Selective Parallel Interference Cancellation scheme for mitigation of MUI in Uplink OFDMA

Transcription

1 Selective Parallel Interference Cancellation scheme for mitigation of MUI in Uplink OFDMA A PROJECT REPORT Submitted by PADMAVATHI.N Register No: 14MCO015 in partial fulfillment for the requirement of award of the degree of MASTER OF ENGINEERING in COMMUNICATION SYSTEMS Department of Electronics and Communication Engineering KUMARAGURU COLLEGE OF TECHNOLOGY (An autonomous institution affiliated to Anna University, Chennai) COIMBATORE ANNA UNIVERSITY: CHENNAI APRIL i

2 i

3 BONAFIDE CERTIFICATE Certified that this project report titled Selective Parallel Interference Cancellation scheme for mitigation of MUI in Uplink OFDMA is the bonafide work of PADMAVATHI.N (Reg. No. 14MCO015) who carried out the project under my supervision. Certified further, that to the best of my knowledge the work reported here in does not form part of any other project or dissertation on the basis of which a degree or award was conferred on an earlier occasion on this or any other candidate. SIGNATURE Dr.G.AMIRTHA GOWRI PROJECT SUPERVISOR Associate Professor Department of ECE Kumaraguru College of Technology SIGNATURE Dr.A.VASUKI PROFESSOR AND HEAD Department of ECE Kumaraguru College of Technology Coimbatore Coimbatore The candidate with Register No. 14MCO015 was examined by us in the project viva-voce examination held on.. INTERNAL EXAMINER EXTERNAL EXAMINER ii

4 iii

5 ACKNOWLEDGEMENT First, I would like to express my praise and gratitude to the Lord, who has showered his grace and blessings enabling me to complete this project in an excellent manner. I express my sincere thanks to the management of Kumaraguru College of Technology and Joint Correspondent Shri Shankar Vanavarayar for his kind support and for providing necessary facilities to carry out the work. I would like to express my sincere thanks to our beloved Principal Dr.R.S.Kumar M.E., Ph.D., Kumaraguru College of Technology, who encouraged me with his valuable thoughts. I would like to thank Dr.A.Vasuki M.E., Ph.D., Head of the Department, Electronics and Communication Engineering, for her kind support and for providing necessary facilities to carry out the project work. In particular, I wish to thank and express my heartfelt gratitude to the project coordinator Dr.M.Alagumeenaakshi M.E., Ph.D., AP(III), Department of Electronics and Communication Engineering, for her expert counselling and guidance to make this project to a great deal of success. I am greatly privileged to express my heartfelt thanks to my project guide Dr.G.Amirtha Gowri M.E., Ph.D., Associate Professor, Department of Electronics and communication Engineering, who encouraged me in each and every step of the project work and I wish to convey my deep sense of gratitude to all teaching and nonteaching staff of ECE Department for their help and cooperation. Finally, I thank my parents and my family members for giving me the moral support and abundant blessings in all of my activities and my dear friends who helped me to endure my difficult times with their unfailing support and warm wishes. iii

6 ABSTRACT A demand for high speed mobile wireless communications is quickly mounting. Wireless communication is the key part of research with growing demand of high data rate at low cost. Orthogonal Frequency Division Multiplexing Access (OFDMA) is a promising technique to achieve high data rate. It is a multi-carrier modulation currently used in LTE and Wimax. The problems associated with multicarrier transmission is Carrier Frequency Offset (CFO). The CFO breaks the orthogonality among the subcarriers which causes Multi User Interference (MUI) in the OFDMA symbol and this increases the Bit Error Rate (BER) of OFDMA Symbol. In this project, different CFO Estimation and cancellation techniques are applied to reduce MUI. Here, Estimation of CFO is done in Frequency Domain method: moose and classen. Mean Square Error (MSE) analysis shows that the classen method has better performance in estimating CFOs when compared to moose. After CFO has been estimated, it is compensated. Effect of CFO is minimized to reduce MUI by Time Domain Multi User Interference cancellation (TDMUIC) and Frequency Domain Multi User Interference cancellation (FDMUIC) schemes. TDMUIC used here are Simple Time Domain Multi User Interference cancellation (SIMUIC) and Code Aided Time Domain Multi User Interference cancellation (CAMUIC). To reduce the MUI, CFO compensation is carried out and this can be done by Interference cancellation algorithm. Successive Interference Cancellation (SIC) and Parallel Interference Cancellation (PIC) are well known Cancellation methods used for CFO Compensation.SIC method is applied in both SIMUIC and CAMUIC schemes and the results are compared. CAMUIC gives out the better BER performance than SIMUIC method. Hence, for simplicity it is alone taken for furture comparison with proposed Selective Parallel Interference Cancellation method(spic).spic is obtained by integrating the SIC and PIC algorithm with the help of decision variable calculation. BER performance for SPIC algorithm is compared with SIC and the result shows that proposed SPIC algorithm has better performance in uplink OFDMA system and it out performs the advantage of both Cancellation algorithm together. iv

7 TABLE OF CONTENT CHAPTER TITLE PAGE NO ABSTRACT LIST OF FIGURES LIST OF TABLES LIST OF ABBREVIATIONS 1 INTRODUCTION OFDM TO OFDMA OBJECTIVE OF THE PROJECT OFDMA OFDMA SYSTEM Modulation and Demodulation Addition and Removal of CP Interleaving and De-interleaving 8 iv vi vii viii Serial to Parallel Converter/ Parallel to serial Converter DAC and ADC FFT and IFFT Merits of OFDMA Demerits of OFDMA WIRELESS CHANNEL AWGN and Fading channel Rayleigh fading channel CARRIER FREQUENCY OFFSET Frequency offset ICI and MUI Doppler Effect OUTLINE OF THE PROJECT 16 vi

8 2 LITERATURE SURVEY 17 3 TIME DOMAIN MUIC SCHEME CFO ESTIMATION Time Domain Estimation Frequency Domain Estimation MULTI-FFT RECEIVER Channel Estimation Channel Correction Symbol mapping/demapping MUI CANCELLATION SCHEMES SI MUIC SCHEME CA-MUIC SCHEME Convolutional coder MUIC ALGORITHM Successive Interference Cancellation Parallel Interference Cancellation Selective Parallel Interference Cancellation (SPIC) 32 4 RESULTS AND DISCUSSION CFO Estimation MUI cancellation 37 5 CONCLUSION 40 REFERENCES 41 LIST OF PUBLICATIONS 42 vi

9 LIST OF FIGURES Figure No. Caption Page No. 1.1 Uplink and Downlink Communication OFDM and OFDMA Time and Frequency view of OFDMA symbol OFDMA Block Diagram QPSK Modulation Cyclic Prefix insertion Comparison of Bandwidth utilization in FDMA and OFDMA Doppler Effect Flow of CFO Estimation technique Flow chart for Classen method General Receiver Structure for TDMUIC Demodulation Block for SI-MUIC Scheme Demodulation Block of CA-MUIC Scheme Block diagram of convolution encoder Flowchart of SIC Algorithm Flowchart of PIC Algorithm Flowchart for SPIC Frequency Domain CFO estimation SNR vs BER Frequency Offset vs BER Number of user vs BER Number of user vs Correlation 39 vii

10 LIST OF TABLES Table No. Caption Page No 4.1 System Parameter and its specification Computational time and complexity comparison 39 viii

11 LIST OF ABBREVIATION 4G - 4 th Generation ADC - Analogue to Digital Converter BER - Bit Error Rate BPSK - Binary Phase Shift Keying BS - Base Station CAMUIC - Code Aided Time Domain Multi User Interference Cancellation CFO - Carrier Frequency Offset CIR - Carrier to Interference Ratio CP - Cyclic Prefix DAC - Digital to Analogue Converter DSBCS - Double-Side Band Suppressed carrier transmission EKF - Extended Kalman Filter FDMA - Frequency Division Multiple Access FDMUIC - Frequency Domain Multi User Interference Cancellation FEC - Forward Error Correction FFT - Fast Fourier Transform GII - Guard Interval Insertion GIR - Guard Interval Removal HIC - Hybrid Interference Cancellation ICI - Inter Carrier Interference IFFT - Inverse Fast Fourier Transform ISI - Inter Symbol Interference LMMSE - Linear Maximum Mean Square Error LTE - Long Term Evolution MAI - Multiple Access Interference MATLAB - Mathematical Laboratory ix

12 ML - Maximum Likelihood MS - Mobile Station MSE - Mean Square Error MUI - Multi User Interference NLOS Non Line of Sight PIC - Parallel Interference Cancellation OFDM - Orthogonal Frequency Division Multiplexing OFDMA - Orthogonal Frequency Division Multiple Access QAM - Quadrature Amplitude Modulation QPSK - Quadrature Phase Shift Keying RSS - Received Signal Strength SC - Self Cancellation SIC - Successive Interference Cancellation SIMUIC - Simple Time Domain Multi User Interference Cancellation SNR - Signal to Noise Ratio SPIC - Selective Parallel Interference Cancellation TDMA - Time Division Multiple Access TDMUIC - Time Domain Multi User Interference Cancellation UT - User Terminals ix

13 CHAPTER 1 INTRODUCTION Orthogonal Frequency Division Multiplexing (OFDM) is one of the most popular techniques for wireless communication due to its robustness against fast fading and Inter Symbol Interference (ISI) and its spectrum efficiency is attained by orthogonality of subcarriers. OFDM structure can be extended to multiple-access scenarios to include multiple user transmission. Orthogonal Frequency Division Multiple Access (OFDMA) is a multiple access technique based on OFDM which has been adopted in different standards such as IEEE and Long Term Evolution (LTE). 1.1 OFDM to OFDMA In recent years, OFDMA has emerged as the primary transmission scheme and it is a popular multiple access scheme for broadband wireless networks. Today s mobile communication aims at providing voice and high data rate services at low cost. However, they cannot efficiently meet the growing demand for mobile services such as multimedia broadband services. Because multimedia communication has a rather large demand on bandwidth and Quality of Service (QoS) compared to what is available today. OFDMA is an efficient modulation scheme to transmit large bit stream through the channel with higher Band width and sufficient QOS. OFDMA is a multiple carrier modulation technique which distributes data over a large number of subcarriers spaced apart at precise frequencies, such that they are orthogonal to one another. OFDM has been successfully applied to a wide variety of digital communication applications over the past several years including digital TV broadcasting, digital audio broadcasting, Asynchronous Digital Subscriber Line (ADSL) modems and wireless networking worldwide. Its application in mobile communication is more complex especially because of the mobility of the mobile users; thus more exact symbol timing and frequency-offset control must be used to ensure that sub-carriers remain orthogonal. 1

14 However, the difference between the frequency of the oscillator in the transmitter and the receiver causes frequency offset. If this offset is not estimated and compensated, the orthogonality of the sub-carriers is ruined and thereby causing large bit errors in the received signal. Also, the distortion of the signals while travelling through the channel and the movement of the mobility user causes synchronisation problems. OFDMA is used in both downlink as well as uplink transmissions due to several of its favorable characteristics inherited from OFDM, such as efficient usage of spectrum, robustness against frequency selective fading and flexible resource allocation. Figure 1.1 shows representation of the uplink and downlink communication and if both are done at same time then it is called two-way communication. Figure 1.1 Uplink and Downlink Communication In OFDMA System, the multiple user signals are separated in the time and frequency domains. Typically, a burst in an OFDMA system will consists of several OFDM Symbols. The subcarriers and the OFDM symbol period are the finest allocation units in the frequency and time domain, respectively. Hence, multiple users are allocated different slots in the time and frequency domain, i.e, different group of subcarriers and OFDM symbols are used for transmitting the signals to/from multiple users. OFDMA is also now a part of the Long Term Evolution (LTE), the international standard for the 4 th Generation (4G) cellular mobile communications. Being a multiuser version of the famous Orthogonal Frequency Division Multiplexing (OFDM) technique, FDMA operates by separating data into multiple lower-rate streams and transmitting them in parallel over orthogonal carrier frequencies, or subcarriers. Due to the orthogonality, those subcarriers are allowed to overlap in the frequency domain. 2

15 Therefore, high spectral efficiency can be achieved. Figure 1.2 shows the user allocation in OFDM and OFDMA. In OFDMA, available subcarriers are grouped into sub-channels, which are assigned to different users operating simultaneously. This allows a finer granularity for multiple access when compared to Orthogonal Frequency Division Multiplexing - Time Division Multiple Access (OFDM-TDMA) scheme, in which all subcarriers are given to only one user at any given time. The carrier frequency misalignment destroys the orthogonality of the subcarriers, which causes Inter Carrier Interference (ICI) and consequently produces Multi User Interference (MUI) among users. Figure 1.2 OFDM and OFDMA OFDMA is used in both downlink as well as uplink transmissions due to several of its favorable characteristics inherited from OFDM, such as efficient usage of spectrum, robustness against frequency selective fading and flexible resource allocation. While the CFO can be estimated and corrected relatively easily in the downlink, the primary challenge in usage of OFDMA in uplink is achieving synchronization between several User Terminals (UTs) and Base Stations (BS) in both time and frequency domain. In the uplink, the received signal is the sum of multiple signals coming from different users, each of which experiences a different CFO due mainly to oscillator instability and/or Doppler shift. These relative CFOs among users must be estimated and corrected, otherwise the system performance degrades severely. Downlink CFO correction methods, which are designed for single-user scenario, are unable to correct multiple CFOs in the uplink, as correction to one user s CFO would misalign the other users. 3

16 1.2 OBJECTIVE OF THE PROJECT The main objective of this project is to investigate effective carrier frequency offset estimation and compensation methods in OFDMA uplink communications systems. The aim of the project is to achieve CFO free performance which in turn reduces the MUI and ICI in uplink OFDMA System. Different CFO Estimation schemes are analyzed and Multi User Interference Cancellation Schemes are applied to reduce the CFO. The MSE of different frequency domain CFO estimation schemes is analyzed to choose the better estimation scheme. The estimated CFO is compensated by MUI cancellation schemes. The effective CFO compensation scheme is chosen based on BER analysis and it is compared with the proposed Cancellation algorithm. The MATLAB is used for the estimation and compensation process. 1.3 ORTHOGONAL FREQUENCY DIVISION MULTIPLE ACCESS OFDMA is very similar to OFDM in function, with the main difference being that instead of being allocated all of the available subcarriers, the base station allocates a subset of carriers to each user in order to accommodate multiple transmissions simultaneously. In OFDM, all the frequencies of the subcarriers were generated by one transmitter and maintaining orthogonality of the subcarriers is relatively easy when compared to OFDMA. Here, many users transmit simultaneously, each with their own estimates of the subcarrier frequencies, a frequency offset is inevitable and multiple access interference occurs as users power leaks into subcarriers bands. In OFDMA system, both time and/or frequency resources are used to separate the multiple user signals. Group of symbols and/or group of subcarriers are the units to separate the transmissions to/from multiple users. In Fig 1.2,the time frequency view of a typical OFDMA signal is shown for a case where there are 3 users. It can be seen from Fig 1.3. Those users signals are separated either in the time-domain by using different OFDM symbols and/or in the subcarrier domain. Thus, both the time and frequency resources are used to support multiuser transmission. 4

17 Figure 1.3 Time and Frequency view of OFDMA symbol OFDMA SYSTEM OFDMA stands for Orthogonal Frequency Division Multiplexing Access. It is a multi-carrier modulation technique. It is used to modulate multiple low bit rate data streams onto multiple closely spaced carriers. It is similar to OFDM but occupies less frequency spectrum compare to OFDM for the same number of users. In OFDMA carriers are densely packed compare to the OFDM technique. Figure 1.4 depicts OFDMA system block diagram. As shown generic OFDMA transmitter consists of FEC encoder,interleaver, data modulation or mapping, IFFT and cyclic prefix addition. OFDMA receiver consists of front end synchronization to correct for any time, frequency or channel impairments. This is followed by reverse modules of the one mentioned in the transmitter. It includes Cyclic Prefix removal, FFT, data demodulation, deinterleaver, FEC decoder. 5

18 Figure 1.4 OFDMA Block Diagram Modulation and Demodulation Modulation is the technique by which the signal is transformed in order to send it over the communication channel in order to minimize the effect of noise. This is done in order to ensure that the received data can be demodulated to give back the original data. In an OFDMA system, the high data rate information is divided into small packets of data which are placed orthogonal to each other. This is achieved by modulating the data by a desirable modulation technique like Quadrature Amplitude Modulation (QAM) or Quadrature Phase Shift Keying (QPSK). After this, IFFT is performed on the modulated signal which is further processed by passing through a parallel to serial converter. Guard Interval Insertion (GII) is done in order to avoid ISI. QPSK is type of phase shift keying based modulation method. Unlike BPSK which is a Double-Side Band Suppressed carrier transmission (DSBCS) modulation scheme with digital information for the message, QPSK is also a DSBCS modulation scheme but it sends two bits of digital information a time (without the use of another carrier frequency). The amount of radio frequency spectrum required to transmit QPSK reliably is half that of required for BPSK signals, which in turn makes chance for more users on the channel. The figure 1.5 below shows a QPSK modulated waveform. 6

19 Figure 1.5 QPSK Modulation Demodulation is the technique by which the original data is recovered from the modulated signal which is received at the receiver end. In this case, the received data is first made to pass through a low pass filter and the Guard Interval Removal (GIR) is done. FFT of the signal is done after it is made to pass through a serial to parallel converter. A demodulator is used, to get back the original signal. The bit error rate and the signal to noise ratio is calculated by taking into consideration the un-modulated signal data and the data at the receiving end Addition and Removal of Cyclic Prefix Wireless communications systems are susceptible to multi-path channel reflections. Hence a Cyclic Prefix (CP) is added to reduce ISI. A cyclic prefix is a repetition of the last section of a symbol that is appended to the front part of the symbol. Figure 1.6 shows the insertion of CP in beginning of each symbol.in addition, it is important because it enables multi-path representations of the original signal to fade so that they do not interfere with the subsequent symbol. 7

20 Figure 1.6 Cyclic Prefix insertion In order to preserve the subcarrier orthogonality and the independence of subsequent OFDMA symbols, a cyclic guard interval is introduced. The guard period is specified in terms of the fraction of the number of samples that make up an OFDMA symbol. The cyclic prefix contains a copy of the end of the forthcoming symbol. Addition of cyclic prefix results in circular convolution between the transmitted signal and the Channel Impulse Response (CIR). Frequency domain equivalent of circular convolution is simply the multiplication of transmitted signal s frequency response and channel frequency response, therefore received signal is only a scaled version of transmitted signal (in frequency domain), hence distortions due to severe channel conditions are eliminated. Removal of cyclic prefix is then done at the receiver end and the cyclic prefix free signal is passed through the various blocks of the receiver Interleaving and De-interleaving OFDMA subcarriers generally have different amplitudes because of the frequency fading of typical radio channels. The deep fades in the spectrum may cause groups of subcarriers to be less reliable than others, thereby causing bit error to occur in bursts rather than being randomly scattered. Interleaving is applied to randomize the occurrence of bit errors prior to decoding. At the transmitter, the coded bits are permuted in a certain way, which makes sure that adjacent bits are separated by several bits after interleaving. Interleaving is done to protect the data from burst errors during transmission. Conceptually, the incoming bit stream is re-arranged so that adjacent bits are no more adjacent to each other. The data is broken into blocks and 8

21 the bits within a block are rearranged. In terms of OFDMA, the bits within an OFDMA symbol are rearranged in such a fashion so that adjacent bits are placed on non-adjacent subcarriers. As far as De-Interleaving is concerned, it again rearranges the bits into original form during reception Serial to Parallel Converter/Parallel to Serial converter In an OFDMA system, each channel can be broken into various sub-carriers. The use of sub-carriers makes optimal use out of the frequency spectrum but also requires additional processing by the transmitter and receiver. This additional processing is necessary to convert a serial bitstream into several parallel bitstreams to be divided among the individual carriers. Once the bitstream has been divided among the individual sub-carriers, each sub-carrier is modulated as if it was an individual channel before all channels are combined back together and transmitted as a whole. For example, in case a subcarrier modulation of 16-QAM each subcarrier carries 4 bits of data, and so for a transmission using 100 subcarriers the number of bits per symbol would be 400. The receiver performs the reverse process to divide the incoming signal into appropriate sub-carriers and then demodulating these individually before reconstructing the original bitstream. Once the cyclic prefix has been added to the subcarrier channels, they must be transmitted as one signal. Thus, the parallel to serial conversion stage is the process of summing all sub-carriers and combining them into one signal DAC and ADC A Digital-to-Analogue Converter (DAC) is used to transform the time domain digital data to time domain analogue data. RF modulation is performed and from the transmitter antenna, the signals go through all the anomaly in wireless channel. The received signal is converted to digital domain using Analogue-to-Digital Converter (ADC). At the time of down-conversion of received signal, carrier frequency synchronization is performed. After ADC conversion, symbol timing synchronization 9

22 is achieved. An FFT block is used to demodulate the OFDMA signal. After that, channel estimation is performed using the demodulated pilots FFT and IFFT OFDMA systems are implemented using a combination of Fast Fourier Transform (FFT) and Inverse Fast Fourier Transform (IFFT) blocks that are mathematically equivalent versions of the DFT and IDFT respectively but more efficient to implement. An OFDMA system treats the source symbols (e.g., the QPSK or QAM symbols that would be present in a single carrier system) at the transmitter as though they are in the frequency-domain. These symbols are used as the inputs to an IFFT block that brings the signal into the time domain. The IFFT takes in N number of symbols at a time where N is the number of subcarriers in the system. Each of these N input symbols has a symbol period of T seconds. The basic functions for an IFFT are N orthogonal sinusoids subcarriers. These subcarrier each have a different frequency and the lowest frequency is DC. Each input symbol acts like a complex weight for the corresponding sinusoidal basis function. Since the input symbols are complex, the value of the symbol determines both the amplitude and phase of the sinusoid for that subcarrier. The IFFT output is the summation of all N sinusoids. Thus, the IFFT block provides a simple way to modulate data onto N number of orthogonal subcarriers. The block of N_output samples from the IFFT makeup a single OFDMA symbol. The length of the OFDMA symbol is product of N and T (NT) where T is the IFFT input symbol period mentioned above. After some additional processing, the time-domain signal that results from the IFFT is transmitted across the channel. At the receiver, an FFT block is used to process the received signal and bring it into the frequency domain. Ideally, the FFT output will be the original symbols that were sent to the IFFT at the transmitter. Thus the role of FFT and IFFT can be summarized as follows: It is IFFT that basically gives OFDM its orthogonality. The IFFT transform a spectrum (amplitude and phase of each component) into a time domain signal. It converts a number of complex data points into the same number of points in time domain. Similarly, FFT at the receiver side performs the reverse task i.e. conversion from time domain back to frequency domain. 10

23 1.3.2 MERITS OF OFDMA OFDMA offers increased resistance to intra-cell interference, while providing remarkable flexibility in resource management and simplified channel equalization. All these features justify the adoption of OFDMA as a physical layer technique in emerging broadband wireless communications, including the IEEE e metropolitan area network (WMAN) standard. It is chosen as primary transmission scheme because of the advantages listed below. 1. As OFDMA is a parallel transmission system which converts the problem of frequency selective fading to flat fading by distributing data to sub channels, It seems to be the better candidate to combat multipath fading and randomizing the errors in burst. 2. In OFDMA system equalization is made very simple and reduces the complexity at the receiver, equalization is only applied to effected sub channel to reduce the error rate. 3. Delay profile of channel is nicely handled by insertion of the appropriate size guard band. 4. OFDMA provides a higher spectral efficiency and bandwidth utilization due to the orthogonality amongst the sub carriers which is depicted in Figure It is attractive for broadcast applications by using single frequency. 6. OFDMA is major role playing in the development of standards of a broadband access and compatibility with existing infrastructure. 7. Subcarrier spacing could be adjusted according to the requirements of an applications and data rate. it supports different modulation schemes for different sub channels DEMERITS OF OFDMA A fundamental weakness of OFDMA is its remarkable sensitivity to frequency and timing errors. A CFO between the receiver and transmitter destroys the subcarrier orthogonality and causes ICI as well as MUI. Along with this problem it has below 11

24 demerits and this leads to the development of SC-FDMA (Single Carrier- Frequency Division Multiple Access) which can overtake some of the problems. Figure 1.7 Comparison of Bandwidth utilization in FDMA and OFDMA 1. There exists a high peak to average power ratio which could drift the system into the region of non linearity and saturation, which reduces the power efficiency of systems. 2. The insertion of guard band reduces the spectral efficiency and thus total channel capacity is decreased. 3. In mobile environment the Doppler shift, Frequency Mismatch, CFO in case of higher number of carriers and spreading of OFDMA symbol. 1.4 WIRELESS CHANNEL In Non Line-of-sight (NLOS) wireless communication, reflection of the signal from the surroundings results in a received signal which is a combination of the multipath signals. Because these multi paths have different amplitude and phase, they may add-up either constructively or destructively leading to a complex envelope, i.e., fading AWGN and Fading Channel The Additive White Gaussian Noise (AWGN) channel is the simplest channel model used in most communication systems. The thermal noise in the receivers can be characterized as an additive white Gaussian process. Other factors inducing channel noise, such as antenna temperature, receiver filter, and multipath fading are also taken for analysis when considering AWGN channel [1]. Channel fading is generally 12

25 categorized into large-scale and small-scale fadings, which often occur simultaneously. Large-scale fading, due to path loss of signal as a function of distance and shadowing by large objects such as buildings and hills. This occurs as the mobile moves through a distance of the order of the cell size, and is typically frequency independent. Small-scale fading, due to the constructive and destructive interference of the multiple signal paths between the transmitter and receiver. This occurs at the spatial scale of the order of the carrier wavelength, and is frequency dependent. Small-scale fading, also known as Rayleigh fading, is not determined by the distance in communication. The small-scale fading is manifest in two ways: the signal spreading and the time variation. Rayleigh fading channel is considered her for analysis Rayleigh Fading Channels When information is transmitted in an environment with obstacles (Non Line of sight - NLOS), more than one transmission paths will appear as result of the reflection. The receiver will then have to process a signal which is a superposition of several different transmission paths. If there exists a large number of transmission paths they may be modelled as statistically independent; the central limit theorem will give the channel the statistical characteristics of a Rayleigh Distribution. p(α) = α σ 2 e α2 2σ 2, α (1.1) where p(α) is the rayleigh distribution,α is the envelope amplitude of the received signal and 2σ 2 is the pre-detection mean power of the multipath signal. 1.5 CARRIER FREQUENCY OFFSET Carrier Frequency Offset is the most significant term with reference to OFDMA systems. CFO occurs due to many reasons. Mainly this is caused due to the mismatch in frequencies between a transmitter and a receiver and also this can be caused due to the Doppler shift.the carrier frequency misalignment destroys the orthogonality of the subcarriers, which causes inter-carrier interference and consequently produces multi- 13

26 user interference among users. While the CFO can be estimated and corrected relatively easily in the downlink, preserving orthogonality in the uplink is much more demanding. In the uplink, the received signal is the sum of multiple signals coming from different users, each of which experiences a different CFO due mainly to oscillator instability and/or Doppler shift. These relative CFOs among users must be corrected; otherwise the system performance degrades severely. Various studies have been carried out to tackle the multiple CFO problems in the OFDMA uplink scenario, which can be divided into two categories: Interference avoidance and interference cancellation methods. Examples of interference avoidance schemes include windowing, self-ici cancellation and feedback-and-adjust approaches. The self-ici cancellation approach is performed in frequency-domain, where a set of code words with low ICI is used to reduce the side-lobes of the output spectrum. The existing system consists of two stages: Prior to uplink transmission, Mobile Station (MS) performs a coarse synchronization to Base Station (BS), using the well-known singleuser CFO technique. In second stage MUI cancellation are employed at the BS s side to estimate and correct the CFOs Frequency Offset The sensitivity of OFDMA systems to frequency offset compared with single carrier systems is a major disadvantage. In general, Frequency offset is defined as the difference between the nominal frequency and actual output frequency. In OFDMA, the uncertainty in carrier frequency, which is due to a difference in the frequencies of the local oscillators in the transmitter and receiver, gives rise to a shift in the frequency domain. This shift is also referred to as frequency offset. It can also be caused due to the Doppler shift in the channel. The demodulation of a signal with an offset in the carrier frequency can cause large bit error rate and may degrade the performance of a symbol synchronizer. It is therefore important to estimate the frequency offset and minimize/eliminate its impact [2]. If frequency offset is denoted as f c, the OFDM signal generated by the transmitter denoted as s(t) and y(t) is the signal received by the receiver with the duration of T, s(t) = e jwt x(t) (1.2) 14

27 y(t) = e j(w w )t x(t) (1.3) w = w w = 2π f c (1.4) Then the received signal has phase offset equal to y(nt) = e j wnt x(nt) (1.5) φ(n) = j wnt (1.6) Where w is the frequency difference between the transmitted and received signal which has the phase shift φ(n). The frequency response of each sub-channel should be zero at all other sub-carrier frequencies, i.e., the sub-channels shouldn t interfere with each other. The effect of frequency offset is a translation of these frequency responses resulting in loss of orthogonality between the sub-carriers and leading to ICI and MUI ICI AND MUI The effect of channel fading on OFDMA symbols is the destruction of the orthogonality of the sub-carriers. The broadening of the signal spectrum of the subcarriers by the Doppler shift makes them (sub-carriers) to overlap, thereby causing ICI and it in turns produces MUI among users. Due to this phenomenon, OFDM becomes very sensitive to frequency offset Doppler Effect The relative motion between receiver and transmitter, or mobile medium among them, would result in the Doppler effect, a frequency shift in narrow band communications. For example, the Doppler effect would influence the quality of a cell phone conversation in a moving car. In general, the Doppler frequency shift can be formulated in a function of the relative velocity, the angle between the velocity direction and the communication link, and the carrier frequency, as shown in Fig

28 Figure 1.8 Doppler Effect The value of Doppler shift could be given as, f d = v 2π cos(θ) (1.7) λ where θ is the angle between the velocity and the communication link, which is generally modeled as a uniform distribution between 0 and 2π, v is the receiver velocity, and the λ is the carrier wavelength. 1.6 OUTLINE OF THE PROJECT Chapter 1 discussed the general introduction to the OFDMA System, objectives as well as the outline. Background of OFDMA is addressed which includes the differences between OFDM and OFDMA modulation schemes. A basic OFDMA architecture and description of its various blocks are explained in this chapter. QPSK is introduced as an appropriate modulation technique for the OFDMA subcarrier. The channel model description and frequency offset generation are also depicted.chapter 2 portrays the Literature Survey.Estimation of the carrier frequency offset as well as the compensation is detailed in chapter 3. It includes the various CFO estimation and compensation techniques as well as algorithms used for cancellation of MUI. In Chapter 4, time and frequency domain estimation techniques are explained in conjunction with their relative merits on MSE performance, ccompensation method s result are analyzed based on SNR performance. The parameters used for simulations are tabulated then Simulation aims, results (plots) and analysis are also detailed in this chapter. Chapter 5 contains the inference and conclusion of the project obtained of the simulation results. 16

29 CHAPTER 2 LITERATURE SURVEY OFDMA is a technique used for radio transmission and reception in LTE. Frequency division multiple access (FDMA) method that assigns sets of subcarriers to different users and this is a popular basic multiple access scheme for OFDMA. In OFDMA downlinks, distinct subcarriers are assigned to different users for simultaneous transmission. Multiple users share the bandwidth simultaneously. Hence the users and the base station in the OFDMA downlink are required to be synchronous in frequency domain. In the uplink of the OFDMA downlinks, offsets in the frequency assigned between users occur whenever their local oscillators are misadjusted and due to a frequency shift in their carrier frequency in offset estimation channels. Asynchronous and synchronous transmission of data does not need the transmission of the training sequence of repeated symbols and differently from other technique the frequency estimator has performance independent of channel zero locations. The frequency synchronization is very much necessary for proper transmission in uplink OFDMA technique since the synchronization between the transmitter and the receiver in the channel reduces the data loss addressed in [1]. Robust timing synchronization algorithm for OFDM systems that utilize pilotaided channel estimation depicted in [2] and the impact of timing errors on the performance of a pilot-aided OFDM system is characterized in the presence of timing errors in high delay spread fading environments. The pilot-aided channel estimators are considerably sensitive to timing synchronization errors due to the impact of rotations in different bases. The algorithm used for synchronization proposed by Schmidt is a cross-block design that uses channel estimation information to improve timing synchronization. Algorithm discussed in [2] presents symbol-timing and carrier-frequency synchronization method in multipath fading environments. Algorithm proposed here was focuses the timing Synchronization for only lowmobility case. Effects of ICI is combated by effective interference cancellation techniques such as ICI Self Cancellation (SC), Maximum Likelihood (ML) estimation, and 17

30 Extended Kalman Filter (EKF) method. [3] analyse the comparisons in terms of bit error rate performance and bandwidth efficiency. It is shown that the three techniques are effective in mitigating the modulation schemes, the ML and EKF methods perform better than the SC method. The works presented in this paper concentrates on a quantitative ICI power analysis of the ICI cancellation scheme, which has not been studied previously. The average carrier-to interference power ratio is used as the ICI level indicator, and a theoretical CIR expression is derived for the proposed scheme. This method is Effective in mitigating the modulation Schemes where as ICI could not be removed in multi-fading channels. A residual CFO estimation scheme is proposed for the uplink of orthogonal frequency division multiple access (OFDMA) systems. which is Multiple access interference caused by CFOs in the uplink is investigated, as it severely affects the performance of a classical maximum likelihood (ML) frequency estimator. By the use of the estimated CFOs of the active users, the linear maximum mean square error (LMMSE) equalization is performed before the ML frequency estimator for the interference cancellation, which can help to sufficiently improve the estimation accuracy for the residual CFO of the incoming user which is clearly observed from [5]. Analysis and simulations show that the modified ML estimator provides a tradeoff between estimation accuracy and computational complexity caused by the LMMSE interference cancellation, and the proposed method allows OFDMA systems flexibly allocating subcarriers to users. In [6] Two effective CFO estimation schemes used for performance analysis i.e., time domain Cyclic Prefix (CP) based estimation scheme and frequency domain based Moose scheme. They Assess the effects of CFO upon Signal to Noise Ratio (SNR) of OFDM system and orthoganality of subcarriers. Where one of the demerit is Doppler fading performance of time domain CP. In [9] a hybrid interference cancellation receiver is designed by combining the successive and parallel interference cancellation algorithms, so that the advantages of both the schemes are utilized effectively for DS-CDMA system. 18

31 CHAPTER 3 TIME DOMAIN MUIC SCHEME OFDMA is used in both downlink as well as uplink transmissions where CFO can be estimated and corrected relatively easily in the downlink. In the uplink, the received signal is the sum of multiple signals coming from different users, each of which experiences a different CFO due mainly to oscillator instability and Doppler shift. Downlink CFO Correction methods, which are designed for single-user scenario, are unable to correct multiple CFOs in the uplink, as correction to one user s CFO would misalign the other users. The performance of the uplink of the OFDMA is severely degraded when the different carrier frequency offsets occur. The offset of the desired user can be compensated but the offset of the others users carriers are always misaligned and the interference due to this misalignment affects the reception. These relative CFOs among users must be estimated and corrected, otherwise the system performance degrades severely. CFO compensation in time domain with effective Selective Parallel Interference Cancellation (SPIC) has been proposed. 3.1 CFO Estimation The sensitivity of OFDMA systems to frequency offset compared with single carrier systems is a major disadvantage. In general, Frequency offset is the difference between the nominal frequency and actual output frequency. In OFDMA, the uncertainty in carrier frequency, which is due to a difference in the frequencies of the local oscillators in the transmitter and receiver, gives rise to a shift in the frequency domain. This shift is also referred to as CFO. It can also be caused due to the Doppler shift in the channel. The demodulation of a signal with an offset in the carrier frequency can cause large bit error rate and may degrade the performance of a symbol synchronizer. It is therefore necessary to estimate the CFO, which explains distortion in the transmitted symbols and hence at the receiver it can be compensated using some estimation techniques where flow chart of CFO estimation is shown in Fig 3.1.CFO estimation can be done effectively using time and frequency domain techniques. 19

32 3.1.1 Time Domain Estimation Figure 3.1 Flow of CFO Estimation technique Time domain estimation proves a fruitful technique in the case of CFO. The use of training symbols and the Cyclic Prefix (CP) are the ways through which we can perform time domain estimation for CFO efficiently.the CFO cause the phase distortion in the OFDMA symbol which causes ICI and MUI. If channel effect is minimal and can be neglected then the phase difference of the CP and the OFDM symbol which is the victim of CFO.The amount of CFO can be found by multiplying the OFDM symbol with CP and after that taking their phase angle measurements. This method is only useful for the fractional CFO. Perfect symbol synchronization done by phase rotation of 2πnε/N in the received signal. Under the assumption of negligible channel effect, the phase difference between CP and the corresponding rear part of an OFDMA symbol (spaced N samples apart) is 2πε. Then, the CFO can be found from the phase angle of the product of CP and the corresponding rear part of an OFDM symbol. Estimated CFO based on CP is, 1 ε = ( 1 2π ) arg { n= Ng y 1 [n]y 1 [n + N] } (3.1) 20

33 where, n = -1, -2,.... -Ng. In order to reduce the noise effect, its average can be taken over the samples in a CP interval Frequency Domain Estimation The problem in the time domain estimation of CFO can be addresed well by the use frequency domain estimation techniques. The well-known technique for CFO estimation is Moose and Classen. Moose gives the Maximal Like hood (ML) for carrier frequency offset estimation. In the Classen technique pilot tones can be inserted in the frequency domain and transmitted in each OFDM symbol for CFO tracking. Symbol gets extracted at the receiver side after synchronization. Moose method uses Maximum likelihood approach to determine the relative frequency offset ε as, ε = 1 2π tan 1 [ ( k im[y 2kY 2k k ( k= K Re[Y 2k Y 2k k= K ]) ] ] (3.2) where ε is the maximum likelihood estimate of the relative frequency offset defined as ε = N f/b, where B is bandwidth, N is the number of subcarriers and is the frequency offset in Hz. The estimation range of the Moose's method is equal to half ±0.5 sub-carrier spacing. Moose increased this range by using shorter training symbols, but that reduced the estimation accuracy. In classen method, two different modes for CFO estimation: acquisition and tracking modes are implemented as shown in Fig 3.2. In the acquisition mode, a large range of CFO including an integer part of CFO is estimated. In the tracking mode, only fine CFO is estimated. Figure 3.2 Flow chart for Classen method 21

34 The integer CFO is estimated by, L 1 ε acq = 1 2πT arg { Y(p(j), ε)y (p(j), ε) } (3.3) j=0 where L, p(j) and denote the number of pilot tones, the location of the pilot tone at the symbol period, respectively. Meanwhile, the fine CFO is estimated by L 1 ε = 1 2πT arg { Y(p(j), ε acq)y (p(j), ε acq) } (3.4) j=0 In the acquisition mode, ε acq and ε are estimated and then the CFO is compensated by their sum. In the tracking mode,ε only is estimated and then compensated. MSE performed by, 3.2 Multi-FFT receiver MSE = ε ε (3.5) In multi-fft receiver structure, each active user has separate demodulator block, so that their CFOs can be compensated independently. Demodulation block for multiple user is depicted in Figure 3.3. After CFO compensation, the output of the OFDMA demodulator belonging to the U th user can be expressed: Z u mfft = S u F N c ( εfu )r (3.6) U 1 Z mfft u = S u (Y u + C (εfu mfft 1 u ) Y u 1 + C ( εfu )V) (3.7) u 1 =0;u 1 u WhereC ( εfu ) represents the time-domain CFO correction by εf u and [. ] mfft in superscript refers to the fact that multiple FFT blocks are employed. εf mfft u1 u = εf u1 εf u denotes the relative CFO between the u th andu 1 th users. Cross-interference term sometimes become larger due to the fact that the new CFO,εf mfft u1 u, might be larger than the original one,εf u1. This tends to cause significant performance degradation. 22

35 Figure 3.3 General Receiver Structure for TDMUIC Channel Estimation Channels are fading both in time and in frequency. Channel estimator has to estimate time-varying amplitudes and phases of all subcarriers. 2D channel estimator estimates the reference values based on a known pilot values. To interpolate the channel estimates both in time and frequency from the available pilots, the pilot spacing has to fulfil the sampling theorem, which states that the sampling interval must be smaller than the inverse of the double-sided bandwidth of the sampled signal. For the case of OFDMA, this means that there exist both minimum subcarrier spacing and minimum symbol spacing between pilots. By choosing the pilot spacing much smaller than these minimum requirements, a good channel estimation can be made with a relatively easy algorithm. Even though more pilots are used the effective SNR obtained is lesser. Hence, the pilot density is a trade-off between channel estimation performance and SNR loss Channel Correction The receiver estimates the channel using the mean square value of channel state information. The user with the strongest Received Signal Strength (RSS) is demodulated first and given as input. The channel correction is done as the next step on where the estimated CFO (wrong channel) is compensated. This is done by subtracting the feedback from all the users except itself. The result of the channel correction would be the compensation of the CFO of the particular user. 23

36 3.2.3 Symbol Mapping and De-mapping During symbol mapping the input data is converted into complex value constellation points, according to a given constellation. Typical constellations for wireless applications are, BPSK, 16 QAM, and QPSK. The amount of data transmitted on each subcarrier depends on the constellation. Channel condition is the deciding factor for the type of constellation to be used. In a channel with high interference a small constellation like QPSK is favourable as the required Signal-to-Noise Ratio (SNR) in the receiver is low. For interference free channel a larger constellation is more beneficial due to the higher bit rate. Known pilot symbols mapped with known mapping schemes can be inserted at this moment. When channel estimation is done the complex received data is obtained which are demapped according to the transmission constellation diagram. At this moment, Forward Error Correction (FEC) decoding and de-interleaving are used to recover the originally transmitted bit stream. 3.3 MULTI USER INTERFERENCE CANCELLATION SCHEMES Each user in the uplink of an OFDMA-based system experiences an independent CFO. These CFOs, if not corrected, destroy the orthogonality among subcarriers, causing ICI and MUI, which could degrade the system s performance severely. In this chapter, the time and frequency techniques to compact the negative effects of the multiple CFOs in the uplink of OFDMA-based wireless communications are addressed SI MUIC SCHEME SI-MUIC stands for Simple Time Domain Multi-User Interference Cancellation Scheme. This scheme is based on multi-fft receiver. In this each active user is assigned one OFDMA demodulator block so that the CFOs can be compensated independently in time domain. The Frequency-Domain Multi-User interference Cancellation scheme has the disadvantage of power loss. This could be corrected by the SI-MUIC scheme. Assume that the users are sorted in order of their Received Signal Strength (RSS) and the BS processes from the user with the strongest received 24

37 power to the one with lowest power, thus increases the chances of correct estimation and decoding. The block diagram of the SI-MUIC Scheme consists of a frequency offset block followed by an IFFT block at the transmitter side which is shown in Figure 3.4. Here, transmitted signal of serial bits is converted to parallel bits and given to IFFT. Data symbols modulate the spectrum and time domain symbols are obtained using the IFFT. Cyclic Prefix is added to the start of each bit. The main advantage of using the cyclic prefix is to maintain the orthogonality. We know that each user my experience a CFO and in order to eliminate it we employ a method by which certain CFO is added with the transmitted signals and the receiver determines the strongest received signal and the weakest signal is eliminated with the strongest signal. Receiver demodulates the signal with strongest RSS first. The time domain symbols are then sent on the channel. Fast Fourier Transform is used in the receiver side where the time domain symbols are converted to frequency domain symbols and data is obtained. Cyclic Prefix is removed here. The output is sent to the decoder section. Figure 3.4 Demodulation Block for SI-MUIC Scheme Successive Interference Cancellation (SIC) algorithm is used in which the feedback from each iteration is taken for process the current user. On each iteration the CFO gets cancelled giving the reduced rate of estimation errors. SIC algorithm in the SI-MUIC scheme is superior to that of the Parallel Interference Cancellation (PIC) algorithm which is used in the FD-MUIC scheme. It has less interference and longer delay and causing the possible interference and error values. The most important advantage of the scheme is the Channel Estimation Information, instead only the 25

38 CFOs of each user is only needed. The most noted disadvantage in the method is the occurrence of residual noise. For removing such a noise the next scheme is used CA-MUIC SCHEME CA-MUIC scheme stands for Code Aided Time Domain Multi-User Interference Cancellation Scheme. It is similar to the SI-MUIC scheme, however differs in the fact that the feedback is added in the scheme at the receiver in order to avoid the residual noise that is appeared in the SI-MUIC Scheme. The main difference between the two schemes is that, instead of output of the OFDMA demodulator, the received signal itself is used to calculate the feedback. Figure 3.5 shows the demodulation block of CAMUIC scheme. For to estimate the received signal and output of the demodulator the carrier transfer function and the transmitted data symbols has to be perfectly estimated in prior. This method removes all the cross interference terms thus enabling a CFO free performance at the output of the demodulator. In order to achieve the correct estimate of the data symbols, channel coding is applied in the CA-MUIC scheme, thus it is named as code aided TD-MUIC scheme. The main difference in the SI-MUIC scheme with the latter is that the symbol mapping is applied to the CA-MUIC scheme. This provides the reduction in provision for multiple symbols for modulation and increased performance with the reduced CFO. Channel Estimation in both the above schemes is based on the pilot signal that is sent along with the transmitted signal. The delay subspace is estimated and delay subspace using the Least Mean Square Estimates. Pilot signals are multiplexed along with the data streams to estimate the channels. When the channel is slow fading the channel estimation inside the block can be updated using the decision feedback equalizer at the each subcarrier. The main idea behind such equalization is the use of channel estimate of the previous symbol for the data detection of current estimation and thereafter using the newly detected data for the estimation of current channel. 26

39 Figure 3.5 Demodulation Block of CA-MUIC Scheme Convolutional coder Convolutional coding is a bit-level encoding technique. Convolutional codes are used in applications that require good performance with low implementation cost. Using convolutional codes a continuous sequence of information bits is mapped into a continuous sequence of encoder output bits. The encoded bits depend not only on current k input bits but also on past input bits. This mapping is highly systematic so that decoding is possible. As compared with the block codes, convolutional codes have a larger coding gain. Encoding of convolutional codes can be accomplished using simple registers. The convolutional codes are denoted by (n,k,l),where n is number of output bits(coded),k is the number of input bits (uncoded), and L is the code memory depth, which represents the number of register stages. The number of the registers used in the encoding process is called the constraint length and it is indicated with C= (L+1).The efficiency or data rate of a convolutional code is measured by the ratio of the number of bits in the input (k) and the number of bits in the output (n), therefore Bit Rate: r = k/ n. To convolutionally encode the data, start with L memory elements shift register, each element holding one input bit. The encoder has modulo-2 adders, and n generator polynomials one for each adder. The Fig 3.6 illustrates a (2, 1) convolutional encoder with constraint length C=4. The selection of bits is to be added (uses XOR operation) to produce the output bits is called generator polynomial. The 27

40 generator polynomials O1 and O2 are of (111, 101) i.e. O1 = mod2 (Reg.1 + Reg.2+ Reg.3), O2 = mod2 (Reg.1 + Reg.3). Figure 3.6 Block diagram of convolution encoder If the initial contents of the register are all zeros, for the input bit stream m= , the output code word sequence obtained is MUIC ALGORITHM The performance of the uplink of the OFDMA is severely degraded when the different carrier frequency offsets occur. The offset of the desired user can be compensated but the other user carriers offset are always misaligned and the interference due to this misalignment affects the reception. Offset compensation of all the users at the base station is not possible because of the complexity involved in this process. CFO compensation in Time and frequency domain can be done by Parallel Interference Cancellation (PIC) and Successive Interference Cancellation (SIC) schemes respectively for cancelling the Multiple User Interference. The performance is better when the interference is weighted and then removed from the desired users signal. The cancellation is performed to remove the interference due to the others carriers of the same user, affected by the same frequency offset Selective Interference Cancellation (SIC) The difference in frequency between the transmitter and the receiver generally causes the carrier frequency offset that inturn causes Inter Carrier Interference and Multiuser Interference. Thus there is an effective algorithm needed to compensate this frequency offset. SIC is applied to compensate for frequency offset. The Multiuser 28

41 interference due to frequency offset is reduced by reconstructing and removing the interfering signals in the time domain.sic method is used for frequency offset compensation assuming that the frequency offsets of all uplink users are known at the receiver. Figure 3.7 shows the flow chart of SIC algorithm.different subcarriers are assigned to different users in OFDMA system this makes the signal to separate easier and the subcarriers coming from different users will have independent attenuations. As different users are assigned to neighbouring subcarriers, where most of the interference comes from, and their power levels are separable, SIC can be used to remove the interference due to frequency offset. Figure 3.7 Flowchart of SIC Algorithm The SIC is effective interference cancellation scheme for mitigate an error propagation. when errors are made during detection after that of the demodulation step, these errors will propagate to other users as MUI which is reconstructed incorrectly. It is especially dominant in low SNR and large frequency offset cases. In order to decrease the amount of error propagation, the detected bits can be decoded and coded again. This will reduce the number erroneous decisions increasing the performance of the algorithm in the expense of larger complexity and delay. SIC Algorithm is given as, Initialization: Set, Stepi = 0 Feedbacksignal, r u,i SI MUIC = 0 foru = 0,1,. U 1 29

42 LoopA i = i + 1 anduser, u = 0 LoopB Calculationofdemodulationandnewfeedbacksignal. Feedback: FromoutputofOFDMA Demodulator(SIMUIC), u 1 Z SI MUIC u,i = S u F N c (εfu ) (r r u SI MUIC 1,i u 1 =0 U 1 u 2 =u+1 SI MUIC r u 2,i 1 ) (3.8) r u,i SI MUIC = C ( εfu )F H SI MUIC N Z u,i (3.9) FromEstimationofreceivedSignal(CAMUIC), u 1 Z CA MUIC u,i = S u F N c (εfu ) (r r u CA MUIC 1,i u 1 =0 U 1 u 2 =u+1 CA MUIC r u 2,i 1 ) (3.10) u = u + 1 Untilu U 1, gobacktoloopb Untili N loop (No. ofiteration), gobacktoloopa Where, r u,i CA MUIC = C ( εfu )F H N Y u,i (3.11) c ( εfu )-Feedback signal from the u th user at the i th step. F N -Size-N FFT matrix. S u -Diagonal matrix of size N, acting as a filter to select only subcarriers belonging to the u th user. r u,i SI/CA MUIC -Feedback signal from the u th user at the i th step. Z SI/CA MUIC u,i -Output of demodulator block. Y u,i -Channel estimated Signal. 30

43 3.4.2 Parallel Interference Cancellation (PIC) Parallel Interference Cancellation is a cancellation scheme that is more or less same with that of an SIC algorithm. This algorithm is also very effective in mitigating the effects of multiuser interference that is occurring due to various users in a multiaccess environment. PIC performs the operations in parallel unlike the SIC algorithm. This could be noted as the main difference between the two. Flow chart of PIC algorithm is given in Figure 3.8. Figure 3.8 Flowchart of PIC Algorithm Estimation of the data coming from different users in the system can be done by various matched filters at the receiver end. The estimates for each user can then be used to reduce the interference to and from the other signals by subtracting the estimate of each interference from the desired user s signal. Ideally, this would allow the elimination of all interference from the desired user. Parallel Interference can be also realized if we move the block Subtract the received signal with residual estimation error obtained from adjacent users out of the user selection loop. In the algorithm instead of sending the feedback of demodulator to the next user we choose to send only the estimated channel and data symbols. In order to achieve correct estimate of data symbols, channel coding is done in this algorithm. PIC Algorithm is given as, Calculationofdemodulated signaland residual estimation error. 31

44 I CA MUIC,PIC u1,i = 0 (For First user) FromEstimationofreceivedSignal(CAMUIC), u 1 Z CA MUIC,PIC u,i = S u (Y u + C (εfu mfft 1 u ) I CA MUIC,PIC u 1,i u 1 =0 U 1 + C (εfu mfft 1 u ) I u 2,i 1 u 2 =u+1 CA MUIC,PIC + C ( εfu )) (3.12) C (εfu mfft 1 u ) = εfu 1 εfu (3.13) Y u,i = channelcoding(z CA MUIC,PIC u,i ) (3.14) Where, I CA MUIC,PIC u1,i = { Y u i = 0 Y u S u Y u,i i > 0 c ( εfu )-Feedback signal from the u th user at the i th step. (3.15) F N -Size-N FFT matrix. S u -Diagonal matrix of size N, acting as a filter to select only subcarriers belonging to the u th user. Z CA MUIC,PIC u,i -Output of demodulator block. Y u,i -Channel estimated Signal. I CA MUIC,PIC u1,i - Residual Estimation error Selective Parallel Interference Cancellation (SPIC) The proposed SPIC combines both SIC and PIC to the correct proportion so that the receiver performance is enhanced to reach the near optimal level. The successive interference cancellation receiver is simple, but requires high computational time, whereas the parallel interference cancellation receiver is more complex, but computational time is less. The best solution to get the desired improvement in performance is to have a perfect trade-off between the computational time and receiver complexity. The proposed receiver has PIC as the first stage of cancellation, 32

45 where the users having the decision statistic above a certain threshold, are cancelled using this method which is explained in Fig 3.9. The remaining users are given to the second stage of cancellation i.e SIC. The SIC process all other users who having decision statistics lesser than threshold, this process is continued until the desired user s signal is separated from adjacent user signal. Finally, the subtracted signal is given to a conventional receiver to get the desired user information. Decision variable can be calculated using, Figure 3.9 Flowchart for SPIC N U z u = Eα 2 m ρ m (j) + ( I i,iu i (i) + n(j) ) (3.16) j=1 i=1 j= Decision variable value for all user is calculated from the formula given in equation (3.16) and they are compared with the threshold value and the users are divided as per the condition shown below. Decision variable,b(u) = { z u > d ; PIC z u < d ; SIC (3.17) E = s i 2 (t) (3.18) t 0 33