to carry mes

    with this kind of signal that is wide band and has communication security. Numerous engineering applicatio

    secure communication with chaos have been developed [8]


    A chaotic sequence is non-converging and non-pe

    sequence that exhibits noise-like behavior through its sendependence on its initial condition [1]. A large numbuncorrelated, random-like, yet deterministic and reprodu

    signals can be generated by changing initial value. Tsequences so generated by chaotic systems are called ch

    sequences [8].

    Chaotic sequences have been proven easy to genand store. Merely a chaotic map and an initial conditioneeded for their generation, which means that there is nofor storage of long sequences. Moreover, a large numb

    different sequences can be generated by simply changin

    initial condition. More importantly, chaotic sequences can bbasis for very secure communication. The secrecy o

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 18

    transmission is important in many applications. The chaotic

    sequences help achieve security from unwanted reception in

    several ways. First of all, the chaotic sequences make thetransmitted signal look like noise; therefore, it does not attract

    the attention of an unfriendly receiver. That is, an ear-dropperwould have a much larger set of possibilities to search through

    in order to obtain the code sequences [4][8].Chaotic sequences are created using discrete, chaoticmaps. The sequences so generated even though are completely

    deterministic and initial sensitive, have characteristics similar tothose of random noise. Surprisingly, the maps can generate large

    numbers of these noise-like sequences having low cross-correlations. The noise-like feature of the chaotic spreadingcode is very desirable in a communication system. This feature

    greatly enhances the LPI (low probability of intercept)performance of the system [4].

    These chaotic maps are utilized to generate infinitesequences with different initial parameters to carry different userpaths, as meaning that the different user paths will spread

    spectrum based on different initial condition [8].


    A. Correlation delay shift keying system

    CDSK system has an adder in transmitter. Existing

    modulation system than CDSK system consists switch intransmitter, and problem of power waste and eavesdropping

    occurs by twice transmission. Technique that has been proposedfor overcoming these problems is CDSK system. And,transmitted signal does not repeat by replacing an adder with a

    switch in the transmitter [9].


    d +1, 1Figure 1: Transmitter of CDSK system

    CDSK transmitter is composed of sum in which

    delayed chaos signal multiplied with information bit is added togenerated chaos signal from chaos signal generator. Here,information bit that is spread as much as spreading factor is

    multiplied by delay chaos signal.

    s = x+ dx (1)

    Above equation (1) indicates transmitted signal


    r y d


    Figure 2: Receiver of CDSK system

    CDSK receiver is correlator based receiver, and

    performed in order to recover the symbol. Received signa

    delay received signal are multiplied, and this signal is as added as spreading factor. Afterward the signal pass througthreshold, and information signal recover through decoding

    Information bits are possible to recover when

    time and spreading factor have to use exact value that is us

    transmitted signal.

    B. Chaos maps

    In this paper, types of chaos map used are Tentand Boss map. At existing study, Boss map means a novel that we proposed for BER performance improvement [8].

    Figure 3: Trajectory of tent map

    Figure (3) shows trajectory of Tent map. The x

    and the y-axis of figure (3) mean xn and xn+1, and Tent ma

    trajectory of triangular shape.

    x = bx c Fx (2)

    Equation (2) of tent map is expressed as aEquation (2) of Tent map uses existing output value as cu






  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 19

    input value, and it is indicated as figure when initial value is 0.1

    and parameter alpha is 1.9999.

    Figure 4: Trajectory of boss map

    Figure (4) shows trajectory of Boss map, a novel map

    that is proposed in order to improve the BER performance. Thex-axis and the y-axis of Boss map mean xn and yn unlike the

    Tent map, it draws trajectory like pyramid shape.

    = 0.450.503 = 0.3 (3)

    Equation (3) of Boss map is expressed as above.

    Equation (3) form of Boss map is similar to Tent map because

    Boss map was proposed by transforming from Tent map. And,trajectory of Boss map is indicated as figure (4) when initial

    value is 0.1 and parameter alpha is 2.5.


    In this paper, the BER performance of chaotic CDSKsystem in AWGN (adaptive white Gaussian noise) channel andRayleigh fading channel is evaluated for Tent map and Boss

    map.In AWGN channel, figure (5) shows BER performance

    of chaotic CDSK system is evaluated. Looking at the figure (5),

    the BER performance of chaotic CDSK system with tent mapand boss map is observed. Here, we observe that the BER

    performance of Boss map is better than Tent map at each stagei.e. at different values of SNR we observe that the Boss map

    shows better performance than Tent map. We also observe that

    at initial values the BER is same for both maps but as

    increases the BER for Boss map is less than Tent map.

    Figure 5: BER analysis in AWGN channel

    In Rayleigh fading channel, figure (6) shows the

    performance of chaotic CDSK system. Here, the performaevaluated for both Tent map and Boss map. We observe t

    initial values of SNR the BER performance is the same formaps. But as SNR value increases the BER performance of

    map is better than Tent map.

    Figure 6: BER performance in Rayleigh fading channel


    In this paper, a new type of communication susing chaos is proposed. Chaos sequences are non pe

    sequences which are sensitive to their initial conditions. Csequences are generated using chaos map. CDSK system

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 20

    chaos has many advantages over other systems. But the BER

    performance of chaos communication system is bad. In order to

    improve this, we proposed a new chaos map that has better BERperformance than existing map. In AWGN and Rayleigh fading

    channel the chaotic CDSK system is evaluated and we observedthat the BER performance of chaos system with Boss map has

    better than with Tent map which improves BER of CDSKcommunication system.


    Chaos communication system increases the number oftransmitted symbols by spreading and transmitting informationbits according to characteristic of chaos maps. So the research

    that improves data transmission speed is necessary for chaoscommunication system. If many antennas are applied to chaos

    communication system, the capacity of data is proportional tothe number of antenna. So it is good way applying multiple-input and multiple-output (MIMO) to the chaos communication



    [1] M. Sushchik, L.S. Tsimring and A.R. Volkovskii,

    "Performance analysis of correlation- based communication

    schemes utilizing chaos," Circuits and Systems I:Fundamental Theory and Applications, IEEE Transactions

    on, vol. 47, no. 12, pp. 1684-1691, Dec. 2000.[2] Q. Ding and J. N. Wang, "Design of frequency-modulated

    correlation delay shift keying chaotic communication

    system," Communications, IET, vol. 5, no. 7, pp. 901-905,

    May 2011.[3] Chen Yi Ping, Shi Ying and Zhang Dianlun, "Performance

    of differential chaos-shift-keying digital communication

    systems over several common channels,"Future Computerand Communication (ICFCC), 2010 2


    Conference on, vol. 2, pp. 755- 759, May 2010.

    [4] Suwa Kim, Junyeong Bok and Heung-Gyoon Ryu,"Performance evaluation of DCSK system with chaotic

    maps," Information Networking (ICOIN), 2013International Conference on, pp. 556-559, Jan. 2013.

    [5] S. Arai and Y. Nishio, Noncoherent correlation-based

    communication systems choosing different chaotic maps,

    Proc. IEEE Int. Symp. On Circuits and Systems, NewOrleans, USA, pp. 1433-1436, June 2007.

    [6] Jun-Hyun Lee and Heung-Gyoon Ryu, "New Chaos Mapfor CDSK Based Chaotic Communication System," The

    28th International Technical Conference on Circuit/System,

    Computers and Communication (ITC-CSCC 2013), Y

    Korea, pp. 775-778, July2013.

    [7] M.A. Ben Farah, A. Kachouri and M. Samet, "Desisecure digital communication systems using DCSK ch

    modulation," Design and Test of Integrated SystemNano-scale Technology, 2006. DTIS 2006.Interna

    Conference on, pp. 200-204, Sept. 2006.[8] Ned J. Corron, and Daniel W. Hahs A new approacommunication using chaotic signals, IEEE transac

    on circuits and systemsI: fundamental theoryapplications, VOL. 44, NO. 5, MAY 1997.

    [9] Wai M. Tam, Francis C. M. Lau, and Chi K. Generalized Correlation-Delay-Shift-Keying SchemNon - coherent Chaos-Based Communication Sys

    IEEE transactions on circuits and systemsI: repapers, VOL. 53, NO. 3, MARCH 2006.

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 21

    Design and Analysis of Water Hammer Effect

    in a Network of PipelinesV. Sai Pavan Rajesh

    Department of Control Systems, St. Marys Group of Institutions,Jawaharlal Nehru Technological University- Hyderabad,Main Road, Kukatpally Housing Board Colony, Kukatpally, Hyderbad, Telangana, India.

    Abstract-There wil l be a chance for the destruction of the system due to transient if it i s not provided with adequate

    protection devices. Generally, transient takes place when parameters involving in conventional flow are distorted with

    respect to the time. Rapid closing of valve in a pipe network wil l be resul ting i nto hydrau li c transient known as water

    hammer occurs due to sudden change in pressure and veloci ty of f low wi th respect to time. Due to impulsive action, pressur e

    surges are induced in the system tr avel along the pipe network with the rapid f lu id acceleration leading to the dramatic

    effects li ke pipe li ne fail ure, damage to the system etc. Consideri ng the impor tance of hydrau li c transient anal ysis, we design

    a system capable of verif ying pipe network contain ing flui d flow.Thi s paper demonstrates design of dif ferent pipe structures

    in pi pe li ne network and analysis of various parameters li ke excess pressure distr ibu tion , veloci ty variati ons and water

    hammer amplitude with respect to time using COMSOL M ulti physics v 4.3. The magnitude of water transient in pipe line

    network at dif ferent pressure points has been discussed in detail.

    Keywords- COMSOL, Pressure distributi on, Velocity variati on, Water H ammer.


    The key to the conservation of water is good water

    measurement practices. As fluid will be running in water

    distribution system, system flow control is dependent based

    on the requirement for opening or closing of valves, and

    starting and stopping of pumps. When these operations are

    performed very quickly, they convert the kinetic energy

    carried by the fluid into strain energy in pipe walls, causing

    hydraulic transient[1]

    phenomena to come into existence in the

    water distribution system i.e., a pulse wave of abnormal

    pressure is generated which travels through the pipe network.

    Pressure surges that are formed or fluid transients in pipelines

    are referred to as Water hammer. This oscillatory form of

    unsteady flow generated by sudden changes results in system

    damage or failure if the transients are not minimized. So now

    the steady state flow conditions are altered by this effect[2]

    resulting in the disturbance to the initial flow conditions of the

    system. Where the system will tend to obtain a static flow rate

    by introduction of new steady state condition. The intensity of

    water hammer effects will depend upon the rate of change in

    the velocity or momentum. Conventional water hammer

    analyses provide information under operational conditions on

    two unknown parameters i.e., pressure and velocity within a

    pipe system. Generally effects such as unsteady friction,

    acoustic radiation to the surroundings or fluid structure

    interaction are not taken into account in the standard theory of

    water hammer, but were considered in general approach[3]


    But mechanisms acting all along the entire pipe section such

    as axial stresses in the pipe and at specific points in the pipe

    system such as unrestrained valves will fall under fluid

    structure interaction extension theory for conventional water

    hammer method.

    Figure 1. Pipe connected to control valve at the end with water

    inlet from reservoir.

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 22

    In the past three decades, since a large number of water

    hammer events occurred in the light-water- reactor power


    , a number of comprehensive studies on the

    phenomena associated with water hammer events have been

    performed. Generally water hammer can occur in any thermal-

    hydraulic systems and it is extremely dangerous for the

    thermal-hydraulic system since, if the pressure induced

    exceeds the pressure range of a pipe given by the

    manufacturer, it can lead to the failure of the pipeline

    integrity. Water hammers occurring at power plants are due to

    rapid valve operation[5]

    , void induced operation, and

    condensation induced water hammer[6]

    . In existing Nuclear

    Power Plants water hammers can occur in case of an inflow of

    sub-cooled water into pipes or other parts of the equipment,

    which are filled with steam or steam-water mixture[7]


    The water hammer theory has been proposed to account for a

    number of effects in biofluids under mechanical stress, as in

    the case of the origin of Korotkoff sounds during blood

    pressure measurement[8, 9]

    , or the development of a fluid-

    filled cavity within the spinal cord[10]

    . In the voice production

    system, the human vocal folds act as a valve[11

    which induces

    pressure waves at a specific point in the airways (the glottis),

    through successive compressing and decompressing actions

    (the glottis opens and closes repeatedly). Ishizaka was

    probably the first to advocate in 1976 the application of the

    water hammer theory, when discussing the input acoustic

    impedance looking into the trachea[12]

    . More recently, the

    water hammer theory was invoked in the context of tracheal

    wall motion detection[13]

    . Generally Water utilities, Industrial

    Pipeline Systems, Hydropower plants, chemical industries,

    Food, pharmaceutical industries face this water transient


    The present work reports the design of different pipe

    channels and analysis of the pressure distribution and velocity

    variation produced all along the pipe flow network when

    subjected to one pressure measuring point. Various parameters

    like inlet input pressure, wall thickness and measurement

    point are changed for analysis.


    The software package selected to model and simulate

    the pipe flow module was COMSOL Multiphysics Version

    4.3. It is a powerful interactive environment for modelling and

    Multiphysics were selected because there was previous

    experience and expertise regarding its use as well as

    confidence in its capabilities. A finite element method based

    commercial software package, COMSOL Multiphysics, is

    used to produce a model and study the flow of liquid in

    different channels. This software provides the flexibility for

    selecting the required module using the model library, whichconsists of COMSOL Multiphysics, MEMS module, micro

    fluidics module etc. Using tools like parameterized geometry,

    interactive meshing, and custom solver sequences, you can

    quickly adapt to the ebbs and flows of your requirements,

    particle tracing module along with the live links for the

    MATLAB. At present this software can solve almost

    problems in multi physics systems and it creates the real world

    of multi physics systems without varying there material

    properties. The operation of this software is easier to

    understand and easier to implement in various aspects for

    designers, in the form of finite element analysis system.

    Figure 2. Multiphysics modelling and simulation software-


  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 23

    In this model as the valve is assumed to close instantaneously

    the generated water hammer pulse has a step function like

    shape. To correctly solve this problem requires a well posed

    numerics. The length of the pipe is meshed with N elements

    giving a mesh size dx = L/N. For the transient solver to be

    well behaved requires that changes in a time step dt are made

    on lengths less than the mesh size. This gives the CFL number


    CFL= 0.2= c.dt/dx (1)

    Meaning that changes during the time dt maximally move 20

    % of the mesh length dx. Thus increasing the mesh resolution

    also requires decreasing the time stepping. This advanced

    version of software helps in designing the required geometry

    using free hand and the model can be analysed form multiple

    angles as it provides the rotation flexibility while working

    with it.


    Water hammer theory dates to 19th century, where several

    authors have contributed their work in analyzing this effect.

    Among them, Joukowsky[14]

    conducted a systematic study of

    the water distribution system in Moscow and derived a

    formula that bears his name, that relates to pressure changes,

    p, to velocity changes, v, according to the equation

    P = cU (2)

    Where is the fluid mass density and c is the speed of sound.

    This relation is commonly known as the Joukowsky

    equation, but it is sometimes referred to as either the

    Joukowsky-Frizell or the Allievi equation.

    For a compressible fluid in an elastic tube, c depends on the

    bulk elastic modulus of the fluid K on the elastic modulus of

    the pipe E, on the inner radius of the pipe D, and on its wall

    thickness. The water hammer equations are some version of

    the compressible fluid flow equations. The choice of the

    version is problem-dependent: basic water hammer neglects

    friction and damping mechanisms, classic water hammer takes

    into account fluid wall friction, extended water hammer

    allows for pipe motion and dynamic Fluid Structure

    Interaction[15, 16]


    In water hammer at static condition pressure wave is a

    disturbance that propagates energy and momentum from one

    point to another through a medium without significant

    displacement of the particles of that medium. A transient

    pressure wave, subjects system piping and other facilities to

    oscillating at high pressures and low pressures. This cyclic

    loads and pressures can have a number of adverse effects on

    the hydraulic system. Hydraulic transients can cause hydraulic

    equipments in a pipe network to fail if the transient pressures

    are excessively high. If the pressures are excessively higher

    than the pressure ratings of the pipeline, failure through pipe

    or joint rupture, or bend or elbow movement may occur.

    Conversely, excessive low pressures (negative pressures) can

    result in buckling, implosion and leakage at pipe joints during

    sub atmospheric phases. Low pressure transients are normally

    experienced on the down streamside of a closing valve. But

    when the valve is closed energy losses are introduced in the

    system and are normally prescribed by means of an

    empirical law in terms of a loss coefficient. This

    coefficient, ordinarily determined under steady flow

    conditions, is known as the valve discharge coefficient,

    especially when the pipeline is terminated by the valve. It

    enables to quantify the flow response in terms of the

    valve action through a relationship between the flow rate

    and pressure for each opening position of the valve. The

    discharge coefficient provides the critical piece of missing

    information for the water hammer analysis. Because the

    existing relationship between pressure and flow rate is often

    a quadratic law type, the empirical coefficient is defined in

    terms of the squared flow rate. When water distribution

    system comprising a short length of pipes (i.e.,

  • 8/10/2019 Merged File 1


    www. Special Issue-1, October 2014

    SJ Impact Factor-3.995 ISSN: 2321-9653

    International Journal for Research in Applied Science & Engineering


    Page 24

    tanks, reservoirs, junctions tend to limit further changes in

    pressure and counteract the initial transient effects. An

    important consideration is dead ends, which may be caused by

    closure of check valves that lock pressure waves into the

    system in cumulative fashion. Wave reflections will be both

    positive and negative pressures; as a result the effect of dead

    ends must be carefully evaluated in transient analysis.

    These pressure surges provide the most effective and viable

    means of identifying weak spots, predicting potentially

    negative effects of hydraulic transient under a number of

    worst case scenarios, and evaluating how they may possibly

    be avoided and controlled. The basic pressure surge modeling

    is based on the numerical conservation of mass and linear

    momentum equations. For this Arbitary Lagrangian


    numerical solution helps in providing the exact

    analytical solution. On the other hand when poorly calibrated

    hydraulic network models results in poor prediction of

    pressure surges thus leading to more hydraulic transients. In

    more complex systems especially, the cumulative effect of

    several types of devices which influence water hammer may

    have an adverse effect. However, even in simple cases, for

    example in pumping water into a reservoir, manipulations

    very unfavorable with regard to water hammer may take

    place. For example, after the failure of the pump, the operator

    may start it again. Much depends on the instant of this

    starting. If it is done at a time when the entire water hammer

    effect has died down, it is an operation for which the system

    must have been designed.


    The design and analysis of the hydraulic transient in a

    pipe flow includes geometry, defining the parameters for the

    required geometry, providing mesh & inputs. The 3D model is

    constructed in the drawing mode of COMSOL Multiphysics.

    In this, a pipe of length L = 20 m is constructed assuming that

    one end is connected to a reservoir, where a valve is placed at

    the other end. The pipe with inner radius of 398.5mm, the

    thickness of the wall about 8 mm and Youngs modulus of

    210GPa was designed. In order to verify pressure distribution,

    a pressure sensor measurement point at a distance of z0= 11.15

    m from the reservoir was arranged and flow has been sent into

    pipe with an initial flow rate of Q0= 0.5 m3/sec.

    Figure 3. Single pipe line