Dissertations / Theses on the topic 'Iterative detection'

[Truncated abstract] The transmission of digital information over a wireless communication channel gives rise to a number of issues which can detract from the system performance. Propagation effects such as multipath fading and intersymbol interference (ISI) can result in significant performance degradation. Recent developments in the field of iterative detection have led to a number of powerful strategies that can be effective in mitigating the detrimental effects of wireless channels. In this thesis, iterative detection is considered for use in two distinct areas of wireless communications. The first considers the iterative decoding of concatenated block codes over slow flat fading wireless channels, while the second considers the problem of detection for a coded communications system transmitting over highly-dispersive frequency-selective wireless channels. The iterative decoding of concatenated codes over slow flat fading channels with coherent signalling requires knowledge of the fading amplitudes, known as the channel state information (CSI). The CSI is combined with statistical knowledge of the channel to form channel reliability metrics for use in the iterative decoding algorithm. When the CSI is unknown to the receiver, the existing literature suggests the use of simple approximations to the channel reliability metric. However, these works generally consider low rate concatenated codes with strong error correcting capabilities. In some situations, the error correcting capability of the channel code must be traded for other requirements, such as higher spectral efficiency, lower end-to-end latency and lower hardware cost. ... In particular, when the error correcting capabilities of the concatenated code is weak, the conventional metrics are observed to fail, whereas the proposed metrics are shown to perform well regardless of the error correcting capabilities of the code. The effects of ISI caused by a frequency-selective wireless channel environment can also be mitigated using iterative detection. When the channel can be viewed as a finite impulse response (FIR) filter, the state-of-the-art iterative receiver is the maximum a posteriori probability (MAP) based turbo equaliser. However, the complexity of this receiver's MAP equaliser increases exponentially with the length of the FIR channel. Consequently, this scheme is restricted for use in systems where the channel length is relatively short. In this thesis, the use of a channel shortening prefilter in conjunction with the MAP-based turbo equaliser is considered in order to allow its use with arbitrarily long channels. The prefilter shortens the effective channel, thereby reducing the number of equaliser states. A consequence of channel shortening is that residual ISI appears at the input to the turbo equaliser and the noise becomes coloured. In order to account for the ensuing performance loss, two simple enhancements to the scheme are proposed. The first is a feedback path which is used to cancel residual ISI, based on decisions from past iterations. The second is the use of a carefully selected value for the variance of the noise assumed by the MAP-based turbo equaliser. Simulations are performed over a number of highly dispersive channels and it is shown that the proposed enhancements result in considerable performance improvements. Moreover, these performance benefits are achieved with very little additional complexity with respect to the unmodified channel shortened turbo equaliser.

In this thesis, Multiple-Input Multiple-Output (MIMO) techniques designed for transmission over narrowband Rayleigh fading channels are investigated. Specifically, in order to provide a diversity gain while eliminating the complexity of MIMO channel estimation, a Differential Space-Time Spreading (DSTS) scheme is designed that employs non-coherent detection. Additionally, in order to maximise the coding advantage of DSTS, it is combined with Sphere Packing (SP) modulation. The related capacity analysis shows that the DSTS-SP scheme exhibits a higher capacity than its counterpart dispensing with SP. Furthermore, in order to attain additional performance gains, the DSTS system invokes iterative detection, where the outer code is constituted by a Recursive Systematic Convolutional (RSC) code, while the inner code is a SP demapper in one of the prototype systems investigated, while the other scheme employs a Unity Rate Code (URC) as its inner code in order to eliminate the error floor exhibited by the system dispensing with URC. EXIT charts are used to analyse the convergence behaviour of the iteratively detected schemes and a novel technique is proposed for computing the maximum achievable rate of the system based on EXIT charts. Explicitly, the four-antenna-aided DSTSSP system employing no URC precoding attains a coding gain of 12 dB at a BER of 10−5 and performs within 1.82 dB from the maximum achievable rate limit. By contrast, the URC aided precoded system operates within 0.92 dB from the same limit. On the other hand, in order to maximise the DSTS system’s throughput, an adaptive DSTSSP scheme is proposed that exploits the advantages of differential encoding, iterative decoding as well as SP modulation. The achievable integrity and bit rate enhancements of the system are determined by the following factors: the specific MIMO configuration used for transmitting data from the four antennas, the spreading factor used and the RSC encoder’s code rate. Additionally, multi-functional MIMO techniques are designed to provide diversity gains, multiplexing gains and beamforming gains by combining the benefits of space-time codes, VBLAST and beamforming. First, a system employing Nt=4 transmit Antenna Arrays (AA) with LAA number of elements per AA and Nr=4 receive antennas is proposed, which is referred to as a Layered Steered Space-Time Code (LSSTC). Three iteratively detected near-capacity LSSTC-SP receiver structures are proposed, which differ in the number of inner iterations employed between the inner decoder and the SP demapper as well as in the choice of the outer code, which is either an RSC code or an Irregular Convolutional Code (IrCC). The three systems are capable of operating within 0.9, 0.4 and 0.6 dB from the maximum achievable rate limit of the system. A comparison between the three iteratively-detected schemes reveals that a carefully designed two-stage iterative detection scheme is capable of operating sufficiently close to capacity at a lower complexity, when compared to a three-stage system employing a RSC or a two-stage system using an IrCC as an outer code. On the other hand, in order to allow the LSSTC scheme to employ less receive antennas than transmit antennas, while still accommodating multiple users, a Layered Steered Space-Time Spreading (LSSTS) scheme is proposed that combines the benefits of space-time spreading, V-BLAST, beamforming and generalised MC DS-CDMA. Furthermore, iteratively detected LSSTS schemes are presented and an LLR post-processing technique is proposed in order to improve the attainable performance of the iteratively detected LSSTS system. Finally, a distributed turbo coding scheme is proposed that combines the benefits of turbo coding and cooperative communication, where iterative detection is employed by exchanging extrinsic information between the decoders of different single-antenna-aided users. Specifically, the effect of the errors induced in the first phase of cooperation, where the two users exchange their data, on the performance of the uplink in studied, while considering different fading channel characteristics.

Multiuser detection (MUD) has been regarded as an effective technique for combating cochannel interference (CCI) in time-division multiple access (TDMA) systems and multiple access interference (MAI) in code-division multiple access (CDMA) systems. An optimal multiuser detector for coded multiuser systems is usually practically infeasible due to the associated complexity. An iterative receiver consisting of a soft-input soft-output (SISO) multiuser detector and a bank of SISO single user decoders can provide a system performance which approaches to that of single user system after a few iterations. In this thesis, MUD and LDPC decoding are combined to improve the multiuser receiver performance. The soft output of the LDPC decoder is fed back to the multiuser detector to improve the detection. This leads to decision variables that have a smaller MAI component. These decision variables are then returned to the decoder and the decoding process benefits from the improvement to the decision variables. The process can be repeated many times. The resulting iterative multiuser receiver is designed based on the soft parallel interference cancellation (PIC) algorithm. For the interference reconstruction, the LDPC decoder is improved to produce the log-likelihood ratios (LLR) of the information bits as well as the parity bits. A sub-optimal approach is proposed to output the LLR of the parity bits with very low complexity. Thanks to the powerful error-correction ability of the LDPC decoder, the LDPC multiuser receiver can achieve a satisfactory convergence, and substantially outperforms non-iterative receivers. Three types of SISO multiuser detectors are provided. They are: Soft Interference Cancellation (SIC) detector, SISO decorrelating detector and SISO minimum mean square error (MMSE) detector. The resulting system performance converges very quickly. The comparison of these three types of detectors is also shown in this thesis.

Turbo codes are a class of forward error correction (FEC) codes that offer energy efficiencies close to the limits predicted by information theory. The features of turbo codes include parallel code concatenation, recursive convolutional encoding, nonuniform interleaving, and an associated iterative decoding algorithm. Although the iterative decoding algorithm has been primarily used for the decoding of turbo codes, it represents a solution to a more general class of estimation problems that can be described as follows: a data set directly or indirectly drives the state transitions of two or more Markov processes; the output of one or more of the Markov processes is observed through noise; based on the observations, the original data set is estimated. This dissertation specifically describes the process of encoding and decoding turbo codes. In addition, a more general discussion of iterative decoding is presented. Then, several new applications of iterative decoding are proposed and investigated through computer simulation. The new applications solve two categories of problems: the detection of turbo codes over time-varying channels, and the distributed detection of coded multiple-access signals. Because turbo codes operate at low signal-to-noise ratios, the process of phase estimation and tracking becomes difficult to perform. Additionally, the turbo decoding algorithm requires precise estimates of the channel gain and noise variance. The first significant contribution of this dissertation is a study of several methods of channel estimation suitable specifically for turbo coded systems. The second significant contribution of this dissertation is a proposed method for jointly detecting coded multiple-access signals using observations from several locations, such as spatially separated base stations. The proposed system architecture draws from the concepts of macrodiversity combining, multiuser detection, and iterative decoding. Simulation results show that when the system is applied to the time division multiple-access cellular uplink, a significant improvement in system capacity results.
Ph. D.

Wireless multi-user communication systems that operate in a low signal to interference noise ratio (SINR) region are studied in this thesis. This thesis examines a class of wireless communication systems that employs an adaptive receiver for multi-user symbol detection that operates in a low SINR (< 5 dB) region. Since the knowledge of channel-parameter estimates is unavailable at the receiver, a pilot (training) sequence is applied in the communication system, to learn the channel state information (CSI) at the receiver. In studying the classical view of a DFE, the mean square error (MSE) behaviour follows the bit error rate (BER) performance. Certain original results are obtained using the classical adaptive DFE to achieve minimum MSE, employing the least mean square (LMS) algorithm. The results thus obtained for an uncoded adaptive receiver system are applied to a coded system, transmitting either recursive systematic code (RSC) or turbo-code through a spread-spectrum multiuser multiple-path channel, which are referred to as two-stage and three-stage systems respectively in this thesis. The following claims are made based on the findings of this thesis: 1. It is known that a receiver implementing DFE can mitigate symbol-interference completely at high SINR. An adaptive LMS DFE realizes this by adapting the forward and backward filter coefficients with respective step-size constants. The classical approach to realizing interference mitigation was to set the forward and backward adaptation constants as the same. While this approach has provided interference mitigation at high SINR, it has been shown in this thesis that such an approach does not yield complete interference mitigation, even at high SINR. Instead, using different step-size constants at the backward and forward step-size constants provides the required optimality. 2. A decision feedback detector (DFD) mitigates the effects of interference on the information symbols that are transmitted through this communication channel. This thesis shows that an adaptive (LMS) DFD, using unequal compared to equal step-size constants to update the forward and backward filter coefficients, has a steady-state MSE improvement for an uncoded frequency selective communication channel. This thesis shows that, when the knowledge of CSI is not assumed to be known at a wireless receiver, a three-stage receiver has a BER performance improvement and operates at a lower SINR, without any additional computational complexity compared to a two-stage receiver.

Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2006.
Includes bibliographical references (leaves 81-85).
Orthogonal Frequency Division Multiple Access (OFDMA) systems have built-in mechanisms to mitigate the effects of the wireless multipath channel but are limited in system capacity to available bandwidth. This shortcoming can be worked around through the process of "overloading," where users are additionally multiplexed in the spatial domain to each frequency resource. To efficiently resolve non-orthogonally multiplexed users within the system, sophisticated multiple antenna receivers with multiuser detection methods are necessary. The focus of this thesis will be the formulation of an iterative multiple antenna receiver framework for overloaded uplink OFDMA systems. Specifically, we formulate optimal MAP and reduced complexity MMSE symbol detection algorithms for the multiuser detection and single user decoding turbo loop. We verify the performance of each algorithm through Monte Carlo simulation with randomly generated multipath MIMO channels. From the results we determine the tradeoffs of algorithm complexity with performance and the effect of channel correlation on the supportable user load. Our MMSE algorithm with soft interference cancellation is observed to closely approach single user performance in low to moderately correlated MIMO channels after turbo loop iteration. Additionally, we observe that increasing the number of antennas relative to the number of overloaded users can mitigate the effects of moderate correlation to provide acceptable error performance.
by Eugene Baik.
M.Eng.

Inspired by multiuser detection (MUD) and the ‘Turbo principle’, this thesis deals with iterative interference cancellation (IIC) in overloaded multiuser multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems. Linear detection schemes, such as zero forcing (ZF) and minimum mean square error (MMSE) cannot be used for the overloaded system because of the rank deficiency of channel matrix, while the optimal approach, the maximum likelihood (ML) detection has high computational complexity. In this thesis, an iterative interference cancellation (IIC) multiuser detection scheme with matched filter and convolutional codes is considered. The main idea of this combination is a low complexity receiver. Parallel interference cancellation (PIC) is employed to improve the multiuser receiver performance for overloaded systems. A log-likelihood ratio (LLR) converter is proposed to further improve the reliability of the soft value converted from the output of the matched filter. Simulation results show that the bit error rate (BER) performance of this method is close to the optimal approach for a two user system. However, for the four user or more user system, it has an error floor of the BER performance. For this case, a channel selection scheme is proposed to distinguish whether the channel is good or bad by using the mutual information based on the extrinsic information transfer (EXIT) chart. The mutual information can be predicted in a look-up table which greatly reduces the complexity. For those ‘bad’ channels identified by the channel selection, we introduce two adaptive transmission methods to deal with such channels: one uses a lower code rate, and the other is multiple transmissions. The use of an IIC receiver with the interleave-division multiple access (IDMA) to further improve the BER performance without any channel selection is also investigated. It has been shown that this approach can remove the error floor. Finally, the influence of channel accuracy on the IIC is investigated. Pilot-based Wiener filter channel estimation is used to test and verify how much the IIC is influenced by the channel accuracy.

Abstract Iterative receiver techniques, multiple-input – multiple-output (MIMO) processing, and orthogonal frequency division multiplexing (OFDM) are amongst the key physical layer technologies when aiming at higher spectral efficiency for a wireless communication system. Special focus is put on iterative detection, decoding, and channel estimation for a MIMO-OFDM system. After designing separately efficient algorithms for the detection, channel decoding, and channel estimation, the objective is to optimize them to work together through optimizing the activation schedules for soft-in soft-out (SfISfO) components. A list parallel interference cancellation (PIC) detector is derived to approximate an a posteriori probability (APP) algorithm with reduced complexity and minimal loss of performance. It is shown that the list PIC detector with good initialization outperforms the K-best list sphere detector (LSD) in the case of small list sizes, whereas the complexities of the algorithms are of the same order. The convergence of the iterative detection and decoding is improved by using a priori information to also recalculate the candidate list, aside from the log-likelihood ratios (LLRs) of the coded bits. Unlike in pilot based channel estimation, the least-squares (LS) channel estimator based on symbol decisions requires a matrix inversion in MIMO-OFDM. The frequency domain (FD) space-alternating generalized expectation-maximization (SAGE) channel estimator calculates the LS estimate iteratively, avoiding the matrix inversion with constant envelope modulation. The performance and computational complexity of the FD-SAGE channel estimator are compared to those of pilot based LS channel estimation with minimum mean square error (MMSE) post-processing exploiting the time correlation of the channel. A time domain (TD) SAGE channel estimator is derived to avoid the matrix inversion in channel estimation based on symbol decisions for MIMO-OFDM systems also with non-constant envelope modulation. An obvious problem, with more than two blocks in an iterative receiver, is to find the optimal activation schedule of the different blocks. It is proposed to use extrinsic information transfer (EXIT) charts to characterize the behavior of the receiver blocks and to find out the optimal activation schedule for them. A semi-analytical expression of the EXIT function is derived for the LS channel estimator. An algorithm is proposed to generate the EXIT function of the APP algorithm as a function of the channel estimate’s mutual information (MI). Surface fitting is used to get closed form expressions for the EXIT functions of the APP algorithm and the channel decoder. Trellis search algorithms are shown to find the convergence with the lowest possible complexity using the EXIT functions. With the proposed concept, the activation scheduling can be adapted to prevailing channel circumstances and unnecessary iterations will be avoided.

This thesis proposes an idea to selectively detect the code bits in an iterative soft interference cancellation multiuser receiver. It is of a great interest to reduce the complexity of the multiuser detectors in order to achieve faster multiuser communication systems. Although the suboptimum detector has much less complexity than the optimum, the detections are made on each code bit of all users through-out every iteration. Selective detection greatly reduces the amount of calculation by re-detecting only the unreliably detected code bits from the second iteration. Simulation results show that the number of detections is significantly reduced, while the performance is maintained. Necessary background information to understand the working principles of the iterative soft cancellation receiver is presented as well. Selective detection may also be used in any other receiver structures with iterative procedures to provide much less complexity. Hence, it is able to handle much more complicated receiver structures, or implement the system to a mobile device where the computational ability is much less than at the base station.

Wireless networks comprising multiple relays are very common and it is important that all users are able to exchange messages via relays in the shortest possible time. A promising technique to achieve this is physical layer network coding (PNC), where the time taken to exchange messages between users is achieved by exploiting the interference at the relay due to the multiple incoming signals from the users. At the relay, the interference is demapped to a binary sequence representing the exclusive-OR of both users' messages. The time to exchange messages is reduced because the relay broadcasts the network coded message to both users, who can then acquire the desired message by applying the exclusive-OR of their original message with the network coded message. However, although PNC can increase throughput it is at the expense of performance degradation due to errors resulting from the demapping of the interference to bits. A number of papers in the literature have investigated PNC with an iterative channel coding scheme in order to improve performance. However, in this thesis the performance of PNC is investigated for end-to-end (E2E) the three most common iterative coding schemes: turbo codes, low-density parity-check (LDPC) codes and trellis bit-interleaved coded modulation with iterative decoding (BICM-ID). It is well known that in most scenarios turbo and LDPC codes perform similarly and can achieve near-Shannon limit performance, whereas BICM-ID does not perform quite as well but has a lower complexity. However, the results in this thesis show that on a two-way relay channel (TWRC) employing PNC, LDPC codes do not perform well and BICM-ID actually outperforms them while also performing comparably with turbo codes. Also presented in this thesis is an extrinsic information transfer (ExIT) chart analysis of the iterative decoders for each coding scheme, which is used to explain this surprising result. Another problem arising from the use of PNC is the transfer of reliable information from the received signal at the relay to the destination nodes. The demapping of the interference to binary bits means that reliability information about the received signal is lost and this results in a significant degradation in performance when applying soft-decision decoding at the destination nodes. This thesis proposes the use of traditional angle modulation (frequency modulation (FM) and phase modulation (PM)) when broadcasting from the relay, where the real and imaginary parts of the complex received symbols at the relay modulate the frequency or phase of a carrier signal, while maintaining a constant envelope. This is important since the complex received values at the relay are more likely to be centred around zero and it undesirable to transmit long sequences of low values due to potential synchronisation problems at the destination nodes. Furthermore, the complex received values, obtained after angle demodulation, are used to derive more reliable log-likelihood ratios (LLRs) of the received symbols at the destination nodes and consequently improve the performance of the iterative decoders for each coding scheme compared with conventionally coded PNC. This thesis makes several important contributions: investigating the performance of different iterative channel coding schemes combined with PNC, presenting an analysis of the behaviour of different iterative decoding algorithms when PNC is employed using ExIT charts, and proposing the use of angle modulation at the relay to transfer reliable information to the destination nodes to improve the performance of the iterative decoding algorithms. The results from this thesis will also be useful for future research projects in the areas of PNC that are currently being addressed, such as synchronisation techniques and receiver design.

ITC/USA 2005 Conference Proceedings / The Forty-First Annual International Telemetering Conference and Technical Exhibition / October 24-27, 2005 / Riviera Hotel & Convention Center, Las Vegas, Nevada
We investigate the performance of Feher-patented quadrature phase-shift keying (FQPSK) and shaped-offset QPSK (SOQPSK) when serially concatenated with an outer code. We show that the receiver complexity for FQPSK and SOQPSK can be greatly reduced by viewing them as continuous phase modulation (CPM) waveforms. We use the pulse amplitude modulation (PAM) representation of CPM, which allows near-optimum detection of both modulations using a simple 4-state trellis. We compare the performance of the PAM-based approximation with another common approximation known as frequency/phase pulse truncation (PT).We use both of these reduced-complexity designs in serially concatenated coding schemes with iterative detection. In the end, we show that the PAM approximation has a slight performance advantage over PT, but both approximations achieve large coding gains in the proposed serially concatenated systems.

Multiple input multiple output (MIMO) techniques have been widely employed by different wireless systems with many advantages. By using multiple antennas, the system is able to transmit multiple data streams simultaneously and within the same frequency band. The methods known as spatial multiplexing (SM) and spatial diversity (SD) improves the high spectral efficiency and link reliability of wireless communication systems without requiring additional transmitting power. By introducing channel coding in the transmission procedure, the information redundancy is introduced to further improve the reliability of SM links and the quality of service for the next generation communication systems. However, the throughput performance of these systems is limited by interference. A number of different interference suppression techniques have been reported in the literature. Theses techniques can be generally categorised into two aspects: the preprocessing techniques at the transmitter side and the decoding techniques at the receiver side. Generally speaking, in the ideal case, the preprocessing techniques orthogonalize the interfering channels, and therefore, the receiver experiences interference free transmission. However, a feedback channel is required to provide the channel information. On the other hand, in this thesis we are interested in the decoding part which uses various techniques to improve the signal-to-interference-and-noise ratio (SINR) of the desired symbols. To achieve this goal, a number of low-complexity iterative detection algorithms have been investigated. In the context of the thesis, the mainly focus is on the interference cancellation techniques. Firstly, we investigate the traditional successive interference cancellation (SIC) algorithm. SIC has the ability to separate the spatially multiplexed signals on a MIMO channel. However, the low detection diversity order as well as the error propagation effect restrict the bit error performance of such detectors. We propose a multiple feedback SIC (MF-SIC) method to enhance the performance of conventional SIC detection by introducing feedback candidates and reliability checking. This algorithm is able to provide significant performance gains with little additional complexity without the protection from channel codes. The MF-SIC algorithm is then incorporated into an iterative detection and decoding (IDD) scheme to process soft information. Secondly, in the case that the MIMO channel is time-varying, the conventional detection algorithms generally bring about expensive complexity in the time domain. In order to address this problem, a decision feedback algorithm is introduced and adaptive algorithms are derived to update the forward and backward filters to perform the detection in each time instant. A constellation based estimation refinement scheme is also introduced in the system and the performance is significantly improved. The proposed decision feedback algorithm is incorporated into an IDD scheme that performs iterative (turbo) interference cancellation. At last, the inter-cell interference is considered in a multi-cell, high frequency reuse scenario. The distributed iterative detection (DID) algorithms are investigated. A large amount of information need to be transmitted via a wired backhaul network where optimal distributed detection exchange all the soft estimates among adjacent base stations (BSs). To address this problem we consider a reduced message passing (RMP) technique in which each BS generates a detection list with the probabilities for the desired symbol that are sorted according to the calculated probability density. RMP introduces low backhaul overhead compared with the hard bit exchange and outperforms the previously reported hard/soft information exchange algorithms.

Multiple-input multiple-output (MIMO) wireless technology is an emerging cost- effective approach to offer multiple-fold capacity improvement relative to the conven- tional single-antenna systems. To achieve the capacities of MIMO channels, MIMO bit-interleaved-coded-modulation (BICM) systems with iterative detection and decod- ing (IDD) are studied in this thesis. The research for this dissertation is conducted based on the iterative receivers with convolutional codes and turbo codes. A variety of MIMO detectors, such as a maximum a posteriori probability (MAP) detector, a list sphere detector (LSD) and a parallel interference canceller (PIC) together with a decision statistic combiner (DSC), are studied. The performance of these iterative receivers is investigated via bounding techniques or Monte-Carlos simulations. Moreover, the computational complexities of the components are quantified and compared. The convergence behaviors of the iterative receivers are analyzed via variance trans- fer (VTR) functions and variance exchange graphs (VEGs). The analysis of conver- gence behavior facilitates the finding of components with good matching. For a fast fading channel, we show that the "waterfall region" of an iterative receiver can be predicted by VEG. For a slow fading channel, it is shown that the performance of an iterative receiver is essentially limited by the early interception ratio (ECR) which is obtained via simulations. After the transfer properties of the detectors are unveiled, a detection switching (DSW) methodology is proposed and the switching criterion based on cross entropy (CE) is derived. By employing DSW, the performance of an iterative receiver with a list sphere detector (LSD) of a small list size is considerably improved. It is shown that the iterative receiver achieves a performance very close to that with a maximum a posteriori probability (MAP) detector but with a significantly reduced complexity. For an iterative receiver with more than two components, various iteration sched- ules are explored. The schedules are applied in an iterative receiver with PIC-DSC. It is shown that the iterative receiver with a periodic scheduling outperforms that with the conventional scheduling at the same level of complexity.

Multiple antenna systems are an appealing candidate for emerging fourth-generation wireless networks due to its potential to exploit space diversity for increasing conveyed throughput without wasting bandwidth and power resources. Particularly, layered space-time architecture (LST) proposed by Foschini, is a technique to achieve a significant fraction of the theoretical capacity with a reasonable implementation complexity. There has been a great deal of challenges in the detection of space-time signal; especially to design a low-complexity detector, which can efficiently remove multi-layer interference and approach the interference free bound. The application of iterative principle to joint detection and decoding has been a promising approach. It has been shown that, the iterative receiver with parallel interference canceller (PIC) has a low linear complexity and near interference free performance. Furthermore, it is widely accepted that the performance of digital communication systems can be considerably improved once the channel state information (CSI) is used to optimize the transmit signal. In this thesis, the problem of the design of a power allocation strategy in LST architecture to simultaneously optimize coding, diversity and weighting gains is addressed. A more practical scenario is also considered by assuming imperfect CSI at the receiver. The effect of channel estimation errors in LST architecture with an iterative PIC receiver is investigated. It is shown that imperfect channel estimation at an LST receiver results in erroneous decision statistics at the very first iteration and this error propagates to the subsequent iterations, which ultimately leads to severe degradation of the overall performance. We design a transmit power allocation policy to take into account the imperfection in the channel estimation process. The transmit power of various layers is optimized through minimization of the average bit error rate (BER) of the LST architecture with a low complexity iterative PIC detector. At the receiver, the PIC detector performs both interference regeneration and cancellation simultaneously for all layers. A convolutional code is used as the constituent code. The iterative decoding principle is applied to pass the a posteriori probability estimates between the detector and decoders. The decoder is based on the maximum a posteriori (MAP) algorithms. A closed-form optimal solution for power allocation in terms of the minimum BER is obtained. In order to validate the effectiveness of the proposed schemes, substantial simulation results are provided.

Master of Engineering (Research)
Multiple antenna systems are an appealing candidate for emerging fourth-generation wireless networks due to its potential to exploit space diversity for increasing conveyed throughput without wasting bandwidth and power resources. Particularly, layered space-time architecture (LST) proposed by Foschini, is a technique to achieve a significant fraction of the theoretical capacity with a reasonable implementation complexity. There has been a great deal of challenges in the detection of space-time signal; especially to design a low-complexity detector, which can efficiently remove multi-layer interference and approach the interference free bound. The application of iterative principle to joint detection and decoding has been a promising approach. It has been shown that, the iterative receiver with parallel interference canceller (PIC) has a low linear complexity and near interference free performance. Furthermore, it is widely accepted that the performance of digital communication systems can be considerably improved once the channel state information (CSI) is used to optimize the transmit signal. In this thesis, the problem of the design of a power allocation strategy in LST architecture to simultaneously optimize coding, diversity and weighting gains is addressed. A more practical scenario is also considered by assuming imperfect CSI at the receiver. The effect of channel estimation errors in LST architecture with an iterative PIC receiver is investigated. It is shown that imperfect channel estimation at an LST receiver results in erroneous decision statistics at the very first iteration and this error propagates to the subsequent iterations, which ultimately leads to severe degradation of the overall performance. We design a transmit power allocation policy to take into account the imperfection in the channel estimation process. The transmit power of various layers is optimized through minimization of the average bit error rate (BER) of the LST architecture with a low complexity iterative PIC detector. At the receiver, the PIC detector performs both interference regeneration and cancellation simultaneously for all layers. A convolutional code is used as the constituent code. The iterative decoding principle is applied to pass the a posteriori probability estimates between the detector and decoders. The decoder is based on the maximum a posteriori (MAP) algorithms. A closed-form optimal solution for power allocation in terms of the minimum BER is obtained. In order to validate the effectiveness of the proposed schemes, substantial simulation results are provided.

Soft errors caused by transient bit flips have the potential to significantly impactan applicalion's behavior. This has motivated the design of an array of techniques to detect, isolate, and correct soft errors using microarchitectural, architectural, compilation­based, or application-level techniques to minimize their impact on the executing application. The first step toward the design of good error detection/correction techniques involves an understanding of an application's vulnerability to soft errors. This work focuses on silent data e orruption's effects on iterative solvers and efforts to mitigate those effects. In this thesis, we first present the first comprehensive characterizalion of !he impact of soft errors on !he convergen ce characteris tics of six iterative methods using application-level fault injection. We analyze the impact of soft errors In terms of the type of error (single-vs multi-bit), the distribution and location of bits affected, the data structure and statement impacted, and varialion with time. We create a public access database with more than 1.5 million fault injection results. We then analyze the performance of soft error detection mechanisms and present the comparalive results. Molivated by our observations, we evaluate a machine-learning based detector that takes as features that are the runtime features observed by the individual detectors to arrive al their conclusions. Our evalualion demonstrates improved results over individual detectors. We then propase amachine learning based method to predict a program's error behavior to make fault injection studies more efficient. We demonstrate this method on asse ssing the performance of soft error detectors. We show that our method maintains 84% accuracy on average with up to 53% less cost. We also show, once a model is trained further fault injection tests would cost 10% of the expected full fault injection runs.
“Soft errors” causados por cambios de estado transitorios en bits, tienen el potencial de impactar significativamente el comportamiento de una aplicación. Esto, ha motivado el diseño de una variedad de técnicas para detectar, aislar y corregir soft errors aplicadas a micro-arquitecturas, arquitecturas, tiempo de compilación y a nivel de aplicación para minimizar su impacto en la ejecución de una aplicación. El primer paso para diseñar una buna técnica de detección/corrección de errores, implica el conocimiento de las vulnerabilidades de la aplicación ante posibles soft errors. Este trabajo se centra en los efectos de la corrupción silenciosa de datos en soluciones iterativas, así como en los esfuerzos para mitigar esos efectos. En esta tesis, primeramente, presentamos la primera caracterización extensiva del impacto de soft errors sobre las características convergentes de seis métodos iterativos usando inyección de fallos a nivel de aplicación. Analizamos el impacto de los soft errors en términos del tipo de error (único vs múltiples-bits), de la distribución y posición de los bits afectados, las estructuras de datos, instrucciones afectadas y de las variaciones en el tiempo. Creamos una base de datos pública con más de 1.5 millones de resultados de inyección de fallos. Después, analizamos el desempeño de mecanismos de detección de soft errors actuales y presentamos los resultados de su comparación. Motivados por las observaciones de los resultados presentados, evaluamos un detector de soft errors basado en técnicas de machine learning que toma como entrada las características observadas en el tiempo de ejecución individual de los detectores anteriores al llegar a su conclusión. La evaluación de los resultados obtenidos muestra una mejora por sobre los detectores individualmente. Basados en estos resultados propusimos un método basado en machine learning para predecir el comportamiento de los errores en un programa con el fin de hacer el estudio de inyección de errores mas eficiente. Presentamos este método para evaluar el rendimiento de los detectores de soft errors. Demostramos que nuestro método mantiene una precisión del 84% en promedio con hasta un 53% de mejora en el tiempo de ejecución. También mostramos que una vez que un modelo ha sido entrenado, las pruebas de inyección de errores siguientes costarían 10% del tiempo esperado de ejecución.

Timing recovery is the processing of synchronizing the sampler with the received analog signal. Sampling at the wrong times can have a devastating impact on performance. Conventional timing recovery techniques are based on a decision-directed phase-locked loop (PLL). They are adequate only when the operating signal-to-noise ratio (SNR) is sufficiently high, but recent advances in error-control coding have made it possible to communicate reliably at very low SNR, where conventional techniques fail. This thesis develops new techniques for timing recovery that are capable of working at low SNR. We propose a new timing recovery scheme based on per-survivor processing (PSP), which jointly performs timing recovery and equalization, by embedding a separate PLL into each survivor of a Viterbi algorithm. The proposed scheme is shown to perform better than conventional scheme, especially when the SNR is low and the timing error is large. An important advantage of this technique is its amenability to real-time implementation. We also propose a new iterative timing recovery scheme that exploits the presence of the error-control code; in doing so, it can perform even better than the PSP scheme described above, but at the expense of increased complexity and the requirement of batch processing. This scheme is realized by embedding the timing recovery process into a trellis-based soft-output equalizer using PSP. Then, this module iteratively exchanges soft information with the error-control decoder, as in conventional turbo equalization. The resulting system jointly performs the functions of timing recovery, equalization, and decoding. The proposed iterative timing recovery scheme is shown to perform better than previously reported iterative timing recovery schemes, especially when the timing error is severe. Finally, performance analysis of iterative timing recovery schemes is difficult because of their high complexity. We propose to use the extrinsic information transfer (EXIT) chart as a tool to predict and compare their performances, considering that the bit-error rate computation takes a significant amount of simulation time. Experimental results indicate that the system performance predicted by the EXIT chart coincides with that obtained by simulating data transmission over a complete iterative receiver, especially when the coded block length is large.

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2002.
Includes bibliographical references (leaves 67-69). Also available in electronic version. Access restricted to campus users.

Con la crescita vertiginosa delle telecomunicazioni, è divenuto sempre più importante un uso efficiente dello spettro a disposizione: l’ottimizzazione del piano di riutilizzo delle frequenze è quindi fondamentale per le attuali reti satellitari, in quanto la banda è ormai una risorsa estremamente preziosa. Lo scopo del presente lavoro di Dottorato è quello di sviluppare una nuova metodologia per migliorare l’utilizzo della banda disponibile nel Return Link (RL) di un satellite multibeam. Il metodo proposto permette l’uso della stessa frequenza per beam adiacenti (mentre solitamente beam adiacenti utilizzano bande differenti per non interferire tra loro alle antenne del satellite) e mitiga l’Interferenza Co-Canale (CCI) risultante adottando un appropriato algoritmo Multiuser Joint Detection (MUD) alla gateway, ovvero, per sopprimere l’interferenza CCI, tutti i canali sono demodulati assieme. Lo scenario di riferimento è quello di un sistema satellitare in cui la gateway supporta una comunicazione bidirezionale verso terminali utente equipaggiati con antenne di apertura relativamente piccola e che trasmettono a bassa potenza. Il RL implementa uno standard DVB-RCS ottimizzato, operante con utenti TDMA a basso bit-rate (fino a qualche Mbit/s), essendo questo lo schema maggiormente in uso per lo scenario in esame. La tecnica MUD selezionata si basa sull’adattamento di strategie iterative (o turbo) di cancellazione d’interferenza già presenti in letteratura, ma analizzate principalmente per sistemi CDMA. Sono state effettuate simulazioni complete del demodulatore in scenari generali, rappresentativi di casi reali, le quali hanno dimostrato una ragionevole complessità dell’algoritmo di mitigazione proposto ed un’efficace riduzione dell’interferenza CCI. Linee guida dettagliate sono riportate alla fine della tesi.
With the explosive growth of telecommunications, the efficient use of available spectrum is becoming increasingly important: tightened frequency reuse planning is a pressing need in satellite networks, as the frequency spectrum has become an extremely precious resource. The purpose of this work is to develop a novel methodology for improving the bandwidth utilization on the Reverse Link (RL) of a multibeam satellite. The proposed method permits the frequency reuse for adjacent beams (usually adjacent beams employ different modulation bands so as not to interfere at the satellite antennas) and mitigates the resulting Co-Channel Interference (CCI) using a suitable Multiuser Joint Detection (MUD) algorithm at the gateway side, that is all the channels are demodulated together in order to suppress the CCI interference. The reference scenario will be a satellite system where the gateway station supports a two-way communication with several interactive user terminals equipped with a relatively small aperture antenna and transmitting low power. The RL implements an enhanced DVB-RCS standard operating with low-rate TDMA users (data rates up to few Mbit/s), as this scheme is currently the most popular for the considered scenario. The selected MUD technique is based on the adaption of iterative (turbo) interference cancellation strategies presented in literature for CDMA systems. Complete demodulator simulations were performed in general scenarios representative of real cases. The simulations have demonstrated the proposed interference mitigation algorithm reasonable complexity and an effective CCI reduction. Detailed guidelines have been finally reported in the thesis.

Thesis (M.S. in electrical engineering)--Washington State University, August 2010.
Title from PDF title page (viewed on July 28, 2010). "School of Electrical Engineering and Computer Science." Includes bibliographical references (p. 51).

The objective of this research work is to develop reliable and power-efficient synchronization algorithms for continuous phase modulation (CPM). CPM is a bandwidth and power efficient signaling scheme suitable for wireless and mobile communications. Binary CPM schemes have been widely used in many commercial and military systems. CPM with multilevel symbol inputs, i.e., M-ary CPM, can achieve a higher data rate than binary CPM. However, the use of M-ary CPM has been limited due to receiver complexity and synchronization problems. In the last decade, serially concatenated CPM (SCCPM) has drawn more attention since this turbo-like coded scheme can achieve near Shannon-limit performance by performing iterative demodulation/decoding. Note that SCCPM typically operates at a low signal-to-noise ratio, which makes reliable and power-efficient synchronization more challenging. In this thesis, we propose a novel timing and phase recovery technique for CPM. Compared to existing maximum-likelihood estimators, the proposed data-aided synchronizer can achieve a better acquisition performance when a preamble is short or channel model errors are present. We also propose a novel adaptive soft-input soft-output (A-SISO) module for iterative detection with parameter uncertainty. In contrast to the existing A-SISO algorithms using linear prediction, the parameter estimation in the proposed structure is performed in a more general least-squares sense. Based on this scheme, a family of fixed-interval A-SISO algorithms are utilized to implement blind iterative phase synchronization for SCCPM. Moreover, the convergence characteristics of iterative phase synchronization and detection are analyzed by means of density evolution. Particularly, an oscillatory convergence behavior is observed when cycle slips occur during phase tracking. In order to reduce performance degradation due to this convergence fluctuation, design issues, including delay depth of the proposed algorithms, iteration-stopping criteria and interleaver size, are also discussed. Finally, for completeness of the study on phase synchronization, we investigate the error probability performance of noncoherently detected full-response CPM, which does not require channel (or phase) estimation.

Abstract Multiuser demodulation algorithms for centralized receivers of asynchronous direct-sequence (DS) spread-spectrum code-division multiple-access (CDMA) systems in frequency-selective fading channels are studied. Both DS-CDMA systems with short (one symbol interval) and long (several symbol intervals) spreading sequences are considered. Linear multiuser receivers process ideally the complete received data block. The approximation of ideal infinite memory-length (IIR) linear multiuser detectors by finite memory-length (FIR) detectors is studied. It is shown that the FIR detectors can be made near-far resistant under a given ratio between maximum and minimum received power of users by selecting an appropriate memory-length. Numerical examples demonstrate the fact that moderate memory-lengths of the FIR detectors are sufficient to achieve the performance of the ideal IIR detectors even under severe near-far conditions. Multiuser demodulation in relatively fast fading channels is analyzed. The optimal maximum likelihood sequence detection receiver and suboptimal receivers are considered. The parallel interference cancellation (PIC) receiver is demonstrated to achieve better performance in known channels than the decorrelating receiver, but it is observed to be more sensitive to channel coefficient estimation errors than the decorrelator. At high channel loads the PIC receiver suffers from bit error rate (BER) saturation, whereas the decorrelating receiver does not. Choice of channel estimation filters is shown to be crucial if low BER is required. Data-aided channel estimation is shown to be more robust than decision-directed channel estimation, which may suffer from BER saturation caused by hang-ups at high signal-to-noise ratios. Multiuser receivers for dynamic CDMA systems are studied. Algorithms for ideal linear detector computation are derived and their complexity is analyzed. The complexity of the linear detector computation is a cubic function of KL, where K and L are the number of users and multipath components, respectively. Iterative steepest descent, conjugate gradient, and preconditioned conjugate gradient algorithms are proposed to reduce the complexity. The computational requirements for one iteration are a quadratic function of KL. The iterative detectors are also shown to be applicable for parallel implementation. Simulation results demonstrate that a moderate number of iterations yields the performance of the corresponding ideal linear detectors. A quantitative analysis shows that the PIC receivers are significantly simpler to implement than the linear receivers and only moderately more complex than the conventional matched filter bank receiver.

This thesis fits into the Multiple-Input Multiple-Output (MIMO) communication systems. Nowadays, these schemes are the most promising technology in the field of wireless communications. The use of this technology allows to increase the rate and the quality of the transmission through the use of multiple antennas at the transmitter and receiver sides. Furthermore, the MIMO technology can also be used in a multiuser scenario, where a Base Station (BS) equipped with several antennas serves several users that share the spatial dimension causing interference. However, employing precoding algorithms the signal of the multiuser interference can be mitigated. For these reasons, the MIMO technology has become an essential key in many new generation communications standards. On the other hand, Massive MIMO technology or Large MIMO, where the BS is equipped with very large number of antennas (hundreds or thousands) serves many users in the same time-frequency resource. Nevertheless, the advantages provided by the MIMO technology entail a substantial increase in the computational cost. Therefore the design of low-complexity receivers is an important issue which is tackled throughout this thesis. To this end, one of the main contributions of this dissertation is the implementation of efficient soft-output detectors and precoding schemes. First, the problem of efficient soft detection with no iteration at the receiver has been addressed. A detailed overview of the most employed soft detectors is provided. Furthermore, the complexity and performance of these methods are evaluated and compared. Additionally, two low-complexity algorithms have been proposed. The first algorithm is based on the efficient Box Optimization Hard Detector (BOHD) algorithm and provides a low-complexity implementation achieving a suitable performance. The second algorithm tries to reduce the computational cost of the Subspace Marginalization with Interference Suppression (SUMIS) algorithm. Second, soft-input soft-output (SISO) detectors, which are included in an iterative receiver structure, have been investigated. An iterative receiver improves the performance with respect to no iteration, achieving a performance close to the channel capacity. In contrast, its computational cost becomes prohibitive. In this context, three algorithms are presented. Two of them achieve max-log performance reducing the complexity of standard SISO detectors. The last one achieves near max-log performance with low complexity. The precoding problem has been addressed in the third part of this thesis. An analysis of some of the most employed precoding techniques has been carried out. The algorithms have been compared in terms of performance and complexity. In this context, the impact of the channel matrix condition number on the performance of the precoders has been analyzed. This impact has been exploited to propose an hybrid precoding scheme that reduces the complexity of the previously proposed precoders. In addition, in Large MIMO systems, an alternative precoder scheme is proposed. In the last part of the thesis, parallel implementations of the SUMIS algorithm are presented. Several strategies for the parallelization of the algorithm are proposed and evaluated on two different platforms: multicore central processing unit (CPU) and graphics processing unit (GPU). The parallel implementations achieve a significant speedup compared to the CPU version. Therefore, these implementations allow to simulate a scalable quasi optimal soft detector in a Large MIMO system much faster than by conventional simu
La presente tesis se enmarca dentro de los sistemas de comunicaciones de múltiples antenas o sistemas MIMO. Hoy en día, estos sistemas presentan una de las tecnologías más prometedoras dentro de los sistemas comunicaciones inalámbricas. A través del uso de múltiples antenas en ambos lados, transmisor y receptor, la tasa de transmisión y la calidad de la misma es aumentada. Por otro lado, la tecnología MIMO puede ser utilizada en un escenario multiusuario, donde una estación base (BS) la cual está equipada con varias antenas, sirve a varios usuarios al mismo tiempo, estos usuarios comparten dimensión espacial causando interferencias multiusuario. Por todas estas razones, la tecnología MIMO ha sido adoptada en muchos de los estándares de comunicaciones de nueva generación. Por otro lado, la tecnología MIMO Masivo, en la cual la estación base está equipada con un gran número de antenas (cientos o miles) que sirve a muchos usuarios en el mismo recurso de tiempo-frecuencia. Sin embargo, las ventajas proporcionadas por los sistemas MIMO implican un aumento en el coste computacional requerido. Por ello, el diseño de receptores de baja complejidad es una cuestión importante en estos sistemas. Para conseguir esta finalidad, las principales contribuciones de la tesis se basan en la implementación de algoritmos de detección soft y esquemas de precodificación eficientes. En primer lugar, el problema de la detección soft eficiente en un sistema receptor sin iteración es abordado. Una descripción detallada sobre los detectores soft más empleados es presentada. Por otro lado, han sido propuestos dos algoritmos de bajo coste. El primer algoritmo está basado en el algoritmo Box Optimization Hard Detector (BOHD) y proporciona una baja complejidad de implementación logrando un buen rendimiento. El segundo de los algoritmos propuestos intenta reducir el coste computacional del conocido algoritmo Subspace Marginalization with Interference Suppression (SUMIS). En segundo lugar, han sido investidados detectores de entrada y salida soft (SISO, soft-input soft-output) los cuales son ejecutados en estructuras de recepción iterativa. El empleo de un receptor iterativo mejora el rendimiento del sistema con respecto a no realizar realimentación, pudiendo lograr la capacidad óptima. Por el contrario, el coste computacional se vuelve prohibitivo. En este contexto, tres algoritmos han sido presentados. Dos de ellos logran un rendimiento óptimo, reduciendo la complejidad de los detectores SISO óptimos que normalmente son empleados. Por el contrario, el otro algoritmo logra un rendimiento casi óptimo a baja complejidad. En la tercera parte, se ha abordado el problema de la precodificación. Se ha llevado a cabo un análisis de algunas de las técnicas de precodificación más usadas. En este contexto, se ha evaluado el impacto que el número de condición de la matriz de canal tiene en el rendimiento de los precodificadores. Además, se ha aprovechado este impacto para proponer un precodificador hibrido. Por otro lado, en MIMO Masivo, se ha propuesto un esquema precodificador. En la última parte de la tesis, la implementación paralela del algoritmo SUMIS es presentada. Varias estrategias sobre la paralelización del algoritmo han sido propuestas y evaluadas en dos plataformas diferentes: Unidad Central de Procesamiento multicore (multicore CPU) y Unidad de Procesamiento Gráfico (GPU). Las implementaciones paralelas consiguen una mejora de speedup. Estas implementaciones permiten simular para MIMO Masivo y de forma más rápida que por simulación convencional, un algo
La present tesi s'emmarca dins dels sistemes de comunicacions de múltiples antenes o sistemes MIMO. Avui dia, aquestos sistemes presenten una de les tecnologies més prometedora dins dels sistemes de comunicacions inalàmbriques. A través de l'ús de múltiples antenes en tots dos costats, transmissor y receptor, es pot augmentar la taxa de transmissió i la qualitat de la mateixa. D'altra banda, la tecnologia MIMO es pot utilitzar en un escenari multiusuari, on una estació base (BS) la qual està equipada amb diverses antenes serveix a diversos usuaris al mateix temps, aquests usuaris comparteixen dimensió espacial causant interferències multiusuari. Per totes aquestes raons, la tecnologia MIMO ha sigut adoptada en molts dels estàndars de comunicacions de nova generació. D'altra banda, la tecnologia MIMO Massiu, en la qual l'estació base està equipada amb un gran nombre d'antenes (centenars o milers) que serveix a molts usuaris en el mateix recurs de temps-freqüència. No obstant això, els avantatges proporcionats pels sistemes MIMO impliquen un augment en el cost computacional requerit. Per això, el disseny de receptors de baixa complexitat és una qüestió important en aquests sistemes. Per tal d'aconseguir esta finalitat, les principals contribucions de la tesi es basen en la implementació d'algoritmes de detecció soft i esquemes de precodificació eficients. En primer lloc, és abordat el problema de la detecció soft eficient en un sistema receptor sense interacció. Una descripció detallada dels detectors soft més emprats és presentada. D'altra banda, han sigut proposats dos algorismes de baix cost. El primer algorisme està basat en l'algorisme Box Optimization Hard Decoder (BOHD) i proporciona una baixa complexitat d'implementació aconseguint un bon resultat. El segon dels algorismes proposats intenta reduir el cost computacional del conegut algoritme Subspace Marginalization with Interference Suppression (SUMIS). En segon lloc, detectors d'entrada i eixidia soft (SISO, soft-input soft-output) els cuals són executats en estructures de recepció iterativa han sigut investigats. L'ocupació d'un receptor iteratiu millora el rendiment del sistema pel que fa a no realitzar realimentació, podent aconseguir la capacitat òptima. Per contra, el cost computacional es torna prohibitiu. En aquest context, tres algorismes han sigut presentats. Dos d'ells aconsegueixen un rendiment òptim, reduint la complexitat dels detectors SISO òptims que normalment són emprats. Per contra, l'altre algorisme aconsegueix un rendiment quasi òptim a baixa complexitat. En la tercera part, s'ha abordat el problema de la precodificació. S'ha dut a terme una anàlisi d'algunes de les tècniques de precodificació més usades, prestant especial atenció al seu rendiment i a la seua complexitat. Dins d'aquest context, l'impacte que el nombre de condició de la matriu de canal té en el rendiment dels precodificadors ha sigut avaluat. A més, aquest impacte ha sigut aprofitat per a proposar un precodificador híbrid , amb la finalitat de reduir la complexitat d'algorismes prèviament proposats. D'altra banda, en MIMO Massiu, un esquema precodificador ha sigut proposat. En l'última part, la implementació paral·lela de l'algorisme SUMIS és presentada. Diverses estratègies sobre la paral·lelizació de l'algorisme han sigut proposades i avaluades en dues plataformes diferents: multicore CPU i GPU. Les implementacions paral·leles aconsegueixen una millora de speedup quan el nombre d'àntenes o l'ordre de la constel·lació incrementen. D'aquesta manera, aquestes implementacions permeten simular per a MIMO Massiu, i de forma més ràpida que la simulació convencional.
Simarro Haro, MDLA. (2017). Effi cient algorithms for iterative detection and decoding in Multiple-Input and Multiple-Output Communication Systems [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/86186
TESIS

Recent medical studies have led cardiologists to revise theories regarding the cause of heart attacks. Rather than a gradual clogging of the arteries, eruption of a vulnerable plaque is thought to be the cause of approximately 75% of all heart attacks. As a result, traditional risk factors are no longer sufficient indicators of who is at risk for a heart attack. Therefore, this research investigates the willingness to pay (WTP) for a new, hypothetical detection (screening) and treatment method for vulnerable plaque. For this study, two survey instruments were developed that take advantage of the visual and interactive aspects of the Internet. Individuals report their perception of heart attack risk both prior to and after receiving new information on who cardiologists currently believe to be at risk for a heart attack. In addition, respondents are provided with information about the effectiveness and risks associated with screening and treatment. Using webbased surveys, which follow a contingent valuation format, an iterative bidding process is used to elicit the respondents WTP for either the screening or treatment method. Internet, on-line surveys are often prone to coverage bias; however, the survey valuing screening (a simple blood test) used a Knowledge Networks panel and resulted in a sample of 268 adults that is essentially representative of the general population. The survey valuing treatment (a more invasive heart catheterization procedure) was administered only to individuals with doctor-diagnosed heart problems, who are presumably more familiar with these types of medical decisions, and resulted in a sample of 295 adults. The mean for screening is $69 and the mean WTP for treatment that is 85% effective is $5,816. A two-part model is used to identify the factors that influence WTP, as well as the decision to receive the screening/treatment. The data suggests that these factors vary across genders. The data obtained for this study demonstrate construct validity; therefore, the results may provide useful information for policy analysis regarding the screening and treatment of heart attack.

Abstract This thesis concentrates on joint optimization of transmit power allocation and receive filtering in multiuser, multi-antenna communications. Due to the increasing number of wireless devices, the design of energy-efficient communication links is becoming increasingly important. In cellular mobile communications, reducing the average power consumption in uplink transmission is beneficial for users in order to extend battery life and, hence, energy efficiency in general. However, the power consumption of the high power amplifier (HPA) at the transmitter depends on the peak power of the transmission. This thesis focuses on power allocation problems for single-carrier (SC) frequency division multiple access (FDMA) and orthogonal FDMA (OFDMA) transmission assuming iterative reception. The goal in the first scheme presented in this thesis is to reduce the average power consumption by designing a power allocation method that takes into account the convergence properties of an iterative receiver in multiuser uplink communications. The proposed scheme can guarantee that the desired quality of service (QoS) is achieved after a sufficient number of iterations. Reducing the peak-to-average power ratio (PAPR) in any transmission system is beneficial because it allows the use of inexpensive, energy-efficient power amplifiers. The goal in the second scheme presented in this thesis is to control the PAPR of the transmitted signal. Hence, in addition to the QoS constraint, the instantaneous PAPR constraint is derived for SC-FDMA and OFDMA transmission. Moreover, a statistical approach is considered in which the power variance of the transmitted waveform is controlled. The QoS and PAPR constraints are considered jointly and, therefore, the proposed power allocation strategy jointly takes into account the channel quality and the PAPR characteristics of the power amplifier. However, the PAPR constraint can be adopted to any SC-FDMA or OFDMA framework and it is not restricted to the scheme presented in this thesis. The objective of the optimization problems considered throughout the thesis is to minimize the sum power. The majority of the derived constraints are non-convex and therefore, two alternative successive convex approximations (SCAs) are derived for all the non-convex constraints considered. The numerical results show that the proposed power allocation strategies can significantly reduce the average transmission power of users while allowing flexible PAPR control. Hence, the proposed methods can be used to extend battery life for users and especially improve the QoS at the cell edges
Tiivistelmä Väitöskirjassa tutkitaan lähettimessä tapahtuvan tehoallokoinnin sekä vastaanottimessa tapahtuvan signaalin suodatuksen yhteisoptimointia monikäyttöön suunnatussa langattomassa moniantennikommunikaatiossa. Langattomien laitteiden lukumäärän kasvaessa energiatehokkuuden merkitys tiedonsiirtolinkkien suunnittelussa korostuu. Soluihin perustuvassa langattomassa tietoliikenteessä keskimääräisen tehonkulutuksen pienentäminen ylälinkkilähetyksessä (käyttäjältä tukiasemaan) on tärkeää käyttäjän kannalta, sillä se pidentää laitteen akun kestoa. Lähettimen tehovahvistimen (high power amplifier (HPA)) tehonkulutus on kuitenkin verrannollinen lähetyksen huipputehoon. Väitöskirjassa luodaaan uusia menetelmiä sekä vertaillaan tehoallokointia yhden kantoaallon taajuustason monikäyttöön (single carrier frequency division multiple access (SC-FDMA)) ja ortogonaalisen taajuustason monikäyttöön (orthogonal FDMA (OFDMA)) perustuvissa lähetysteknologioissa. Työn ensimmäisessä osiossa tavoitteena on keskimääräisen tehonkulutuksen pienentäminen monen käyttäjän ylälinkkikommunikaatiossa suunnittelemalla tehoallokointimenetelmä, joka ottaa huomioon iteratiivisen vastaanottimen konvergenssiominaisuudet. Työssä ehdotettu menetelmä takaa vastaanotetun informaation halutun laadun (quality of service (QoS)) riittävän monen vastaanottimessa tehdyn iteraation jälkeen. Huipputehon ja keskitehon suhteen (peak to average power ratio (PAPR)) pienentäminen missä tahansa lähetyksessä on hyödyllistä, sillä sen ansiosta voidaan käyttää energiatehokkaampia ja halvempia tehovahvistimia. Työn jälkimmäisessä osiossa tavoitteena on kontrolloida lähetetyn signaalin huipputehon ja keskitehon suhdetta. Työn ensimmäisessä osiossa esitetyn QoS-rajoitteen lisäksi tehoallokointia rajoitetaan symbolisekvenssikohtaisella PAPR-rajoitteella SCFDMA- ja OFDMA-lähetyksessä. Lisäksi esitetään tilastollinen menetelmä, jossa rajoitetaan lähetetyn signaalin tehon varianssia. Kun käytetään yhtäaikaisesti QoS- ja PAPR-rajoitteita, voidaan tiedonsiirtokanavaan suunnitella optimaalinen tehoallokointi ottaen huomioon tehovahvistimen epälineaarisuudet. Työssä esitetty PAPR-rajoite on kuitenkin geneerinen, ja se voidaan sovittaa mihin tahansa SCFDMA- tai OFDMA- optimointikehykseen. Työssä esitettävien optimointiongelmien tavoitteena on käyttäjien summatehon minimointi. Suurin osa työssä esiintyvistä ongelmista on ei-konvekseja, joten siinä esitetään kaksi vaihtoehtoista peräkkäinen konveksi approksimaatio (successive convex approximation (SCA)) -menetelmää kaikille ei-konvekseille rajoitteille. Numeeriset tulokset osoittavat, että esitetyt tehoallokointimenetelmät pienentävät merkittävästi keskimääräistä tehonkulutusta mahdollistaen lisäksi adaptiivisen PAPR-kontrolloinnin. Väitöskirjassa esitettyjen menetelmien avulla voidaan pidentää mobiilikäyttäjien akun kestoa sekä erityisesti parantaa solun reunakäyttäjien palvelun laatua

Maintaining or even improving image quality while lowering patient dose is always the desire in clinical computed tomography (CT) imaging. Iterative reconstruction (IR) algorithms have been designed to allow for a reduced dose while maintaining or even improving an image. However, we have previously shown that the dose-saving capabilities allowed with IR are different for different clinical tasks. The channelized scanning linear observer (CSLO) was applied to study clinical tasks that combine detection and estimation when assessing CT image data. The purpose of this work is to illustrate the importance of task complexity when assessing dose savings and to move toward more realistic tasks when performing these types of studies. Human-observer validation of these methods will take place in a future publication. Low-contrast objects embedded in body-size phantoms were imaged multiple times and reconstructed by filtered back projection (FBP) and an IR algorithm. The task was to detect, localize, and estimate the size and contrast of low-contrast objects in the phantom. Independent signal-present and signal-absent regions of interest cropped from images were channelized by the dense-difference of Gauss channels for CSLO training and testing. Estimation receiver operating characteristic (EROC) curves and the areas under EROC curves (EAUC) were calculated by CSLO as the figure of merit. The one-shot method was used to compute the variance of the EAUC values. Results suggest that the IR algorithm studied in this work could efficiently reduce the dose by similar to 50% while maintaining an image quality comparable to conventional FBP reconstruction warranting further investigation using real patient data. (C) The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

碩士
國立中正大學
電機工程所
94
Exploiting maximum likelihood decoder can obtain the optimal solutions in multiple input multiple output system, however, the system become complicated. Sphere decoder not only obtains the optimal solutions near ML decoder, but also has lower complexity than ML decoder. To find the optimal solutions, the algorithm of sphere decoder applies search tree algorithm mostly, and detects transmitted signals from lower bound to upper bound. In this thesis, different detection techniques include sphere decoder and expectation maximization algorithm comparing to other detection techniques, then applied in the space-time transceivers.

碩士
國立中央大學
通訊工程研究所
92
Orthogonally multiplexed modulation (OMM) signals are investigated in this thesis. The research is focused on the application of the iterative detection of nonlinearly distorted OMM signals. In this thesis, we employ the iterative detection method to reduce the nonlinear distortion in orthogonal multiplexing orthogonal amplitude modulation (OMOAM) and orthogonal multiplexing orthogonal phase modulation (OMOPM). Three nonlinear distortion models are employed, and the results show that the performance trend of OMM distorted by these three models using the iterative detection are the same.

碩士
國立清華大學
工程與系統科學系
93
A benchmark in the development of product code is the invention of the iterative (turbo) decoding algorithm by Pyndiah. Pyndiah's algorithm works for any product code using linear block component codes. The turbo decoded product codes, now often referred to as turbo product codes (TPC) or block turbo codes (BTC), have the distinct merits of simple decoding, fast convergence, high code rate and very low (or no) error floor although their performance at the waterfall region is not as impressive as that of conventional turbo codes. This thesis explore the feasibility of using BTC in conjunction with the orthogonal frequency division multiplexing (OFDM) signals. Our focus is on the major receiver design concerns that dominate the error rate performance of the system, namely, the channel estimation and decoding algorithm. While no original estimation or decoding algorithm is proposed, we examine the performance of a joint iterative channel estimation and decoding approach. As for the transmitted waveform design, we check the impact of the interleaver and present a new class of BTC that we refer to as parallel concatenated product codes (PCPC). The PCPC has a structure similar to the conventional turbo code with the component codes replaced by systematic product codes. Numerical results indicate that it does outperform its BTC counterpart with a comparable code rate.

The title of this thesis is "On iterative detection for channels with memory". The core idea and the leitmotif of this work has been trying to find efficient and simple ways to perform iterative detection in a digital communication system characterized by concatenated coding and transmission over a channel with memory. The key steps involved in the development of this research theme can be summarized as follows. * Individuation of channels with memory, in particular of the source of their memory. * Soft-output algorithms for channels with memory. * Complexity reduction for these algorithms with minimal performance degradation. * Optimization of the reduced complexity algorithms. * An example of application. Each key point indicated above has found a possible development in this thesis. The solutions proposed are by no means the only ones. They are possible solutions.

碩士
國立中興大學
通訊工程研究所
101
With the increasing demand of multimedia services in wireless transmission nowadays, Multiple-Input Multiple-Output (MIMO) technologies are adopted in many standards to enhance the data rate and the link robustness. Among others, in 3GPP LTE (3rd Generation Partnership Project Long Term Evolution), it has been a common method. And the combination of the SC-FDMA modulation scheme for the uplink transmission scheme, to approach the demand for data transmission rates and error performance. Because of, SC-FDMA (Single Carrier Frequency Division Multiple Access) combines the desirable characteristics of OFDMA with the low PAPR (Peak-to-Average Power Ratio) of single-carrier transmission schemes. In this thesis, we using the Space Frequency block code (SFBC) applied on each transmitted block for the multiuser scenario. At receiver side, in order to reduce the search complexity, we can be used the Layered ORthogonal lattice Detector (LORD) algorithm, and combines the Turbo codes for MIMO SC-FDMA systems. Then utilize the Extrinsic Information from Maximum A Posteriori (MAP) decoder to cancel the Multiple Access Interference (MAI). Simulations for an uplink scenario assess the proposed algorithms in several situations.

The ever-increasing demand for higher information-transfer rates in wireless data networks invokes the need to develop more spectrally-efficient communication strategies. Techniques such as MIMO and turbo-coded CDMA are well known and obvious candidates for improving the spectral efficiency of next generation wireless networks, and addressing the limitations of currently implemented technologies. Correspondingly, such methods are finding their way into wireless network standards such as 3GPP and IEEE 802.20.
One measure of the size of a communication system is the number of independent data streams being transmitted simultaneously through a channel, assuming tight constraints on available bandwidth and signal power. Such data streams may originate from different users all wishing to communicate at once. In addition, each user may wish to transmit independent data on more than one antenna simultaneously in order to increase his or her own data rate. Although strategies for such multi-dimensional signalling have seen significant improvements in recent years, most of the techniques proposed in the literature still suffer from either poor performance or prohibitive complexity when the size of the system grows large.
This thesis is concerned primarily with supporting high system spectral-efficiencies in very large systems, while maintaining strong resistance to data errors with manageable complexity.
Iterative decoding, or Bayesian message-passing, is demonstrably able to approach closely the performance of an optical decoder for certain families of single-user error correction codes, with low computational complexity. The topic of this work, iterative list detection, is a technique for jointly decoding many independent data streams from multiple users and/or antennas, using powerful iterative decoding strategies developed for such single-user codes. The receiver strategies presented are based on the premise that iterative Bayesian decoding is capable of achieving performance very close to that of an optimal decoder for a multi-dimensional system, given certain assumptions on the system model. Other than this, iterative list detection makes no assumptions about the statistics of the interfering signals, linearity, or any other simplifying impositions. Rather, the method seeks only to approximate closely the probabilistic quantities dictated by the rules of the iterative decoding paradigm, which is by now well understood.
List detection itself refers to the computationally efficient calculation of signal probabilities conditioned on a noise-and-interference corrupted signal at the receiver, computed for each simultaneously transmitted signal. The calculation is the key step in the implementation of an iterative receiver for such a system.
After introducing the list detection strategy in the context of iterative receivers for multi-user MIMO channels, algorithms for optimal list detection are described. A new optimal list detection algorithm with some superior properties to other implementations in the literature is proposed. While still very computationally complex, performance results for optimal list detection are presented that demonstrate the effectiveness and utility of the paradigm, and provide a performance benchmark for any sub-optimal list detection technique. The performance is also compared with other techniques such as linear filters, providing an appreciation of the benefits of list detection.
An asymptotic large-systems analysis is then undertaken in order to determine the behaviour of a fundamental parameter that determines the complexity of list detection, specifically the number of terms in a certain summation. The minimum number of terms is derived under an accuracy constraint on the signal probabilities. Results demonstrate that the number of terms does not necessarily increase with the size of the system, and the conditions under which this is true are indicated.
The main contribution of the thesis is the development of a number or computationally efficient sub-optimal list detection algorithms. Various strategies are proposed for different system scenarios, resulting in near-optimal performance with complexity that adapts automatically to cope with changing channel conditions and interference. The performance of the new techniques is demonstrated via simulation in channels with various statistics, dimensionality and interference, showing significant improvements in terms of both error resistance and complexity over other proposed methods.
Thesis (PhDTelecommunications)--University of South Australia, 2004.

碩士
中華大學
資訊工程學系碩士班
101
The region completeness of object detection is very crucial to video surveillance, such as the pedestrian and vehicles identifications. However, many conventional object detecting techniques can’t guarantee the object region completeness because the object detections can be affected by lighting or color variations. In order to overcome this problem, we propose the iterative superpixels grouping (ISPG) method to extract the precise object boundary and generate the object region with high completeness after the object detections. By extending the superpixel segmentation method, the proposed ISPG method can improve the inaccurate segmentation problem and guarantee the region completeness on the object regions. The larger the number of object pixels in each superpixel, the more possibility this superpixel belongs to the object region. Hence, the region completeness can be improved by merging the superpixels that are classified into the object. Experimental results show that the proposed method outperforms the some conventional objection method in terms of object completeness evaluations.

碩士
元智大學
電機工程學系甲組
107
With the growth of wireless communication system, the frequency spectrum has been widely occupied by many wireless communication standards. Improving the spectral efficiency with the reasonable hardware complexity is a critical issue for the future communication standards. Many wireless communication standards at nowadays have adopted the multiple-input multiple-output (MIMO) technology which is a solution to support higher data rates and spectral efficiency. However, the transmitted signals will suffer from the inter symbol interference (ISI) effect using multiple antennas. Employing iterative detection and decoding (IDD) receiver in MIMO system was demonstrated to improve the system performance significantly with a reasonable hardware complexity. Based on the MIMO IDD receiver, we proposed a flexible sphere decoding architecture that can support multiple configurable antennas and modulation schemes. In order to achieve the Maximum Likelihood (ML) detection performance and lower computational complexity, we adopt two kinds of sphere decoding codes, K-best and iterative K-best algorithms. Although iterative detection can improve the system performance, it will greatly increase the computational complexity. We further proposed a low-complexity enumeration to reduce the computational complexity in iterative K-best sphere decoding. Using the 40nm CMOS technology, the implementation results show that the proposed enumeration algorithm reduces the computational complexity by 15% and 45% in high channel quality and low channel quality which compared with the conventional iterative K-best algorithm. Also, we use a pipelined design to optimize the architecture, and the operating frequency is hence improved up to 230 MHz with the core area 1.88 mm2. The throughput achieves 408 Mbps and 204 Mbps under 16-QAM and 4×4 antennas for K-best mode and iterative K-best mode, respectively. Consequently, the proposed sphere detector provides a better system performance and hardware efficiency.