Multi user detection in CDMA system

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Chapter 1: Introduction


Code Division Multiple Access (CDMA) is gaining popularity in wireless communications which can provide high quality for the services and capacity for systems. In mobile communication, CDMA accommodates bandwidth for each and every user by utilizing various kinds of coding properties. In CDMA a unique code will be provided to the user with the help of code multiplexing techniques. This unique code can be used in identifying the data and system of the user. Direct Sequence CDMA (DS-CDMA) is a modulation technique in which the unique code sequence will be multiplied with user data signal to generate information through spread spectrum signal. For any communication system particular algorithms or protocols has to be adopted in order to provide access to multiple users through a single medium of channels (Mike Buehrer, 2006). CDMA faces the problem of multi access interface. To face this problem a new concept called Multi User Detection (MUD) is introduced which can improve system capacity at CDMA receiver end. CDMA can obtain greater benefits through MUD in wireless communication systems. In this research Probabilistic Data Association (PDA) algorithm is used for providing multi user detection in CDMA systems. PDA is an alternative for multiuser detection which can makes use of soft inter user interface (IUI) cancellation. Using PDA algorithm a communication system can update the user signals iteratively. PDA algorithm first recognizes the users and then starts demodulating in order to obtain various user signals. Low Bit error rate (BER) and Signal to Noise Ratio (SNR) are used while implementing PDA algorithm in order to provide better results. The signal transmission has to be done through a channel, AWGN channel is used for evaluating this research (Peng Hui Tan and Lars K. Ramussen, 2006). Hence it can be concluded that, for providing multi user access in CDMA systems PDA algorithm can be utilized. The effectiveness of PDA alogorithm can be analyzed through this research.

Aims and Objectives


To implement PDA algorithm for CDMA based multi access detection with usage of AWGN channel.


  • Investigation on CDMA systems and its applications in wireless communications
  • Identifying the importance of Multi Access Detection (MUD) in CDMA systems
  • Evaluating PDA algorithm in CDMA systems

Purpose of Study

The main aim of this research is to implement PDA algorithm for facing the challenges in wireless communication systems. This research can serve the needs of future generation communication systems as it concentrates on improving multi user detection through employing algorithms. As the usage of communication systems is increasing, it is very difficult provide sufficient bandwidths to the users. Introducing Multi User detection (MUD) is an improvement to CDMA which can improve system capacity. Utilization of bandwidths in CDMA systems can be understood based on this research. This research focuses on CDMA problems with multi access and possible solutions using PDA algorithm. Another major intension behind this research is to increase knowledge of the scholar in communication systems and its implementation processes. The ability of practical implementation to PDA algorithm can be obtained by the researcher using this research (Piero Castoldi, 2002). Thus, this research is conducted for the purpose of improving communications in CDMA systems.

Research Context

The research on multi user detection in CDMA systems is in context of wireless communications. Background for the research topic should be identified before starting the actual process of researching. A research or project has to be planned effectively in order to attain maximized outputs. This research can help many wireless communication users in understanding the basic concepts of communication algorithms and its channels. Analyzing and identifying the research objectives is the essential step for conducting a research. The basic available sources must be recognized initially in order to understand the actual topic. Using strategies for collecting and interpreting data is the most needed setting for a research (Gale Miller and Robert Dingwall, 1997). This research can be used by various scholars who interested in CDMA systems and its implementation process. Various modulation and demodulation techniques available for attaining signal can be analyzed very easily by making use of this research. PDA algorithm is implemented in this research through an AWGN channel. Many communication channels previous to CDMA don't have the capability of providing Multi User Detection (MUD). CDMA offered many advantages by providing multi access to the users. This research can be used in mobile or broadband communications in order improve system capacity. Future generation communication systems can utilize this research in improving the system capacities (Francis Swarts and Ian Oppermann, 1999). Hence from the above discussion it can be understood that, research conducted in this report is based on communication systems in wireless environments. This research can be helpful to many users who deal with CDMA systems.

Research method

For evaluating any research or project a scholar has to use various kinds of research methods for collecting the required information. Information gathering must be done from valid materials and professionals in order produce efficient results for the project. A researcher can gather data from available materials or can establish new information by conducting surveys and interviews. As the topic considered in this research is available through many materials there is no need to generate new information. This research on Multi User Detection in CDMA System used secondary data as the information (Alison Mackey and Susan M. Gass, 2005). Already available data is gathered from various websites and books in order to evaluate the actual needs of the project. For secondary data either qualitative or quantitative approaches can be used. Selecting of research methods mostly depends on researcher interest and type of research topic selected by the scholar. For this research on CDMA systems qualitative data is the most suitable method as it provides information based on theoretical approaches. The collected data from various sources has to be analyzed properly so that a solution for the research problems can be obtained (Janet M. Ruane, 2004). PDA algorithm and its implementation process can be understood very easily by gathering information from available materials. Theoretical approaches provide the basic knowledge that is needed to implement the algorithms practically. Various kinds of sources for information are available in the present research environments, out of all available sources websites are the best possible forms for conducting this research (Irving B. Weiner, Donald K. Freedheim and John A. Schinka, 2003). Hence from the above discussion it can be understood that, qualitative data which is provided by various authors and professionals are considered for effective evaluation of this research.

Chapter 2: Literature review

3-G Wireless Networks

Based on the most of the demands for broadband services operators exaggerated the third generation networks. Third generation wireless networks evolved from the second generation wireless networks. 3G wireless networks use different types of technologies including GSM, CDMA and many other mobile services. It is the wireless function which transfers both voice and non-voice data. The third generation wireless networks standard is defined by International Telecommunication Union (ITU). Based on the difficulties from 2G wireless communication systems, 3G wireless systems established. There are three different elements which are involved in wireless networks. They are radio signals, data arrangement and network structure (Reza Vaez-Ghaemi, 2006). Wireless networks provide communication between the computers through radio signal frequency. Now-a-days these wireless networks are becoming popular because no cables are involved in the wireless networks. Wireless adapters and wireless routers are used in the wireless networks. The information is transmitted in wireless network is as follows. Initially the computer sends the data after encoding it to radio frequency and the encoded data is transmitted through routers. Finally the receiver decodes the same data to actual data. Both modems and broadband cables are used to work with wireless networks. Increase in the usage of internet and World Wide Web made to implement the wireless network. As so many people are using more than two PC's which are at the same place wireless network connection is implemented in these computers. Only router component is enough to install wireless networks. There are many advantages in using wireless network. It increases flexibility to the users and the cost is also minimized. Wireless networks are quietly different from fixed networks. But the wireless networks are not replaced (Matthew Gast, 2005). 3G wireless networks allow each and every user to have an immediate access for all its services. Wireless communications transform into real time connectivity. The third generation wireless networks for example a mobile serves both data and voice functions. These networks are based on code division multiple access (CDMA) technology. The characteristics of third generation wireless networks are as follows. It provides much flexibility to the users as it supports broad range of services. It is affordable than second generation networks that means it is reasonable. It is much more compatible with the existing systems than in using second generation technology. Third generation wireless systems can be easily extended to increase in the capacity and also the new services with only initial assets. The main aim of third generation wireless networks is to provide higher band width and it supports all multimedia applications. It helps to communicate with other information networks. These networks have high voice capacity and it supports all the advanced data applications (Moi Choo Chuah and Qinging Zhnag, 2006). Hence from the above discussion it can be said that third generation wireless networks solves the needs of the second generation wireless networks. Broad range of users makes use of these wireless networks. It provides more security, reliability. It provides interaction between the computers without nay cables. New radio spectrum is implemented in wireless networks for the effective communication.

Multiple Access Techniques

Multiple Access is nothing but the efficient allocation of signals between users. This is important for both uplink and downlink channels because single bandwidth is very expensive at the time when dedicated channels are allocated to users and dedicated channels are obtained from channelization method such as time division and frequency division multiple access (Andrea Goldsmith, 2005). Thus multiple access techniques depend upon the acceptance of digital technology as the digital as several advantages over analog.

Frequency Division Multiple Access

FDMA is traditional and famous technique used for the transmitting purpose. FDMA is fundamental technology in Advanced Mobile Phone system. In this the available bandwidth is divided into series of channels and each user is given only one channel at a time the receiver and transmitter is tuned to that frequency then allocated for control. Channels are dedicated to the single user, expect control channel the one which user share and used in conventional analog systems (Hill Associates and Inc, 2002). FDMA is easy and requires very simple algorithm and very efficient when network size is small. In FDMA the entire band is spit into sub-carriers. In FDMA a channel is corresponds to frequency band and assign individual channels to individual users. So has the entire bandwidth is divided in frequency bands no user can share the same frequency band at the same time, guards are maintained between the adjacent signals to minimize the cross talk between channels. Block and interleaved FDMA are different variations possibilities. Block FDMA assigns group of adjacent sub-carriers to each user where as interleaved assigns sub-carriers are divided in the frequency band (Ahmed Bahai R S, Burton Saltzberg R and Mustafa Ergen, 2004). The required FDMA bandwidth assuming single carrier is,

Where is number of FDMA channels supported by carrier and is channel bandwidth and is the guard band. The guard bands cause needless waste of the available bandwidth. The channel interference is important for finding the proper spacing between the FDMA carrier spectra.

FDMA is the technique that can be used for both analog and digital communications. Hence from the above context it can be stated as frequency division multiple access is the transmitting technique used for the communications. In this the available bandwidth is divided into series of channels and each user is assigned only one channel at a time.

Time Division Multiple Access

Time-division multiple access is digital transmission technology that allows number of users to access a single radio frequency channel without the interference and each time slot is assigned to each user within the each channel and also can be sated as, Radio frequency channel is utilized by only one of the stations sharing the channel and each station is allowed to transmit the channel for limited bursts so that the other stations can use same channel in other intervals of time. It uses the same frequency spectrum and that can be used by all the users at different time slots by dividing the time in slots. In TDMA the user is allocated the time slots and the information is transmitted in this slot of time by sharing the same frequency band (Barry George Evans and Institution of Electrical Engineers, 1999). The TDMA generally works as the audio signal is divided into number of small packets i.e. the time is divided into number of milliseconds packets. The single frequency bands are allocated to each channel and moves to the next. So the single transmitter will occupy different time slots in several bands at the same time. This plays vital role in case of advanced cellular communications and best technique used in commercial systems. This can be used for the voice communication for transmission of data. In case of mobile communication it can provides extended battery life and talk time. This also offers the communication services like fax, voice band data and SMS and also in multimedia and voice conferencing and this is the only technique which offers efficient utilization of hierarchical cell structures, by this the system can be achieved and cost effective (Abdelsalam Helal A, 1999). Hence from the above context it can be stated as the Time division multiple access is the digital technique used by dividing the time and assigning each time slot for the each user with in each channel.


Code Division multiple access is one of the method used in the wireless communication. By making use of the CDMA the capacity of the system and also the quality of service will be more increased in the cellular radio system operators. This method has gain most importance in the cellular system. Majority of the users was using this technology (M. R. Karim and M. Sarraf, 2002). Usually CDMA is in the form of the spread spectrum. From the last few years this technology is used in the digital communication systems. CDMA uses the conversion of analog to digital by combining the other technology called spread spectrum. Input audio is first transmitted to the binary elements. The frequency that is executed from the signal that is transmitted will make slight difference according to the given code because of this it only interrupts with the receiver which will combine with the similar frequency so the receiver will follow the transmitter which is having the same frequency (Byeong gi Lee and Byoung Hoon Kim, 2001). To make disturbances in the frequencies many possible frequencies are available. The networks of the CDMA will be using various kind of the methods called as the soft handoff which will reduce the disturbances in the frequency and passes signals from one network to another network. By grouping the modes of the digital and the spectrum, these modes usually supports many times in different signals as the bandwidth as the analog modes. CDMA is most efficient than the other mobile technologies where by allowing the roaming in the worldwide. This technology is normally common in the mobile technologies (Piero Castoldi, 2002). Hence from the above it is identified that code division multiple access is the technique where the electromagnetic frequency is transferred with the wider bandwidth over different directions.

CDMA Generation


Transmitter is an electronic device which broadcasts the radio, television and telecommunications signals with the help of the antennas. Transmitter antenna is used for transmitting the signals to reach the receiver. It allows the power supply, modulators and frequency with amplitude, and carrier. Transmitters are majorly used in wireless communications. Firstly the transmitter receives data and generates alternate current (AC). Frequency of the transmitting data is decided by the AC signals. Here transmitter acquires the data and modifies it using the modulator then encodes the data into signals. This will be the carrier signal containing the data ((David D. Coleman and David A. Westcott, 2009)). Carrier signal is received by the receiver antenna through cable. In this generation of signals, transmitter antenna is responsible of power levels in it and the original message with amplitude. If the amplitude of the signal is very high the wave travels more than the advance powerfully. Antennas focus on the RF signals in the transmission of the data. These are used in CDMA systems (Michael John Ryan and Michael Frater (2002). In Code Division Multiple Access the transmitter carries the signal at same channel. It consumes large bandwidths later modifies the signals. In this CDMA process the signals are spread with particular code in the transmitter antenna. This spreading in transmitter results in increasing the bandwidth and data rate of the signal. Hence spreading the data with allocated code is the technique in CDMA transmitter to improve the spreading and transmission easily.

Block Interleaving

Block interleaving is the tool for reducing the errors and making it performance well. It is generally used in error correcting methods in data transmission, computer and disk memory storage. It is the tool majorly used in digital communications as it helps in error correcting of block codes. It rearranges the encoded signals across the multiple coded block signals. By this the noise bursts look achievable foe the decoder. Error protection established on the length of the noise bursts can decide the depth of the interleaving (Andrea Goldsmith, 2005). Block interleave is classified in periodic interleaving. In which the periodic interleave arranges the data in repeating sequence of bytes. This interleaver takes the symbols in blocks and executes replacing over each and every block of data. It demands in accepting the input data and writing the symbols in rows and columns into matrix. Later this matrix came as x (i,j) block interleaver. Where i is number of rows and j is number of columns. These block interleavers are brought up in hardware (John Terry and Juha Heiskala, 2002). Usage of block interleaver reduces the error rate in digital communication in which burst errors are produced. When two adjacent code words are corrupted by more number of errors interleaver rearranges the bits of code words. These three errors increase the error correction capacity of the hamming code. When the hamming code blended with the interleaver the communication system can recover the original data from the transmitter.

Long PN sequence

The pseudo Noise sequence is one of the device used in the multicarrier CDMA mobile network system. Each and every base station produces PN sequences by including the multiple carriers receiving signal by making use of the local PN sequence which is present in the integrating period. PN sequences are the standard technique in the wireless communication system (Allen Belzer, Allen Kent and Albert George holzman, 1997). The base station separates the traffic by applying the various sequences to various subscribers. For the security purpose every uses various type of long codes, so decoding the content is not easy as the sequences that are using are orthogonal to one other (Lipo Wang, Dr. Ke Chen and Yew soon Ong, 2005). Two methods are involved to create the long PN sequence. In this one method uses the Electronic serial number of the user by producing the long PN sequence, if the electronic serial is identifies then it is easy to identify the PN sequence. Another method used is this produces PN sequence by making use of the keys which are basically known to the user and the base station there by providing the security and the simple de-spreading.

Data scrambler

Data scrambler is one of the technique used in a data communication system, to operate in a given data communication mode and in non data communication mode, the modem in the transmitter make use of the single scrambler. During the process of data communication mode to scramble the data it makes use of scrambler for communication through the transmitter (W. Pam siriwongpairat and k. j. ray liu, 2007). The modem given is in the form of ADSL modem where the data communication mode is in the format of SHOWTIME and the non data mode is in the form of Q-mode. Data bit that is transferred in the data transmission must be independent. Bit sequences can create wide differences in the power level that is received and will create difficulties for the adaptive equalization and clock recovery. All these difficulties are occurring due to bit sequence is in random that is, it has no visible. Most of the modems make use of data scrambler to generate a pseudorandom sequence for the provided input bit sequence (Steven A. trotter, 1995). Usually the scrambler considers shift register along with the feedback connection and the unscrambler is considered as a feed forward connected shift register. These operations clearly mention the process of data scramblers.

Walsh Coding

This is a group of spreading codes with good auto correlation properties and poor cross correlation properties. This code has much importance in CDMA systems and is used to develop the individual channels in CDMA. Apart from this there are some issues with Walsh codes like synchronization of all users is required. This is actually an algorithm that generated statistical unique sets of numbers for use in encryption and cellular communication. Walsh codes are used in frequency hopping spread spectrum systems to select the target frequency. Walsh coding is used in some CDMA systems to provide orthogonal codes for different users. When combined with OFDM as in the case of Multi-Carrier CDMA systems, it can allow exploitation of the diversity of the multiple carriers without the channel knowledge required at the transmitter for differentially loading each carrier. This is used to define individual communication channels uniquely. A Walsh encoded signal appears as a random noise to CDMA capable mobile terminal unless that terminal uses the same code as the one used to encode the incoming signal. The main purpose of Walsh codes in CDMA is to provide orthogonality among all the users in a cell. These Walsh codes are used to identify the data for each individual transmission (Charan Langton, 2002). Hence from the above context it is understood that Walsh coding increases the data rate and provides good autocorrelation.

Quadrature Modulation

It is the simple technique which is a combination of amplitude modulation and phase shift keying. In technical aspects quadrature modulation is a system of modulation in which data is transferred by modulating the amplitude of two separate carrier waves. Quadrature Modulation techniques enable two independent signals to be combined at a transmitter, transmitted on the same transmission band and separated at a receiver. This modulation up converts the information to an intermediate frequency (IF) or directly to the carrier frequency depending on the structure of the transmitter radio frequency chain. This allows combining the two independent signals to be separated again at the receiver. Quadrature modulation usually involves multiplication with a complex exponential and subsequently extracting the real part of the complex product for transmission. Because of its efficiency in power and bandwidth, QM is widely used as modulation technique (Xiaolong Li, 2008). This technique makes use of different phases and refers to QPSK with Amplitude Modulation. This is a very efficient technique for digital communication. It makes use of multiple signal phase and amplitude levels to carry multiple bits per symbol which requires accurate and robust carrier phase and frequency estimation in the receiver (Feng Rice, Bill Cowley, Bill Moran and Mark Rice, 2001). Hence it is understood that this technique allows transfer of data by modulation of two carrier waves.

Multi Path Fading Channel

Multi Path Fading usually occurs when a transmitted signal divides and takes more than one path to a receiver while some of the signals arrive out of phase resulting in a weak signal. It is also known as Multi path Interference (MPI). This type of fading is relatively fast and is responsible for the signal variations. It can be used to describe the rapid fluctuations in the amplitude of a radio signal. In a broadcast television and poorly installed cable television, ghosting may be result of multipath fading. Factors like MPI specter, ghosting contributes to multipath fading. Multipath fading affects most forms of radio communications links in one form or the other. It can be detected on signals across the frequency spectrum from the HF bands. Along with short wave radio communications wherein the signals fade, it is also experienced in other forms of radio communications like cellular telecommunications and many other users of VHF and UHF spectrums. At times there will be changes in the relative path lengths which can be resulted from either the radio transmitter or the receiver moving. Multipath fading can affect radio communications in two ways namely selective and flat fading. It causes serious multipath interference which exists in a mobile communication system. While designing an Automatic Frequency Correction (AFC), the effects caused by the multipath fading channel should be considered. (King.Ngan.N, Chi.Yap.W and Keng.Tan.T, 2001). Hence from the above content it is understood that Multipath fading is responsible for signal variations which can affect radio communications in two ways namely selective and flat fading.

Noise in Communication Channel

A Communication channel can be referred as a medium used to convey information from a sender to a receiver. It is a way in which one component can influence other by causing the other to do something or by simply providing the other with some information. Communication between two entities can be considered either in-band or out-band depending upon the context. Communication may occur through different channels like telephone, electronic, visual or air waves. Communication noise refers to influences on an effective communication. It has a large impact on perception of interactions. Noise in a communication channel makes effective communication difficult (Michael P. Pagano and Michael Pagano, 2009). Noise is explained in detail in further section.

White Noise

By combining the sounds of different frequencies together, a noise is produced. Such type of noise is referred to as White noise. We can have white noise in all the imaginable tomes a human can hear. Because white noise contains all types of frequencies, it is frequently used to mask other sounds. For instance, if three people are talking simultaneously, a human brain can probably identify single voice at a time. Similarly, if 1,000 people are talking simultaneously, there is no way that a human brain can still pick out one voice. Here the instance of 1,000 people talking together can be referred to as White noise. It is a random signal which contains equal power within a fixed bandwidth at any frequency (James Howard Cox, 2006). White noise is commonly used in the production of electronic music usually as an input for a filter to create other types of noise signals. It is also used to generate impulse responses. It is particularly a good source signal for masking other devices as it contains higher frequencies in equal volumes to lower ones and therefore is capable of more effective masking for high pitched ringtones. White noise is regarded as a random process z (t) with mean 0 and variance 8 such that z (t) and z(s) are independent for t ≠ s and

Therefore z (t) can be considered as a continuous analogue of an independent and identically distributed random sequence. Z (t) will have infinite fluctuations at each time (Hui-Hsiung Kuo, 1996). Hence from the above content, it is understood that a noise which is produced by combining different frequencies together is called as white noise.

CDMA Receiver

RAKE Receiver Combining

Rake receiver usage improves the performance of wireless systems such as in Code Division Multiple Access (CDMA). Rake receiver blends different signals that are sent through channel by different paths. This combination of different signals will raise to the increase the signal to noise ratio. Rake receiver signals can improve and increase its energy in fading channel by the diversity techniques. Rake receiver structure can take the power of the received signals and can have the equal number of fingers similar to number of multipath terms. This may advance in problem so it needs a number of rake receivers and infinite correlates. (lan Oppermann, Matti Hamalainen and Jari linatti, 2004). The functioning in Additive White Gaussian Noise (AWGN) which is close to the operation within rake receiver can be satisfied by the maximum ratio combining (MRC). This combining of the signals gain the optimum operations, but many of the channels approximations needs the information channel. Here the transmitted signal comes to the receiver after the multipath broadcast. These signals passing through the fading channel are detected from delay spreads. Each of the receiver signals are multiplied by the coefficient of channel estimator, later all these signals are combined in the process called rake combining (Minoru Etoh, 2005). Rake receivers are mainly used to reduce the distortions in CDMA. As the signal passes through many channels it undergoes thermal noise, multipath fading, co-channel interference and others results in signal noise. Because of this rake receiver follows the diversity techniques to overcome the noise distortions. Some of the techniques include time, frequency and space. Rake receiver after combining and with its diversity techniques uses its signals with minimum noise distortion. Hence from the above context it is understood that the signals which are propagated through the many fading channels experiences the signal to noise ratio and distortions so the rake receiver combines its signals to improve the performance. This combining of the rake receiver helps in minimizing the distortions in signals.

Long PN Sequence

PN sequence is known as the Pseudo-random Noise sequence of the binary numbers. It is the sequence of chips used to random the spreading of the signals. This sequence is generated by mixing the outputs of the shift registers, which can increases the length of the sequence. The chip sequence is strained by the chip pulse which can decide the bandwidth of the CDMA system. Whenever the spreading sequence is long, the signal differs from symbol to symbol. If the PN sequence is long the spreading sequence is non periodic and the signal can be random with some limit of hours (Piero Castoldi, 2002). This type of PN sequence in CDMA is known as pseudo random CDMA. Here the sequence is non periodic and signal changes from symbol to symbol. Every CDMA will hold the logical channels that are modulated and multiplexed by the PN sequences. CDMA performs its function with two short PN sequences and one long PN sequence. All these long pseudo random noise sequences used in CDMA systems are synchronized to a referred time as the beginning of the CDMA systems (Leonhard Korowajczuk, 2004). This PN sequence is longer than the Walsh codes in CDMA. Phase shift between the two same PN sequences is known as the phase offset in number of chips. This phase offset is used to separate the channels which uses the correlates, and also used when the source and destination are not perfectly synchronized. There are two types of long PN sequence. They are: private and public sequence. Public PN sequence is created by the mobile ESN user where as the private PN sequence is created during procedure. Using public ling PN code calls are started. Later the authentication starts using the private long PN code. Hence these sequences are used to synchronize the multiplexed and modulated signal in CDMA system. PN codes of the CDMA shows the characteristics of the randomness. . PN sequences of chips will provide the separation and improves the characteristics of the signals in CDMA.

Viterbi Decoder

It is an algorithm used for decoding the particular convolution code. These perform very effectively to decode the wireless communications in cell phone and satellite communications. It was developed by the Andrew Viterbi. Mainly used in CDMA systems. It estimates the output as 0 or 1 according to the input. Viterbi decoder in CDMA technology used to correct the errors in convolutional code which are brought by the base station. This viterbi decoder is used in many systems which face the errors and transmits the data (Rudolf Tanner and Janson P. Woodard, 2004). Viterbi decoder consists of two architectures: one of them uses small area; it transmits the data very fast along serial implementation. Second it may break the data transmission rate that can reduce the bandwidth needed for the channel. Viterbi decoder in CDMA systems supports multi channel applications with high speed. The two basic architectures allow the users to modify the resource usage. Convolutional code operation increases in decreasing the code rate of signal and with increasing the length (Luca Fanucci, Filippo Giannetti Luise and Massimo Rovini, 2004). Decoders are used at the receivers that impact to a great extent in increasing the constraint length. On this basis, in some typical communication cases the viterbi decoders are used in CDMA systems. These viterbi decoders can experience the trellis representation of convultional code and finds out the trellis based receiver signals (M. R. Karim and M. Sarraf, 2002). Some of the multi user detection structures replace the viterbi decoders by different devices to reduce the complexity. Hence from the above context is said that viterbi algorithm decodes the convolution code which results to decode effectively. It will reduce the errors created from the base station then transmits the data to the receiver. Viterbi decoder backups the CDMA system with high speed and transmission rate at low bandwidth.


Code division multiple access is one of the different way to transmit the information by its techniques and advanced technology between the base station and a mobile. As it is a multiplexing method it puts all the calls in a single channel. In FDMA division multiplexing occurs with changes in frequency, same as in TDMA the time changes in time division multiplexing access. In all these multiplexing accesses the traffic involves the voice, audio and noise signals. It advances its technology on basis of 3G networks. Transmitter, interleave and scrambler are used in generating the CDMA system. All these scrambler, sequence will code the signals and transmits to the receiver. All these Walsh codes, data descrambler and modulation techniques are implemented in the transmission process to reduce the noise distortions to improve the fastness and receiving power of the signals from the transmitter through base station. In the same way the rake receiver uses its diversity techniques to minimize the distortions and can reduce the bandwidth which involves time and frequency. All these include for implementing the new techniques for CDMA development and performing effectively in communications.

Chapter - 3: Basic Building Blocks in the proposed CDMA System

Multi User Detection

Multi user detection is the process of detecting the preferred signal from various types of communication and noise. It considers as signals for each other. It demodulates more than one signal with the existence of multi user interference. The problems of code division multiple access are been solved by implementing different multi user detection techniques. One of the features of multi user detection is it designs the receiver filters to conquer the intervention of other users (Raghuveer M.Rao and Sohail A.Dianat, 2005). To improve the capacity of CDMA systems multi user detection technique is implemented. The purpose of multi user detection is increasingly rapidly in non-orthogonal CDMA systems. It is proposed mainly for the CDMA systems. Multi user detection helps to solve the near far problem. This problem occurs when the users tries to submerge the signals which are far away from them resulting in less power. Hence this is solved by the multi user detection by adjusting the different transmission levels of different users (Christian Schlegel and Alex Grant, 2006). The different schemes in multi user detection are optimum detector, conventional detector, multistage interference cancellation and many others. It figures out the transmitted bits through multiple access interference. It became the most important task for signal processing in wireless communication. It detects the signals for the effective communication at the both ends of signal transmission. Multi user detection and multi carrier signaling techniques are combined in direct sequence code division multiple access. Hence from the above discussion it can be said that multi user detection is helpful in detecting the signals for the efficient communication in between transmitter and receiver. It is designed mainly for the code division multiple access to increase the performance of the system. Many strategies are implemented in the multi user detection.

Adaptive Filters

Adaptive filter is the filter which adjusts itself network function. It varies with time. Based on the complications of optimizing algorithms, most of the adaptive filters are used as digital filters which perform digital signal processing. In some applications adaptive coefficients are required since the properties of the signals are not known earlier. So, in these cases it is recommended to use adaptive filters as it gives the proper feedback about the signals. This process makes use of cost function to modify the contents in the filter. It helps in minimizing the disturbances or any noise in the given input. Adaptive filters are used rapidly in many devices such as mobile phones, cameras and in many other electronic appliances. The process of adaptive filtering is done effectively even in lower speeds. But it is mostly preferred in higher speeds when power consumption is low (Yichuang Sun, 2002). Adaptive filters are mainly used in statistical signal processing. It is used in both end points of the communication. Even though the knowledge of signal is not known, it is possible for the adaptive filter to develop the useful filter. Different parameters of an adaptive filter is been modified time to time after the specified data flows through it. There are two types of adaptive filters. They are linear adaptive filter and non-linear adaptive filter. Linear adaptive filter is the filter in which the output is the combination of input and between adaptive operations. Non-linear adaptive filter is the filter of which non-linear relationship between input and output (Anthony Zaknich, 2005). It is mainly required when the actual specifications are not known or that specifications are not been specified by filter. The characteristics of adaptive filter are dependent on input signal. So, the adaptive filter is known as non-linear filter. Sometimes the parameters of filter are suspended. In these cases the adaptive filter is known as linear filter. To increase the performance of the system, adaptive filters vary from time to time. It is complicated than fixed filters since it is said to be non-linear system. Adaptive filters are implemented using adaptive algorithms (Rulph Chassaing, 2005). Hence from the above discussion it can be said that adaptive filters are widely implemented in many applications. It is used even in radar communications. It is applicable in cancellation of noise, echo and many others. It is updated continuously based on the requirements of the system.

Adaptive Filters Structure

The adaptive filters are carried out in different number of structures or recognitions. These structures are implemented to increase the performance of the system. Adaptive filters structure is divided into two main modules which are renowned by the form of impulse response. They are finite-duration impulse response (FIR) filter and infinite-duration impulse response (IIR) filter. FIR filters are enforced through non-recursive structures and IIR filters enforce recursive structures.

Adaptive FIR filters structure:

The adaptive FIR filter structure which is used broadly is transversal filter (Paulo Serigo Ramirez Diniz, 2008). This filter divides the input signal into individual filter by using some phases. For this filter structure, the output signal is the combination of the set of filter coefficients. Different FIR filter structure are also used in compared with transversal filter in terms of speed and so on.

Adaptive IIR filter structure:

The structure which is used in adaptive IIR filter structure is the direct form of structure since it easily implemented with all the basic analysis. Even though it is designed simply but still there are some limitations of it since the speed is slow and so on. Different structures are suggested to solve the problems of this direct form structure. It is used in infinite duration. The main advantage of direct form structure is it uses very less number of coefficients to realize the network function. To overcome the problems of this structure alternate adaptive filter structures are proposed.

There are three major components which are involved in adaptive filter structure. They are filter structure, adaptation algorithm and criterion of performance. Filter structure is the accomplishment of the FIR filter. Based on the input signal, filter structure blocks the output signal. All the coefficients of the filter are updated by adaptive algorithm. A performance criterion of the adaptive filter structure analyzes the output signal and compares it with the other signal (Jose Mira and Juan V.Sanchez-Andres, 2000). After analyzing the signal it is sent to the adaptive algorithm for modification. The main division of the adaptive filter structure is the adaptive algorithm. It decides how to change the signals based on the feedback sent by performance criteria. It helps in designing the system. Hence from the above discussion it can be said that adaptive filter structures are implemented to increase the performance of the system. It generates the signals which are unrelated to each other. These filters are used in many consumer appliances.

System Model

System model in CDMA combines the following schemes namely Selection Diversity (SD), Equal Gain Combining (EGC) and Maximum Ratio Combining (MRC). This model is used in Rake receiver combining for its better achievement of diversity techniques in propagating the information through the multipath.

Selection diversity: It is used in reverse link of IS -95. When a number of mobile signals are passed to different base stations, the base station that which receives the strongest signal is selected to serve in return to the mobile server (Kim-Chyan Gan, 2004). In some of the cases signals are received with the L-multipath, here the combining scheme of the receiver selects the high signal to noise ratio and keep away the L-1 paths. These reverse links of the signals in multiple base stations are used to implement the diversity in CDMA receiver.

Equal Gain Combining: The rake receiver of the CDMA system gets the energy of the received signals. There received signals gain the equal number of the weight like the number of multipath. The power of transmitted signals earned at receiver after correcting the phase rotation and noise distortions in propagation. And all these distortions and phase of the rotation are aroused because of the multipath fading channel (Tolga M. Duman and Ali Ghrayeb, 2007). As the signals are propagated through fading channels are detected using delay spreads. Since the signals are broadcasted through many paths it will get through the signals to noise ratio, white Gaussian noise and errors.

Maximum Ratio Combining: It is the process of decoding the additive white Gaussian channel using the information from many paths. Errors and phase rotation caused by the fading channel are corrected by the receiver and combines the received signals relative to strength of each and every signal. As of all the received signals undergoes different fading channels and then by the maximum combining of these signals can generate a good solution for AWGN channel. This maximum ratio combining works better in Rayleigh fading channel (Andrea Goldsmith, 2005). The performance of the Maximum ratio combining will be better by following the EGC and SD since the signals strength may be equal. In this system model the signals are correlated differently because the extra data is extracted from each path. Hence from the above context it is understood that signals received at the receiver are propagated through multipath fading channels and it undergoes many errors and noise. These can be reduced and makes the performance of the system better after following the schemes of the system model in CDMA communications system. These are known as diversity techniques implemented to acquire the power of the received signals depending on the environment.

Matched Filter

Matched filters are the basic tools in the electrical field which is used for extracting the wavelets from a signal that is affected by noise. This is processed and overcome by cross correlating the signal with the wavelet. Actually this is a process for detecting a sample of signal or wavelet that is embedded in noise. This filter will maximize the signal to noise ratio of the signal being detected with respect to the noise. Matched filters are designed to extract the maximum SNR of a signal that is hidden in noise (Springer, 2005). The noise may be Gaussian or any other form of noise. In the absence of noise the output is just the signal energy with the matched filter. The matched filter maximizes the SNR at the output of an FIR filter. Its frequency response is designed to exactly match the frequency spectrum of the input signal. Its operation is same as that of correlating a signal. These filters are widely used in many applications and are also used in signal processor in communication receivers to calculate the correlation between the transmitted signal and the received signal. There are two prototypes of matched filters that are presented by FPAA device where one is a chirp compression filter where the matched filter has same structure as FIR, and the other is direct matched filter that is implemented using analog models (CAMS) with lookup tables (LUT). In CDMA systems the matched filters are tuned to match the code sequence which is expected to be contained within the digital samples entering the systems receiver. When this code is detected, the matched filter indicates it in the input data stream and the output of this filter will be a score value (Vincent K. Mc Donald, Paul Hursky and KauaiEx Group, 1999). Higher value indicates more tuned match with the code sequence of the received data stream which is called as correlation and hence a high score value represents a good correlation of input with the code sequence of interest. The output of the matched filter is called as matched filter response which is a signal that may not look like a transmitted signal but has a value at the moment of decision (Charan Langton, 2003). Hence it can be understood that the matched filter has unique property of recognizing the signal to noise ratio of a signal that is hidden in a noise and has wide applications in electronics field.

Optimal Multi-user Detection

Multiple user detection is used to tackle the near far problems. It actually detects the data signals from all the users. The receivers based on multiuser detection usually outperform but are usually more complex than receivers based on single user detection. Design issues like security of joint detection, implementation complexity and availability of information required to perform multiuser section. This detection improves the DS-CDMA detection through the use of multi-user detectors (Zhi Ding, Xiaodong Wang and Macro Lops, 2002). In this the code and timing information of multiple users are jointly used to detect each individual user. It actually refers to the joint decoding of user's signals in wireless systems. Instead of viewing the users as interference noise, MUD tries to cancel the effect that each user has on the other to achieve significant capacity increase and resistance. MUD provides significant benefits in Code Division Multiple Access (CDMA) system. The system capacity increases with the increase of MUD performance as interference can be tolerated. Not only that this allows the weaker users to be detected in presence of strong interfere, less stringent power control is needed and a better near far resistance is achieved. Even though different limitations delay the application of MUD, there are many improvements in this technology (Rami Abdullah, 2007). As optimum multiuser techniques can't be realized, more attentions are paid to suboptimum detection schemes. This offers optimal performance with a complexity. Several methods have been suggested to reduce the effects of multiple access interference and near far interference. Optimal multiuser detector is proposed to mitigate the effect of Multiple Access Interference (MAI). This OMUD computational complexity grows exponentially with the number of active users (Yanping Li, Chengrui Zhang, Xiaohong Xie and Huakui Wang, 2003). Hence from the above context it can be understood that optimal multiuser performs optimal performance in the CDMA system.

Maximum Likelihood Criterion

Maximum Likelihood (ML) Criterion is an alternative techniques used for allocating the resources to different people in a group. Maximum likelihood is the best principle which helps in identifying parameters from the data. For identification of parameter with the help of maximum likelihood two main things have to be considered which are given as (Malcolm Haddon, 2001)

Theories have considered depending upon the models

If the considered theories are correct, than a function is defined for calculating the probability density

Consider sequence of observation, which belong to a class w, then the probability of total likelihood of the observation can be given by the mathematical expression

Maximum likelihood Criterion can also be referred as Generalized Maximum Likelihood (GML) (Narada Warakagoda, 1996). The main principle of PDA algorithm is to select the probability density for the selected data. Maximum Likelihood (ML) provides benefits are it eliminates interference and improves the speech. Hence from the above context it can be concluded that Maximum Likelihood criterion is used for identification of parameters within the data.

Gradient Projection Method

Gradient projection method was given by Rosen in year 1961 where it reduces nonlinear function to linear functions. This method gives procedure for calculating direction vector in an easy way. This method provides good results when compared with the results of feasible direction approach. Some of the features of this method are movement through the polyhedron feasible region, generating objective values and generation of dual variables. In this method, if initial point is within the feasible set the steepest descent direction foe cost function is used and if initial point is infeasible then correction steps have to be taken (Singires S. Rao, 2009).

At point and -are the steepest descent direction and is the negative projection gradient. The constraint correction steps are executed from point to to reach the feasible point. With usage of this method search direction can be calculated easily. The Gradient projection method and constrained steepest method have same direction when constraints are active.

Kuhn-Tucker Conditions

In the field of non-linear programming, the most important theoretical results are the conditions of Kuhn and Tucker. The Kuhn-Tucker conditions are usually necessary if the objective function is concave provided each constraint is linear. In other words, it can be said that the problems belong to a class called as Convex Programming problems. For both the unconstrained optimization problem as well as optimization problem with an equality constraint, first-order constraints are usually sufficient and the constraint functions satisfy the concavity and convexity conditions. The same can be true in the case of optimization problem with inequality constraints. The Kuhn-Tucker conditions present a combined management of the constrained optimization in which the constraints may be binding or not binding at the particular solution, boundary solutions are allowed, any number of constraints can be permitted, the non-negativity as well as structural constraints are treated alike. These conditions can be referred as the first-order conditions for a constrained optimization problem. Linear Programming is a special case which is covered by Kuhn-tucker conditions (Alfio Quarteroni, Riccardo Sacco and Fausto Saleri, 2007). Hence from the above content, it is understood that, Kuhn-Tucker conditions are necessary if the objective function is concave provided that each constraint is linear. Sign convention has to be strictly followed in order for the Kuhn-Tucker conditions to be applicable. Solving a Support Vector Machine (SVM) problem is similar to solving a Kuhn-Tucker problem. Some of the Kuhn-Tucker conditions are that the gradient of the Lagrangian and complementary slackness should be zero. Feasibility is also considered as an important Kuhn-Tucker condition. If a given optimization problem is a convex programming problem, the relative minima will be negligible and hence the extreme point found by applying the Kuhn-Tucker conditions is considered to be an absolute minimum of the objective function. The derivation of these conditions was based on the development given for the equality constraints. One major requirement for these conditions was that at least one of the total constraints should be non-zero (Rao S.S, 2007). Hence from the above content it is understood that, Sign convention has to be followed strictly in order for the Kuhn-tucker conditions to be applicable and Feasibility is considered as an important Kuhn-tucker condition which should be followed.

Stopping Criteria

Stopping Criteria is usually used to terminate the execution of the optimization algorithms. It is impossible to define stopping criteria without introducing one or more parameters. These parameter settings generally depend upon the given optimization problem. In place of using maximum number of function evaluations as a stopping condition, another criterion usually has the advantage of reacting adaptively to the state of optimization as a result of which function evaluations can be saved. It is important to check whether there is a stopping criteria present for which the parameter can be subjected to change or if they can be set depending upon the certain aspects of the given problem. It is also important to note that by limiting the number of function evaluations as a stopping criterion by also including a problem dependent parameter. The two stopping criteria of an adaptive algorithm are maximum allowed adaptive iteration steps along with maximum allowable mesh refinement (Hai Jin and International Federation for Information Processing, 2004). Hence from the above content it is understood that stopping criteria is generally used to terminate the execution of the optimization algorithms. Algorithms which generate streamlines based on maximum directions have tended to require harsh streamline stopping criteria based on the fractional anisotropy as well as local curvature which is nothing but the angle between the successive steps. Fractional anisotropy usually tends to be in the range of 0.2-0.4 whereas the curvature thresholds have successive steps to remain within 45 degrees. These criteria are in place in order to reduce the sensitivity of the streamlining to noise in the image, partial volume effects and some other related problems. The final aim is to reduce the possibility of seeing false positives in the results by only progressing when there is high confidence in fiber direction and also when the direction is possible. Several schemes such as Cross Entropy, Sign Change Ratio, Hard Decision-Aided, Sign Decision Ratio and Improved Hard Decision-Aided have been proposed to control the termination in CDMA systems (Khaled Fazel and Stefan Kaiser, 2003). Hence from the above content, it is understood that the aim is to reduce the possibility of seeing false positives in the results. It is important to select a correct stopping criterion for any given algorithm. In general, a CDMA mobile communication system refers to a mobile communication that performs radio communication by adopting the CDMA technique. Such type of mobile communication performs Forward Error Correction in order to correct the errors that are caused by the noises which are generated in the transport channels. For the forward error correction, the CDMA communication system generally uses a turbo code which is also called as a conventional code. Turbo code in particular has been adopted as an error correction code in both the synchronous as well as asynchronous systems (William H. Tranter, 2004). Hence from the above content, it is understood that selecting a perfect stopping criteria for any given algorithm is very important.

Computational complexity

CDMA is nothing but code division multiple access which involves in multiple access interface and the near-far effect which cause the limitation of capacity. Number of users exponentially grows from other side of the computational complexity of the optimum multiuser detector. With less complexity and reasonable performance there is more interest in suboptimal multiuser detectors. For multiuser detector of DS/CDMA signals uses classic and modified new genetic algorithm. In high signal to noise ratios while the proposed method has higher performance than the classic one, it is shown that the classic genetic algorithm (GA) reaches an error floor. Indoor wireless, cellular mobile and personal communication systems are considered as the third generation in DS/CDMA. Frequency reuse, soft handoff, increased capacity, and multipath combating are best features by CDMA. Simultaneously transmit information over a common channel using pre-assigned signature codes by several users in a CDMA system. Bank of filters to matched for spreading codes and deciding on the sign of the outputs in the conventional single user detector. These spreading codes should be orthogonal in CDMA. In filter demodulation codes are non-orthogonal, it creates interface in conventional in practice (Bernd Reusch 2001). In CDMA number of users exponentially grows from other side of the computational complexity of the optimum multiuser detector. There are various different types of cdma technology signals Time Hopped Direct Sequence Code Division Multiple Access (TH/DS-CDMA) is also known as pulsed DS-CDMA, Bust CDMA, Pulsed Direct Sequence Spread Spectrum or Burst Pseudolite signal. Multiple input multiple output and code division multiple access systems are the interference mitigation. There are huge data rates in emerging Internet and multimedia services in order to meet increasing demands in wireless communication systems. For available spectrum in CDMA, efficient use of multiuser detection and space diversity techniques are the main principles. Low computational complexities are novel design of interference cancellation receivers which are adaptive and iterative. Non-adaptive iterative receiver is compared with adaptive least mean square algorithm. In the time and frequency domain adaptive detection scheme employs an adaptive LMS algorithm operating ((John Wiley 2002). While presenting a satisfactory system performance by performing a convolution in the frequency domain, reduce the computational complexity of the system. Compared to time domain approach it reduces the complexity.

Chapter 4: Proposed Algorithm

PDA Algorithm

Probabilistic Data Association Algorithm has been proposed to address the detection in a variety of communication channels. This algorithm gives the concepts of both to give a more reliable inference to CDMA. Probabilistic Data Association filter has a good performance in dense targets and also in clutter environment. Data association problem between target and measurements, many methods are proposed among which Joint PDA algorithm has a very good tracking performance especially in case of some measurements. The general idea of probabilistic data association is that instead of using one measurement among multiple received ones and discarding the others preferring all of the validated measurements with different probabilities is given preference. A new algorithm on the idea of PDA is proposed for the multiuser detection in synchronous CDMA communications. Simulation results show that the PDA detector provides near optimal performance with the overall computational cost. The PDA algorithm calculates in real time probability that each validated measurement is attributable to the target of interest. The Probabilistic Data Algorithm (PDA) and its multi target version which is joint probabilistic data association (JPDA) Algorithm are the association techniques that are based on appropriately modified PDAF and JPDAF.

This algorithm was first proposed by Bar-Shalom and Tse who considered the case of all target trajectories that are sufficiently far from one another to be considered (Anthony K. Hyder, E. Shahbazian and Edward Waltz, 2002).

Probabilistic Data Association is the algorithm that applies the inter user interference (IUI) cancellation in CDMA system. Here the probability means the each user signal should be updated regularly. This PDA detector may provide the better performance and responsibility that merges the low Signal to Noise Ratio (SNR) and higher Bit Error Rate (BER). Mainly use of this PDA algorithm is to detect and cancel the error in the signal. This is available at receiver of the CDMA system for identifying the multi users and signals (Keijji Tachikawa, 2002). These signals are modulated to analyze the efficiency of the PDA algorithm by its techniques. Some cases signals are demodulated while identifying the users. This detection and identifying users and cancelling the error must since it undergoes the additive white Gaussian noise or multipath fading channels. This PDA avoids the local minima and probability of error.

Where the variable bi represents the ith element of the b vector, ei is the column vector whose ith componen