Kevin M. Buckley

On MLSE algorithms for unknown fast time-varying channels

We consider maximum-likelihood sequence estimation (MLSE) algorithms for unknown, time-varying intersymbol interference communication channels. We assume a statistical channel model, and marginalize over model parameters to derive expectation-maximization (EM) algorithms for both time-independent Gaussian and Gauss-Markov models, and we contrast these with direct MLSE and computationally efficient per-survivor processing implementations. We identify a general concern associated with the convergence of EM-based discrete parameter (e.g., symbol) estimators.

Time-recursive maximum likelihood based sequence estimation for unknown ISI channels

In this paper MLSE (maximum likelihood sequence estimation) is applied to unknown ISI (inter-symbol interference) fast time-varying channels. Assuming a probabilistic model for an FIR channel, a maximum a-posterior (MAP) or ML cost function for the communication symbols is obtained by integration over the channel FIR parameters. A time-recursive (i.e. on-line) sequence estimation algorithm is then identified.

Multitarget list Viterbi tracking algorithm

We present an approach to multitarget tracking algorithm development. The approach is based on a trellis diagram which depicts the possible progressions of sequences of location measurements, over time. Resulting algorithms are sequential and very flexible in that the approach can handle multiple tracks, track initiation, missed detections, false alarms, and various performance cost functions, while managing computation cost by pruning based on track feasibility. The output of a resulting algorithm can be either a single best set of K tracks, or a list of L best best sets of K tracks. The latter is useful, for example, in data fusion where information from other platforms can be used to select one set from the list.

Direct and EM-based map sequence estimation with unknown time-varying channels

We address sequence estimation when the intersymbol interference (ISI) communication channel is unknown and time varying. We employ a maximum a posteriori (MAP) approach, in which the unknown channel parameters are assigned a distribution and integrated out. For several channel models of interest we describe both the exact MAP estimator and Viterbi algorithm based implementations. We also present EM algorithms for solving these MAP sequence estimation problems, and we contrast these EM solutions with direct MAP algorithms.

Time-recursive number-of-tracks estimation for MHT

In this paper we address the issue of measurement-to-track association within the framework of multiple hypothesis tracking (MHT). Specifically, we generate a maximum a posterior (MAP) cost as a function of the number of tracks K. This cost is generated, for each K, as a marginalization over the set of hypothesized track-sets. The proposed algorithm is developed based on a trellis diagram representation of MHT, and a generalized list-Viterbi algorithm for pruning and merging hypotheses. Compared to methods of pruning hypotheses for either MHT or Bayesian multitarget tracking, the resulting Viterbi MHT algorithm is less likely to incorrectly drop tracks in high clutter and high missed- detection scenarios. The proposed number-of-tracks estimation algorithm provides a time-recursive estimate of the number of tracks. It also provides track estimates, allows for the deletion and addition of tracks, and accounts for false alarms and missed detections.

Soft output decoding algorithm for spacetime block codes over unknown time varying channels with intersymbol interference

In this paper we describe and evaluate a soft input soft output (SlSO) decoding algorithm for space time block codes (STBCs) over unknown, time-varying intersymbol interference (ISI) channels. We consider a per-survivor processing (PSP) approach with Kalman filter tracking and forward-backward processing. We also provide a space-time coding example, which employs a serially concatenated recursive systematic code (RSC) and a STBC scheme, with interleaving, and a soft decision iterative (turbo) receiver for unknown time-varying ISI channels.

Multiple user maximum likelihood based sequence estimation for unknown, frequency selective, fast time varying channels

In this paper maximum likelihood sequence estimation (MLSE) is applied to unknown, frequency selective, fast time-varying channels. Assuming a first order Gauss-Markov probabilistic model for the FIR intersymbol interference (ISI) channels, a maximum a-posterior (MAP) or ML cost function for the communication symbols is obtained by integration over the channel FIR parameters. A reduced-complexity, time-recursive sequence estimation algorithm is then described which is based on per survivor precessing (PSP) and reduced state sequence estimation (RSSE).

Fault monitoring using acoustic emissions

Automatic monitoring techniques are a means to safely relax and simplify preventive maintenance and inspection procedures that are expensive and necessitate substantial down time. Acoustic emissions (AEs), that are ultrasonic waves emanating from the formation or propagation of a crack in a material, provide a possible avenue for nondestructive evaluation. Though the characteristics of AEs have been extensively studied, most of the work has been done under controlled laboratory conditions at very low noise levels. In practice, however, the AEs are buried under a wide variety of strong interference and noise. These arise due to a number of factors that, other than vibration, may include fretting, hydraulic noise and electromagnetic interference. Most of these noise events are transient and not unlike AE signals. In consequence, the detection and isolation of AE events from the measured data is not a trivial problem. In this paper we present some signal processing techniques that we have proposed and evaluated for the above problem. We treat the AE problem as the detection of an unknown transient in additive noise followed by a robust classification of the detected transients. We address the problem of transient detection using the residual error in fitting a special linear model to the data. Our group is currently working on the transient classification using neural networks.