Rajarshi Das

FatBrain.ai

A mechanism is provided for automatically detecting and locating equipment within an intelligent equipment rack. The intelligent equipment rack comprises a rack controller that determines whether a signal has been received indicating that a rack space in a plurality of rack spaces in the intelligent equipment rack has been occupied by a piece of el...

A system method and computer program product for managing readiness states of a plurality of computing devices. In response to a request, a computer system operates to either: provide one or more computing devices from an inactive pool to an active pool, or accept one or more active computing devices into the inactive pool. An Inactive Pool Manager...

Managing readiness states of a plurality of computing devices. A programmed processor unit operates, upon receipt of a request, to: provide one or more computing devices from an inactive pool to an active pool, or accept one or more active computing devices into the inactive pool. The system proactively manages the inactive states of each computing...

One of the main driving forces of the growing adoption of virtualization is its dramatic simplification of the provisioning and dynamic management of IT resources. By decoupling running entities from the underlying physical resources, and by providing easy-to-use controls to allocate, deallocate and migrate virtual machines (VMs) across physical bo...

One of the main driving forces of the growing adoption of virtualization is its dramatic simplification of the provisioning and dynamic management of IT resources. By decoupling running entities from the underlying physical resources, and by providing easy-to-use controls to allocate, deallocate and migrate virtual machines (VMs) across physical bo...

In one embodiment, the present invention is a method for allocation of finite computational resources amongst multiple entities, wherein the method is structured to optimize the business value of an enterprise providing computational services. One embodiment of the inventive method involves establishing, for each entity, a service level utility ind...

Decentralized resource allocation is a key problem for large-scale autonomic (or self-managing) computing systems. Motivated by a data center scenario, we explore efficient techniques for resolving resource conflicts via cooperative negotiation. Rather than computing in advance the functional dependence of each element's utility upon the amount of...

Energy consumption has become a critical issue for data centers, triggered by the rise in energy costs, volatility in the supply and demand of energy and the widespread proliferation of power-hungry information technology (IT) equipment. In response, researchers are developing energy-efficient data centers by incorporating energy-aware systems both...

Energy consumption has become a critical issue for data centers, triggered by the rise in energy costs, volatility in the supply and demand of energy and the wide spread proliferation of power-hungry information technology (IT) equipment. Since nearly half the energy consumed in a data center (DC) goes towards cooling, much of the efforts in minimi...

This paper describes our experiences deploying a recommender system for a mobile phone-based knowledge sharing application to farmers in rural India. Users of the system record questions and call back for answers left by other users and experts. We used collaborative filtering to derive relevant content for each user based on historical navigation...

The sharp rise in energy usage in data centers, fueled by increased IT workload and high server density, and coupled with a concomitant increase in the cost and volatility of the energy supply, have triggered urgent calls to improve data center energy efficiency. In response, researchers have developed energy-aware IT systems that slow or shut down...

Server farms today consume more than 1.5% of the total electricity in the U.S. at a cost of nearly $4.5 billion. Given the rising cost of energy, many industries are now seeking solutions for how to best make use of their available power. An important question which arises in this context is how to distribute available power among servers in a serv...

As data-center energy consumption continues to rise, efficient power management is becoming in- creasingly important. In this work, we examine the use of a novel market mechanism for finding the right balance between power and performance. The market enables a separation between a 'buyer side' that strives to maximize performance and a 'seller side...

Server farms today consume more than 1.5% of the total electricity in the U.S. at a cost of nearly $4.5 billion. Given the rising cost of energy, many industries are now seeking solutions for how to best make use of their available power. An important question which arises in this context is how to distribute available power among servers in a serv...

We introduce a novel power capping technique, IdleCap, that achieves higher effective server frequency for a given power constraint than existing techniques. IdleCap works by repeatedly alternating between the highest performance state and a low-power idle state, maintaining a fixed av-erage power budget, while significantly increasing the av-erage...

The rapidly rising cost and environmental impact of energy consumption in data centers has become a multi-billion dol-lar concern globally. In response, the IT Industry is ac-tively engaged in a first-to-market race to develop energy-conserving hardware and software solutions that do not sac-rifice performance objectives. In this work we demonstrat...

Reinforcement Learning (RL) provides a promising new approach to systems performance management that differs radically from standard queuing-theoretic approaches making use of explicit system performance models. In principle, RL can automatically learn high-quality management policies without an explicit performance model or traffic model, and with...

Getting multiple autonomic managers to work together towards a common goal is a significant architectural and algorithmic challenge, as noted in the ICAC 2006 panel discussion regarding "Can we build effective multi-vendor autonomic systems?" We address this challenge in a real small-scale system that processes web transactions. An administrator us...

Self-management in accordance with high-level objectives that users can specify is a hallmark of autonomic computing systems. The authors advocate utility functions as a principled, practical, and general way of representing such objectives. In an effort to bring the promise of utility-based frameworks to the marketplace, they describe how they've...

Electrical power management in large-scale IT systems such as commercial datacenters is an application area of rapidly growing interest from both an economic and ecological perspective, with billions of dollars and millions of metric tons of CO 2 emissions at stake annually. Businesses want to save power without sacrificing performance. This paper...

This paper presents an autonomic system in which two managers with different responsibilities collaborate to achieve an overall objective within a cluster of server computers. The first, a node group manager, uses modeling and optimization algorithms to allocate server processes and individual requests among a set of server machines grouped into no...

Reinforcement Learning (RL) holds particular promise in an emerging application domain of performance management of computing systems. In recent work, online RL yielded effective server allocation policies in a prototype Data Center, without explicit system models or built-in domain knowledge. This paper presents a substantially improved and more p...

Reinforcement Learning (RL) provides a promising new approach to systems performance management that differs radically from standard queuing-theoretic approaches making use of explicit system performance models. In principle, RL can automatically learn high-quality management policies without an explicit performance model or traffi c model, and wit...

Previous experience with a data center prototype called Unity established that utility functions provide a natural framework for self-optimization in distributed autonomic computing systems [1]. In an effort to bring the promise of utility-based resource allocation to the marketplace, we have infused methods prototyped in Unity into two interacting...

Autonomic computing, a proposed solution to the looming complexity crisis in IT, is a realm in which software agents and multi-agent systems can play a critically important role. Conversely, given its importance to a multi-billion dollar industry, it is fair to say that autonomic computing is a killer app for agents. Two years ago, we introduced Un...

We present a new hybrid approach to performance man- agement, combining disparate strengths of Reinforcment Learning (RL) with model-based (e.g. queuing-theoretic) ap- proaches. Our method trains nonlinear function approxima- tors using offline RL on data collected while a model-based policy controls the system. By training offline we avoid po- ten...

We study autonomic resource allocation among multiple applications based on optimizing the sum of utility for each application. We compare two methodologies for estimating the utility of resources: a queuing-theoretic performance model and model-free reinforcement learning. We evaluate them empirically in a distributed prototype data center and hig...

Autonomic (self-managing) computing systems face the critical problem of resource allocation to different computing elements. Adopting a recent model, we view the problem of provisioning re- sources as involving utility elicitation and opti- mization to allocate resources given imprecise util- ity information. In this paper, we propose a new algori...

The goal of autonomic computing is to create computing systems capable of managing themselves to a far greater extent than they do today. This paper presents Unity, a decentralized architecture for autonomic computing based on multiple interacting agents called autonomic elements.We illustrate how the Unity architecture realizes a number of desired...

Utility functions provide a natural and advantageous framework for achieving self-optimization in distributed autonomic computing systems. We present a distributed architecture, implemented in a realistic prototype data center, that demonstrates how utility functions can enable a collection of autonomic elements to continually optimize the use of c...

Engineering and Applied Sciences Auctions define games of incomplete information for which it is often too hard to compute the exact Bayesian-Nash equilibrium. Instead, the infinite strategy space is often populated with heuristic strategies, such as myopic best-response to prices. Given these heuristic strategies, it can be useful to evaluate the...

Decentralized resource allocation is a key problem for large-scale autonomic (or self-managing) computing systems. Motivated by a data center scenario, we explore efficient techniques for resolving resource conflicts via cooperative negotiation.

Resource allocation is a key problem in autonomic computing. In this paper we use a data center sce­ nario to motivate the need for decentralization and cooperative negotiation, and describe a promising approach that employs preference elicitation. 1 Resource Allocation in an Autonomic Computing System An autonomic computing system is designed to d...

Decentralized resource allocation is a key prob- lem for large-scale autonomic (or self-managing) computing systems. Motivated by a data center scenario, we explore efficient techniques for re- solving resource conflicts via cooperative nego- tiation. Rather than computing in advance the functional dependence of each element's utility upon the amou...

Digital information goods potentially provide information producers with a new set of strategies, or price schedules, for offering these goods to a consumer population. If consumer preferences are known, then a producer can choose from the available schedules according to the profits they are able to extract.

In an economy in which a producer must learn the preferences of a consumer population, it is faced with a classic decision problem: when to explore and when to exploit. If the producer has a limited number of chances to experiment, it must explicitly consider the cost of learning (in terms of foregone profit) against the value of the information ac...

We develop a model for analyzing complex games with re-peated interactions, for which a full game-theoretic analy-sis is intractable. Our approach treats exogenously specified, heuristic strategies, rather than the atomic actions, as primi-tive, and computes a heuristic-payoff table specifying the ex-pected payoffs of the joint heuristic strategy s...

Electronic goods are exible and have negligible marginal costs. These features allow a producer of electronic goods to explore pricing schemes, and in particular bundling, that would not be feasible with physical goods. However, they can also make it more difficult for a producer to differentiate itself from competitors offering identical goods. Pr...

We develop two bidding algorithms for real-time Continuous Double Auctions (CDAs) using a variety of market rules that o#er what we believe to be the strongest known performance of any published bidding strategy. Our algorithms are based on extensions of the "ZIP" (Cli#, 1997) and "GD" (Gjerstad and Dickhaut, 1998) strategies: we have made essentia...

An increasingly important focus in agent-based electronic commerce is the design of robust heuristic bidding algorithms for a variety of auctions, including the Continuous Double Auction institution (CDA), which is pervasive in real-world markets and which is known for its high market eciency. A signi cant competitive analysis of agent trading stra...

The Continuous Double Auction (CDA) is the dominant market institution for real-world trading of equities, commodities, derivatives, etc. We describe a series of laboratory experiments that, for the first time, allow human subjects to interact with software bidding agents in a CDA. Our bidding agents use strategies based on extensions of the Gjerst...

Electronic goods are exible and have negligible marginal costs. These features allow a producer of electronic goods to explore pricing schemes, and in particular bundling, that would not be feasible with physical goods. However, they can also make it more dicult for a producer to dierentiate itself from competitors oering identical goods. Previous...

Markets for digital information goods provide the possibility of exploring new and more complex pricing schemes, due to information goods' flexibility and negligible marginal cost. In this paper we compare the dynamic performance of price schedules of varying complexity under two different specifications of consumer demand shifts. A monopolist prod...

We explore a scenario in which a monopolist producer of information goods seeks to maximize its profits in a market where consumer demand shifts frequently and unpredictably. The producer is free to set an arbitrarily complex price schedule-a function that maps the set of purchased items to a price-but without direct knowledge of consumer demand it...

We study the price dynamics in a multi-agent economy consisting of buyers and competing sellers, where each seller has limited information about its competitors ’ prices. In this economy, buyers use shopbots while the sellers employ automated pricing agents or pricebots. A pricebot resets its seller’s price at regular intervals with the objective o...

We review recent work done by our group on applying genetic algorithms (GAs) to the design of cellular automata (CAs) that can perform computations requiring global coordination. A GA was used to evolve CAs for two computational tasks: density classification and synchronization. In both cases, the GA discovered rules that gave rise to sophisticated...

How does an evolutionary process interact with a decentralized, distributed system in order to produce globally coordinated behavior? Using a genetic algorithm (GA) to evolve cellular automata (CAs), we show that the evolution of spontaneous synchronization, one type of emergent coordination, takes advantage of the underlying medium's potential to...

Studies the price dynamics in a multi-agent economy consisting of buyers and competing sellers, where each seller has limited information about its competitors' prices. In this economy, buyers use comparison shopping agents (shopbots) while the sellers employ automated pricing agents (pricebots). Derivative following (DF) provides a simple, albeit...

Markets for electronic goods provide the possibility of exploring new and more complex pricing schemes, due to the flexibility of information goods and negligible marginal cost. In this paper we compare dynamic performance across price schedules of varying complexity. We provide a monopolist producer with two machine learning methods which implemen...

In an automated market for electronic goods problems arise that have not been well studied previously. For example, information goods are very exible. In contrast to physical goods, marginal costs are negligible and nearly limitless bundling and unbundling of these items are possible. Consequently, producers can offer complex pricing schemes. Howev...

Commerce in information goods is one of the earliest emerging applications for intelligent agents in commerce. However, the fundamental characteristics of information goods mean that they can and likely will be offered in widely varying configurations. Participating agents will need to deal with uncertainty about both prices and location in multi-d...

We study the ability of a genetic algorithm to design cellular automata exhibiting spontaneous global synchronization. We show that in order to produce emergent coordination, the evolutionary process can take advantage of the underlying medium's potential to form embedded particles. The particles, typically phase defects between synchronous regions...

We investigate the ability of a genetic algorithm to design cellular automata that perform computations. The computational strategies of the resulting cellular automata can be understood using a framework in which ``particles'' embedded in space-time configurations carry information and interactions between particles effect information processing....

How does evolution produce sophisticated emergent computation in systems composed of simple components limited to local interactions? To model such a process, we used a genetic algorithm (GA) to evolve cellular automata to perform a computational task requiring globally-coordinated information processing. On most runs a class of relatively unsophis...

Moments after a baseball batter has hit a fly ball, an outfielder has to decide whether to run forward or backward to catch the ball. Judging a fly ball is a difficult task, especially when the fielder is in the plane of the ball's trajectory. There exists several alternative hypotheses in the literature which identify different perceptual features...

Empirical tests indicate that at least one class of genetic algorithms yields good performance for neural network weight optimization in terms of learning rates and scalability. The successful application of these genetic algorithms to supervised learning problems sets the stage for the use of genetic algorithms in reinforcement learning problems....

Holland's fundamental theorem of genetic algorithms (the “schema theorem”) provides a lower bound on the sampling rate of a single hyperplane during genetic search. However, the theorem tracks the change in representation for a single hyperplaneas if its representation is independent of other hyperplanes. Hyperplane samples are clearly interdepende...

Genetic cascade learning is a new constructive algorithm for connectionist learning which combines genetic algorithms and the architectural feature of the cascade-correlation learning algorithm. Like the cascade-correlation learning architecture, this new algorithm also starts with a minimal network and dynamically builds a suitable cascade structu...

Kanerva's `sparse distributed memory' (SDM) is a type of self-organizing neural network which is able to extract a statistical summary from large volumes of data as it is being processed online. Genetic algorithms have been used to optimize the `location address space' which corresponds to the mapping from the input layer to the hidden units in the...

It is pointed out that the genetic algorithms which have been shown to yield good performance for neural network weight optimization are really genetic hill-climbers, with a strong reliance on mutation rather than hyperplane sampling. Neural control problems are more appropriate for these genetic hill-climbers than supervised learning applications...

Typescript. Thesis (M.S.E.E.)--University of Alabama, 1989. Includes bibliographical references (leaves 42-44).

Thesis (Ph. D.)--Colorado State University, 1998. Includes bibliographical references.