The ongoing convergence of spectrum and hardware requirements for wireless sensing and communication applications has fueled the integrated sensing and communication (ISAC) paradigm in next-generation networks. Neural-network-based ISAC leverages data-driven learning techniques to add sensing capabilities to existing communication infrastructure. This paper presents a novel signal-processing framework for such neural ISAC systems based on the multiple-input multiple-output (MIMO) and orthogonal frequency-division multiplexing (OFDM) downlink. Our approach enables generalized sensing functionality without modifying the MIMO-OFDM communication link. Specifically, our neural ISAC pipeline measures the backscattered communication signals to generate discrete map representations of spatial occupancy, formulated as multiclass or multilabel classification problems, which can then be utilized by specialized downstream tasks. To improve sensing performance in closed or cluttered environments, our neural ISAC pipeline relies on features specifically designed to mitigate strong reflective paths. Extensive simulations using ray-tracing models demonstrate that our neural ISAC framework reliably reconstructs scene maps without altering the MIMO-OFDM communication pipeline or reducing data rates.

Massive multiuser multiple-input multiple-output (MIMO) and millimeter-wave (mmWave) communication are key physical layer technologies in future wireless systems. Their deployment, however, is expected to incur excessive baseband processing hardware cost and power consumption. Beamspace processing leverages the channel sparsity at mmWave frequencies to reduce baseband processing complexity. In this paper, we review existing beamspace data detection algorithms and propose new algorithms as well as corresponding VLSI architectures that reduce data detection power. We present VLSI implementation results for the proposed architectures in a 22nm FDSOI process. Our results demonstrate that a fully-parallelized implementation of the proposed complex sparsity-adaptive equalizer (CSPADE) achieves up to 54% power savings compared to antenna-domain equalization. Furthermore, our fully-parallelized designs achieve the highest reported throughput among existing massive MIMO data detectors, while achieving better energy and area efficiency. We also present a sequential multiply-accumulate (MAC)-based architecture for CSPADE, which enables even higher power savings, i.e., up to 66%, compared to a MAC-based antenna-domain equalizer.

Channel charting (CC) is a self-supervised positioning technique whose main limitation is that the estimated positions lie in an arbitrary coordinate system that is not aligned with true spatial coordinates. In this work, we propose a novel method to produce CC locations in true spatial coordinates with the aid of a digital twin (DT). Our main contribution is a new framework that (i) extracts large-scale channel-state information (CSI) features from estimated CSI and the DT and (ii) matches these features with a cosine-similarity loss function. The DT-aided loss function is then combined with a conventional CC loss to learn a positioning function that provides true spatial coordinates without relying on labeled data. Our results for a simulated indoor scenario demonstrate that the proposed framework reduces the relative mean distance error by 29% compared to the state of the art. We also show that the proposed approach is robust to DT modeling mismatches and a distribution shift in the testing data.

The semiconductor industry is pivotal to Europe's economy, especially within the industrial and automotive sectors. However, Europe faces a significant shortfall in chip design capabilities, marked by a severe skilled labor shortage and lagging contributions in the design value chain segment. This paper explores the role of European universities and academic initiatives in enhancing chip design education and research to address these deficits. We provide a comprehensive overview of current European chip design initiatives, analyze major challenges in recruitment, productivity, technology access, and design enablement, and identify strategic opportunities to strengthen chip design capabilities within academic institutions. Our analysis leads to a series of recommendations that highlight the need for coordinated efforts and strategic investments to overcome these challenges.

We introduce a novel class of regularization functions, called Cauchy–Schwarz (CS) regularizers, which can be designed to induce a wide range of properties in solution vectors of optimization problems. To demonstrate the versatility of CS regularizers, we derive regularization functions that promote discrete-valued vectors, eigenvectors of a given matrix, and orthogonal matrices. The resulting CS regularizers are simple, differentiable, and can be free of spurious stationary points, making them suitable for gradient-based solvers and large-scale optimization problems. In addition, CS regularizers automatically adapt to the appropriate scale, which is, for example, beneficial when discretizing the weights of neural networks. To demonstrate the efficacy of CS regularizers, we provide results for solving underdetermined systems of linear equations and weight quantization in neural networks. Furthermore, we discuss specializations, variations, and generalizations, which lead to an even broader class of new and possibly more powerful regularizers.

Guessing random additive noise decoding (GRAND) is a code-agnostic decoding method that iteratively guesses the noise pattern affecting the received codeword. The number of noise sequences to test depends on the noise realization. Thus, GRAND exhibits random runtime which results in nondeterministic throughput. However, real-time systems must process the incoming data at a fixed rate, necessitating a fixed-throughput decoder in order to avoid losing data. We propose a first-in first-out (FIFO) scheduling architecture that enables a fixed throughput while improving the block error rate (BLER) compared to the common approach of imposing a maximum runtime constraint per received codeword. Moreover, we demonstrate that the average throughput metric of GRAND-based hardware implementations typically provided in the literature can be misleading as one needs to operate at approximately one order of magnitude lower throughput to achieve the BLER of an unconstrained decoder.

We present a 22nm FD-SOI (fully depleted silicon-on-insulator) application-specific integrated circuit (ASIC) implementation of a novel soft-output Gram-domain block coordinate descent (GBCD) data detector for massive multi-user (MU) multiple-input multiple-output (MIMO) systems. The ASIC simultaneously addresses the high throughput requirements for millimeter wave (mmWave) communication, stringent area and power budget per subcarrier in an orthogonal frequency-division multiplexing (OFDM) system, and error-rate performance challenges posed by realistic mmWave channels. The proposed GBCD algorithm utilizes a posterior mean estimate (PME) denoiser and is optimized using deep unfolding, which results in superior error-rate performance even in scenarios with highly correlated channels or where the number of user equipment (UE) data streams is comparable to the number of basestation (BS) antennas. The fabricated GBCD ASIC supports up to 16 UEs transmitting QPSK to 256-QAM symbols to a 128-antenna BS, and achieves a peak throughput of 7.1Gbps at 367mW. The core area is only 0.97mm2 thanks to a reconfigurable array of processing elements that enables extensive resource sharing. Measurement results demonstrate that the proposed GBCD data-detector ASIC achieves best-in-class throughput and area efficiency.

Guessing random additive noise decoding (GRAND) is establishing itself as a universal method for decoding linear block codes, and ordered reliability bits GRAND (ORBGRAND) is a hardware-friendly variant that processes soft-input information. In this work, we propose an efficient hardware implementation of ORBGRAND that significantly reduces the cost of querying noise sequences with slight frame error rate (FER) performance degradation. Different from logistic weight order (LWO) and improved LWO (iLWO) typically used to generate noise sequences, we introduce a reduced-complexity and hardware-friendly method called shift LWO (sLWO), of which the shift factor can be chosen empirically to trade the FER performance and query complexity well. To effectively generate noise sequences with sLWO, we utilize a hardware-friendly lookup-table (LUT)-aided strategy, which improves throughput as well as area and energy efficiency. To demonstrate the efficacy of our solution, we use synthesis results evaluated on polar codes in a 65-nm CMOS technology. While maintaining similar FER performance, our ORBGRAND implementations achieve 53.6-Gbps average throughput ($1.26\times $higher), 4.2-Mbps worst case throughput ($8.24\times $higher), 2.4-Mbps/mm2 worst case area efficiency ($12\times $higher), and$4.66\times 10 ^{{4}}$pJ/bit worst case energy efficiency ($9.96\times $lower) compared with the synthesized ORBGRAND design with LWO for a (128, 105) polar code and also provide$8.62\times $higher average throughput and$9.4\times $higher average area efficiency but$7.51\times $worse average energy efficiency than the ORBGRAND chip for a (256, 240) polar code, at a target FER of$10^{-7}$.

Channel charting is an emerging self-supervised method that maps channel-state information (CSI) to a low-dimensional latent space (the channel chart) that represents pseudo-positions of user equipments (UEs). While channel charts preserve local geometry, i.e., nearby UEs are nearby in the channel chart (and vice versa), the pseudo-positions are in arbitrary coordinates and global geometry is typically not preserved. In order to embed channel charts in real-world coordinates, we first propose a bilateration loss for distributed multiple-input multiple-output (D-MIMO) wireless systems in which only the access point (AP) positions are known. The idea behind this loss is to compare the received power at pairs of APs to determine whether a UE should be placed closer to one AP or the other in the channel chart. We then propose a line-of-sight (LoS) bounding-box loss that places the UE in a predefined LoS area of each AP that is estimated to have a LoS path to the UE. We demonstrate the efficacy of combining both of these loss functions with neural-network-based channel charting using ray-tracing-based and measurement-based channel vectors. Our proposed approach outperforms several baselines and maintains the self-supervised nature of channel charting as it neither relies on geometrical propagation models nor on any ground-truth UE position information.

We propose new low-fidelity (LoFi) user equipment (UE) scheduling algorithms for multiuser multiple-input multiple-output (MIMO) wireless communication systems. The proposed methods rely on an efficient guess-and-check procedure that, given an objective function, performs paired comparisons between random subsets of UEs that should be scheduled in certain time slots. The proposed LoFi scheduling methods are computationally efficient, highly parallelizable, and gradient-free, which enables the use of almost arbitrary, non-differentiable objective functions. System simulations in a millimeter-wave (mmWave) multiuser MIMO scenario demonstrate that the proposed LoFi schedulers outperform a range of state-of-the-art user scheduling algorithms in terms of bit error-rate and/or computational complexity.

We analyze the uplink performance of a distributed massive multiple-input multiple-output (MIMO) architecture in which the remotely located access points (APs) are connected to a central processing unit via a fiber-optical fronthaul carrying a dithered and 1-bit quantized version of the received radio-frequency (RF) signal. The innovative feature of the proposed architecture is that no down-conversion is performed at the APs. This eliminates the need to equip the APs with local oscillators, which may be difficult to synchronize. Under the assumption that a constraint is imposed on the amount of data that can be exchanged across the fiber-optical fronthaul, we investigate the tradeoff between spatial oversampling, defined in terms of the total number of APs, and temporal oversampling, defined in terms of the oversampling factor selected at the central processing unit, to facilitate the recovery of the transmitted signal from 1-bit samples of the RF received signal. Using the so-called error-vector magnitude (EVM) as performance metric, we shed light on the optimal design of the dither signal, and quantify, for a given number of APs, the minimum fronthaul rate required for our proposed distributed massive MIMO architecture to outperform a standard co-located massive MIMO architecture in terms of EVM.

Localization services for wireless devices play an increasingly important role in our daily lives and a plethora of emerging services and applications already rely on precise position information. Widely used on-device positioning methods, such as the global positioning system, enable accurate outdoor positioning and provide the users with full control over what services and applications are allowed to access their location information. In order to provide accurate positioning indoors or in cluttered urban scenarios without line-of-sight satellite connectivity, powerful off-device positioning systems, which process channel state information (CSI) measured at the infrastructure base stations or access points with deep neural networks, have emerged recently. Such off-device wireless positioning systems inherently link a user’s data transmission with its localization, since accurate CSI measurements are necessary for reliable wireless communication—this not only prevents the users from controlling who can access this information but also enables virtually everyone in the device’s range to estimate its location, resulting in serious privacy and security concerns. We therefore propose on-device attacks against off-device wireless positioning systems in multi-antenna orthogonal frequency-division multiplexing systems while remaining standard compliant and minimizing the impact on quality-of-service, and we demonstrate their efficacy using real-world measured datasets for cellular outdoor and wireless LAN indoor scenarios. We also investigate defenses to counter such attack mechanisms, and we discuss the limitations and implications on protecting location privacy in existing and future wireless communication systems.

Channel charting is an emerging self-supervised method that maps channel state information (CSI) to a low-dimensional latent space, which represents pseudo-positions of user equipments (UEs). While this latent space preserves local geometry, i.e., nearby UEs are nearby in latent space, the pseudo-positions are in arbitrary coordinates and global geometry is not preserved. In order to enable channel charting in real-world coordinates, we propose a novel bilateration loss for multipoint wireless systems in which only the access point (AP) locations are known—no geometrical models or ground-truth UE position information is required. The idea behind this bilateration loss is to compare the received power at pairs of APs in order to determine whether a UE should be placed closer to one AP or the other in latent space. We demonstrate the efficacy of our method using channel vectors from a commercial ray-tracer.

Even though machine learning (ML) techniques are being widely used in communications, the question of how to train communication systems has received surprisingly little attention. In this paper, we show that the commonly used binary cross-entropy (BCE) loss is a sensible choice in uncoded systems, e.g., for training ML-assisted data detectors, but may not be optimal in coded systems. We propose new loss functions targeted at minimizing the block error rate and SNR deweighting, a novel method that trains communication systems for optimal performance over a range of signal-to-noise ratios. The utility of the proposed loss functions as well as of SNR deweighting is shown through simulations in NVIDIA Sionna.

Multi-antenna (MIMO) processing is a promising solution to the problem of jammer mitigation. Existing methods mitigate the jammer based on an estimate of its subspace (or receive statistics) acquired through a dedicated training phase. This strategy has two main drawbacks: (i) it reduces the communication rate since no data can be transmitted during the training phase and (ii) it can be evaded by smart or multi-antenna jammers that are quiet during the training phase or that dynamically change their subspace through time-varying beamforming. To address these drawbacks, we propose joint jammer mitigation and data detection (JMD), a novel paradigm for MIMO jammer mitigation. The core idea is to estimate and remove the jammer interference subspace jointly with detecting the transmit data over multiple time slots. Doing so removes the need for a dedicated rate-reducing training period while enabling the mitigation of smart and dynamic multi-antenna jammers. We instantiate our paradigm with SANDMAN, a simple and practical algorithm for multi-user MIMO uplink JMD. Extensive simulations demonstrate the efficacy of JMD, and of SANDMAN in particular, for jammer mitigation.

Analog subtractive sound synthesis is widely used in music production since the 1960s, with popular synthesizers from Moog Music, Sequential, and ARP Instruments. While analog synthesizers are considered superior in terms of sound quality compared to their digital counterparts, analog circuitry typically suffers from temperature instabilities, component variations, and lack of flexibility. Digital music synthesizers can avoid all of these drawbacks, but it is challenging to design digital oscillators and filters that do not cause aliasing artifacts and sound as impressive as their analog counterparts. In this live demonstration, we show a polyphonic hybrid digital-analog music synthesizer prototype in which the oscillator signals are aliasing-free and generated by an FPGA, whereas the filters and amplifiers are implemented with analog circuits, thus combining the best of both worlds.

Orthogonal frequency-division multiplexing (OFDM) time-domain signals exhibit high peak-to-average (power) ratio (PAR), which requires linear radio-frequency chains to avoid an increase in error-vector magnitude (EVM) and out-of-band (OOB) emissions. In this paper, we propose a novel joint PAR reduction and precoding algorithm that relaxes these linearity requirements in massive multiuser (MU) multiple-input multiple- output (MIMO) wireless systems. Concretely, we develop a novel alternating projections method, which limits the PAR and transmit power increase while simultaneously suppressing MU interference. We provide a theoretical foundation of our algorithm and provide simulation results for a massive MU-MIMO-OFDM scenario. Our results demonstrate significant PAR reduction while limiting the transmit power, without causing EVM or OOB emissions.

Systolic array architectures have recently emerged as successful accelerators for deep convolutional neural network (CNN) inference. Such architectures can be used to efficiently execute general matrix–matrix multiplications (GeMMs), but computing convolutions with this primitive involves transforming the 3-D input tensor into an equivalent matrix, which can lead to an inflation of the input data, increasing the OFF-chip memory traffic which is critical for energy efficiency. In this work, we propose a GeMM-based systolic array accelerator that uses a novel data feeder architecture to perform ON-chip, on-the-fly convolution lowering (also known as im2col), supporting arbitrary tensor and kernel sizes as well as strided and dilated (or atrous) convolutions. By using our data feeder, we reduce memory transactions and required bandwidth on state-of-the-art CNNs by a factor of two, while only adding an area and power overhead of 4% and 7%, respectively. Application specific integrated circuit (ASIC) implementation of our accelerator in 22-nm technology fits in less than 1.1 mm 2 and reaches an energy efficiency of 1.10 TFLOP/sW with 16-bit floating-point arithmetic.

Baseband processing algorithms often require knowledge of the noise power, signal power, or signal-to-noise ratio (SNR). In practice, these parameters are typically unknown and must be estimated. Furthermore, the mean-square error (MSE) is a desirable metric to be minimized in a variety of estimation and signal recovery algorithms. However, the MSE cannot directly be used as it depends on the true signal that is generally unknown to the estimator. In this paper, we propose novel blind estimators for the average noise power, average receive signal power, SNR, and MSE. The proposed estimators can be computed at low complexity and solely rely on the large-dimensional and sparse nature of the processed data. Our estimators can be used (i) to quickly track some of the key system parameters while avoiding additional pilot overhead, (ii) to design low-complexity nonparametric algorithms that require such quantities, and (iii) to accelerate more sophisticated estimation or recovery algorithms. We conduct a theoretical analysis of the proposed estimators for a Bernoulli complex Gaussian (BCG) prior, and we demonstrate their efficacy via synthetic experiments. We also provide three application examples that deviate from the BCG prior in millimeter-wave multi-antenna and cell-free wireless systems for which we develop nonparametric denoising algorithms that improve channel-estimation accuracy with a performance comparable to denoisers that assume perfect knowledge of the system parameters.

Wireless systems must be resilient to jamming attacks. Existing mitigation methods require knowledge of the jammer’s transmit characteristics. However, this knowledge may be difficult to acquire, especially for smart jammers that attack only specific instants during transmission in order to evade mitigation. We propose a novel method that mitigates attacks by smart jammers on massive multi-user multiple-input multiple-output (MU-MIMO) basestations (BSs). Our approach builds on recent progress in joint channel estimation and data detection (JED) and exploits the fact that a jammer cannot change its subspace within a coherence interval. Our method, called MAED (short for MitigAtion, Estimation, and Detection), uses a novel problem formulation that combines jammer estimation and mitigation, channel estimation, and data detection, instead of separating these tasks. We solve the problem approximately with an efficient iterative algorithm. Our simulation results show that MAED effectively mitigates a wide range of smart jamming attacks without having any a priori knowledge about the attack type.

We propose a novel beam-tracking algorithm based on channel charting (CC) which maintains the communication link between a base station (BS) and a mobile user equipment (UE) in a millimeter wave (mmWave) mobile communications system. Our method first uses large-scale channel state information at the BS in order to learn a CC. The points in the channel chart are then annotated with the signal-to-noise ratio (SNR) of best beams. One can then leverage this CC-to-SNR mapping in order to track strong beams between UEs and BS efficiently and robustly at very low beam-search overhead. Simulation results in a mmWave scenario show that the performance of the CC-assisted beam tracking method approaches that of an exhaustive beam-search approach while requiring significantly lower beam-search overhead than conventional tracking methods.

Channel state information (CSI)-based fingerprinting via neural networks (NNs) is a promising approach to enable accurate indoor and outdoor positioning of user equipment (UE), even under challenging propagation conditions. In this paper, we propose a positioning pipeline for wireless LAN MIMO-OFDM systems which uses uplink CSI measurements obtained from one or more unsynchronized access points (APs). For each AP receiver, novel features are first extracted from the CSI that are robust to system impairments arising in real-world transceivers. These features are the inputs to a NN that extracts a probability map indicating the likelihood of a UE being at a given grid point. The NN output is then fused across multiple APs to provide a final position estimate. We provide experimental results with real-world indoor measurements under line-of-sight (LoS) and non-LoS propagation conditions for an 80MHz bandwidth IEEE 802.11ac system using a two-antenna transmit UE and two AP receivers each with four antennas. Our approach is shown to achieve centimeter-level median distance error, an order of magnitude improvement over a conventional baseline.

We propose a joint channel estimation and data detection (JED) algorithm for densely-populated cell-free massive multiuser (MU) multiple-input multiple-output (MIMO) systems, which reduces the channel training overhead caused by the presence of hundreds of simultaneously transmitting user equipments (UEs). Our algorithm iteratively solves a relaxed version of a maximum a-posteriori JED problem and simultaneously exploits the sparsity of cell-free massive MU-MIMO channels as well as the boundedness of QAM constellations. In order to improve the performance and convergence of the algorithm, we propose methods that permute the access point and UE indices to form so-called virtual cells, which leads to better initial solutions. We assess the performance of our algorithm in terms of root-mean-squared-symbol error, bit error rate, and mutual information, and we demonstrate that JED significantly reduces the pilot overhead compared to orthogonal training, which enables reliable communication with short packets to a large number of UEs.

We propose blind estimators for the average noise power, receive signal power, signal-to-noise ratio (SNR), and mean-square error (MSE), suitable for multi-antenna millimeter wave (mmWave) wireless systems. The proposed estimators can be computed at low complexity and solely rely on beamspace sparsity, i.e., the fact that only a small number of dominant propagation paths exist in typical mmWave channels. Our estimators can be used (i) to quickly track some of the key quantities in multi-antenna mmWave systems while avoiding additional pilot overhead and (ii) to design efficient nonparametric algorithms that require such quantities. We provide a theoretical analysis of the proposed estimators, and we demonstrate their efficacy via synthetic experiments and using a nonparametric channel-vector denoising task with realistic multi-antenna mmWave channels.

Linearization of massive MIMO arrays is a significant computational challenge that typically scales with the number of antennas. In this work, we introduce a beam-oriented digital predistortion (DPD) scheme for OFDM-based massive MIMO systems that applies predistortion before the precoder in the OFDM guard-band subcarriers. Using simulation results, we show that, for a 64 antenna massive MIMO array, our proposed method can achieve the same DPD performance as a conventional DPD method while requiring an order of magnitude fewer multiplications.

Beamspace processing is an emerging technique to reduce baseband complexity in massive multiuser (MU) multiple-input multiple-output (MIMO) communication systems operating at millimeter-wave (mmWave) and terahertz frequencies. The high directionality of wave propagation at such high frequencies ensures that only a small number of transmission paths exist between user equipments and basestation (BS). In order to resolve the sparse nature of wave propagation, beamspace processing traditionally computes a spatial discrete Fourier transform (DFT) across a uniform linear antenna array at the BS where each DFT output is associated with a specific beam. In this paper, we study optimality conditions of the DFT for sparsity-based beamspace processing with idealistic mmWave channel models and realistic channels. To this end, we propose two algorithms that learn unitary beamspace transforms using an$\ell^{4}$-norm-based sparsity measure, and we investigate their optimality theoretically and via simulations.

We propose a novel soft-output joint channel estimation and data detection (JED) algorithm for multiuser (MU) multiple-input multiple-output (MIMO) wireless communication systems. Our algorithm approximately solves a maximum a-posteriori JED optimization problem using deep unfolding and generates soft-output information for the transmitted bits in every iteration. The parameters of the unfolded algorithm are computed by a hyper-network that is trained with a binary cross entropy (BCE) loss. We evaluate the performance of our algorithm in a coded MU-MIMO system with 8 basestation antennas and 4 user equipments and compare it to state-of-the-art algorithms separate channel estimation from soft-output data detection. Our results demonstrate that our JED algorithm outperforms such data detectors with as few as 10 iterations.

We consider the network-side mobile localization problem in future 5G and beyond wireless networks with distributed multi-antenna base stations (BSs). For this application, we propose a semi-supervised multi-point channel charting (SS-MPCC) framework, which consists of (i) collaborative collection of channel state information (CSI) and other side-information by distributed BSs; (ii) local CSI feature extraction and self-learning of a dissimilarity metric, and (iii) global graph construction and constrained manifold learning. We show that side-information from routine network operations, including timestamps, channel qualities, and a small set of labeled samples, can be exploited to construct a consistent global graph. The graph is then mapped to a 2D channel chart using constrained manifold learning for localization purposes. We evaluate the performance of SS-MPCC in a simulated urban outdoor scenario with realistic user motion. Our results show that SS-MPCC achieves a mean localization error of 5.6 m with only 10% of labeled CSI samples. SS-MPCC does not require accurate synchronization among multiple BSs and is promising for future cellular localization.

For massive multiple-input multiple-output (MIMO) systems, linear minimum mean-square error (MMSE) detection has been shown to achieve near-optimal performance but suffers from excessively high complexity due to the large-scale matrix inversion. Being matrix inversion free, detection algorithms based on theGauss–Seidel(GS) method have been proved more efficient than conventionalNeumannseries expansion-based ones. In this paper, an efficient GS-based soft-output data detector for massive MIMO and a corresponding VLSI architecture are proposed. To accelerate the convergence of the GS method, a new initial solution is proposed. Several optimizations on the VLSI architecture level are proposed to further reduce the processing latency and area. Our reference implementation results on a Xilinx Virtex-7 XC7VX690T FPGA for a 128 base-station antenna and eight user massive MIMO system show that our GS-based data detector achieves a throughput of 732 Mb/s with close-to-MMSE error-rate performance. Our implementation results demonstrate that the proposed solution has advantages over the existing designs in terms of complexity and efficiency, especially under challenging propagation conditions.

In this article, we propose a novel data detection algorithm and a corresponding VLSI design for massive multiuser (MU) multiple-input-multiple-output (MIMO) wireless systems. Our algorithm uses alternating direction method of multipliers (ADMM)-based infinity-norm-constrained equalization and is called ADMIN. ADMIN is an iterative algorithm that outperforms linear detectors by a large margin when the ratio between the numbers of base-station (BS) and user antennas is small. In the first iteration, ADMIN computes the linear minimum mean-square error (MMSE) solution, which is sufficient when the ratio between the numbers of BS and user antennas is large. We develop time-shared and iterative VLSI architectures for LDL-decomposition-based soft-output ADMIN supporting 16- and 32-user systems. We present application-specific integrated circuit (ASIC) designs for 16-64 antenna base stations in 28-nm CMOS that supports up to 64 quadrature amplitude modulation (QAM). The 16-user ADMIN ASIC achieves 303 Mb/s while dissipating 85 mW. The 32-user ADMIN ASIC achieves 287 and 241 Mb/s while dissipating 121 and 135 mW for 32 and 64 BS antennas, respectively. ADMIN has also been implemented on a Xilinx Virtex-7 field-programmable gate array (FPGA) and is compared with state-of-the-art massive MIMO data detectors.

Transfer learning facilitates the training of task-specific classifiers using pre-trained models as feature extractors. We present a family of transferable adversarial attacks against such classifiers, generated without access to the classification head; we call these headless attacks. We first demonstrate successful transfer attacks against a victim network using only its feature extractor. This motivates the introduction of a label-blind adversarial attack. This transfer attack method does not require any information about the class-label space of the victim. Our attack lowers the accuracy of a ResNet18 trained on CIFAR10 by over 40%.

Nxt-generation wireless systems are expected to combine millimeter-wave (mmWave) and massive multi-user multiple-input multiple-output (MU-MIMO) technologies to deliver high data-rates. These technologies require the basestations (BSs) to process high-dimensional data at extreme rates, which results in high power dissipation and system costs. Finite-alphabet equalization has been proposed recently to reduce the power consumption and silicon area of uplink spatial equalization circuitry at the BS by coarsely quantizing the equalization matrix. In this work, we improve upon finite-alphabet equalization by performing unbiased estimation and soft-output computation for coded systems. By simulating a massive MU-MIMO system that uses orthogonal frequency-division multiplexing and per-user convolutional coding, we show that soft-output finite-alphabet equalization delivers competitive error-rate performance using only 1 to 3 bits per entry of the equalization matrix, even for challenging mmWave channels.

We propose equalization-based data detection algorithms for all-digital millimeter-wave (mmWave) massive multiuser multiple-input multiple-out (MU-MIMO) systems that exploit sparsity in the beamspace domain to reduce complexity. We provide a condition on the number of users, basestation antennas, and channel sparsity for which beamspace equalization can be less complex than conventional antenna-domain processing. We evaluate the performance-complexity trade-offs of existing and new beamspace equalization algorithms using simulations with realistic mmWave channel models. Our results reveal that one of our proposed beamspace equalization algorithms achieves up to 8× complexity reduction under line-of-sight conditions, assuming a sufficiently large number of transmissions within the channel coherence interval.

We propose finite-alphabet equalization, a new paradigm that restricts the entries of the spatial equalization matrix to low-resolution numbers, enabling high-throughput, low-power, and low-cost hardware equalizers. To minimize the performance loss of this paradigm, we introduce FAME, short for finite-alphabet minimum mean-square error (MMSE) equalization, which is able to significantly outperform a naïve quantization of the linear MMSE matrix. We develop efficient algorithms to approximately solve the NP-hard FAME problem and showcase that near-optimal performance can be achieved with equalization coefficients quantized to only 1-3 bits for massive multi-user multiple-input multiple-output (MU-MIMO) millimeter-wave (mmWave) systems. We provide very-large scale integration (VLSI) results that demonstrate a reduction in equalization power and area by at least a factor of 3.9× and 5.8×, respectively.

Processing in memory (PIM) moves computation into memories with the goal of improving throughput and energy-efficiency compared to traditional von Neumann-based architectures. Most existing PIM architectures are either general-purpose but only support atomistic operations, or are specialized to accelerate a single task. We propose the Parallel Processor in Associative Content-addressable memory (PPAC), a novel in-memory accelerator that supports a range of matrix-vector-product (MVP)-like operations that find use in traditional and emerging applications. PPAC is, for example, able to accelerate low-precision neural networks, exact/approximate hash lookups, cryptography, and forward error correction. The fully-digital nature of PPAC enables its implementation with standard-cell-based CMOS, which facilitates automated design and portability among technology nodes. To demonstrate the efficacy of PPAC, we provide post-layout implementation results in 28nm CMOS for different array sizes. A comparison with recent digital and mixed-signal PIM accelerators reveals that PPAC is competitive in terms of throughput and energy-efficiency, while accelerating a wide range of applications and simplifying development.

In this paper, we explore clean-label poisoning attacks on deep convolutional networks with access to neither the network’s output nor its architecture or parameters. Our goal is to ensure that after injecting the poisons into the training data, a model with unknown architecture and parameters trained on that data will misclassify the target image into a specific class. To achieve this goal, we generate multiple poison images from the base class by adding small perturbations which cause the poison images to trap the target image within their convex polytope in feature space. We also demonstrate that using Dropout during crafting of the poisons and enforcing this objective in multiple layers enhances transferability, enabling attacks against both the transfer learning and end-to-end training settings. We demonstrate transferable attack success rates of over 50% by poisoning only 1% of the training set.

Massive multiuser (MU) multiple-input multiple-output (MIMO) promises significant improvements in spectral efficiency compared to small-scale MIMO. Typical massive MU-MIMO base-station (BS) designs rely on centralized linear data detectors and precoders which entail excessively high complexity, interconnect data rates, and chip input/output (I/O) bandwidth when executed on a single computing fabric. To resolve these complexity and bandwidth bottlenecks, we propose new decentralized algorithms for data detection and precoding that use coordinate descent. Our methods parallelize computations across multiple computing fabrics, while minimizing interconnect and I/O bandwidth. The proposed decentralized algorithms achieve near-optimal error-rate performance and multi-Gbps throughput at sub-1 ms latency when implemented on a multi-GPU cluster with half-precision floating-point arithmetic.

Adversarial training, in which a network is trained on adversarial examples, is one of the few defenses against adversarial attacks that withstands strong attacks. Unfortunately, the high cost of generating strong adversarial examples makes standard adversarial training impractical on large-scale problems like ImageNet. We present an algorithm that eliminates the overhead cost of generating adversarial examples by recycling the gradient information computed when updating model parameters. Our "free" adversarial training algorithm achieves comparable robustness to PGD adversarial training on the CIFAR-10 and CIFAR-100 datasets at negligible additional cost compared to natural training, and can be 7 to 30 times faster than other strong adversarial training methods. Using a single workstation with 4 P100 GPUs and 2 days of runtime, we can train a robust model for the large-scale ImageNet classification task that maintains 40% accuracy against PGD attacks.

We consider the downlink of a massive multiuser (MU) multiple-input multiple-output (MIMO) system in which the base station (BS) is equipped with low-resolution digital-to-analog converters (DACs). In contrast to most existing results, we assume that the system operates over a frequency-selective wideband channel and uses orthogonal frequency division multiplexing (OFDM) to simplify equalization at the user equipments (UEs). Furthermore, we consider the practically relevant case of oversampling DACs. We theoretically analyze the uncoded bit error rate (BER) performance with linear precoders (e.g., zero forcing) and quadrature phase-shift keying using Bussgang's theorem. We also develop a lower bound on the information-theoretic sum-rate throughput achievable with Gaussian inputs, which can be evaluated in closed form for the case of 1-bit DACs. For the case of multi-bit DACs, we derive approximate, yet accurate, expressions for the distortion caused by low-precision DACs, which can be used to establish the lower bounds on the corresponding sum-rate throughput. Our results demonstrate that, for a massive MU-MIMO-OFDM system with a 128-antenna BS serving 16 UEs, only 3-4 DAC bits are required to achieve an uncoded BER of 10-4with a negligible performance loss compared to the infinite-resolution case at the cost of additional out-of-band emissions. Furthermore, our results highlight the importance of considering the inherent spatial and temporal correlations caused by low-precision DACs.

Future wireless communication systems will rely on large antenna arrays at the infrastructure base stations (BSs) to serve multiple users with high data rates in a single cell. We demonstrate that the availability of high-dimensional channel state information (CSI) acquired at such multi-antenna BSs enables one to learn a chart of the radio geometry, which captures the spatial geometry of the users so that points close in space are close in the channel chart, using no other information than wireless channels of users. Specifically, we propose a novel unsupervised framework that first extracts channel features from CSI which characterize large-scale fading effects of the channel, and then uses specialized dimensionality reduction tools to construct the channel chart. The channel chart can, for example, be used to perform (relative) user localization, predict cell hand-overs, or guide scheduling tasks, without accessing location information from global navigation satellite systems.

In this paper, we solve the sum mean-squared error (MSE)-optimal 1-bit quantized precoding problem exactly for small-to-moderate sized multiuser multiple-input multiple-output (MU-MIMO) systems via branch and bound. To this end, we reformulate the original NP-hard precoding problem as a tree search and deploy a number of strategies that improve the pruning efficiency without sacrificing optimality. We evaluate the error-rate performance and the complexity of the resulting 1-bit branch-and-bound (BB-1) precoder, and compare its efficacy to that of existing, suboptimal algorithms for 1-bit precoding in MU-MIMO systems.

Phase retrieval refers to the problem of recovering real- or complex-valued vectors from magnitude measurements. The best-known algorithms for this problem are iterative in nature and rely on so-called spectral initializers that provide accurate initialization vectors. We propose a novel class of estimators suitable for general nonlinear measurement systems, called linear spectral estimators (LSPEs), which can be used to compute accurate initialization vectors for phase retrieval problems. The proposed LSPEs not only provide accurate initialization vectors for noisy phase retrieval systems with structured or random measurement matrices, but also enable the derivation of sharp and nonasymptotic mean-squared error bounds. We demonstrate the efficacy of LSPEs on synthetic and real-world phase retrieval problems, and we show that our estimators significantly outperform existing methods for structured measurement systems that arise in practice.

The Rasch model is widely used for item response analysis in applications ranging from recommender systems to psychology, education, and finance. While a number of estimators have been proposed for the Rasch model over the last decades, the associated analytical performance guarantees are mostly asymptotic. This paper provides a framework that relies on a novel linear minimum mean-squared error (L-MMSE) estimator which enables an exact, nonasymptotic, and closed-form analysis of the parameter estimation error under the Rasch model. The proposed framework provides guidelines on the number of items and responses required to attain low estimation errors in tests or surveys. We furthermore demonstrate its efficacy on a number of real-world collaborative filtering datasets, which reveals that the proposed L-MMSE estimator performs on par with state-of-the-art nonlinear estimators in terms of predictive performance.

Fifth-generation (5G) cellular systems will build on massive multi-user (MU) multiple-input multiple-output (MIMO) technology to attain high spectral efficiency. However, having hundreds of antennas and radio-frequency (RF) chains at the base station (BS) entails prohibitively high hardware costs and power consumption. This paper proposes a novel nonlinear precoding algorithm for the massive MU-MIMO downlink in which each RF chain contains an 8-phase (3-bit) constant-modulus transmitter, enabling the use of low-cost and power-efficient analog hardware. We present a high-throughput VLSI architecture and show implementation results on a Xilinx Virtex-7 FPGA. Compared to a recently-reported nonlinear precoder for BS designs that use two 1-bit digital-to-analog converters per RF chain, our design enables up to 3.75 dB transmit power reduction at no more than a 2.7× increase in FPGA resources.

Neural network training relies on our ability to find "good" minimizers of highly non-convex loss functions. It is well known that certain network architecture designs (e.g., skip connections) produce loss functions that train easier, and well-chosen training parameters (batch size, learning rate, optimizer) produce minimizers that generalize better. However, the reasons for these differences, and their effect on the underlying loss landscape, is not well understood. In this paper, we explore the structure of neural loss functions, and the effect of loss landscapes on generalization, using a range of visualization methods. First, we introduce a simple "filter normalization" method that helps us visualize loss function curvature, and make meaningful side-by-side comparisons between loss functions. Then, using a variety of visualizations, we explore how network architecture affects the loss landscape, and how training parameters affect the shape of minimizers.

Data poisoning is an attack on machine learning models wherein the attacker adds examples to the training set to manipulate the behavior of the model at test time. This paper explores poisoning attacks on neural nets. The proposed attacks use ``clean-labels''; they don't require the attacker to have any control over the labeling of training data. They are also targeted; they control the behavior of the classifier on a specific test instance without degrading overall classifier performance. For example, an attacker could add a seemingly innocuous image (that is properly labeled) to a training set for a face recognition engine, and control the identity of a chosen person at test time. Because the attacker does not need to control the labeling function, poisons could be entered into the training set simply by putting them online and waiting for them to be scraped by a data collection bot. We present an optimization-based method for crafting poisons, and show that just one single poison image can control classifier behavior when transfer learning is used. For full end-to-end training, we present a ``watermarking'' strategy that makes poisoning reliable using multiple (approx. 50) poisoned training instances. We demonstrate our method by generating poisoned frog images from the CIFAR dataset and using them to manipulate image classifiers.

To perform multi-user multiple-input and multipleoutput transmission in millimeter-wave (mmWave) cellular systems, the high-dimensional channels need to be estimated for designing the multi-user precoder. Conventional grid-based Compressed Sensing (CS) methods for mmWave channel estimation suffer from the basis mismatch problem, which prevents accurate channel reconstruction and degrades the precoding performance. This paper formulates mmWave channel estimation as an Atomic Norm Minimization (ANM) problem. In contrast to grid-based CS methods which use discrete dictionaries, ANM uses a continuous dictionary for representing the mmWave channel. We consider a continuous dictionary based on sub-sampling in the antenna domain via a small number of radio frequency chains. We show that mmWave channel estimation using ANM can be formulated as a Semidefinite Programming (SDP) problem, and the channel can be accurately estimated via off-the-shelf SDP solvers in polynomial time. Simulation results indicate that ANM can achieve much better estimation accuracy compared to grid-based CS, and significantly improves the spectral efficiency provided by multi-user precoding.

We consider the recovery of a (real- or complex-valued) signal from magnitude-only measurements, known as phase retrieval. We formulate phase retrieval as a convex optimization problem, which we call PhaseMax. Unlike other convex methods that use semidefinite relaxation and lift the phase retrieval problem to a higher dimension, PhaseMax is a “non-lifting” relaxation that operates in the original signal dimension. We show that the dual problem to PhaseMax is basis pursuit, which implies that the phase retrieval can be performed using algorithms initially designed for sparse signal recovery. We develop sharp lower bounds on the success probability of PhaseMax for a broad range of random measurement ensembles, and we analyze the impact of measurement noise on the solution accuracy. We use numerical results to demonstrate the accuracy of our recovery guarantees, and we showcase the efficacy and limits of PhaseMax in practice.

Generation and evaluation of crowdsourced content is commonly treated as two separate processes, performed at different times and by two distinct groups of people: content creators and content assessors. As a result, most crowdsourcing tasks follow this template: one group of workers generates content and another group of workers evaluates it. In an educational setting, for example, content creators are traditionally students that submit open-response answers to assignments (e.g., a short answer, a circuit diagram, or a formula) and content assessors are instructors that grade these submissions. Despite the considerable success of peer-grading in massive open online courses (MOOCs), the process of test-taking and grading are still treated as two distinct tasks which typically occur at different times, and require an additional overhead of grader training and incentivization. Inspired by this problem in the context of education, we propose a general crowdsourcing framework that fuses open-response test-taking (content generation) and assessment into a single, streamlined process that appears to students in the form of an explicit test, but where everyone also acts as an implicit grader. The advantages offered by our framework include: a common incentive mechanism for both the creation and evaluation of content, and a probabilistic model that jointly models the processes of contribution and evaluation, facilitating efficient estimation of the quality of the contributions and the competency of the contributors. We demonstrate the effectiveness and limits of our framework via simulations and a real-world user study.

Many modern computer vision and machine learning applications rely on solving difficult optimization problems that involve non-differentiable objective functions and constraints. The alternating direction method of multipliers (ADMM) is a widely used approach to solve such problems. Relaxed ADMM is a generalization of ADMM that often achieves better performance, but its efficiency depends strongly on algorithm parameters that must be chosen by an expert user. We propose an adaptive method that automatically tunes the key algorithm parameters to achieve optimal performance without user oversight. Inspired by recent work on adaptivity, the proposed adaptive relaxed ADMM (ARADMM) is derived by assuming a Barzilai-Borwein style linear gradient. A detailed convergence analysis of ARADMM is provided, and numerical results on several applications demonstrate fast practical convergence.

Massive multiuser (MU) multiple-input multiple- output (MIMO) is foreseen to be a key technology in future wireless communication systems. In this paper, we analyze the downlink performance of an orthogonal frequency division multiplexing (OFDM)-based massive MU-MIMO system in which the base station (BS) is equipped with 1-bit digital-to-analog converters (DACs). Using Bussgang's theorem, we characterize the performance achievable with linear precoders (such as maximal-ratio transmission and zero forcing) in terms of bit error rate (BER). Our analysis accounts for the possibility of oversampling the time-domain transmit signal before the DACs. We further develop a lower bound on the information-theoretic sum-rate throughput achievable with Gaussian inputs. Our results suggest that the performance achievable with 1-bit DACs in a massive MU-MIMO- OFDM downlink are satisfactory provided that the number of BS antennas is sufficiently large.

One-bit quantization at the base-station (BS) of a massive multiple-input multiple-output (MIMO) wireless system enables significant power and cost savings. While the 1-bit uplink (users communicate to BS) has gained significant attention, the downlink (BS transmits to users) is far less studied. In this paper, we propose a novel, computationally-efficient 1-bit beamforming algorithm referred to as POKEMON (short for PrOjected downlinK bEaMfOrmiNg), which-after its convolution with the MIMO channel matrix-minimizes multi-user interference (a.k.a. spatial leakage). Our algorithm builds upon the biconvex relaxation (BCR) framework, which efficiently approximates the optimal 1-bit beamforming problem that is of combinatorial nature. Our simulation results show that POKEMON significantly outperforms linear beamformers followed by 1-bit quantization in terms of error-rate performance and recent non-linear beamformers in terms of complexity.

Semidefinite relaxation methods transform a variety of non-convex optimization problems into convex problems, but square the number of variables. We study a new type of convex relaxation for phase retrieval problems, called PhaseMax, that convexifies the underlying problem without lifting. The resulting problem formulation can be solved using standard convex optimization routines, while still working in the original, low-dimensional variable space. We prove, using a random spherical distribution measurement model, that PhaseMax succeeds with high probability for a sufficiently large number of measurements. We compare our approach to other phase retrieval methods and demonstrate that our theory accurately predicts the success of PhaseMax.

Joint channel estimation and data detection (JED) enables near-optimal error-rate performance in realistic wireless communication systems that suffer from channel estimation errors. In this paper, we propose a new JED algorithm and a corresponding FPGA design for large single-input multiple-output (SIMO) wireless systems that use constant-modulus constellations. Our algorithm, referred to as PrOX (short for PRojection Onto conveX hull), relies on biconvex relaxation (BCR) in order to efficiently compute an approximate solution of the maximum-likelihood JED problem that exhibits prohibitive complexity. PrOX is a simple and hardware-friendly algorithm that achieves near-optimal error-rate performance for a wide-range of system configurations. To demonstrate the efficacy of PrOX, we develop a scalable VLSI architecture and present reference implementation results on a Xilinx Virtex-7 FPGA. Compared to a recently-reported reference JED design, PrOX achieves 3 × higher throughput, 20 × better hardware-efficiency (in terms of throughput per look-up tables), and 8 × improved energy-efficiency.

We propose a novel data detection algorithm and a corresponding VLSI design for large multi-user (MU) multiple-input multiple-output (MIMO) wireless receiver. Our algorithm, referred to as ADMIN, performs alternating direction method of multipliers (ADMM)-based infinity norm constrained equalization. ADMIN is an iterative algorithm that outperforms linear detectors if the number of users is small compared to that of the antennas in base station (BS). ADMIN computes the linear minimum mean-square error (MMSE) solution in the first iteration. It is sufficient when the ratio between the numbers of BS antennas and users is rather large. We develop a time-shared and iterative VLSI architecture for LDL-decomposition based soft-output ADMIN. Our architecture achieves 685.71 Mb/s for linear MMSE and 212.38 Mb/s for ADMIN for a 16-user system that employs 64-QAM in a 28 nm CMOS technology.

This paper deals with linear equalization in massive multi-user multiple-input multiple-output (MU-MIMO) wireless systems. We first provide simple conditions on the antenna configuration for which the well-known linear minimum mean-square error (L-MMSE) equalizer provides near-optimal spectral efficiency, and we analyze its performance in the presence of parameter mismatches in the signal and/or noise powers. We then propose a novel, optimally-tuned NOnParametric Equalizer (NOPE) for massive MU-MIMO systems, which avoids knowledge of the transmit signal and noise powers altogether. We show that NOPE achieves the same performance as that of the L-MMSE equalizer in the large-antenna limit, and we demonstrate its efficacy in realistic, finite-dimensional systems. From a practical perspective, NOPE is computationally efficient and avoids dedicated training that is typically required for parameter estimation.

Massive multi-user (MU) multiple-input multiple-output (MIMO) promises significant gains in spectral efficiency compared to traditional, small-scale MIMO technology. Linear equalization algorithms, such as zero forcing (ZF) or minimum mean-square error (MMSE)-based methods, typically rely on centralized processing at the base station (BS), which results in (i) excessively high interconnect and chip input/output data rates, and (ii) high computational complexity. In this paper, we investigate the achievable rates of decentralized equalization that mitigates both of these issues. We consider two distinct BS architectures that partition the antenna array into clusters, each associated with independent radio-frequency chains and signal processing hardware, and the results of each cluster are fused in a feed forward network. For both architectures, we consider ZF, MMSE, and a novel, non-linear equalization algorithm that builds upon approximate message passing (AMP), and we theoretically analyze the achievable rates of these methods. Our results demonstrate that decentralized equalization with our AMP-based methods incurs no or only a negligible loss in terms of achievable rates compared to that of centralized solutions.

Currently, deep neural networks are deployed on low-power portable devices by first training a full-precision model using powerful hardware, and then deriving a corresponding low-precision model for efficient inference on such systems. However, training models directly with coarsely quantized weights is a key step towards learning on embedded platforms that have limited computing resources, memory capacity, and power consumption. Numerous recent publications have studied methods for training quantized networks, but these studies have mostly been empirical. In this work, we investigate training methods for quantized neural networks from a theoretical viewpoint. We first explore accuracy guarantees for training methods under convexity assumptions. We then look at the behavior of these algorithms for non-convex problems, and show that training algorithms that exploit high-precision representations have an important greedy search phase that purely quantized training methods lack, which explains the difficulty of training using low-precision arithmetic.

Massive multiuser (MU) multiple-input multiple-output (MIMO) is foreseen to be one of the key technologies in fifth-generation wireless communication systems. In this paper, we investigate the problem of downlink precoding for a narrowband massive MU-MIMO system with low-resolution digital-to-analog converters (DACs) at the base station (BS). We analyze the performance of linear precoders, such as maximal-ratio transmission and zero-forcing, subject to coarse quantization. Using Bussgang's theorem, we derive a closed-form approximation on the rate achievable under such coarse quantization. Our results reveal that the performance attainable with infinite-resolution DACs can be approached using DACs having only 3-4 bits of resolution, depending on the number of BS antennas and the number of user equipments (UEs). For the case of 1-bit DACs, we also propose novel nonlinear precoding algorithms that significantly outperform linear precoders at the cost of an increased computational complexity. Specifically, we show that nonlinear precoding incurs only a 3 dB penalty compared with the infinite-resolution case for an uncoded bit-error rate of 10-3, in a system with 128 BS antennas that uses 1-bit DACs and serves 16 single-antenna UEs. In contrast, the penalty for linear precoders is about 8dB.

We investigate the uplink throughput achievable by a multiple-user (MU) massive multiple-input multiple-output (MIMO) system, in which the base station is equipped with a large number of low-resolution analog-to-digital converters (ADCs). Our focus is on the case where neither the transmitter nor the receiver have any a priori channel state information. This implies that the fading realizations have to be learned through pilot transmission followed by channel estimation at the receiver, based on coarsely quantized observations. We propose a novel channel estimator, based on Bussgang's decomposition, and a novel approximation to the rate achievable with finite-resolution ADCs, both for the case of finite-cardinality constellations and of Gaussian inputs, that is accurate for a broad range of system parameters. Through numerical results, we illustrate that, for the 1-bit quantized case, pilot-based channel estimation together with maximal-ratio combing, or zero-forcing detection enables reliable multi-user communication with high-order constellations, in spite of the severe nonlinearity introduced by the ADCs. Furthermore, we show that the rate achievable in the infinite-resolution (no quantization) case can be approached using ADCs with only a few bits of resolution. We finally investigate the robustness of low-ADC-resolution MU-MIMO uplink against receive power imbalances between the different users, caused for example by imperfect power control.

Conventional algorithms for sparse signal recovery and sparse representation rely on l1-norm regularized variational methods. However, when applied to the reconstruction of sparse images, i.e., images where only a few pixels are non-zero, simple l1-norm-based methods ignore potential correlations in the support between adjacent pixels. In a number of applications, one is interested in images that are not only sparse, but also have a support with smooth (or contiguous) boundaries. Existing algorithms that take into account such a support structure mostly rely on nonconvex methods and-as a consequence-do not scale well to high-dimensional problems and/or do not converge to global optima. In this paper, we explore the use of new block l1-norm regularizers, which enforce image sparsity while simultaneously promoting smooth support structure. By exploiting the convexity of our regularizers, we develop new computationally-efficient recovery algorithms that guarantee global optimality. We demonstrate the efficacy of our regularizers on a variety of imaging tasks including compressive image recovery, image restoration, and robust PCA.

Statistical models of student responses on assessment questions, such as those in homeworks and exams, enable educators and computer-based personalized learning systems to gain insights into students’ knowledge using machine learning. Popular student-response models, including the Rasch model and item response theory models, represent the probability of a student answering a question correctly using an affine function of latent factors. While such models can accurately predict student responses, their ability to interpret the underlying knowledge structure (which is certainly nonlinear) is limited. In response, we develop a new, nonlinear latent variable model that we call the dealbreaker model, in which a student’s success probability is determined by their weakest concept mastery. We develop efficient parameter inference algorithms for this model using novel methods for nonconvex optimization. We show that the dealbreaker model achieves comparable or better prediction performance as compared to affine models with real-world educational datasets. We further demonstrate that the parameters learned by the dealbreaker model are interpretable—they provide key insights into which concepts are critical (i.e., the “dealbreaker”) to answering a question correctly. We conclude by reporting preliminary results for a movie-rating dataset, which illustrate the broader applicability of the dealbreaker model.

We propose a novel, near-optimal data detection algorithm and a corresponding FPGA design for large multiple-input multiple-output (MIMO) wireless systems. Our algorithm, referred to as TASER (short for triangular approximate semidefinite relaxation), relaxes the maximum-likelihood (ML) detection problem to a semidefinite program and solves a non-convex approximation using a preconditioned forward-backward splitting procedure. We show that TASER achieves near-ML performance at low computational complexity, even for large-dimensional MIMO systems. We develop a systolic array that implements TASER and achieves high throughput at low hardware complexity. To demonstrate the effectiveness of our solution, we develop reference designs on a Xilinx Virtex-7 FPGA for various antenna configurations. One of our TASER designs achieves up to 98 Mb/s for a 32-user system that employs QPSK, while consuming only 150 k FPGA look-up tables.

We propose a new, low-complexity data-detection algorithm and a corresponding high-throughput FPGA design for 3GPP LTE-based large-scale (or massive) multi-user (MU) multiple-input multiple-output (MIMO) wireless communication systems. Our algorithm performs approximate minimum mean-square error (MMSE) data detection using coordinate descent (CD), which enables near-MMSE performance at low computational complexity, even for systems with hundreds of antennas at the base station (BS). We design a high-throughput VLSI architecture for 3GPP LTE wideband systems with a deep and interleaved pipeline, which can be parametrized at design time to support various antenna configurations. Our CD-based data detector achieves 379Mb/s throughout, while using 24 k LUTs and 771 DSP units on a Xilinx Virtex-7 FPGA for a 128 BS antenna, 8 user large-scale MU-MIMO system.

We investigate the performance of mismatched data detection in large multiple-input multiple-output (MIMO) systems, where the prior distribution of the transmit signal used in the data detector differs from the true prior. To minimize the performance loss caused by this prior mismatch, we include a tuning stage into our recently-proposed large MIMO approximate message passing (LAMA) algorithm, which allows us to develop mismatched LAMA algorithms with optimal as well as sub-optimal tuning. We show that carefully-selected priors often enable simpler and computationally more efficient algorithms compared to LAMA with the true prior while achieving near-optimal performance. A performance analysis of our algorithms for a Gaussian prior and a uniform prior within a hypercube covering the QAM constellation recovers classical and recent results on linear and non-linear MIMO data detection, respectively.

Difficult multiple-choice (MC) questions can be made easy by providing a set of answer options of which most are obviously wrong. In the education literature, a plethora of instructional guides exist for crafting a suitable set of wrong choices (distractors) that enable the assessment of the students' understanding. The art of MC question design thus hinges on the question-maker's experience and knowledge of the potential misconceptions. In contrast, we advocate a data-driven approach, where correct and incorrect options are assembled directly from the students' own past submissions. Large-scale online classroom settings, such as massively open online courses (MOOCs), provide an opportunity to design optimal and adaptive multiple-choice questions that are maximally informative about the students' level of understanding of the material. In this work, we (i) develop a multinomial-logit discrete choice model for the setting of MC testing, (ii) derive an optimization objective for selecting optimally discriminative option sets, (iii) propose an algorithm for finding a globally-optimal solution, and (iv) demonstrate the effectiveness of our approach via synthetic experiments and a user study. We finally showcase an application of our approach to crowd-sourcing tests from technical online forums.

Difficult test questions can be made easy by providing a set of possible answer options of which most are obviously wrong. In the education literature, a plethora of instructional guides exist for crafting a suitable set of wrong choices (distractors) in order to probe the students' understanding of the tested concept. The art of multiple-choice question design thus hinges on the question-maker's experience and knowledge of the potential misconceptions. In contrast, we advocate a data-driven approach, where correct and incorrect options are assembled directly from the students' own past submissions. Large-scale online classroom settings, such as massively open online courses (MOOCs), provide an opportunity to design optimal and adaptive multiple-choice questions that are maximally informative about the students' level of understanding of the material. We deploy a multinomial-logit discrete choice model for the setting of multiple choice testing, derive an optimization objective for selecting optimally discriminative option sets, and demonstrate the effectiveness of our approach via a user study.

Coarse quantization at the base station (BS) of a massive multi-user (MU) multiple-input multiple-output (MIMO) wireless system promises significant power and cost savings. Coarse quantization also enables significant reductions of the raw analog-to-digital converter data that must be transferred from a spatially separated antenna array to the baseband processing unit. The theoretical limits as well as practical transceiver algorithms for such quantized MU-MIMO systems operating over frequency-flat, narrowband channels have been studied extensively. However, the practically relevant scenario where such communication systems operate over frequency-selective, wideband channels is less well understood. This paper investigates the uplink performance of a quantized massive MU-MIMO system that deploys orthogonal frequency-division multiplexing (OFDM) for wideband communication. We propose new algorithms for quantized maximum a posteriori channel estimation and data detection, and we study the associated performance/quantization tradeoffs. Our results demonstrate that coarse quantization (e.g., four to six bits, depending on the ratio between the number of BS antennas and the number of users) in massive MU-MIMO-OFDM systems entails virtually no performance loss compared with the infinite-precision case at no additional cost in terms of baseband processing complexity.

We propose an FPGA design for soft-output data detection in orthogonal frequency-division multiplexing (OFDM)-based large-scale (multi-user) MIMO systems. To reduce the high computational complexity of data detection, our design uses a modified version of the conjugate gradient least square (CGLS) algorithm. In contrast to existing linear detection algorithms for massive MIMO systems, our method avoids two of the most complex tasks, namely Gram-matrix computation and matrix inversion, while still being able to compute soft-outputs. Our architecture uses an array of reconfigurable processing elements to compute the CGLS algorithm in a hardware-efficient manner. Implementation results on Xilinx Virtex-7 FPGA for a 128 antenna, 8 user large-scale MIMO system show that our design only uses 70% of the area-delay product of the competitive method, while exhibiting superior error-rate performance.

Optimal data detection in multiple-input multiple-output (MIMO) communication systems with a large number of antennas at both ends of the wireless link entails prohibitive computational complexity. In order to reduce the computational complexity, a variety of sub-optimal detection algorithms have been proposed in the literature. In this paper, we analyze the optimality of a novel data-detection method for large MIMO systems that relies on approximate message passing (AMP). We show that our algorithm, referred to as individually-optimal (IO) large-MIMO AMP (short IO-LAMA), is able to perform IO data detection given certain conditions on the MIMO system and the constellation set (e.g., QAM or PSK) are met.

Spatial multiplexing cameras (SMCs) acquire a (typically static) scene through a series of coded projections using a spatial light modulator (e.g., a digital micromirror device) and a few optical sensors. This approach finds use in imaging applications where full-frame sensors are either too expensive (e.g., for short-wave infrared wavelengths) or unavailable. Existing SMC systems reconstruct static scenes using techniques from compressive sensing (CS). For videos, however, existing acquisition and recovery methods deliver poor quality. In this paper, we propose the CS multiscale video (CS-MUVI) sensing and recovery framework for high-quality video acquisition and recovery using SMCs. Our framework features novel sensing matrices that enable the efficient computation of a low-resolution video preview, while enabling high-resolution video recovery using convex optimization. To further improve the quality of the reconstructed videos, we extract optical-flow estimates from the low-resolution previews and impose them as constraints in the recovery procedure. We demonstrate the efficacy of our CS-MUVI framework for a host of synthetic and real measured SMC video data, and we show that high-quality videos can be recovered at roughly $60\times$ compression.

Massive multiple-input multiple-output (MIMO) promises improved spectral efficiency, coverage, and range, compared to conventional (small-scale) MIMO wireless systems. Unfortunately, these benefits come at the cost of significantly increased computational complexity, especially for systems with realistic antenna configurations. To reduce the complexity of data detection (in the uplink) and precoding (in the downlink) in massive MIMO systems, we propose to use conjugate gradient (CG) methods. While precoding using CG is rather straightforward, soft-output minimum mean-square error (MMSE) detection requires the computation of the post-equalization signal-to-interference-and-noise-ratio (SINR). To enable CG for soft-output detection, we propose a novel way of computing the SINR directly within the CG algorithm at low complexity. We investigate the performance/complexity trade-offs associated with CG-based soft-output detection and precoding, and we compare it to existing exact and approximate methods. Our results reveal that the proposed algorithm is able to outperform existing methods for massive MIMO systems with realistic antenna configurations.

This paper deals with the recovery of an unknown, low-rank matrix from quantized and (possibly) corrupted measurements of a subset of its entries. We develop statistical models and corresponding (multi-)convex optimization algorithms for quantized matrix completion (Q-MC) and quantized robust principal component analysis (Q-RPCA). In order to take into account the quantized nature of the available data, we jointly learn the underlying quantization bin boundaries and recover the low-rank matrix, while removing potential (sparse) corruptions. Experimental results on synthetic and two real-world collaborative filtering datasets demonstrate that directly operating with the quantized measurements - rather than treating them as real values - results in (often significantly) lower recovery error if the number of quantization bins is less than about 10.

This paper proposes - to the best of our knowledge - the first ASIC design for high-throughput data detection in single carrier frequency division multiple access (SC-FDMA)-based large-scale MIMO systems, such as systems building on future 3GPP LTE-Advanced standards. In order to substantially reduce the complexity of linear soft-output data detection in systems having hundreds of antennas at the base station (BS), the proposed detector builds upon a truncated Neumann series expansion to compute the necessary matrix inverse at low complexity. To achieve high throughput in the 3GPP LTE-A uplink, we develop a systolic VLSI architecture including all necessary processing blocks. We present a corresponding ASIC design that achieves 3.8 Gb/s for a 128 antenna, 8 user 3GPP LTE-A based large-scale MIMO system, while occupying 11.1 mm2in a TSMC 45nm CMOS technology.

We propose SPARFA-Trace, a new machine learning-based framework for time-varying learning and content analytics for educational applications. We develop a novel message passing-based, blind, approximate Kalman filter for sparse factor analysis (SPARFA) that jointly traces learner concept knowledge over time, analyzes learner concept knowledge state transitions (induced by interacting with learning resources, such as textbook sections, lecture videos, etc., or the forgetting effect), and estimates the content organization and difficulty of the questions in assessments. These quantities are estimated solely from binary-valued (correct/incorrect) graded learner response data and the specific actions each learner performs (e.g., answering a question or studying a learning resource) at each time instant. Experimental results on two online course datasets demonstrate that SPARFA-Trace is capable of tracing each learner's concept knowledge evolution over time, analyzing the quality and content organization of learning resources, and estimating the question--concept associations and the question difficulties. Moreover, we show that SPARFA-Trace achieves comparable or better performance in predicting unobserved learner responses compared to existing collaborative filtering and knowledge tracing methods.

Unions of subspaces have recently been shown to provide a compact nonlinear signal model for collections of high-dimensional data, such as large collections of images or videos. In this paper, we introduce a novel data-driven algorithm for learning unions of subspaces directly from a collection of data; our approach is based upon forming minimum ℓ2-norm (least-squares) representations of a signal with respect to other signals in the collection. The resulting representations are then used as feature vectors to cluster the data in accordance with each signal's subspace membership. We demonstrate that the proposed least-squares approach leads to improved classification performance when compared to state-of-the-art subspace clustering methods on both synthetic and real-world experiments. This study provides evidence that using least-squares methods to form data-driven representations of collections of data provide significant advantages over current methods that rely upon sparse representations.

In this paper, we present a novel algorithm to learn phase-invariant dictionaries, which can be used to efficiently approximate a variety of signals, such as audio signals or images. Our approach relies on finding a small number of generating atoms that can be used-along with their phase-shifts-to sparsely approximate a given signal. Our method is inspired by the K-SVD algorithm, but imposes an extra constraint that the dictionaries we learn are phase-invariant. We show that the learned dictionaries achieve competitive approximation performance compared to that of state-of-the-art methods for audio signals and images, while substantially reducing the storage requirements and computational complexity.

Light curtains are presence detection devices, typically employed to detect when an object enters (or passes through) a region to initiate emergency security procedures. However, most existing light curtain implementations are not designed to detect the precise location where the barrier was broken. In this paper, we present a hardware prototype implementation that is able to identify the locations where the light curtain was broken, even if multiple objects penetrate the barrier simultaneously. To this end, we deploy techniques from sparse signal recovery and compressive sensing to perform detection and localization with only a few infrared transmitters and sensors. The proposed prototype implementation is scalable to large physical sizes and can be tailored to suit a particular application in terms of the number and size of objects to detect, as well as the desired spatial resolution.

We develop a new model and algorithm for machine learning-based learning analytics, which estimate a learner's knowledge of the concepts underlying a domain. Our model represents the probability that a learner provides the correct response to a question in terms of three factors: their understanding of a set of underlying concepts, the concepts involved in each question, and each question's intrinsic difficulty. We estimate these factors given the graded responses to a set of questions. We develop a bi-convex algorithm to solve the resulting SPARse Factor Analysis (SPARFA) problem. We also incorporate user-defined tags on questions to facilitate the interpretability of the estimated factors. Experiments with synthetic and real-world data demonstrate the efficacy of our approach.

In this paper, we analyze the VLSI implementation tradeoffs for linear data detection in the uplink of large-scale multiple-input multiple-output (MIMO) wireless systems. Specifically, we analyze the error incurred by using the sub-optimal, low-complexity matrix inverse proposed in Wu et al., 2013, ISCAS, and compare its performance and complexity to an exact matrix inversion algorithm. We propose a Cholesky-based reference architecture for exact matrix inversion and show corresponding implementation results on an Virtex-7 FPGA. Using this reference design, we perform a performance/complexity trade-off comparison with an FPGA implementation for the proposed approximate matrix inversion, which reveals that the inversion circuit of choice is determined by the antenna configuration (base-station antennas vs. number of users) of large-scale MIMO systems.

We demonstrate the restoration of audio signals corrupted by clicks and pops using techniques from sparse signal recovery and compressive sensing. The demonstration features real-time signal restoration using the approximate message passing algorithm on an FPGA prototyping board. To highlight the restoration performance of our implementation, we remove clicks and pops from old phonograph recordings in real time.

The high processing complexity of data detection in the large-scale multiple-input multiple-output (MIMO) uplink necessitates high-throughput VLSI implementations. In this paper, we propose - to the best of our knowledge - first matrix inversion implementation suitable for data detection in systems having hundreds of antennas at the base station (BS). The underlying idea is to carry out an approximate matrix inversion using a small number of Neumann-series terms, which allows one to achieve near-optimal performance at low complexity. We propose a novel VLSI architecture to efficiently compute the approximate inverse using a systolic array and show reference FPGA implementation results for various system configurations. For a system where 128 BS antennas receive data from 8 single-antenna users, a single instance of our design processes 1.9M matrices/s on a Xilinx Virtex-7 FPGA, while using only 3.9% of the available slices and 3.6% of the available DSP48 units.

We investigate an orthogonal frequency-division multiplexing (OFDM)-based downlink transmission scheme for large-scale multi-user (MU) multiple-input multiple-output (MIMO) wireless systems. The use of OFDM causes a high peak-to-average (power) ratio (PAR), which necessitates expensive and power-inefficient radio-frequency (RF) components at the base station. In this paper, we present a novel downlink transmission scheme, which exploits the massive degrees-of-freedom available in large-scale MU-MIMO-OFDM systems to achieve low PAR. Specifically, we propose to jointly perform MU precoding, OFDM modulation, and PAR reduction by solving a convex optimization problem. We develop a corresponding fast iterative truncation algorithm (FITRA) and show numerical results to demonstrate tremendous PAR-reduction capabilities. The significantly reduced linearity requirements eventually enable the use of low-cost RF components for the large-scale MU-MIMO-OFDM downlink.

We consider the recovery of sparse signals subject to sparse interference, as introduced by Studer , IEEE T-IT, 2012. We present novel probabilistic recovery guarantees for this framework, covering varying degrees of knowledge of the signal and interference support, which are relevant for a large number of practical applications. Our results assume that the sparsifying dictionaries are characterized by coherence parameters and we require randomness only in the signal and/or interference. The obtained recovery guarantees show that one can recover sparsely corrupted signals with overwhelming probability, even if the sparsity of both the signal and interference scale (near) linearly with the number of measurements.

This paper deals with the recovery of signals that admit an approximately sparse representation in some known dictionary (possibly over-complete) and are corrupted by additive noise. In particular, we consider additive measurement noise with bounded ℓp-norm for p ≥ 2, and we minimize the ℓqquasi-norm (with q ∈ (0, 1]) of the signal vector. We develop coherence-based recovery guarantees for which stable recovery via generalized basis-pursuit de-quantizing (BPDQp,q) is possible. We finally show that depending on the measurement-noise model and the choice of the ℓp-norm used in the constraint, (BPDQp,q) significantly outperforms classical basis pursuit de-noising (BPDN).

In this paper, we investigate dictionary learning (DL) from sparsely corrupted or compressed signals. We consider three cases: I) the training signals are corrupted, and the locations of the corruptions are known, II) the locations of the sparse corruptions are unknown, and III) DL from compressed measurements, as it occurs in blind compressive sensing. We develop two efficient DL algorithms that are capable of learning dictionaries from sparsely corrupted or compressed measurements. Empirical phase transitions and an in-painting example demonstrate the capabilities of our algorithms.

Compressive sensing (CS)-based spatial-multiplexing cameras (SMCs) sample a scene through a series of coded projections using a spatial light modulator and a few optical sensor elements. SMC architectures are particularly useful when imaging at wavelengths for which full-frame sensors are too cumbersome or expensive. While existing recovery algorithms for SMCs perform well for static images, they typically fail for time-varying scenes (videos). In this paper, we propose a novel CS multi-scale video (CS-MUVI) sensing and recovery framework for SMCs. Our framework features a co-designed video CS sensing matrix and recovery algorithm that provide an efficiently computable low-resolution video preview. We estimate the scene's optical flow from the video preview and feed it into a convex-optimization algorithm to recover the high-resolution video. We demonstrate the performance and capabilities of the CS-MUVI framework for different scenes.

We formulate a unified framework for the separation of signals that are sparse in “morphologically” different redundant dictionaries. This formulation incorporates the so-called “analysis” and “synthesis” approaches as special cases and contains novel hybrid setups. We find corresponding coherence-based recovery guarantees for an ℓ1-norm based separation algorithm. Our results recover those reported in Studer and Baraniuk, ACHA, submitted, for the synthesis setting, provide new recovery guarantees for the analysis setting, and form a basis for comparing performance in the analysis and synthesis settings. As an aside our findings complement the D-RIP recovery results reported in Candès et al., ACHA, 2011, for the “analysis” signal recovery problem minimizex||Ψx̃||1subject to ||y - Ax̃||2≤ ϵ by delivering corresponding coherence-based recovery results.

In this paper, we present novel probabilistic recovery guarantees for sparse signals subject to sparse interference, covering varying degrees of knowledge of the signal and interference support. Our results assume that the sparsifying dictionaries are characterized by coherence parameters and we require randomness only in the signal and/or interference. The obtained recovery guarantees show that one can recover sparsely corrupted signals with overwhelming probability, even if the sparsity of both the signal and interference scale (near) linearly with the number of measurements.

We investigate the recovery of signals exhibiting a sparse representation in a general (i.e., possibly redundant or incomplete) dictionary that are corrupted by additive noise admitting a sparse representation in another general dictionary. This setup covers a wide range of applications, such as image inpainting, super-resolution, signal separation, and recovery of signals that are impaired by, e.g., clipping, impulse noise, or narrowband interference. We present deterministic recovery guarantees based on a novel uncertainty relation for pairs of general dictionaries and we provide corresponding practicable recovery algorithms. The recovery guarantees we find depend on the signal and noise sparsity levels, on the coherence parameters of the involved dictionaries, and on the amount of prior knowledge about the signal and noise support sets.

Low-density parity-check (LDPC) codes are key ingredients for improving reliability of modern communication systems and storage devices. On the implementation side however, the design of energy-efficient and high-speed LDPC decoders with a sufficient degree of reconfigurability to meet the flexibility demands of recent standards remains challenging. This survey paper provides an overview of the state-of-the-art in the design of LDPC decoders using digital integrated circuits. To this end, we summarize available algorithms and characterize the design space. We analyze the different architectures and their connection to different codes and requirements. The advantages and disadvantages of the various choices are illustrated by comparing state-of-the-art LDPC decoder designs.

We investigate the recovery of signals exhibiting a sparse representation in a general (i.e., possibly redundant or incomplete) dictionary that are corrupted by additive noise admitting a sparse representation in another general dictionary. This setup covers a wide range of applications, such as image inpainting, super-resolution, signal separation, and the recovery of signals that are corrupted by, e.g., clipping, impulse noise, or narrowband interference. We present deterministic recovery guarantees based on a recently developed uncertainty relation and provide corresponding recovery algorithms. The recovery guarantees we find depend on the signal and noise sparsity levels, on the coherence parameters of the involved dictionaries, and on the amount of prior knowledge on the support sets of signal and noise.

We analyze the (computational) complexity distribution of sphere decoding (SD) for random infinite lattices. In particular, we show that under fairly general assumptions on the statistics of the lattice basis matrix, the tail behavior of the SD complexity distribution is fully determined by the inverse volume of the fundamental regions of the underlying lattice. Particularizing this result to${N} \times {M},$${N} \geq {M}$, i.i.d. circularly symmetric complex Gaussian lattice basis matrices, we find that the corresponding complexity distribution is of Pareto-type with tail exponent given by${N}-{M}+1$. A more refined analysis reveals that the corresponding average complexity of SD is infinite for${N} = {M}$and finite for${N} > {M}$. Finally, for i.i.d. circularly symmetric complex Gaussian lattice basis matrices, we analyze SD preprocessing techniques based on lattice-reduction (such as the LLL algorithm or layer-sorting according to the V-BLAST algorithm) and regularization. In particular, we show that lattice-reduction does not improve the tail exponent of the complexity distribution while regularization results in a SD complexity distribution with tails that decrease faster than polynomial.

Lattice-reduction (LR)-aided successive interference cancellation (SIC) is able to achieve close-to optimum error-rate performance for data detection in multiple-input multiple-output (MIMO) wireless communication systems. In this work, we propose a hardware-efficient VLSI architecture of the Lenstra-Lenstra-Lovász (LLL) LR algorithm for SIC-based data detection. For this purpose, we introduce various algorithmic modifications that enable an efficient hardware implementation. Comparisons with existing FPGA implementations show that our design outperforms state-of-the-art LR implementations in terms of hardware-efficiency and throughput. We finally provide reference ASIC implementation results for 130nm CMOS technology.

The real-time acoustic camera prototype. consists of a field-programmable gate array (FPGA) base-board and an A/D-conversion board with 32 attached microphone modules arranged in a rectangular array. The FPGA base-board is connected to an external TFT display. Two loudspeakers will be connected to a laptop and play audio test-signals. The external display visualizes the location and intensity of the generated sound sources in real-time.

Acoustic cameras visualize the origin and intensity of sound waves using an array of microphones and sophisticated signal-processing algorithms. Due to the high memory-bandwidth and signal-processing complexity required by such algorithms, current available devices either compute the corresponding intensity-images off-line or require large and expensive hardware equipment. In this paper, we describe a low-complexity field-programmable gate array (FPGA)-based prototype, which computes and visualizes acoustic intensity-images in real-time. The system consists of 32 microphones and performs all signal processing tasks on a low-cost Xilinx Spartan 3E FPGA. The prototype computes intensity-images with a resolution of 320×240 pixels at 10 frames-per-second.

Sphere decoding (SD) is a promising means for implementing high-performance data detection in multiple-input multiple-output (MIMO) wireless communication systems. In this paper, we focus on the register transfer level implementation of SD with minimum area-delay product for application in wideband MIMO communication systems, such as IEEE 802.11n, where multiple SD cores need to be instantiated. The basic architectural considerations and the proposed optimizations are explained based on hard-output SD, but are also applicable to soft-output SD. Corresponding VLSI implementation results (for both hard-output and soft-output SD) show an improvement in the area-delay product by almost 50% compared to that of other SD implementations reported in the literature.

Soft-input soft-output (SISO) detection algorithms form the basis for iterative decoding. The computational complexity of SISO detection often poses significant challenges for practical receiver implementations, in particular in the context of multiple-input multiple-output (MIMO) wireless communication systems. In this paper, we present a low-complexity SISO sphere-decoding algorithm, based on the single tree-search paradigm proposed originally for soft-output MIMO detection in Studer (“Soft-output sphere decoding: Algorithms and VLSI implementation,” IEEE J. Sel. Areas Commun., vol. 26, no. 2, pp. 290-300, Feb. 2008). The new algorithm incorporates clipping of the extrinsic log-likelihood ratios (LLRs) into the tree-search, which results in significant complexity savings and allows to cover a large performance/complexity tradeoff region by adjusting a single parameter. Furthermore, we propose a new method for correcting approximate LLRs - resulting from sub-optimal detectors - which (often significantly) improves detection performance at low additional computational complexity.

We analyze the (computational) complexity distribution of sphere-decoding (SD) for random infinite lattices. In particular, we show that under fairly general assumptions on the statistics of the lattice basis matrix, the tail behavior of the SD complexity distribution is solely determined by the inverse volume of a fundamental region of the underlying lattice. Particularizing this result to N × M, N ¿ M, i.i.d. Gaussian lattice basis matrices, we find that the corresponding complexity distribution is of Pareto-type with tail exponent given by N - M + 1. We furthermore show that this tail exponent is not improved by lattice-reduction, which includes layer-sorting as a special case.

Beamforming (BF) improves the error rate performance of multiple-input multiple-output (MIMO) wireless communication systems by spatial separation of the transmitted data streams. Spatial separation is achieved by multiplication of the transmit vector by a steering matrix, which is obtained through the singular value decomposition (SVD) of the channel matrix. In this paper, we describe a hardware-efficient VLSI architecture for steering matrix computation using a hardware- optimized SVD algorithm. Our architecture contains a high-speed Givens rotation unit which achieves high processing throughput at low area. The resulting VLSI implementation requires 3.3 mus per steering matrix computation at an expense of 41.3 kGEs and shows a 3.5-fold hardware-efficiency gain compared to a reference SVD implementation.

Beamforming (BF) for multiple-input multiple-output (MIMO) wireless communications systems can improve the error rate performance by spatial separation of the transmitted data streams. BF requires to feed back steering matrices from the receiver to the transmitter. The usually large amount of feedback data asks for data reduction schemes. In this paper, we investigate the error rate performance/feedback rate trade-off associated with steering matrix data-reduction schemes and present a corresponding hardware-optimized compression/decompression architecture. Our VLSI implementation achieves up to 50% data reduction for 4times4-dimensional steering matrices without a significant decrease in terms of error rate performance at a circuit complexity of only 7 k gate equivalents.

Soft-input soft-output (SISO) detection algorithms form the basis for iterative decoding. The associated computational complexity often poses significant challenges for practical receiver implementations, in particular in the context of multiple- input multiple-output wireless systems. In this paper, we present a low-complexity SISO sphere decoder which is based on the single tree search paradigm, proposed originally for soft-output detection in Studer et al., IEEE J-SAC, 2008. The algorithm incorporates clipping of the extrinsic log-likelihood ratios in the tree search, which not only results in significant complexity savings, but also allows to cover a large performance/complexity trade-off region by adjusting a single parameter.

Multiple-input multiple-output (MIMO) detection algorithms providing soft information for a subsequent channel decoder pose significant implementation challenges due to their high computational complexity. In this paper, we show how sphere decoding can be used as an efficient tool to implement soft-output MIMO detection with flexible trade-offs between computational complexity and (error rate) performance. In particular, we provide VLSI implementation results which demonstrate that single tree-search, sorted QR-decomposition, channel matrix regularization, log-likelihood ratio clipping, and imposing runtime constraints are the key ingredients for realizing soft-output MIMO detectors with near max-log performance at a chip area that is only 58% higher than that of the best-known hard-output sphere decoder VLSI implementation.

Maximum likelihood (ML) detection provides optimum error rate performance for uncoded multiple-input multiple-output (MIMO) systems. However, circuit complexity of a straightforward implementation of ML detection is uneconomic for high-rate systems. This paper addresses the VLSI implementation trade-offs of a MIMO detection algorithm that achieves close-to ML error rate performance with reduced computational complexity. The described implementations in a 0.25 μm CMOS technology for 4-4 MIMO systems feature a simple data-path, achieve high throughput, and use small silicon area. Important contributing factors to these results are efficient enumeration strategies and the application of simplified norms and sophisticated scheduling techniques together with a new low-complexity preprocessing scheme.

Multiple-input multiple-output (MIMO) communication systems require well-designed synchronization schemes in the receiver to meet stringent QoS requirements. In particular, OFDM modulation is very sensitive to timing synchronization errors, which cause inter-symbol interference. This paper describes a frame-start detection algorithm, which relies on received signal power increase and does not require any special properties of the transmitted signal. The performance is analyzed and then, verified through simulations in a MIMO system employing orthogonal frequency division multiplexing. Finally, a low-complexity FPGA implementation of the presented algorithm is described in detail.

We describe the VLSI implementation of MIMO detectors that exhibit close-to optimum error-rate performance, but still achieve high throughput at low silicon area. In particular, algorithms and VLSI architectures for sphere decoding (SD) and K-best detection are considered, and the corresponding trade-offs between uncoded error-rate performance, silicon area, and throughput are explored. We show that SD with a per-block run-time constraint is best suited for practical implementations.