arXiv subjects

Veeraruna Kavitha

Publications and source records attributed to Veeraruna Kavitha.

52 records · Page 3Linked to original sources

arXiv2021

Random Fixed Points, Limits and Systemic risk

We consider vector fixed point (FP) equations in large dimensional spaces involving random variables, and study their realization-wise solutions. We have an underlying directed random graph, that defines the connections between various components of the FP equations. Existence of an edge between nodes i, j implies the i th FP equation depends on the j th component. We consider a special case where any component of the FP equation depends upon an appropriate aggregate of that of the random neighbor components. We obtain finite dimensional limit FP equations (in a much smaller dimensional space), whose solutions approximate the solution of the random FP equations for almost all realizations, in the asymptotic limit (number of components increase). Our techniques are different from the traditional mean-field methods, which deal with stochastic FP equations in the space of distributions to describe the stationary distributions of the systems. In contrast our focus is on realization-wise FP solutions. We apply the results to study systemic risk in a large financial heterogeneous network with many small institutions and one big institution, and demonstrate some interesting phenomenon.

math.PR↗

arXiv2021

Coalition Formation in Constant Sum Queueing Games

We analyse a coalition formation game between strategic service providers of a congestible service. The key novelty of our formulation is that it is a constant sum game, i.e., the total payoff across all service providers (or coalitions of providers) is fixed, and dictated by the total size of the market. The game thus captures the tension between resource pooling (to benefit from the resulting statistical economies of scale) and competition between coalitions over market share. In a departure from the prior literature on resource pooling for congestible services, we show that the grand coalition is in general not stable, once we allow for competition over market share. Instead, the stable configurations are duopolies, where the dominant coalition exploits its economies of scale to corner a disproportionate market share. We analyse the stable duopolies that emerge from this interaction, and also study a dynamic variant of this game.

cs.GT↗

arXiv2021

Evolutionary Vaccination Games with premature vaccines to combat ongoing deadly pandemic

We consider a vaccination game that results with the introduction of premature and possibly scarce vaccines introduced in a desperate bid to combat the otherwise ravaging deadly pandemic. The response of unsure agents amid many uncertainties makes this game completely different from the previous studies. We construct a framework that combines SIS epidemic model with a variety of dynamic behavioral vaccination responses and demographic aspects. The response of each agent is influenced by the vaccination hesitancy and urgency, which arise due to their personal belief about efficacy and side-effects of the vaccine, disease characteristics, and relevant reported information (e.g., side-effects, disease statistics etc.). Based on such aspects, we identify the responses that are stable against static mutations. By analysing the attractors of the resulting ODEs, we observe interesting patterns in the limiting state of the system under evolutionary stable (ES) strategies, as a function of various defining parameters. There are responses for which the disease is eradicated completely (at limiting state), but none are stable against mutations. Also, vaccination abundance results in higher infected fractions at ES limiting state, irrespective of the disease death rate.

math.DS↗

arXiv2021

Financial replicator dynamics: emergence of systemic-risk-averting strategies

We consider a random financial network with a large number of agents. The agents connect through credit instruments borrowed from each other or through direct lending, and these create the liabilities. The settlement of the debts of various agents at the end of the contract period can be expressed as solutions of random fixed point equations. Our first step is to derive these solutions (asymptotically), using a recent result on random fixed point equations. We consider a large population in which agents adapt one of the two available strategies, risky or risk-free investments, with an aim to maximize their expected returns (or surplus). We aim to study the emerging strategies when different types of replicator dynamics capture inter-agent interactions. We theoretically reduced the analysis of the complex system to that of an appropriate ordinary differential equation (ODE). We proved that the equilibrium strategies converge almost surely to that of an attractor of the ODE. We also derived the conditions under which a mixed evolutionary stable strategy (ESS) emerges; in these scenarios the replicator dynamics converges to an equilibrium at which the expected returns of both the populations are equal. Further the average dynamics (choices based on large observation sample) always averts systemic risk events (events with large fraction of defaults). We verified through Monte Carlo simulations that the equilibrium suggested by the ODE method indeed represents the limit of the dynamics.

q-fin.GN↗

arXiv2020

Cooperative Ressource Sharing With Adamant Player

Cooperative game theory deals with systems where players want to cooperate to improve their payoffs. But players may choose coalitions in a non-cooperative manner, leading to a coalition-formation game. We consider such a game with several players (willing to cooperate) and an adamant player (unwilling to cooperate) involved in resource-sharing. Here, the strategy of a player is the set of players with whom it wants to form a coalition. Given a strategy profile, an appropriate partition of coalitions is formed; players in each coalition maximize their collective utilities leading to a non-cooperative resource-sharing game among the coalitions, the utilities at the resulting equilibrium are shared via Shapley-value; these shares define the utilities of players for the given strategy profile in coalition-formation game. We also consider the utilitarian solution to derive the price of anarchy (PoA). We considered a case with symmetric players and an adamant player; wherein we observed that players prefer to stay alone at Nash equilibrium when the number of players (n) is more than 4. In contrast, in the majority of the cases, the utilitarian partition is grand coalition. Interestingly the PoA is smaller with an adamant player of intermediate strength. Further, PoA grows like O(n).

cs.GT↗

arXiv2020

Branching Process with Attack: Viral Competing Markets

The marked increase in advertisements over online social networks (OSNs) necessitates the study of content propagation. We analyse the viral markets with content providers competing for the propagation of similar posts over OSNs. Towards this, we required a new variant of the branching process (BP), which we named as "Branching process with attack"; the entities upon wake up attempt to attack and acquire the opposite population; furthermore, each entity produces its offsprings as is usually considered in BPs. In addition to providing expressions for the growth rates of individual posts, dichotomy etc., we explore the co-existence of posts; can the competing content spread and explode (number of unread/live copies of both posts grow significantly with time) simultaneously over the network? We prove that either one or both populations/posts get extinct or the populations settle to a unique co-existence equilibrium and derive the corresponding asymptotic ratios of the two populations/posts. Our analysis applies to large population networks focusing on mass behaviour, rather than micro details. Our study provides insights into two crucial design aspects, the number of seed users and the quality of the post.

cs.SI↗

arXiv2020

Asymmetric Information Acquisition Games

We consider a stochastic game with partial, asymmetric and non-classical information, where the agents are trying to acquire as many available opportunities/locks as possible. Agents have access only to local information, the information updates are asynchronous and our aim is to obtain relevant equilibrium policies. Our approach is to consider optimal open-loop control until the information update, which allows managing the belief updates in a structured manner. The agents continuously control the rates of their Poisson search clocks to acquire the locks, and they get rewards at every successful acquisition; an acquisition is successful if all the previous stages are successful and if the agent is the first one to complete. However, none of them have access to the acquisition status of the other agents, leading to an asymmetric information game. Using standard tools of optimal control theory and Markov decision process (MDP) we solved a bi-level control problem; every stage of the dynamic programming equation of the MDP is solved using optimal control tools. We finally reduced the game with an infinite number of states and infinite-dimensional actions to a finite state game with one-dimensional actions. We provided closed-form expressions for Nash Equilibrium in some special cases and derived asymptotic expressions for some more.

math.OC↗

arXiv2020

On Optimal Control of Discounted Cost Infinite-Horizon Markov Decision Processes Under Local State Information Structures

This paper investigates a class of optimal control problems associated with Markov processes with local state information. The decision-maker has only local access to a subset of a state vector information as often encountered in decentralized control problems in multi-agent systems. Under this information structure, part of the state vector cannot be observed. We leverage ab initio principles and find a new form of Bellman equations to characterize the optimal policies of the control problem under local information structures. The dynamic programming solutions feature a mixture of dynamics associated unobservable state components and the local state-feedback policy based on the observable local information. We further characterize the optimal local-state feedback policy using linear programming methods. To reduce the computational complexity of the optimal policy, we propose an approximate algorithm based on virtual beliefs to find a sub-optimal policy. We show the performance bounds on the sub-optimal solution and corroborate the results with numerical case studies.

eess.SY↗

arXiv2019

A Viral Timeline Branching Process to study a Social Network

Bio-inspired paradigms are proving to be useful in analyzing propagation and dissemination of information in networks. In this paper we explore the use of multi-type branching processes to analyse viral properties of content in a social network, with and without competition from other sources. We derive and compute various virality measures, e.g., probability of virality, expected number of shares, or the rate of growth of expected number of shares etc. They allow one to predict the emergence of global macro properties (e.g., viral spread of a post in the entire network) from the laws and parameters that determine local interactions. The local interactions, greatly depend upon the structure of the timelines holding the content and the number of friends (i.e., connections) of users of the network. We then formulate a non-cooperative game problem and study the Nash equilibria as a function of the parameters. The branching processes modelling the social network under competition turn out to be decomposable, multi-type and continuous time variants. For such processes types belonging to different sub-classes evolve at different rates and have different probabilities of extinction etc. We compute content provider wise extinction probability, rate of growth etc. We also conjecture the content-provider wise growth rate of expected shares.

math.PR↗

arXiv2019

Acquisition Games with Partial-Asymmetric Information

We consider an example of stochastic games with partial, asymmetric and non-classical information. We obtain relevant equilibrium policies using a new approach which allows managing the belief updates in a structured manner. Agents have access only to partial information updates, and our approach is to consider optimal open loop control until the information update. The agents continuously control the rates of their Poisson search clocks to acquire the locks, the agent to get all the locks before others would get reward one. However, the agents have no information about the acquisition status of others and will incur a cost proportional to their rate process. We solved the problem for the case with two agents and two locks and conjectured the results for $N$-agents. We showed that a pair of (partial) state-dependent time-threshold policies form a Nash equilibrium.

cs.GT↗

arXiv2019

Continuous-Time Markov Decision Processes with Controlled Observations

In this paper, we study a continuous-time discounted jump Markov decision process with both controlled actions and observations. The observation is only available for a discrete set of time instances. At each time of observation, one has to select an optimal timing for the next observation and a control trajectory for the time interval between two observation points. We provide a theoretical framework that the decision maker can utilize to find the optimal observation epochs and the optimal actions jointly. Two cases are investigated. One is gated queueing systems in which we explicitly characterize the optimal action and the optimal observation where the optimal observation is shown to be independent of the state. Another is the inventory control problem with Poisson arrival process in which we obtain numerically the optimal action and observation. The results show that it is optimal to observe more frequently at a region of states where the optimal action adapts constantly.

math.OC↗

arXiv2019

Decomposable Branching Processes and Viral Marketing

The decomposable branching processes are relatively less studied objects, particularly in the continuous time framework. In this paper, we consider various variants of decomposable continuous time branching processes. As usual practice in the theory of decomposable branching processes, we group various types into irreducible classes. These irreducible classes evolve according to the well-studied nondecomposable/ irreducible branching processes. And we investigate the time evolution of the population of various classes when the process is initiated by the other class particle(s). We obtained class-wise extinction probability and the time evolution of the population in the different classes. We then studied another peculiar type of decomposable branching process where any parent at the transition epoch either produces a random number of offspring, or its type gets changed (which may or may not be regarded as new offspring produced depending on the application). Such processes arise in modeling the content propagation of competing contents in online social networks. Here also, we obtain various performance measures. Additionally, we conjecture that the time evolution of the expected number of shares (different from the total progeny in irreducible branching processes) is given by the sum of two exponential curves corresponding to the two different classes.

math.PR↗

arXiv2018

Speed Based Optimal Power Control in Small Cell Networks

Small cell networks promise good quality of service (QoS) even for cell edge users, however pose challenges to cater to the high-speed users. The major difficulty being that of frequent handovers and the corresponding handover losses, which significantly depend upon the speed of the user. It was shown previously that the optimal cell size increases with speed. Thus, in scenarios with diverse users (speeds spanning over large ranges), it would be inefficient to serve all users using common cell radius and it is practically infeasible to design different cell sizes for different speeds. Alternatively, we propose to allocate power to a user based on its speed, e.g., higher power virtually increases the cell size. We solve well known Hamiltonian Jacobi equations under certain assumptions to obtain a power law, optimal for load factor and busy probability, for any given average power constraint and cell size. The optimal power control turns out to be linear in speed. We build a system level simulator for small cell network, using elaborate Monte-Carlo simulations, and show that the performance of the system improves significantly with linear power law. The power law is tested even for the cases, for which the system does not satisfy the assumptions required by the theory. For example, the linear power law has significant improvement in comparison with the 'equal power' system, even in presence of time varying and random interference. We observe good improvement in almost all cases with improvements up to 89\% for certain configurations.

cs.PF↗

arXiv2018

Controlling Packet Drops to Improve Freshness of information

Many systems require frequent and regular updates of a certain information. These updates have to be transferred regularly from the source to the destination. We consider scenarios in which an old packet becomes completely obsolete, in the presence of a new packet. In this context, if a new packet arrives at the source while it is transferring a packet, one needs to decide the packet to be dropped. New packet has recent information, but might require more time to transfer. Thus it is not clear as to which packet to be dis- carded, and this is the main focus of the paper. Recently introduced performance metrics, called average age of information (AAoI) and peak age of information (PAoI) of the information available at the destination, are the relevant performance measures. These type of systems do not require storage buffers, of size more than one, at the source queue. We consider single source / multiple sources regularly updating information to a single destination possibly over wireless channels to derive optimal drop policies that optimize the AAoI. We showed that the state independent (static) policies like dropping always the old packets or dropping always the new packets is optimal in many scenarios, among an appropriate set of stationary Markov policies. We consider relevant games when multiple sources compete. In many scenarios, the non-cooperative solution almost minimizes the social objective, the sum of AAoIs of all the sources.

cs.NI↗

arXiv2018

Queuing with Heterogeneous Users: Block Probability and Sojourn times

Communication networks need to support voice and data calls simultaneously. This results in a queueing system with heterogeneous agents. One class of agents demand immediate service, would leave the system if not provided. The second class of customers have longer job requirements and can wait for their turn. We discuss the achievable region of such a two class system, which is the region of all possible pairs of performance metrics. Blocking probability is the relevant performance for eager/impatient class while the expected sojourn time is appropriate for the second tolerant class. We obtain the achievable region, considering static policies that do not depend upon the state of the second class. We conjecture a pseudo conservation law, in a fluid limit for eager customers, which relates the blocking probability of eager customers with the expected sojourn time of the tolerant customers. Using this conjecture we obtain the static achievable region. We validate the pseudo conservation law using two example families of static schedulers, both of which achieve all the points on the achievable region. Along the way we obtain smooth control (sharing) of resources between voice and data calls. We also consider an example dynamic policy to establish that the dynamic achievable region is strictly bigger than the static region, for this heterogeneous queueing system.

math.PR↗

arXiv2013

Mixed Polling with Rerouting and Applications

Queueing systems with a single server in which customers wait to be served at a finite number of distinct locations (buffers/queues) are called discrete polling systems. Polling systems in which arrivals of users occur anywhere in a continuum are called continuous polling systems. Often one encounters a combination of the two systems: the users can either arrive in a continuum or wait in a finite set (i.e. wait at a finite number of queues). We call these systems mixed polling systems. Also, in some applications, customers are rerouted to a new location (for another service) after their service is completed. In this work, we study mixed polling systems with rerouting. We obtain their steady state performance by discretization using the known pseudo conservation laws of discrete polling systems. Their stationary expected workload is obtained as a limit of the stationary expected workload of a discrete system. The main tools for our analysis are: a) the fixed point analysis of infinite dimensional operators and; b) the convergence of Riemann sums to an integral. We analyze two applications using our results on mixed polling systems and discuss the optimal system design. We consider a local area network, in which a moving ferry facilitates communication (data transfer) using a wireless link. We also consider a distributed waste collection system and derive the optimal collection point. In both examples, the service requests can arrive anywhere in a subset of the two dimensional plane. Namely, some users arrive in a continuous set while others wait for their service in a finite set. The only polling systems that can model these applications are mixed systems with rerouting as introduced in this manuscript.

cs.PF↗