Siva Theja Maguluri
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Siva Theja Maguluri has a total of 67 co-authors. Here are the top co-authors by number of shared publications:
- Sushil Mahavir Varma
School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332
16 shared publications - Daniela Hurtado-Lange
Kellogg School of Management, Northwestern University, Evanston, Illinois, USA
6 shared publications - Francisco Castro
Anderson School of Management, University of California, Los Angeles, Los Angeles, California 90095
6 shared publications - Thinh T. Doan
Department of Aerospace Engineering and Engineering Mechanics, University of Texas at Austin, Austin, TX, USA
6 shared publications - Daniela Hurtado-Lange
Georgia Institute of Technology Georgia Institute of Technology
5 shared publications - Justin Romberg
Georgia Institute of Technology
5 shared publications
Latest Publications
- 22 September 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
A Heavy Traffic Theory of Matching Queues
Motivated by emerging applications in online matching platforms and marketplaces, we study a matching queue. Customers and servers that arrive in a matching queue depart as soon as they are matched. It is known that a matching queue without an external control is unstable, and so we study its behavior for a general state-dependent control. Whereas state-dependent control is an effective lever to regulate the throughput and delay, it often comes at a cost for matching platforms in practice. Optimizing this fundamental trade-off motivates the use of small amounts of control, so we study a matching queue in an asymptotic regime in which the state-dependent control decreases to zero. Unlike the heavy traffic regime in classical queues, there are two different ways the control can be sent to zero: via a magnitude scaling parameter [Formula: see text] that goes to zero and a time scaling parameter [Formula: see text] that goes to infinity. Depending on the cost of control, we show that the rates of [Formula: see text] and [Formula: see text] that optimize the trade-off between delay and cost of control could correspond to three different regimes. As we traverse these regimes, we observe a phase transition in the limiting distribution of the matching queue. We show that a low cost of control corresponds to the regime [Formula: see text], and we call it the delay-driven regime. The limiting behavior in this regime is an asymmetrical Laplace distribution. On the other hand, [Formula: see text] is the cost-driven regime corresponding to a high cost of control in which the limiting behavior is either a uniform or a truncated exponential distribution. We christen the in-between regime of [Formula: see text] the hybrid regime in which the limiting behavior is a Gibbs distribution. These results are obtained by novel generalizations of the transform method, in which each regime requires new ideas. The hybrid regime employs inverse Fourier transforms, whereas the other two regimes engineer multiple complex exponential test functions. Funding: This work was supported by the National Science Foundation [Grants CMMI-2140534 and EPCN-2144316].
- 22 September 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
A Heavy Traffic Theory of Matching Queues
Motivated by emerging applications in online matching platforms and marketplaces, we study a matching queue. Customers and servers that arrive in a matching queue depart as soon as they are matched. It is known that a matching queue without an external control is unstable, and so we study its behavior for a general state-dependent control. Whereas state-dependent control is an effective lever to regulate the throughput and delay, it often comes at a cost for matching platforms in practice. Optimizing this fundamental trade-off motivates the use of small amounts of control, so we study a matching queue in an asymptotic regime in which the state-dependent control decreases to zero. Unlike the heavy traffic regime in classical queues, there are two different ways the control can be sent to zero: via a magnitude scaling parameter [Formula: see text] that goes to zero and a time scaling parameter [Formula: see text] that goes to infinity. Depending on the cost of control, we show that the rates of [Formula: see text] and [Formula: see text] that optimize the trade-off between delay and cost of control could correspond to three different regimes. As we traverse these regimes, we observe a phase transition in the limiting distribution of the matching queue. We show that a low cost of control corresponds to the regime [Formula: see text], and we call it the delay-driven regime. The limiting behavior in this regime is an asymmetrical Laplace distribution. On the other hand, [Formula: see text] is the cost-driven regime corresponding to a high cost of control in which the limiting behavior is either a uniform or a truncated exponential distribution. We christen the in-between regime of [Formula: see text] the hybrid regime in which the limiting behavior is a Gibbs distribution. These results are obtained by novel generalizations of the transform method, in which each regime requires new ideas. The hybrid regime employs inverse Fourier transforms, whereas the other two regimes engineer multiple complex exponential test functions. Funding: This work was supported by the National Science Foundation [Grants CMMI-2140534 and EPCN-2144316].
- 22 September 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
Erratum: “Dynamic Pricing and Matching for Two-Sided Queues”
Corrections to the Publication “Dynamic Pricing and Matching for Two-Sided Queues” In the context of two-sided queues, Theorem 2 of the paper [Dynamic Pricing and Matching for Two-Sided Queues. Operations Research 2023;71(1):83–100] establishes optimality guarantees for the two-price policy combined with the max-weight matching policies. As stated, this theorem is not valid. In this erratum, we show that the conclusions of Theorem 2 hold under a slight modification of the max-weight policy, which does not use some of the edges in the graph to match. Moreover, we point out some technical issues in this paper and outline how to resolve them.
- 22 September 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
Erratum: “Dynamic Pricing and Matching for Two-Sided Queues”
Corrections to the Publication “Dynamic Pricing and Matching for Two-Sided Queues” In the context of two-sided queues, Theorem 2 of the paper [Dynamic Pricing and Matching for Two-Sided Queues. Operations Research 2023;71(1):83–100] establishes optimality guarantees for the two-price policy combined with the max-weight matching policies. As stated, this theorem is not valid. In this erratum, we show that the conclusions of Theorem 2 hold under a slight modification of the max-weight policy, which does not use some of the edges in the graph to match. Moreover, we point out some technical issues in this paper and outline how to resolve them.
- 18 September 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
A Nonasymptotic Theory of Seminorm Lyapunov Stability: From Deterministic to Stochastic Iterative Algorithms
We study the problem of solving fixed-point equations for seminorm-contractive operators and establish foundational results on the nonasymptotic behavior of iterative algorithms in both deterministic and stochastic settings. In the deterministic setting, we present a fixed-point theorem for seminorm-contractive operators, showing that the iterates converge geometrically to the kernel of the seminorm. In the stochastic setting, which is our main focus, we analyze stochastic approximation (SA) algorithms under seminorm-contractive operators and Markovian noise, providing a finite-sample analysis for various step size choices. A benchmark for equation solving is linear systems of equations, in which the convergence behavior of fixed-point iteration is closely tied to the stability of linear dynamical systems. In this special case, our results provide a characterization of system stability with respect to a seminorm, linking it to the solution of a Lyapunov equation in terms of positive semidefinite matrices. In the stochastic setting, we establish a finite-sample analysis for linear Markovian SA without requiring the Hurwitzness assumption. Our theoretical results offer a unified framework for deriving finite-sample bounds for reinforcement learning algorithms in the average reward setting, including TD([Formula: see text]) for policy evaluation (which is a special case of solving a Poisson equation) and Q-learning for control. Funding: This work was partially supported by the National Science Foundation [Grants EPCN-2144316, CPS-2240982, CMMI-2112533], a seed grant from Georgia Tech, and an award from Raytheon Technologies. Supplemental Material: The online appendix is available at https://doi.org/10.1287/moor.2025.0920 .
- 18 September 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
A Nonasymptotic Theory of Seminorm Lyapunov Stability: From Deterministic to Stochastic Iterative Algorithms
We study the problem of solving fixed-point equations for seminorm-contractive operators and establish foundational results on the nonasymptotic behavior of iterative algorithms in both deterministic and stochastic settings. In the deterministic setting, we present a fixed-point theorem for seminorm-contractive operators, showing that the iterates converge geometrically to the kernel of the seminorm. In the stochastic setting, which is our main focus, we analyze stochastic approximation (SA) algorithms under seminorm-contractive operators and Markovian noise, providing a finite-sample analysis for various step size choices. A benchmark for equation solving is linear systems of equations, in which the convergence behavior of fixed-point iteration is closely tied to the stability of linear dynamical systems. In this special case, our results provide a characterization of system stability with respect to a seminorm, linking it to the solution of a Lyapunov equation in terms of positive semidefinite matrices. In the stochastic setting, we establish a finite-sample analysis for linear Markovian SA without requiring the Hurwitzness assumption. Our theoretical results offer a unified framework for deriving finite-sample bounds for reinforcement learning algorithms in the average reward setting, including TD([Formula: see text]) for policy evaluation (which is a special case of solving a Poisson equation) and Q-learning for control. Funding: This work was partially supported by the National Science Foundation [Grants EPCN-2144316, CPS-2240982, CMMI-2112533], a seed grant from Georgia Tech, and an award from Raytheon Technologies. Supplemental Material: The online appendix is available at https://doi.org/10.1287/moor.2025.0920 .
- Open Access
- 9 September 2026
- Published by Cambridge University Press (CUP) in Journal of Applied Probability
- p. 1-10
Convergence rate of the join-the-shortest-queue system
The join-the-shortest-queue (JSQ) policy is among the most widely used load-balancing algorithms and has been extensively studied. However, an exact characterization of the system behavior remains challenging. Most prior research has focused on analyzing its performance in the steady state in certain asymptotic regimes, such as the heavy-traffic regime. However, convergence to the steady state in these regimes is often slow, so steady-state and heavy-traffic characterizations may be less informative over practical time horizons. To address this limitation, we provide a finite-time convergence rate analysis of a JSQ system with two symmetric servers. In sharp contrast to the existing literature, we directly study the original system rather than an approximate limiting system such as a diffusion approximation. Our results demonstrate that for such a system, the convergence rate to its steady state, measured in the total variation distance, is upper O left parenthesis left parenthesis 1 divided by left parenthesis 1 minus rho right parenthesis cubed right parenthesis left parenthesis 1 divided by t right parenthesis right parenthesis $O((1/(1-\rho)^3)(1/t))$ , where rho element of left parenthesis 0 comma 1 right parenthesis $\rho \in (0,1)$ is the traffic intensity.- 9 September 2026
- Published by Cambridge University Press (CUP) in Journal of Applied Probability
- p. 1-10
Open AccessConvergence rate of the join-the-shortest-queue system
The join-the-shortest-queue (JSQ) policy is among the most widely used load-balancing algorithms and has been extensively studied. However, an exact characterization of the system behavior remains challenging. Most prior research has focused on analyzing its performance in the steady state in certain asymptotic regimes, such as the heavy-traffic regime. However, convergence to the steady state in these regimes is often slow, so steady-state and heavy-traffic characterizations may be less informative over practical time horizons. To address this limitation, we provide a finite-time convergence rate analysis of a JSQ system with two symmetric servers. In sharp contrast to the existing literature, we directly study the original system rather than an approximate limiting system such as a diffusion approximation. Our results demonstrate that for such a system, the convergence rate to its steady state, measured in the total variation distance, is upper O left parenthesis left parenthesis 1 divided by left parenthesis 1 minus rho right parenthesis cubed right parenthesis left parenthesis 1 divided by t right parenthesis right parenthesis $O((1/(1-\rho)^3)(1/t))$ , where rho element of left parenthesis 0 comma 1 right parenthesis $\rho \in (0,1)$ is the traffic intensity. - Open Access
- 20 July 2026
- Published by Association for Computing Machinery (ACM) in ACM SIGMETRICS Performance Evaluation Review
- Vol. 54 (1) , 55-57
Higher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
We study the stationary sojourn time distribution in an M/G/1 queue operating under heavy traffic. It is known that the sojourn time converges to an exponential distribution in the limit. Our focus is on obtaining pre-asymptotic, higher-order approximations that go beyond the classical exponential limit. Using Stein's method, we develop an approach based on higher-order expansions of the generator of the underlying Markov process. The key technical step is to represent higher-order derivatives in terms of lower-order ones and control the resulting error via derivative bounds of the Stein equation. Under suitable moment-matching conditions on the service distribution, we show that the approximation error decays as a high-order power of the slack parameter ε=1-ρ. Error bounds are established in the Zolotarev metric, which further imply bounds on the Wasserstein distance as well as the moments. Our results demonstrate that the accuracy of the exponential approximation can be systematically improved by matching progressively more moments of the service distribution.
- 20 July 2026
- Published by Association for Computing Machinery (ACM) in ACM SIGMETRICS Performance Evaluation Review
- Vol. 54 (1) , 55-57
Open AccessHigher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
We study the stationary sojourn time distribution in an M/G/1 queue operating under heavy traffic. It is known that the sojourn time converges to an exponential distribution in the limit. Our focus is on obtaining pre-asymptotic, higher-order approximations that go beyond the classical exponential limit. Using Stein's method, we develop an approach based on higher-order expansions of the generator of the underlying Markov process. The key technical step is to represent higher-order derivatives in terms of lower-order ones and control the resulting error via derivative bounds of the Stein equation. Under suitable moment-matching conditions on the service distribution, we show that the approximation error decays as a high-order power of the slack parameter ε=1-ρ. Error bounds are established in the Zolotarev metric, which further imply bounds on the Wasserstein distance as well as the moments. Our results demonstrate that the accuracy of the exponential approximation can be systematically improved by matching progressively more moments of the service distribution.
- Open Access
- 8 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 55-57
Higher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
- 8 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 55-57
Open AccessHigher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
- Top Cited PapersOpen Access
- 26 March 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Measurement and Analysis of Computing Systems
- Vol. 10 (1) , 1-36
Higher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
We study the stationary sojourn time distribution in an M/G/1 queue operating under heavy traffic. It is known that the sojourn time converges to an exponential distribution in the limit. Our focus is on obtaining pre-asymptotic, higher-order approximations that go beyond the classical exponential limit. Using Stein's method, we develop an approach based on higher-order expansions of the generator of the underlying Markov process. The key technical step is to represent higher-order derivatives in terms of lower-order ones and control the resulting error via derivative bounds of the Stein equation. Under suitable moment-matching conditions on the service distribution, we show that the approximation error decays as a high-order power of the slack parameter ? = 1 ? ?. Error bounds are established in the Zolotarev metric, which further imply bounds on the Wasserstein distance as well as the moments. Our results demonstrate that the accuracy of the exponential approximation can be systematically improved by matching progressively more moments of the service distribution.
- 26 March 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Measurement and Analysis of Computing Systems
- Vol. 10 (1) , 1-36
Top Cited PapersOpen AccessHigher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
We study the stationary sojourn time distribution in an M/G/1 queue operating under heavy traffic. It is known that the sojourn time converges to an exponential distribution in the limit. Our focus is on obtaining pre-asymptotic, higher-order approximations that go beyond the classical exponential limit. Using Stein's method, we develop an approach based on higher-order expansions of the generator of the underlying Markov process. The key technical step is to represent higher-order derivatives in terms of lower-order ones and control the resulting error via derivative bounds of the Stein equation. Under suitable moment-matching conditions on the service distribution, we show that the approximation error decays as a high-order power of the slack parameter ? = 1 ? ?. Error bounds are established in the Zolotarev metric, which further imply bounds on the Wasserstein distance as well as the moments. Our results demonstrate that the accuracy of the exponential approximation can be systematically improved by matching progressively more moments of the service distribution.
- Open Access
Steady-State Behavior of Constant-Stepsize Stochastic Approximation: Gaussian Approximation and Tail Bounds
Constant-stepsize stochastic approximation (SA) is widely used in learning for computational efficiency. For a fixed stepsize, the iterates typically admit a stationary distribution that is rarely tractable. Prior work shows that as the stepsize $α\downarrow 0$, the centered-and-scaled steady state converges weakly to a Gaussian random vector. However, for fixed $α$, this weak convergence offers no usable error bound for approximating the steady-state by its Gaussian limit. This paper provides explicit, non-asymptotic error bounds for fixed $α$. We first prove general-purpose theorems that bound the Wasserstein distance between the centered-scaled steady state and an appropriate Gaussian distribution, under regularity conditions for drift and moment conditions for noise. To ensure broad applicability, we cover both i.i.d. and Markovian noise models. We then instantiate these theorems for three representative SA settings: (1) stochastic gradient descent (SGD) for smooth strongly convex objectives, (2) linear SA, and (3) contractive nonlinear SA. We obtain dimension- and stepsize-dependent, explicit bounds in Wasserstein distance of order $α^{1/2}\log(1/α)$ for small $α$. Building on the Wasserstein approximation error, we further derive non-uniform Berry--Esseen-type tail bounds that compare the steady-state tail probability to Gaussian tails. We achieve an explicit error term that decays in both the deviation level and stepsize $α$. We adapt the same analysis for SGD beyond strongly convexity and study general convex objectives. We identify a non-Gaussian (Gibbs) limiting law under the correct scaling, which is validated numerically, and provide a corresponding pre-limit Wasserstein error bound.Open AccessSteady-State Behavior of Constant-Stepsize Stochastic Approximation: Gaussian Approximation and Tail Bounds
Constant-stepsize stochastic approximation (SA) is widely used in learning for computational efficiency. For a fixed stepsize, the iterates typically admit a stationary distribution that is rarely tractable. Prior work shows that as the stepsize $α\downarrow 0$, the centered-and-scaled steady state converges weakly to a Gaussian random vector. However, for fixed $α$, this weak convergence offers no usable error bound for approximating the steady-state by its Gaussian limit. This paper provides explicit, non-asymptotic error bounds for fixed $α$. We first prove general-purpose theorems that bound the Wasserstein distance between the centered-scaled steady state and an appropriate Gaussian distribution, under regularity conditions for drift and moment conditions for noise. To ensure broad applicability, we cover both i.i.d. and Markovian noise models. We then instantiate these theorems for three representative SA settings: (1) stochastic gradient descent (SGD) for smooth strongly convex objectives, (2) linear SA, and (3) contractive nonlinear SA. We obtain dimension- and stepsize-dependent, explicit bounds in Wasserstein distance of order $α^{1/2}\log(1/α)$ for small $α$. Building on the Wasserstein approximation error, we further derive non-uniform Berry--Esseen-type tail bounds that compare the steady-state tail probability to Gaussian tails. We achieve an explicit error term that decays in both the deviation level and stepsize $α$. We adapt the same analysis for SGD beyond strongly convexity and study general convex objectives. We identify a non-Gaussian (Gibbs) limiting law under the correct scaling, which is validated numerically, and provide a corresponding pre-limit Wasserstein error bound. - Open Access
Higher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
We study the stationary sojourn time distribution in an M/G/1 queue operating under heavy traffic. It is known that the sojourn time converges to an exponential distribution in the limit. Our focus is on obtaining pre-asymptotic, higher-order approximations that go beyond the classical exponential limit. Using Stein's method, we develop an approach based on higher-order expansions of the generator of the underlying Markov process. The key technical step is to represent higher-order derivatives in terms of lower-order ones and control the resulting error via derivative bounds of the Stein equation. Under suitable moment-matching conditions on the service distribution, we show that the approximation error decays as a high-order power of the slack parameter $\varepsilon=1-ρ$. Error bounds are established in the Zolotarev metric, which further imply bounds on the Wasserstein distance as well as the moments. Our results demonstrate that the accuracy of the exponential approximation can be systematically improved by matching progressively more moments of the service distribution.Open AccessHigher-Order Approximations of Sojourn Times in M/G/1 Queues via Stein's Method
We study the stationary sojourn time distribution in an M/G/1 queue operating under heavy traffic. It is known that the sojourn time converges to an exponential distribution in the limit. Our focus is on obtaining pre-asymptotic, higher-order approximations that go beyond the classical exponential limit. Using Stein's method, we develop an approach based on higher-order expansions of the generator of the underlying Markov process. The key technical step is to represent higher-order derivatives in terms of lower-order ones and control the resulting error via derivative bounds of the Stein equation. Under suitable moment-matching conditions on the service distribution, we show that the approximation error decays as a high-order power of the slack parameter $\varepsilon=1-ρ$. Error bounds are established in the Zolotarev metric, which further imply bounds on the Wasserstein distance as well as the moments. Our results demonstrate that the accuracy of the exponential approximation can be systematically improved by matching progressively more moments of the service distribution. - 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 71 (6) , 3942-3957
Stochastic Approximation for Nonlinear Discrete Stochastic Control: Finite-Sample Bounds
In this article, we consider a nonlinear discrete stochastic control system, and our goal is to design a feedback control policy in order to lead the system to a prespecified state. We adopt a stochastic approximation viewpoint of this problem. It is known that by solving the corresponding continuous-time deterministic system and using the resulting feedback control policy, one ensures almost sure convergence to the prespecified state in the discrete system. In this article, we adopt such a control mechanism and provide its finite-sample convergence bounds whenever a Lyapunov function is known for the continuous system. In particular, we consider four cases based on whether the Lyapunov function for the continuous system gives exponential or subexponential rates and based on whether it is smooth or not. We provide the finite-time bounds in all cases. Our proof relies on constructing a Lyapunov function for the discrete system based on the given Lyapunov function for the continuous system. We do this by appropriately smoothing the given function using the Moreau envelope. We present numerical experiments corresponding to the various cases, which validate the rates we establish.
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 71 (6) , 3942-3957
Stochastic Approximation for Nonlinear Discrete Stochastic Control: Finite-Sample Bounds
In this article, we consider a nonlinear discrete stochastic control system, and our goal is to design a feedback control policy in order to lead the system to a prespecified state. We adopt a stochastic approximation viewpoint of this problem. It is known that by solving the corresponding continuous-time deterministic system and using the resulting feedback control policy, one ensures almost sure convergence to the prespecified state in the discrete system. In this article, we adopt such a control mechanism and provide its finite-sample convergence bounds whenever a Lyapunov function is known for the continuous system. In particular, we consider four cases based on whether the Lyapunov function for the continuous system gives exponential or subexponential rates and based on whether it is smooth or not. We provide the finite-time bounds in all cases. Our proof relies on constructing a Lyapunov function for the discrete system based on the given Lyapunov function for the continuous system. We do this by appropriately smoothing the given function using the Moreau envelope. We present numerical experiments corresponding to the various cases, which validate the rates we establish.