The problem addressed is source localization via time-difference-of-arrival estimation in a multipath channel. Solving this localization problem typically implies cross-correlating the noisy signals received at pairs of sensors deployed within reception range of the source. Correlation-based localization is severely degraded by the presence of multipath. The proposed method exploits the sparsity of the multipath channel for estimation of the line-of-sight component. The time-delay estimation problem is formulated as an ℓ1-regularization problem, where the ℓ1-norm is used as a channel sparsity constraint. The proposed method requires knowledge of the pulse shape of the transmitted signal, but it is blind in the sense that information on the specific transmitted symbols is not required at the sensors. Simulation results show that the proposed method delivers higher accuracy and robustness to noise compared to conventional or even super-resolution MUSIC time-difference-of-arrival source localization methods.

A game-theoretic model is proposed to study the cross-layer problem of joint power and rate control with quality of service (QoS) constraints in multiple-access networks. In the proposed game, each user seeks to choose its transmit power and rate in a distributed manner in order to maximize its own utility while satisfying its QoS requirements. The user's QoS constraints are specified in terms of the average source rate and an upper bound on the average delay where the delay includes both transmission and queuing delays. The utility function considered here measures energy efficiency and is particularly suitable for wireless networks with energy constraints. The Nash equilibrium solution for the proposed non-cooperative game is derived and a closed-form expression for the utility achieved at equilibrium is obtained. It is shown that the QoS requirements of a user translate into a "size" for the user which is an indication of the amount of network resources consumed by the user. Using this competitive multiuser framework, the tradeoffs among throughput, delay, network capacity and energy efficiency are studied. In addition, analytical expressions are given for users' delay profiles and the delay performance of the users at Nash equilibrium is quantified.

A game-theoretic approach for studying energy efficiency-delay tradeoffs in multiple-access networks is proposed. Focusing on the uplink of a code-division multiple-access (CDMA) network, a noncooperative game is considered in which each user seeks to choose a transmit power that maximizes its own utility while satisfying its (transmission) delay requirements. The utility function measures the number of reliable bits transmitted per joule of energy and the user's delay constraint is modeled as an upper bound on the delay outage probability. The Nash equilibrium for the proposed game is derived, and its existence and uniqueness are proved. Using a large-system analysis, explicit expressions for the utilities achieved at equilibrium are obtained for the matched filter, decorrelating and (linear) minimum-mean-square-error (MMSE) multiuser detectors. The effects of delay quality-of-service (QoS) constraints on the users' utilities (in bits per joule) and network capacity (i.e., the maximum number of users that can be supported) are quantified. Using the proposed framework, the tradeoffs between energy efficiency and delay are quantified in a competitive multiuser setting.

A unified approach to power control is proposed for maximizing utility in terms of energy efficiency in code-division multiple access (CDMA) networks. The approach is applicable to a large family of multiuser receivers including the matched filter, the decorrelator, the linear minimum mean-square error (MMSE) receiver, and the (nonlinear) optimal detectors. It exploits the linear relationship between the transmit power and the output signal-to-interference-plus-noise ratio (SIR) for each user in the large-system limit. Suppose that each user seeks to selfishly maximize its own energy efficiency, a unique Nash equilibrium is shown to exist and be SIR-balanced, thus extending a previous result on linear receivers. A unified power control algorithm for reaching the Nash equilibrium is proposed, which adjusts transmit powers iteratively by computing the large-system multiuser efficiency, which is independent of instantaneous spreading sequences. The convergence of the algorithm is proved for linear receivers, and is demonstrated via simulation for the multiuser maximum likelihood detector. Moreover, the performance of the algorithm in finite-size systems is studied and compared with that of a conventional power control scheme, in which user powers depend on the instantaneous spreading sequences.

The general problem of object tracking can be modeled as a Markov process and solved by computing probability distributions of the possible object states, followed by MAP estimation. This paper presents a new framework for the efficient estimation of the probability distribution of the states. In contrast to particle filters, where the possible states are numerous and random, we limit the possible states to a finite candidate set which is guaranteed with high probability to contain the true state of the object. After the problem is reduced to a finite-state Markov process (FSM), forward filtering is used to estimate the distribution of the object state. Moreover, the Viterbi algorithm can also be used to estimate the most likely state sequence. We test the new framework by both these methods and compare the tracking results. Experimental results show the effectiveness and efficacy of the proposed algorithm.

A game-theoretic framework is used to study the effect of constellation size on the energy efficiency of wireless networks for M-QAM modulation. A non-cooperative game is proposed in which each user seeks to choose its transmit power (and possibly transmit symbol rate) as well as the constellation size in order to maximize its own utility while satisfying its delay quality-of-service (QoS) constraint. The utility function used here measures the number of reliable bits transmitted per joule of energy consumed, and is particularly suitable for energy-constrained networks. The best-response strategies and Nash equilibrium solution for the proposed game are derived. It is shown that in order to maximize its utility (in bits per joule), a user must choose the lowest constellation size that can accommodate the user's delay constraint. This strategy is different from one that would maximize spectral efficiency. Using this framework, the tradeoffs among energy efficiency, delay, throughput and constellation size are also studied and quantified. In addition, the effect of trellis-coded modulation on energy efficiency is discussed.

This paper presents a novel real-time DCT-based object tracking algorithm that operates on I frames in compressed MPEG video. Discrete Cosine Transform (DCT) coefficients that are provided from video sequences are exploited to find motion and color cues. The temporal motion and spatial color information are then fused by means of a posterior probability framework which allows the information from different measurement sources to be fused in a principled manner. The DCT-based method and probabilistic framework ensure robustness with respect to noise, occlusion, local scene changes, global illumination changes and inconsistent movement. The tracker is also able to recognize the object when it re-appears after it has left the scene. After tracker selection, the background and object model are selectively updated to adapt to environmental changes. Experimental results show the effectiveness and efficacy of the proposed method.

In this paper, a novel robust tracking algorithm in compressed video is proposed. Within the framework of video compression standards, we consider how to accurately estimate motion of an object by utilizing motion vectors available in compressed video together with derived confidence measures. These confidence measures are based on DCT coefficients, spatial continuity of motion and texture measure of the object. We perform tracking directly on the compressed data and also consider tracking of an object with image scale change. In order to achieve robust tracking, we develop a system which enables us to detect object appearance change such as illumination change and occlusion by exploring the confidence measures derived above. Preliminary results indicate that our tracking algorithm works well with a variety of video sequences

A game-theoretic model is proposed to study the cross-layer problem of joint power and rate control with quality of service (QoS) constraints in multiple-access networks. In the proposed game, each user seeks to choose its transmit power and rate in a distributed manner in order to maximize its own utility and at the same time satisfy its QoS requirements. The user's QoS constraints are specified in terms of the average source rate and average delay. The utility function considered here measures energy efficiency and the delay includes both transmission and queueing delays. The Nash equilibrium solution for the proposed non-cooperative game is derived and a closed-form expression for the utility achieved at equilibrium is obtained. It is shown that the QoS requirements of a user translate into a "size" for the user which is an indication of the amount of network resources consumed by the user. Using this framework, the tradeoffs among throughput, delay, network capacity and energy efficiency are also studied.

A game-theoretic model for studying power control in multicarrier code-division multiple-access systems is proposed. Power control is modeled as a noncooperative game in which each user decides how much power to transmit over each carrier to maximize its own utility. The utility function considered here measures the number of reliable bits transmitted over all the carriers per joule of energy consumed and is particularly suitable for networks where energy efficiency is important. The multidimensional nature of users' strategies and the nonquasi-concavity of the utility function make the multicarrier problem much more challenging than the single-carrier or throughput-based-utility case. It is shown that, for all linear receivers including the matched filter, the decorrelator, and the minimum-mean-square-error detector, a user's utility is maximized when the user transmits only on its "best" carrier. This is the carrier that requires the least amount of power to achieve a particular target signal-to-interference-plus-noise ratio at the output of the receiver. The existence and uniqueness of Nash equilibrium for the proposed power control game are studied. In particular, conditions are given that must be satisfied by the channel gains for a Nash equilibrium to exist, and the distribution of the users among the carriers at equilibrium is characterized. In addition, an iterative and distributed algorithm for reaching the equilibrium (when it exists) is presented. It is shown that the proposed approach results in significant improvements in the total utility achieved at equilibrium compared with a single-carrier system and also to a multicarrier system in which each user maximizes its utility over each carrier independently.

This paper examines the uplink user capacity in a two-tier code division multiple access (CDMA) system with hotspot microcells when user terminal power is limited and the wireless channel is finitely-dispersive. A finitely-dispersive channel causes variable fading of the signal power at the output of the RAKE receiver. First, a two-cell system composed. of one macrocell and one embedded microcell is studied and analytical methods are developed to estimate the user capacity as a function of a dimensionless parameter that depends on the transmit power constraint and cell radius. Next, novel analytical methods are developed to study the effect of variable fading, both with and without transmit power constraints. Finally, the analytical methods are extended to estimate uplink user capacity for multicell CDMA systems, composed of multiple macrocells and multiple embedded microcells. In all cases, the analysis-based estimates are compared with and confirmed by simulation results.

The number of simultaneous users (or user capacity) supportable on the uplink of a multiple-macrocell code division multiple-access (CDMA) system with multiple "hotspot" microcells embedded within is studied. These microcells operate on the same frequency as the macrocells and are installed in regions of high user demand. It is shown that the user capacity depends on how the users are distributed among cells, and that the maximum (called the attainable capacity) occurs when all cells serve roughly the same number of users. The approach builds on a two-cell analysis published previously, for a single microcell embedded in a single macrocell. First, this analysis is expanded upon to estimate the attainable capacity for M macrocells, where the center one contains L microcells. Then the case in which L microcells are distributed randomly among the M macrocells is analyzed. In each case, the formula for attainable capacity is very simple and highly accurate (as demonstrated via simulations) up to reasonably high values of L. For example, with L microcells distributed among M macrocells, the analysis is accurate at least up to eight microcells per macrocell. The analysis and results are general with respect to cell geometries, propagation parameters, and other variables of the two-tier CDMA system

The distributed nature of wireless sensor networks illustrates well classical engineering tradeoffs: how to minimize communication (and possibly computation) cost, and thus energy dissipation, while maintaining acceptable performance levels in estimation and inference applications. We study a simple sensor network under dependent Gaussian noise and develop strategies for parameter estimation in a variety of communication scenarios. From an energy point of view, sending all data to a fusion center is the most costly, but leads to optimum performance results. Processing data at each sensor and sending parameter estimates and associated quality measures is a reasonable communication saving procedure and yet, in some cases, may lead to performance equivalent to sending all data to the fusion center. A sequential procedure is most parsimonious in terms of communication cost and especially effective in large wireless sensor networks. We explore those conditions for which little, or no loss in performance is encountered with this sequential procedure. Specifically, we provide analytical expressions for the maximum likelihood estimator under "geometric" dependent noise. We show, by means of analysis and simulations, that the performance is only marginally degraded when the noise is assumed to be independent.

Accurate foreground segmentation is a difficult task due to factors such as illumination variation, occlusion, background movement, and noise. We present a novel adaptive transform domain approach for foreground segmentation in video sequences. A set of DCT-based features is employed to exploit the spatial and temporal correlation in the video sequences. We maintain an adaptive background model and make a decision based on the distance between the features of the current frame and that of the background model. Additional higher level processing is employed to deal with the variation of the environment and to improve the accuracy of segmentation. The approach is shown to be insensitive to illumination change and to noise. It also overcomes many common difficulties of segmentation such as foreground aperture and moved background objects. The algorithm can perform in real-time.

A game-theoretic approach for studying power control in multiple-access networks with transmission delay constraints is proposed. A non-cooperative power control game is considered in which each user seeks to choose a transmit power that maximizes its own utility while satisfying the user's delay requirements. The utility function measures the number of reliable bits transmitted per joule of energy and the user's delay constraint is modeled as an upper bound on the delay outage probability. The Nash equilibrium for the proposed game is derived, and its existence and uniqueness are proved. Using a large-system analysis, explicit expressions for the utilities achieved at equilibrium are obtained for the matched filter, decorrelating and (linear) minimum mean square error multiuser detectors. The effects of delay constraints on the users' utilities (in bits/Joule) and network capacity (i.e., the maximum number of users that can be supported) are quantified

In the power control game proposed for MC-CDMA systems, each user needs to decide how much power to transmit over each carrier to maximize its overall utility. The utility function considered measures the number of reliable bits transmitted per joule of energy consumed. It is shown that the user's utility is maximized when the user transmits only on the carrier with the best "effective channel". The existence and uniqueness of Nash equilibrium for the proposed game are investigated and the properties of equilibrium are studied. Also, an iterative and distributed algorithm for reaching equilibrium (if it exists) is presented. It is shown that the proposed approach results in a significant improvement in the total utility achieved at equilibrium compared to the case in which each user maximizes its utility over each carrier independently.

In this paper, the cross-layer design problem of joint multiuser detection and power control is studied, using a game-theoretic approach that focuses on energy efficiency. The uplink of a direct-sequence code-division multiple-access data network is considered, and a noncooperative game is proposed in which users in the network are allowed to choose their uplink receivers as well as their transmit powers to maximize their own utilities. The utility function measures the number of reliable bits transmitted by the user per joule of energy consumed. Focusing on linear receivers, the Nash equilibrium for the proposed game is derived. It is shown that the equilibrium is one where the powers are signal-to-interference-plus-noise ratio-balanced with the minimum mean-square error (MMSE) detector as the receiver. In addition, this framework is used to study power-control games for the matched filter, the decorrelator, and the MMSE detector; and the receivers' performance is compared in terms of the utilities achieved at equilibrium (in bits/joule). The optimal cooperative solution is also discussed and compared with the noncooperative approach. Extensions of the results to the case of multiple receive antennas are also presented. In addition, an admission-control scheme based on maximizing the total utility in the network is proposed.

This paper examines the effect of soft handoff on the uplink user capacity of a code division multiple access system consisting of a single macrocell in which a single hotspot microcell is embedded. The users of these two base stations operate over the same frequency band. In the soft-handoff scenario studied here, both macrocell and microcell base stations serve each system user, and the two received copies of a desired user's signal are summed using maximal ratio combining. Exact and approximate analytical methods are developed to compute uplink user capacity. Simulation results demonstrate a 20% increase in user capacity compared to hard handoff. In addition, simple approximate methods are presented for estimating soft-handoff capacity and are shown to be quite accurate.

This paper studies a two-tier code-division multiple-access (CDMA) system in which the microcell base is converted into a data access point (DAP), i.e., is a limited-range base station that provides high-speed access to one user at a time. The microcell (or DAP) user operates on the same frequency as the macrocell users and has the same chip rate. However, it adapts its spreading factor, and thus its data rate, in accordance with interference conditions. By contrast, the macrocell serves multiple simultaneous data users, each with the same fixed rate. The achievable throughput for individual microcell users is examined and a simple accurate approximation for its probability distribution is presented. Computations for average throughputs, both per-user and total, are also presented. The numerical results highlight the impact of a desensitivity parameter used in the base-selection process.

This paper examines the effect of soft handoff on the uplink user capacity of a CDMA system consisting of a single macrocell in which a single hotspot microcell is embedded. The users of these two base stations operate over the same frequency band. In the soft handoff scenario studied here, both macrocell and microcell base stations serve each system user and the two received copies of a desired user's signal are summed using maximal ratio combining. Exact and approximate analytical methods are developed to compute uplink user capacity. Simulation results demonstrate a 20% increase in user capacity compared to hard handoff. In addition, simple, approximate methods are presented for estimating soft handoff capacity and are shown to be quite accurate.

This paper examines the downlink user capacity in a CDMA system composed of a macrocell and a microcell, where both cells use the same frequency channel. The effect of channel dispersion on the number of simultaneous users supported is studied for three different downlink power control methods. Channel dispersion produces in-cell interference and variable fading of the received signal power. Novel analytical methods are developed to account for both effects; these techniques derive user capacity for any delay profile using the results for the uniform delay profile. The downlink user capacities for the three power control methods are compared to the uplink user capacity under the same channel conditions. The results point to the need for fast power control to obtain uplink-limited performance.

This paper examines the uplink user capacity in a CDMA system composed of one macrocell and one microcell when user terminal power is limited and the wireless channel is finitely-dispersive. A finitely-dispersive channel causes variable fading of the signal power at the output of the RAKE receiver. Analytical methods are developed to compute the user capacity as a function of a dimensionless parameter that depends on the transmit power constraint and cell radius. Additionally, a novel analytical method is developed to study the effect of variable fading due to a finite number of multipaths. This technique derives the user capacity for any delay profile using the results for the uniform multipath channel. Simulation results confirm the accuracy of these simple analytical solutions.

We investigate the problem of joint frequency offset and channel estimation for OFDM systems. The complexity of the joint maximum likelihood (ML) estimation procedure motivates us to propose an adaptive MLE algorithm which iterates between estimating the frequency offset and the channel parameters. Pilot tones are used to obtain the initial estimates and then a decision-directed technique provides an effective estimation technique. The joint modified (averaged) Cramer-Rao lower bounds (MCRB) of the channel coefficients and frequency offset estimates are derived and discussed. It is shown that, for the case of a large number of subcarriers in the OFDM system, there is approximately a 6 dB loss in the frequency offset estimate lower bound due to the lack of knowledge of the channel impulse response (CIR). The degradation of the CIR lower bound is less severe and depends on the channel delay spread. We show both analytically and by simulation, that the channel estimate accuracy is less sensitive to unknown frequency offset than the frequency offset estimation is affected by the unknown CIR. Comprehensive simulations have been carried out to validate the effectiveness of the adaptive joint estimation algorithm.

The combination of multiple-antenna and orthogonal frequency division multiplexing (OFDM) provides reliable communications over frequency selective fading channels. We investigate this approach and focus on the application of space-time block codes (STBC) and space-frequency block codes (SFBC) in OFDM systems. We compare the performance of maximum likelihood (ML), zero forcing (ZF) and conventional detection algorithms. We show that ZF provides a good trade-off between computational complexity and performance. The problem of channel estimation in STBC-OFDM and SFBC-OFDM systems is also studied, including the derivation of the Cramer-Rao lower bound (CRLB). Since knowledge of the channel is required to coherently decode STBC-OFDM and SFBC-OFDM, we propose an iterative channel estimation algorithm based on the EM algorithm that requires very few pilot symbols. The CRLB can be achieved by the channel estimation algorithm.

Estimating a channel that is subject to frequency selective Rayleigh fading is a challenging problem in an orthogonal frequency division multiplexing (OFDM) system. We propose an enhanced channel estimation algorithm that combines the EM-based algorithms proposed previously and a least squares polynomial fitting (LSPF) approach. The combined algorithm can efficiently estimate the channel response of an OFDM system operating in an environment with multipath fading and additive white Gaussian noise (AWGN). The algorithm can improve the channel estimate obtained from the EM-based algorithms by polynomial fitting. Simulation results show that the bit error rate (BER) as well as the mean square error (MSE) of the channel can be improved by the algorithm. In particular, with these additional computations and demodulation delay, the MSE can be made smaller than the Cramer-Rao lower bound (CRLB).

The appearance of an object can be severely affected by illumination. Thus, illumination correction is necessary both for human perception and machine recognition. The paper reports on a general approach for fast illumination correction. The approach has been tested for application in face normalization as a preprocessing step in face recognition. The basic idea of the algorithm is to normalize the image contrast locally using an affine transformation lighting model based on local estimation of background and gain. The background is estimated via an efficient multi-resolution low-pass filter and the gain is estimated via homomorphic filtering. This is followed by normalizing the data with the help of a clipped histogram. Experiments on images with different lighting conditions produce results that are better than those from using several popular illumination correction methods.

Performance of both orthogonal frequency division multiplexing (OFDM) and single carrier (SC) systems suffers from a carrier frequency offset (CFO) due to Doppler effect and the carrier instability between the transmitter and the receiver. We investigate the bit error rate (BER) performance degradation of OFDM and SC systems due to the frequency offset in an additive white Gaussian noise (AWGN) channel as well as multipath Rayleigh fading channels. We consider three effects to the BER degradation, i.e., phase shift, useful power decrease and intercarrier interference (ICI). We also derive the approximate expressions of BER under binary phase shift keying (BPSK) and quaternary phase shift keying (QPSK) for both OFDM and SC systems in the presence of CFO. In general, SC is more robust to CFO in the AVVGN channel than OFDM in terms of BER, while both of them suffer similarly from CFO in the multipath Rayleigh fading channels assuming the the same CFO.

To improve compensation to channel or interference changes, we propose adapting an auxiliary feedback filter (FBF) in the receiver of systems which use Tomlinson-Harashima (1971, 1972) precoding. We show how the auxiliary FBF can be adapted in conjunction with the receiver feedforward filter (FFF). Simulations demonstrate the performance advantage of our auxiliary FBF technique relative to FFF updating alone, and how the FFF combines interference suppression with despreading in wideband applications. Error propagation can be effectively avoided by using the auxiliary FBF values to decide when to update the precoder, while transient increases in mean-squared error are avoided by using the FBF values in the update equation.

This paper studies the number of voice users (user capacity) supported on the uplink of a single-macrocell/single-microcell code-division multiple-access system. A "hotspot" microcell is embedded within a larger macrocell and operates over the same bandwidth as the larger cell. Analytic methods are presented for computing user capacity which account for propagation loss, multiple-access interference, power control, and random locations of user terminals, as well as two distinct methods by which users select base stations (tiers). Along with the exact user capacity, a technique for making accurate approximations is also presented. Simulation results verify both the exact and approximate analytical methods. This simulation is also employed to study the capacity gains of a third, more optimal, tier-selection scheme. These results point to differences in capacity performance based on the tier-selection method, as well as on the traffic density within the hotspot region.

We present an adaptive feature selection scheme to jointly optimize the detector performance and the computational efficiency for statistical object detection. From the statistical distribution of wavelet coefficients, we construct an error-bound-tree (EBT) to analyze the error probability of the Bayes test. The wavelet features put into test are adaptively selected to minimize the detection error. The selected features are more discriminative than others and allow the detector to reach a decision faster without jeopardizing its accuracy. The proposed scheme is demonstrated in face detection.

The paper presents a sequential classification approach to improve the efficiency in visual object (face) detection. To reduce the computation while maintaining detection accuracy, a two-level hierarchy of sequential classification is proposed. At the top level, the overall detector is built on a cascade of classifiers at multiple resolution scales produced by a wavelet transform. Classifiers at low-resolution scales quickly rule out the regions likely to be background. Only object-like candidates are passed to subsequent high-resolution scales for more expensive tests. At the bottom level of the hierarchy, each classifier is implemented as a sequential Bayesian test using the features within the scale. The features are ranked adaptively according to their discrimination ability, which also leads to a quick decision. We demonstrate the scheme by an example of frontal view face detection.

This paper studies the number of simultaneous voice users (or user capacity) supportable on the uplink of a multiple macrocell CDMA system with multiple "hotspot" microcells embedded within. These microcells operate over the same bandwidth as the macrocells and are assumed to be installed in regions of high user demand. First, conditions for optimal performance of such a system are established. This analysis shows that the total user capacity is maximum when there are equal numbers of active users in each cell. Along with these conditions, analytical approximations are presented for computing maximum user capacity as a function of the number of microcell bases in the system. In particular, the approximations show the total user capacity as growing linearly with the number of hotspot microcells. Simulation results verify this trend when the number of microcells in the system is not large. The numerical constants for this linear trend are related to basic system and propagation parameters. When the number of microcells gets large, it is shown that the same linear growth in capacity can be sustained by desensitizing the macrocells, i.e., raising the received power requirements at the macrocell bases.

We propose an EM-based algorithm to efficiently detect transmitted data in an OFDM system as well as estimating the channel impulse response (CIR). The maximum likelihood estimate of CIR is obtained by using channel statistics (their means and covariances) via the expectation-maximization (EM) algorithm. This algorithm can improve signal detection and the channel estimation accuracy by making use of pilot symbols to obtain an initial estimate for the iteration. Simulation results show that the bit error rate (BER) can be significantly reduced by this algorithm, and validate its good convergence and robust properties.

Accurate characterization of data distribution is of significant importance for vision problems. In many situations, multivariate visual data often spread into a nonlinear manifold in the high-dimensional space, which makes traditional linear modeling techniques ineffective. This paper proposes a generic nonlinear modeling scheme based on parametric data representations. We build a compact representation for the visual data using a set of parameterized basis (wavelet) functions, where the parameters are randomized to characterize the nonlinear structure of the data distribution. Meanwhile, a new progressive density approximation scheme is proposed to obtain an accurate estimate of the probability density, which imposes discrimination power on the model. Both synthetic and real image data are used to demonstrate the strength of our modeling scheme.

We propose a directional smoothness measure for block-based error concealment through spatial correlation. Image structures revealed by consistent edge profiles are very important for subjective visual quality. We treat the problem of block reconstruction as consistent recovery of local image structures. The directional smoothness measure evaluates structural consistency along an edge elongation and is used as the object function for block reconstruction. Corrupted DCT coefficients are recovered by smoothly extending various edge profiles from surrounding areas to missing blocks. The reconstruction is adaptive to local image structures. Consistent cross-edge sharpness and along-edge smoothness are maximally preserved during the reconstruction. The proposed concealment method demonstrates encouraging improvement both in the subjective image quality and in the reconstruction PSNR over conventional schemes. It is applicable to various spatial and spectral interleaving systems and a fast implementation is also proposed.

We propose a new wavelet domain image interpolation scheme based on statistical signal estimation. A linear composite MMSE estimator is constructed to synthesize the detailed wavelet coefficients as well as to minimize the mean squared error for high-resolution signal recovery. Based on a discrete time edge model, we use low-resolution information to characterize local intensity changes and perform resolution enhancement accordingly. A linear MMSE estimator follows to minimize the estimation error. Local image statistics are involved in determining the spatially adaptive optimal estimator. With knowledge of edge behavior and local signal statistics, the composite estimation is able to enhance important edges and to maintain the intensity consistency along edges. Strong improvement in both the visual quality and the PSNRs of the interpolated images has been achieved by the proposed estimation scheme.

This paper studies the number of voice users supported on the uplink (user capacity) of a single-macrocell/single-microcell CDMA system. A "hotspot" microcell is embedded within a larger macrocell and operates over the same bandwidth as the larger cell. Analytic methods are presented for computing user capacity which account for propagation loss, multiple-access interference, power-control, and random locations of user terminals, as well as two distinct methods by which users select base stations (tiers). Along with the exact user capacity, a technique for approximations is also presented. Simulation results verify both the exact and approximate analytical methods. This simulation is also employed to study capacity gains of a third, more optimal, tier-selection scheme. The results point to differences in capacity performance based on the tier-selection method as well as on the traffic density within the hotspot region.

The performance of OFDM system in mobile wireless channels can be enhanced by introducing adaptive loading schemes: information bits are distributed unequally in subchannels according to the channel frequency response. Under the assumption of perfect channel estimation, considerable gain can be achieved over the non-adaptive loading scheme for typical channels. If there are channel estimation errors in the system, the adaptive loading gain will be adversely affected in proportion to the MSE of the errors. The effect of channel mismatch due to Doppler spread is also studied. The results show that adaptive loading can preserve most of its gain even with imperfect channel information.

We investigate linear and nonlinear space-time minimum mean-square-error (MMSE) multiuser detectors for high data rate wireless code-division multiple-access (CDMA) networks. The centralized reverse-link detectors comprise a space-time feedforward filter and a multiuser feedback filter which processes the previously detected symbols of all in-sector users. The feedforward filter processes chip-rate samples from a bank of chip-matched filters which operate on the baseband outputs from an array of antennas. We present an adaptive multiuser recursive least squares (RLS) algorithm which determines the MMSE adjusted filter coefficients with less complexity than individual adaptation for each user. We calculate the outage probabilities and isolate the effects of antenna, diversity, and interference suppression gains for linear and nonlinear filtering and for CDMA systems with varying levels of system control (e.g., timing control, code assignment, cell layout). For eight users transmitting uncoded 2-Mb/s quadrature phase-shift keying with a spreading gain of eight chips per symbol over a fading channel with a multipath delay spread of 1.25 /spl mu/s, the performance of a three-antenna feedforward/feedback detector was within 1 dB (in signal-to-noise ratio per antenna) of ideal detection in the absence of interference. By training for 10% of a 5-ms frame, RLS adaptation enabled the same detector to suffer less than a 0.5-dB penalty due to the combined effects of imperfect coefficients and error propagation. The advantage of nonlinear feedforward/feedback detection over linear feedforward detection was shown to be significantly larger for a CDMA system with enhanced system control.

Computation complexity is an important issue for current face detection systems. This paper proposes a subspace approach to capture local discriminative features in the space-frequency domain for fast face detection. Based on orthonormal wavelet packet analysis, we develop a discriminant subspace algorithm to search for the "minimum cost" subspace of the high-dimensional signal space, which leads to a set of wavelet features with maximum class discrimination and dimensionality reduction. Detailed (high frequency) information within local facial areas shows noticeable discrimination ability for face detection problem. We demonstrate the algorithm in the context of detecting frontal view faces in a complex background. Discrete pattern distribution functions and fast likelihood ratio detection are adopted by the system. Because of the reduced dimensionality, feature discrimination and the discrete stochastic model, our face detection system consumes much less computation while the performance is comparable with other reported leading systems.

We consider the design and performance of nonlinear minimum mean-square-error multiuser detectors for direct sequence code-division multiple-access (CDMA) networks. With multiple users transmitting asynchronously at high data rates over multipath fading channels, the detectors contend with both multiple-access interference (MAI) and intersymbol interference (ISI). The cyclostationarity of the MAI and ISI is exploited through a feedforward filter (FFF), which processes samples at the output of parallel chip-matched filters, and a feedback filter (FBF), which processes detected symbols. By altering the connectivity of the FFF and FBF, we define four architectures based on fully connected (FC) and nonconnected (NC) filters. Increased connectivity of the FFF gives each user access to more samples of the received signal, while increased connectivity of the FBF provides each user access to previous decisions of other users. We consider three methods for specifying the FFF sampling and propose a nonuniform FFF sampling scheme based on multipath ray tracking that can offer improved performance relative to uniform FFF sampling. For the FC architecture, we capitalize on the sharing of filter contents among users by deriving a multiuser recursive least squares (RLS) algorithm and direct matrix inversion approach, which determine the coefficients more efficiently than single-user algorithms. We estimate the uncoded bit-error rate (BER) of the feedforward/feedback detectors for CDMA systems with varying levels of power control and timing control for multipath channels with quasi-static Rayleigh fading. Simulations of packet-based QPSK transmission validate the theoretical BER analysis and demonstrate that the multiuser RLS adapted detectors train in several hundred symbols and avoid severe error propagation during data transmission mode.

The optimum (maximum likelihood) multi-element global sequence detector is derived for the multiuser communication channel. The resulting integrated array-detector is composed of retrodirective beamformers in the user's directions followed by a bank of matched filters and a processor implementing a dynamic programming (Viterbi) algorithm. Suboptimal realizations of the multi-element detector are considered, offering reduced complexity as compared to the optimum solution. The performance of the multi-element detector is analyzed in terms of error probability, detection asymptotic efficiency and near-far resistance. In particular, the effects of array beamforming on the performance is discussed and shown to generalize existing results for the scalar (single channel) case. Processing which uses the additional dimension of space can result in significant improvements in performance over the scalar case. Following a general performance analysis, a comparison of several suboptimal multi-element detectors is given. In particular, it is shown that a new combined spatial-temporal processing is always uniformly superior to that of using beamformers which result in separation of signals, followed by single-user detectors.>

We propose a distributed reservation protocol for high-speed, star-based, local-communication systems with heterogeneous users. The users have different delay, throughput, and priority requirements, and each user has knowledge of only its own traffic mix. At light loads, the usual delay overhead of reserveration schemes can be eliminated. Also, the probability of correct reservation reception is improved: it is only necessary to detect the presence or absence of a signal since reservation signals do not carry address information or the desired amount of channel time. Users know the value of a global reservation counter, which specifies the amount of transmission time acquired by a reservation. The protocol can support both circuit- and packet-switched traffic (e.g., integrated voice and data).

A method for modeling large errors in correlation-based time-delay estimation is developed in terms of level-crossing probabilities. The level-crossing interpretation for peak ambiguity leads directly to an exact expression for the probability of large error involving the hazard function associated with the level-crossing process. Two models for the distribution of the error over the level-crossing time yield approximations to the mean-square error (MSE) that involve the low-order (>

This correspondence examines the influence of the summation order on the numerical accuracy of the resulting sum when the addition is performed using floating-point, finite-precision arithmetic. A simple statistical model is used to find the relative performance of various addition procedures. It is shown that phrasing the question in terms of finding a best ordering is overly restrictive since the most natural and accurate procedures for performing the addition utilize the storage of intermediate sums as well as performing orderings.

We study statistical modeling by a Gaussian-Gaussian mixture for two different underwater noise samples. We show that one of them can be adequately described by a Gaussian-Gaussian mixture whereas the other one is very close to a Gaussian model and is described by a mixture with a very small perturbating term. The first noise is also studied with emphasis on the optimal receiver structure for the detection of a deterministic signal. The performance of two test-functions are studied. The principal result is that the use of the likelihood ratio receiver associated with the mixture model leads to improvements with respect to the classical matched filter, this improvement being measured in term of R O C curves.

A robust test for detection of narrowband coherent signals of unknown amplitude and phase in contaminated Gaussian noise of unknown scale is proposed. The test preserves the structure of the quadrature matched filter. which is the uniformly most powerful (UMP) test for pure Gaussian noise. In the new test, the sample-mean estimators are replaced by robust α- trimmed estimators, with an adaptive threshold derived from noise-reference samples. The resulting detector is shown to be asymptotically CFAR and maximin optimal robust for various signal amplitude models and for arbitrary signal level Monte-Carlo results demonstrate that robustness is maintained even with small sample sizes, and that the proposed test substantially outperforms the robust test that has been derived from the traditional weak signal assumption.

An alternative method for estimating the time delay between two noisy waveforms containing a common signal is presented. The estimate is obtained by means of an approximation to the center of symmetry of a certain correlation function. For narrowband signals preliminary results indicate that the procedure is less sensitive to peak ambiguity which is inherent in the classical optimal estimator.

Signal-to-interference ratio statistics for mobile telephony systems with hexagonal coverage areas, multiple interferers, and threecorner base stations are investigated. Corner base stations simultaneously reduce total interference (because of antenna directivity) and provide macrodiversity against shadow fading. Our results indicate that for 3 or 7 channel sets, with 2 tiers of interferers and typical system parameters (propagation exponent of 3.7 and lognormal spread of 8 dB), the improvement is on the order of 11 dB over a baseline arrangement which employs centrally located omnibase stations without macrodiversity. Additional considerations of blocking probabilities and average voice activity indicate possible further improvements.

The mobile radio channel is characterized by three important factors: path losses larger than free space, fading typically taken as Rayleigh, and shadowing generally characterized as lognormal. For cellular systems, in order to determine acceptable reuse distances between base stations and to compare modulation methods, the probability of unacceptable cochannel interference (outage probability) has to be determined in the realistic situation where both fading and shadowing occur. In this paper, the average outage probability is computed for centrally located base stations when multiple log-normal interferers are present. This is done for both the mobile-to-base and base-to-mobile communication links. An unexpected result of this study is that the outage probabilities for the two cases do not differ in a significant way. Cumulative probability curves of the short-term average-signal-toaverage-interference ratio (SIR) are presented for a variety of system parameters: channel set number, propagation law exponent (γ), and dB spread (σ) of the log-normal distribution for the signal and interferers. An important observation is the large sensitivity of the performance curves to the propagation parameters: for a system with seven channel sets with a 10 dB SIR threshold, the average outage probability varies from 10 percent for\gamma = 3.7, \sigma = 6dB, to 70 percent for\gamma = 3, \sigma = 14dB. Alternatively, for a fixed outage objective of 10 percent, the required SIR threshold value ranges from -17 dB to 11 dB, depending on the propagation parameters. These variations make it imperative that accurate measurements of these parameters be obtained for the different service areas. Outage probabilities are also easily related to specific modulation methods and diversity approaches; detailed results are given for several representative cases.

A maximin approach is applied to the problem of detecting a point target in white noise and clutter. That is, the reverberation scattering function (which characterizes the clutter) is not assumed to be known exactly. The problem becomes one of finding a signal-filter combination which provides the best possible lower bound on detector performance. A "quasi-universal" solution is obtained for this problem which is applicable in many practical situations. For many other situations a general (but preliminary) theory is developed.

Starting with the canonical discrete-time detection formula for random signals in Gaussian noise, we demonstrate how different noise models lead to different limiting forms Of the detector. A Brownian motion model leads to the well-known estimator-correlator formula. With a band-limited white noise model, the limiting form contains an additional term: the posterior conditional variance. This second form, which is the Radon-Nikodym derivative of two finitely additive Gaussian measures, was derived by Balakrishnan using a different formulation [10]. We thus provide an alternative development and interpretation of this second detector, and illustrate the effect of model assumptions on going from discrete-time to continuous-time problems.

A class of discrete-time detection problems is formulated and a general recursive formula for the likelihood ratio is obtained. The result is analogous to the general likelihood-ratio formula or "estimator-correlator" receiver which has been derived in continuous time for a number of detection problems. As in the continuous-time versions, a conditional-mean estimate again plays a central role in the receiver structure.

For conditional probability density functions (pdf's) drawn from the exponential family, it is shown that the marginal pdf is completely determined by a posterior conditional mean estimate (CME). This result implies that likelihood ratios involving these marginals have the estimator-correlator structure in the following sense: if the noise is drawn from an exponential pdf, then independent of the signal (prior pdf), the optimum detector correlates the estimate with the data. A generalization of Esposito's result on "pseudoestimates" is also given.

This paper appears in the March, 1972, issue of the Communications of the ACM. Its abstract is reproduced below.

A detector that is not nonparametric, but that nevertheless performs well over a broad class of noise distributions is termed a robust detector. One possible way to obtain a certain degree of robustness or stability is to look for a min-max solution. For the problem of detecting a signal of known form in additive, nearly Gaussian noise, the solution to the min-max problem is obtained when the signal amplitude is known and the nearly Gaussian noise is specified by a mixture model. The solution takes the form of a correlator-limiter detector. For a constant signal, the correlator-limiter detector reduces to a limiter detector, which is shown to be robust in terms of power and false alarm. By adding a symmetry constraint to the nearly normal noise and formulating the problem as one of local detection, the limiter-correlator is obtained as the local min-max solution. The limiter-correlator is shown to be robust in terms of asymptotic relative efficiency (ARE). For a pulse train of unknown phase, a limiter-envelope sum detector is also shown to be robust in terms of ARE.

A decision-directed receiver (DDR) uses previous outputs (decisions) to estimate unknown parameters and, on the basis of these estimates, modifies the detector structure for subsequent decisions. Although the DDR is less complex to instrument than other adaptive schemes, inherent in the decision-directed approach is the possibility of a runaway. This occurs when the detector commits a sequence of decision errors resulting in a degradation of parameter estimates, which, in turn, results in a further deterioration of detector performance. Because of the dependencies introduced by the learning process, runaway is difficult to analyze. In this paper, a DDR with unknown a priori probabilities is considered. The priors are estimated by the relative frequency of decisions of that event. For binary detection, it is shown that there is a positive probability of a runaway (the estimates converge to 1 or 0), which equals 1 if the signal-to-noise ratio is below a critical value. A tight bound on the probability of a runaway is obtained by approximating the learning process by a random walk with independent increments. The analysis demonstrates that a runaway is quite improbable even for moderate signal-to-noise ratios. The analysis is extended to multiple signals and to the situation where the estimates of the priors are updated continuously through exponential weighting rather than allowed to converge.

A procedure for the optimum detection of stochastic signals in noise is discussed. The optimum test function is expanded in a point-wise convergent series for which a bound on the convergence properties can be obtained. Knowledge of this bound permits the substitution of a truncated version of the series for the optimum test function. This leads to a test procedure that uses a variable number of terms of the series for each decision and also gives the same decision as the optimum detector. For detection of stochastic signals in Gaussian noise, an expansion is obtained in terms of the eigenfunctions associated with the Gaussian probability density function, which leads to optimum decisions with a moderate number of terms of the series. It is also well suited for adaptive detection in which the distribution function of the stochastic signal is unknown--the coefficients of the expansion factor into two terms, one dependent only on the noise distribution and the other dependent on the distribution of the stochastic signal. Computer results for Gaussian noise are given. For this case, the test procedure can be viewed as a sequence of linear, quadratic, etc., detectors that, when a basic inequality is met, terminates with an optimum decision.