Subhrakanti Dey

Data Rate Theorem for Stabilization Over Time-Varying Feedback Channels

A data rate theorem for stabilization of a linear, discrete-time, dynamical system with arbitrarily large disturbances, over a rate-limited, time-varying communication channel is presented. Necessary and sufficient conditions for stabilization are derived, their implications and relationships with related results in the literature are discussed. The proof techniques rely on both information-theoretic and control-theoretic tools.

Optimal and distributed protocols for cross-layer design of physical and transport layers in MANETs

We seek distributed protocols that attain the global optimum allocation of link transmitter powers and source rates in a cross-layer design of a mobile ad hoc network. Although the underlying network utility maximization is nonconvex, convexity plays a major role in our development. We provide new convexity results surrounding the Shannon capacity formula, allowing us to abandon suboptimal high-SIR approximations that have almost become entrenched in the literature. More broadly, these new results can be back-substituted into many existing problems for similar benefit. Three protocols are developed. The first is based on a convexification of the underlying problem, relying heavily on our new convexity results. We provide conditions under which it produces a globally optimum resource allocation. We show how it may be distributed through message passing for both rate- and power-allocation. Our second protocol relaxes this requirement and involves a novel sequence of convex approximations, each exploiting existing TCP protocols for source rate allocation. Message passing is only used for power control. Our convexity results again provide sufficient conditions for global optimality. Our last protocol, motivated by a desire of power control devoid of message passing, is a near optimal scheme that makes use of noise measurements and enjoys a convergence rate that is orders of magnitude faster than existing methods.

Stochastic consensus over noisy networks with Markovian and arbitrary switches

This paper considers stochastic consensus problems over lossy wireless networks. We first propose a measurement model with a random link gain, additive noise, and Markovian lossy signal reception, which captures uncertain operational conditions of practical networks. For consensus seeking, we apply stochastic approximation and derive a Markovian mode dependent recursive algorithm. Mean square and almost sure (i.e., probability one) convergence analysis is developed via a state space decomposition approach when the coefficient matrix in the algorithm satisfies a zero row and column sum condition. Subsequently, we consider a model with arbitrary random switching and a common stochastic Lyapunov function technique is used to prove convergence. Finally, our method is applied to models with heterogeneous quantizers and packet losses, and convergence results are proved.

On Kalman filtering over fading wireless channels with controlled transmission powers

We study stochastic stability of centralized Kalman filtering for linear time-varying systems equipped with wireless sensors. Transmission is over fading channels where variable channel gains are counteracted by power control to alleviate the effects of packet drops. We establish sufficient conditions for the expected value of the Kalman filter covariance matrix to be exponentially bounded in norm. The conditions obtained are then used to formulate stabilizing power control policies which minimize the total sensor power budget. In deriving the optimal power control laws, both statistical channel information and full channel information are considered. The effect of system instability on the power budget is also investigated for both these cases.

Optimal power control for Rayleigh-faded multiuser systems with outage constraints

How can we achieve the conflicting goals of reduced transmission power and increased capacity in a wireless network, without attempting to follow the instantaneous state of a fading channel? In this paper, we address this problem by jointly considering power control and multiuser detection (MUD) with outage-probability constraints in a Rayleigh fast-fading environment. The resulting power-control algorithms (PCAs) utilize the statistics of the channel and operate on a much slower timescale than traditional schemes. We propose an optimal iterative solution that is conceptually simple and finds the minimum sum power of all users while meeting their outage targets. Using a derived bound on outage probability, we introduce a mapping from outage to average signal-to-interference ratio (SIR) constraints. This allows us to propose a suboptimal iterative scheme that is a variation of an existing solution to a joint power control and MUD problem involving SIR constraints. We further use a recent result that transforms complex SIR expressions into a compact and decoupled form, to develop a noniterative and computationally inexpensive PCA for large systems of users. Simulation results are presented showing the closeness of the optimal and mapped schemes, speed of convergence, and performance comparisons.

Kalman filtering with faded measurements

This paper considers a sensor network where single or multiple sensors amplify and forward their measurements of a common linear dynamical system (analog uncoded transmission) to a remote fusion centre via noisy fading wireless channels. We show that the expected error covariance (with respect to the fading process) of the time-varying Kalman filter is bounded and converges to a steady state value, based on some general earlier results on asymptotic stability of Kalman filters with random parameters. More importantly, we provide explicit expressions for sequences which can be used as upper bounds on the expected error covariance, for specific instances of fading distributions and scalar measurements (per sensor). Numerical results illustrate the effectiveness of these bounds.

A power control game based on outage probabilities for multicell wireless data networks

We present a game-theoretic treatment of distributed power control in CDMA wireless systems using outage probabilities. We prove that the noncooperative power control game considered admits a unique Nash equilibrium (NE) for uniformly strictly convex pricing functions and under some technical assumptions on the SIR threshold levels. We analyze global convergence of continuous-time as well as discrete-time synchronous and asynchronous iterative power update algorithms to the unique NE of the game. Furthermore, a stochastic version of the discrete-time update scheme, which models the uncertainty due to quantization and estimation errors, is shown to converge almost surely to the unique NE point. We further investigate and demonstrate the convergence and robustness properties of these update schemes through simulation studies.

Risk-sensitive filtering and smoothing via reference probability methods

We address the risk-sensitive filtering problem which is minimizing the expectation of the exponential of the squared estimation error multiplied by a risk-sensitive parameter. Such filtering can be more robust to plant and noise uncertainty than minimum error variance filtering. Although optimizing a differently formulated performance index to that of the so-called H/sub /spl infin// filtering, risk-sensitive filtering leads to a worst case deterministic noise estimation problem given from the differential game associated with H/sub /spl infin// filtering. We consider a class of discrete-time stochastic nonlinear state-space models. We present linear recursions in the information state and the result for the filtered estimate that minimizes the risk-sensitive cost index. We also present fixed-interval smoothing results for each of these signal models. In addition, a brief discussion is included on relations of the risk-sensitive estimation problem to minimum variance estimation and a worst case estimation problem in a deterministic noise scenario related to minimax dynamic games. The technique used in this paper is the so-called reference probability method which defines a new probability measure where the observations are independent and translates the problem to the new measure. The optimization problem is solved using simple estimation theory in the new measure, and the results are interpreted as solutions in the original measure.

Compressed Sensing With Prior Information: Information-Theoretic Limits and Practical Decoders

This paper considers the problem of sparse signal recovery when the decoder has prior information on the sparsity pattern of the data. The data vector x=[x1,...,xN]T has a randomly generated sparsity pattern, where the i-th entry is non-zero with probability pi. Given knowledge of these probabilities, the decoder attempts to recover x based on M random noisy projections. Information-theoretic limits on the number of measurements needed to recover the support set of x perfectly are given, and it is shown that significantly fewer measurements can be used if the prior distribution is sufficiently non-uniform. Furthermore, extensions of Basis Pursuit, LASSO, and Orthogonal Matching Pursuit which exploit the prior information are presented. The improved performance of these methods over their standard counterparts is demonstrated using simulations.

Risk-sensitive filtering and smoothing for hidden Markov models

Abstract In this paper, we address the problem of risk-sensitive filtering and smoothing for discrete-time Hidden Markov Models (HMM) with finite-discrete states. The objective of risk-sensitive filtering is to minimise the expectation of the exponential of the squared estimation error weighted by a risk-sensitive parameter. We use the so-called Reference Probability Method in solving this problem. We achieve finite-dimensional linear recursions in the information state, and thereby the state estimate that minimises the risk-sensitive cost index. Also, fixed-interval smoothing results are derived. We show that L2 or risk-neutral filtering for HMMs can be extracted as a limiting case of the risk-sensitive filtering problem when the risk-sensitive parameter approaches zero.