arXiv subjects

Srikrishna Bhashyam

Publications and source records attributed to Srikrishna Bhashyam.

At least 19 recordsLinked to original sources

arXiv2026

Online Clustering of Data Sequences with Bandit Information

We study the problem of online clustering of data sequences in the multi-armed bandit (MAB) framework under the fixed-confidence setting. There are $M$ arms, each providing i.i.d. samples from a parametric distribution whose parameters are unknown. The $M$ arms form $K$ clusters based on the distance between the true parameters. In the MAB setting, one arm can be sampled at each time. The objective is to estimate the clusters of the arms using as few samples as possible from the arms, subject to an upper bound on the error probability. Our setting allows for: arms within a cluster to have non-identical distributions, vector parameter arms, vector observations, and $K \le M$ clusters. We propose and analyze the Average Tracking Bandit Online Clustering (ATBOC) algorithm. ATBOC is asymptotically order-optimal for multivariate Gaussian arms, with expected sample complexity grows at most twice as fast as the lower bound as $δ\rightarrow 0$, and this guarantee extends to multivariate sub-Gaussian arms. For single-parameter exponential family arms, ATBOC is asymptotically optimal, matching the lower bound. We also propose a computationally more efficient alternatives Lower and Upper Confidence Bound based Bandit Online Clustering Algorithm (LUCBBOC), and Bandit Online Clustering-Elimination (BOC-ELIM). We derive the computational complexity of the proposed algorithms and compare their per-sample runtime through simulations. LUCBBOC and BOC-ELIM require lower per-sample runtime than ATBOC while achieving comparable performance. All the proposed algorithms are $δ$-Probably correct, i.e., the error probability of cluster estimate at the stopping time is atmost $δ$. We validate the asymptotic optimality guarantees through simulations, and present the comparison of our proposed algorithms with other related work through simulations on both synthetic and real-world datasets.

cs.LG↗

arXiv2026

Sequential Spectral Clustering of Data Sequences

We study the problem of non-parametric clustering of data sequences, where each data sequence comprises independent and identically distributed (i.i.d.) samples generated from an unknown distribution. The true clusters are the clusters obtained using the Spectral clustering algorithm (SPEC) on the pairwise distance between the true distributions corresponding to the data sequences. Since the true distributions are unknown, the objective is to estimate the clusters by observing the minimum number of samples from the data sequences, given a specified error probability. To solve this problem, we propose the Sequential Spectral clustering algorithm (SEQ-SPEC), and show that it stops in finite time almost surely and is exponentially consistent. We also propose a computationally more efficient algorithm called the Incremental Approximate Sequential Spectral clustering algorithm (IA-SEQ-SPEC). Through simulations, we show that both SEQ-SPEC and IA-SEQ-SPEC perform better than the fixed sample size SPEC, the Sequential $K$-Medoids clustering algorithm (SEQ-KMED), and the Sequential Single Linkage clustering algorithm (SEQ-SLINK). In addition, we propose memory-efficient versions, SEQ-SPEC-B and IA-SEQ-SPEC-B. Unlike other related sequential clustering algorithms, which require storing all past samples, these algorithms require storing only the most recent $B$ samples. Both the computationally efficient and memory-efficient versions of SEQ-SPEC perform comparably to SEQ-SPEC in simulations.

eess.SP↗

arXiv2026

Efficient Clustering in Stochastic Bandits

We study the Bandit Clustering (BC) problem under the fixed confidence setting, where the objective is to group a collection of data sequences (arms) into clusters through sequential sampling from adaptively selected arms at each time step while ensuring a fixed error probability at the stopping time. We consider a setting where arms in a cluster may have different distributions. Unlike existing results in this setting, which assume Gaussian-distributed arms, we study a broader class of vector-parametric distributions that satisfy mild regularity conditions. Existing asymptotically optimal BC algorithms require solving an optimization problem as part of their sampling rule at each step, which is computationally costly. We propose an Efficient Bandit Clustering algorithm (EBC), which, instead of solving the full optimization problem, takes a single step toward the optimal value at each time step, making it computationally efficient while remaining asymptotically optimal. We also propose a heuristic variant of EBC, called EBC-H, which further simplifies the sampling rule, with arm selection based on quantities computed as part of the stopping rule. We highlight the computational efficiency of EBC and EBC-H by comparing their per-sample run time with that of existing algorithms. The asymptotic optimality of EBC is supported through simulations on the synthetic datasets. Through simulations on both synthetic and real-world datasets, we show the performance gain of EBC and EBC-H over existing approaches.

cs.LG↗

arXiv2025

Exponentially Consistent Nonparametric Linkage-Based Clustering of Data Sequences

In this paper, we consider nonparametric clustering of $M$ independent and identically distributed (i.i.d.) data sequences generated from {\em unknown} distributions. The distributions of the $M$ data sequences belong to $K$ underlying distribution clusters. Existing results on exponentially consistent nonparametric clustering algorithms, like single linkage-based (SLINK) clustering and $k$-medoids distribution clustering, assume that the maximum intra-cluster distance ($d_L$) is smaller than the minimum inter-cluster distance ($d_H$). First, in the fixed sample size (FSS) setting, we show that exponential consistency can be achieved for SLINK clustering under a less strict assumption, $d_I < d_H$, where $d_I$ is the maximum distance between any two sub-clusters of a cluster that partition the cluster. Note that $d_I < d_L$ in general. Thus, our results show that SLINK is exponentially consistent for a larger class of problems than previously known. In our simulations, we also identify examples where $k$-medoids clustering is unable to find the true clusters, but SLINK is exponentially consistent. Then, we propose a sequential clustering algorithm, named SLINK-SEQ, based on SLINK and prove that it is also exponentially consistent. Simulation results show that the SLINK-SEQ algorithm requires fewer expected number of samples than the FSS SLINK algorithm for the same probability of error.

stat.ML↗

arXiv2024

Simultaneous Communication and Tracking using Fused Bistatic Measurements

In this paper, we propose a bistatic sensing-assisted beam tracking method for simultaneous communication and tracking of user vehicles navigating arbitrary-shaped road trajectories. Prior work on simultaneous communication and tracking assumes a colocated radar receiver at the transmitter for sensing measurements using the reflected Integrated Sensing and Communication (ISAC) signals in the mmWave band. Full isolation between transmitter and receiver is required here to avoid self-interference. We consider the bistatic setting where the sensing receivers are not colocated and can be realized in practice using traditional half-duplex transmit or receive nodes. First, we process the echoes reflected from the vehicle at multiple multi-antenna nodes at various locations, facilitating estimation of the vehicle's current position. Then, we propose selection criteria for the estimates and a maximum likelihood (ML) fusion scheme to fuse these selected estimates based on the estimated error covariance matrices of these measurements. This fusion scheme is important in bistatic and multistatic settings as the localization error depends significantly on the geometry of the transmitter, target, and receiver locations. Finally, we predict the vehicle's next location using a simple kinematic equation-based model. Through extensive simulation, we study the average spectral efficiency of communication with a moving user using the proposed simultaneous communication and tracking scheme. The proposed fusion-based scheme achieves almost the same average spectral efficiency as an ideal scheme that knows the exact trajectory. We also show that the proposed scheme can be easily extended to systems with Hybrid Digital-Analog architectures and performs similarly even in these systems.

cs.IT↗

arXiv2022

Learning to detect an oddball target with observations from an exponential family

The problem of detecting an odd arm from a set of K arms of a multi-armed bandit, with fixed confidence, is studied in a sequential decision-making scenario. Each arm's signal follows a distribution from a vector exponential family. All arms have the same parameters except the odd arm. The actual parameters of the odd and non-odd arms are unknown to the decision maker. Further, the decision maker incurs a cost for switching from one arm to another. This is a sequential decision making problem where the decision maker gets only a limited view of the true state of nature at each stage, but can control his view by choosing the arm to observe at each stage. Of interest are policies that satisfy a given constraint on the probability of false detection. An information-theoretic lower bound on the total cost (expected time for a reliable decision plus total switching cost) is first identified, and a variation on a sequential policy based on the generalised likelihood ratio statistic is then studied. Thanks to the vector exponential family assumption, the signal processing in this policy at each stage turns out to be very simple, in that the associated conjugate prior enables easy updates of the posterior distribution of the model parameters. The policy, with a suitable threshold, is shown to satisfy the given constraint on the probability of false detection. Further, the proposed policy is asymptotically optimal in terms of the total cost among all policies that satisfy the constraint on the probability of false detection.

cs.IT↗

arXiv2020

Sequential Multi-hypothesis Testing in Multi-armed Bandit Problems:An Approach for Asymptotic Optimality

We consider a multi-hypothesis testing problem involving a K-armed bandit. Each arm's signal follows a distribution from a vector exponential family. The actual parameters of the arms are unknown to the decision maker. The decision maker incurs a delay cost for delay until a decision and a switching cost whenever he switches from one arm to another. His goal is to minimise the overall cost until a decision is reached on the true hypothesis. Of interest are policies that satisfy a given constraint on the probability of false detection. This is a sequential decision making problem where the decision maker gets only a limited view of the true state of nature at each stage, but can control his view by choosing the arm to observe at each stage. An information-theoretic lower bound on the total cost (expected time for a reliable decision plus total switching cost) is first identified, and a variation on a sequential policy based on the generalised likelihood ratio statistic is then studied. Due to the vector exponential family assumption, the signal processing at each stage is simple; the associated conjugate prior distribution on the unknown model parameters enables easy updates of the posterior distribution. The proposed policy, with a suitable threshold for stopping, is shown to satisfy the given constraint on the probability of false detection. Under a continuous selection assumption, the policy is also shown to be asymptotically optimal in terms of the total cost among all policies that satisfy the constraint on the probability of false detection.

cs.IT↗

arXiv2019

Sequential Controlled Sensing for Composite Multihypothesis Testing

The problem of multi-hypothesis testing with controlled sensing of observations is considered. The distribution of observations collected under each control is assumed to follow a single-parameter exponential family distribution. The goal is to design a policy to find the true hypothesis with minimum expected delay while ensuring that the probability of error is below a given constraint. The decision-maker can control the delay by intelligently choosing the control for observation collection in each time slot. We derive a policy that satisfies the given constraint on the error probability. We also show that the policy is asymptotically optimal in the sense that it asymptotically achieves an information-theoretic lower bound on the expected delay.

math.ST↗

arXiv2019

Change Detection with Sparse Signals using Quantum Designs

We consider the change detection problem where the pre-change observation vectors are purely noise and the post-change observation vectors are noise-corrupted compressive measurements of sparse signals with a common support, measured using a sensing matrix. In general, post-change distribution of the observations depends on parameters such as the support and variances of the sparse signal. When these parameters are unknown, we propose two approaches. In the first approach, we approximate the post-change pdf based on the known parameters such as mutual coherence of the sensing matrix and bounds on the signal variances. In the second approach, we parameterize the post-change pdf with an unknown parameter and try to adaptively estimate this parameter using a stochastic gradient descent method. In both these approaches, we employ CUSUM algorithm with various decision statistics such as the energy of the observations, correlation values with columns of the sensing matrix and the maximum value of such correlations. We study the performance of these approaches and offer insights on the relevance of different decision statistics in different SNR regimes. We also address the problem of designing sensing matrices with small coherence by using designs from quantum information theory. One such design, called SIC POVM, also has an additional structure which allows exact computation of the post-change pdfs of some decision statistics even when the support set of the sparse signal is unknown. We apply our detection algorithms with SIC POVM based sequences to a massive random access problem and show their superior performance over conventional Gold codes.

eess.SP↗

arXiv2019

Channel Conditions for the Optimality of Interference Decoding Schemes for K-user Gaussian Interference Channels

The Han-Kobayashi (HK) scheme achieves the best known achievable rate region for the K user interference channel (IC). Simple HK schemes are HK schemes with Gaussian signaling, no time sharing, and no private-common power splitting. The class of simple HK schemes includes the treating interference as noise (TIN) scheme and schemes that involve various levels of interference decoding and cancellation at each receiver. We derive conditions under which simple HK schemes achieve sum capacity for general K user Gaussian ICs. These results generalize existing sum capacity results for the TIN scheme to the class of simple HK schemes.

cs.IT↗

arXiv2017

Almost Budget Balanced Mechanisms with Scalar Bids For Allocation of a Divisible Good

This paper is about allocation of an infinitely divisible good to several rational and strategic agents. The allocation is done by a social planner who has limited information because the agents' valuation functions are taken to be private information known only to the respective agents. We allow only a scalar signal, called a bid, from each agent to the social planner. Yang and Hajek [Jour. on Selected Areas in Comm., 2007] as well as Johari and Tsitsiklis [Jour. of Oper. Res., 2009] proposed a scalar strategy Vickrey-Clarke-Groves (SSVCG) mechanism with efficient Nash equilibria. We consider a setting where the social planner desires minimal budget surplus. Example situations include fair sharing of Internet resources and auctioning of certain public goods where revenue maximization is not a consideration. Under the SSVCG framework, we propose a mechanism that is efficient and comes close to budget balance by returning much of the payments back to the agents in the form of rebates. We identify a design criterion for {\em almost budget balance}, impose feasibility and voluntary participation constraints, simplify the constraints, and arrive at a convex optimization problem to identify the parameters of the rebate functions. The convex optimization problem has a linear objective function and a continuum of linear constraints. We propose a solution method that involves a finite number of constraints, and identify the number of samples sufficient for a good approximation.

cs.GT↗

arXiv2017

On the Sum Capacity of Many-to-one and One-to-many Gaussian Interference Channels

We obtain new sum capacity results for the Gaussian many-to-one and one-to-many interference channels in channel parameter regimes where the sum capacity was known only up to a constant gap. Simple Han-Kobayashi (HK) schemes, i.e., HK schemes with Gaussian signaling, no time-sharing, and no common-private power splitting, achieve sum capacity under the channel conditions for which the new results are obtained. To obtain sum capacity results, we show that genie-aided upper bounds match the achievable sum rate of simple HK schemes under certain channel conditions.

cs.IT↗

arXiv2016

On the Sum Rate of a 2 x 2 Interference Network

In an M x N interference network, there are M transmitters and N receivers with each transmitter having independent messages for each of the 2^N -1 possible non-empty subsets of the receivers. We consider the 2 x 2 interference network with 6 possible messages, of which the 2 x 2 interference channel and X channel are special cases obtained by using only 2 and 4 messages respectively. Starting from an achievable rate region similar to the Han-Kobayashi region, we obtain an achievable sum rate. For the Gaussian interference network, we determine which of the 6 messages are sufficient for maximizing the sum rate within this rate region for the low, mixed, and strong interference conditions. It is observed that 2 messages are sufficient in several cases.

cs.IT↗

arXiv2016

On the Capacity of the Half-Duplex MIMO Gaussian Diamond Channel

In this paper, we analyze the 2-relay multiple-input multiple-output (MIMO) Gaussian diamond channel. We show that a multihopping decode-and-forward with multiple access (MDF-MAC) protocol achieves rates within a constant gap from capacity when a channel parameter $Δ$ is greater than zero. We also identify the transmit covariance matrices to be used by each relay in the multiple-access (MAC) state of the MDF-MAC protocol. As done for the single-antenna 2-relay Gaussian diamond channel, the channel parameter $Δ$ is defined to be the difference between the product of the capacities of the links from the source to the two relays and the product of the capacities of the links from the two relays to the destination.

cs.IT↗

arXiv2015

On the Gaussian Many-to-One X Channel

In this paper, the Gaussian many-to-one X channel, which is a special case of general multiuser X channel, is studied. In the Gaussian many-to-one X channel, communication links exist between all transmitters and one of the receivers, along with a communication link between each transmitter and its corresponding receiver. As per the X channel assumption, transmission of messages is allowed on all the links of the channel. This communication model is different from the corresponding many-to-one interference channel (IC). Transmission strategies which involve using Gaussian codebooks and treating interference from a subset of transmitters as noise are formulated for the above channel. Sum-rate is used as the criterion of optimality for evaluating the strategies. Initially, a $3 \times 3$ many-to-one X channel is considered and three transmission strategies are analyzed. The first two strategies are shown to achieve sum-rate capacity under certain channel conditions. For the third strategy, a sum-rate outer bound is derived and the gap between the outer bound and the achieved rate is characterized. These results are later extended to the $K \times K$ case. Next, a region in which the many-to-one X channel can be operated as a many-to-one IC without loss of sum-rate is identified. Further, in the above region, it is shown that using Gaussian codebooks and treating interference as noise achieves a rate point that is within $K/2 -1$ bits from the sum-rate capacity. Subsequently, some implications of the above results to the Gaussian many-to-one IC are discussed. Transmission strategies for the many-to-one IC are formulated and channel conditions under which the strategies achieve sum-rate capacity are obtained. A region where the sum-rate capacity can be characterized to within $K/2-1$ bits is also identified.

cs.IT↗

arXiv2015

On the Sum Capacity of the Gaussian X Channel in the Mixed Interference Regime

In this paper, we analyze the Gaussian X channel in the mixed interference regime. In this regime, multiple access transmission to one of the receivers is shown to be close to optimal in terms of sum rate. Three upper bounds are derived for the sum capacity in the mixed interference regime, and the subregions where each of these bounds dominate the others are identified. The genie-aided sum capacity upper bounds derived also show that the gap between sum capacity and the sum rate of the multiple access transmission scheme is small for a significant part of the mixed interference region. For any δ> 0, the region where multiple access transmission to one of the receivers is within δfrom sum capacity is determined.

cs.IT↗

arXiv2013

The Gaussian Two-way Diamond Channel

We consider two-way relaying in a Gaussian diamond channel, where two terminal nodes wish to exchange information using two relays. A simple baseline protocol is obtained by time-sharing between two one-way protocols. To improve upon the baseline performance, we propose two compute-and-forward (CF) protocols: Compute-and-forward Compound multiple access channel (CF-CMAC) and Compute-and-forward-Broadcast (CF-BC). These protocols mix the two flows through the two relays and achieve rates better than the simple time-sharing protocol. We derive an outer bound to the capacity region that is satisfied by any relaying protocol, and observe that the proposed protocols provide rates close to the outer bound in certain channel conditions. Both the CF-CMAC and CF-BC protocols use nested lattice codes in the compute phases. In the CF-CMAC protocol, both relays simultaneously forward to the destinations over a Compound Multiple Access Channel (CMAC). In the simpler CF-BC protocol's forward phase, one relay is selected at a time for Broadcast Channel (BC) transmission depending on the rate-pair to be achieved. We also consider the diamond channel with direct source-destination link and the diamond channel with interfering relays. Outer bounds and achievable rate regions are compared for these two channels as well. Mixing of flows using the CF-CMAC protocol is shown to be good for symmetric two-way rates.

cs.IT↗

arXiv2013

Cross-Layer Strategies for Throughput Maximization in Data Aggregating Wireless Networks

We consider a data aggregating wireless network where all nodes have data to send to a single destination node, the sink. We consider a linear placement of nodes with the sink at one end. The nodes communicate directly to the sink (single hop transmission) and we assume that the nodes are scheduled one at a time by a central scheduler (possibly the sink). The wireless nodes are power limited and our network objective (notion of fairness) is to maximize the minimum throughput of the nodes subject to the node power constraints. In this work, we consider network designs that permit adapting node transmission time, node transmission power and node placements, and study cross- layer strategies that seek to maximize the network throughput. Using simulations, we characterize the performance of the dif- ferent strategies and comment on their applicability for various network scenarios.

cs.NI↗