Purdue Research Foundation

Avaya Inc.

International Business Machines Corporation

At&t Intellectual Property I, L.p.

Tata Consultancy Services Limited

Saurabh Bagchi

Somali Chaterji

Rajesh Krishna Panta

Ashraf Mahgoub

Navjot Singh

Timothy Tsai

Sachin Garg

Yu-Sung Wu

James Vincent Krogmeier

Ness B Shroff

Ran Xu

Issa M I Khalil

Mark D Krasniewski

Paul Curtis Wood

Subrata Mitra

Jeffrey Owen Kephart

Marc Adam Kaplan

Yixin Diao

Rajarshi Das

Moo-Ryong Ra

Himadri Sekhar Paul

Swagata Biswas

Vinaitheerthan Sundaram

Patrick Eugster

Shammi R Didla

Aaron C Ault

Matthew Edward Tan Creti

Amiya Kumar Maji

A system may receive video information. The system may select a combination of light-weight features and heavy weight features. The light-weight features may be extracted from the video information and the heavy weight features not extracted. The system may forecast, based on the light-weight features and the heavy weight features, accuracy, and latency metrics for performing the object detection and tracking using a plurality of candidate branch configurations, respectively. The system may select a branch configuration from the plurality of candidate branch configurations in response to satisfaction of an optimization criterion. The system may perform object detection and tracking based on the selected branch configuration.

System for content aware reconfiguration of video object detection

System and methods for content- and contention-aware object detection are provided. A system may receive video information and perform object detection and object tracking based on an execution configuration. The system may approximate an optimized execution configuration. To approximate the optimized execution configuration, the system may identify, based on the video information, a plurality of content features. The system may further measure a contention level of a computer resource or multiple resources. The system may approximate, based on the content features and the utilization metric, latency metrics, for a plurality of execution configuration sets, respectively. The system may also approximate, based on the content features, accuracy metrics for the execution configuration sets, respectively. The system may select the optimized execution configuration set in response to satisfaction of a performance criterion. The system may perform object detection and object tracking based on the optimized execution configuration set.

System and methods for content and contention-aware approximate object detection

A system may forecast a workload for a cluster of nodes in a database management system. The system may generate a reconfiguration plan based on the forecasted workload. The system may obtain a heterogenous configuration set. The heterogenous configuration set may include respective configuration sets for the complete sets of nodes. The system may forecast, based on a first machine learning model, respective performance metrics for nodes in each of the complete sets. The system may forecast a cluster performance metric for the entire cluster of nodes based on a second machine learning model. The system may include, in response to satisfaction of an acceptance criterion, the heterogenous configuration set in the reconfiguration plan. The system may cause the cluster of nodes to be reconfigured based on the reconfiguration plan.

System and methods for heterogeneous configuration optimization for distributed servers in the cloud

A fully tested autonomous system works predictably under ideal or assumed environment. However, behavior of the system is not fully defined when component(s) malfunction or fail (e.g., sensor failures) and the like which leads to inefficient task execution. Present disclosure provides system and method for imperfect sensing-based analysis of agents deployed in environments for traversal task execution, specifically, in unknown environments. The system estimates performance metric (e.g., task execution time) and safety metrics (e.g., number of collisions encountered while executing the task) related to task of traversal of the vehicle/agent from its current position to a target location. The system also incorporates sensitivity of each sensor(s), for given task, when they malfunction. The sensing malfunction can be both independent and/or co-related with malfunction of other sensors. Such a model helps to identify the most critical component in the agent(s), thereby increasing reliability of the system and meet safety standards.

Imperfect sensing based analysis of agents deployed in environments for traversal task execution

A system may forecast a plurality of workload measurements for a database management system (DBMS) at respective times based on a workload model. The system may determine, based on the forecasted workload measurements, configuration parameter sets optimized for the DBMS at the respective times. The system may generate a reconfiguration plan. The system may determine performance that would result from reconfiguring nodes of the DBMS with the configurations parameter sets. The system may select a reconfiguration plan in response to the performance satisfying a fitness criterion. The system may cause, at the reconfiguration times, the nodes to begin reconfiguration with the configuration parameter sets included in the selected reconfiguration plan.

Clustered database reconfiguration system for time-varying workloads

A system may forecast a plurality of workload measurements for a database management system (DBMS) at respective times based on a workload model. The system may determine, based on the forecasted workload measurements, configuration parameter sets optimized for the DBMS at the respective times. The system may generate a reconfiguration plan. The system may determine a performance gain that would result from reconfiguring nodes of the DBMS with the configurations parameter sets. In addition, the system may determine a performance loss that would result from the respective databases of the nodes being inaccessible during reconfiguration with the configuration parameter sets. The system may select a reconfiguration plan in response to the performance gain and the performance loss satisfying a fitness criterion. The system may cause, at the reconfiguration times, the nodes to begin reconfiguration with the configuration parameter sets included in the selected reconfiguration plan.

A reconstruction may be divided into a set of partial operations and scheduled in parallel using a distributed protocol which overlays a reduction tree to aggregate the results. In addition, a scheduling algorithm called multiple partial parallel repair is introduced for handling concurrent failures. Multiple reconstructions may be coordinated in parallel while working to minimize the conflict for shared resources.

Parallel partial repair of storage

A cloud computing system which is configured to monitor operating status of a plurality of virtual computing machines running on a physical computing machine, wherein said monitoring includes monitoring a cycles per instruction (CPI) parameter and a cache miss rate (CMR) parameter of at least one of the plurality of virtual computing machines. The system detects interference in the operation of the at least one virtual machine, with the detection including determining when at least one of the CPI and CMR values exceed a predetermined threshold. When interference is detected, the system reconfigures a load balancing module associated with the virtual machine in question to send fewer requests to the virtual machine.

Integrated configuration engine for interference mitigation in cloud computing

An embedded device includes a processor executing instructions from module(s) in a code memory. The instructions specify: reading data from two non-deterministic registers (NDRs) of different types, compressing the data using respective, different compression algorithms, and storing the compressed data in a nonvolatile medium. A method of enabling debug tracing in a computer program product (CPP) includes locating instructions in the CPP that read NDRs, determining types of the NDRs, and adding instruction(s) to the CPP to compress the values read using compression algorithms corresponding to the respective NDR types. An emulator in a computer-readable medium receives emulation-target instructions (ETIs) and compressed NDR data, and emulates an execution sequence of the ETIs by determining NDR-reading instructions, determining a type of the NDR read by each, decompressing a portion of the NDR data using a type-specific decompressor, and updating emulated-machine state based on the decompressed portion.

Debugging non-deterministic embedded systems

A protocol for optimizing the energy for code upload to sensors in a wireless sensor network and speeding up the dissemination if multiple sources of code are available. Energy optimization is achieved by equipping each node with limited non-local topology information, which it uses to determine the time when it can sleep since code is not being distributed in its vicinity. Another aspect of the invention is a protocol that facilitates secure upload of code in the wireless sensor network. The secure communication protocol divides the sensor field into control groups each with a control node, and manages data exchange between nodes within a control group through the mediation of a control head which provides the common key. The keys are refreshed periodically and the control nodes are changed periodically to enhance security. The protocol facilitates secure communication between sensor nodes despite the compromise of other nodes in the network.

Protocol for secure and energy-efficient reprogramming of wireless multi-hop sensor networks

An encryption method and device employing a modified low-resource AES algorithm. The algorithm in one embodiment has a 128-bit key and a 16-bit data type, along with optimization functions including function inlining, memory move reduction via multiple transformations on a given state during a given iteration of a main loop of the algorithm, pointer-based accessing of the state from a transformation function, and a global key schedule. Another embodiment of the invention is a low-power secure communication device comprising a ZigBee-compliant transceiver having a maximum over-the-air data rate of 250 kbps, and a 16-bit RISC encryption processor configured to implement an AES algorithm adapted to encrypt data at a faster rate than 250 kbps. The AES algorithm only requires about 5000 bytes of ROM and about 250 bytes of RAM. The disclosed implementation of AES in a high-level language (C in this case) is believed to be the first software-based solution for 16-bit microcontrollers capable of matching the communication rate of 250 kbps specified by the Zigbee protocol, while also minimizing RAM and ROM usage.

Communication encryption method and device

A protocol for optimizing the energy for code upload to sensors in a wireless sensor network and speeding up the dissemination if multiple sources of code are available. Energy optimization is achieved by equipping each node with limited non-local topology information, which it uses to determine the time when it can go to sleep since code is not being distributed in its vicinity. Another aspect of the invention is a protocol that facilitates secure upload of code in the wireless sensor network. The secure communication protocol divides the sensor field into control groups each with a control node, and manages data exchange between nodes within a control group through the mediation of a control head which provides the common key. The keys are refreshed periodically and the control nodes are changed periodically to enhance security. The protocol facilitates secure communication between sensor nodes despite the compromise of any number of other nodes in the network.

A method and system performs dynamic online multi-parameter optimization for autonomic computing systems. A simplex is maintained. The system's performance is measured for the particular setting of configuration parameters associated with each point in the simplex. A new sample point is determined using the geometric transformations of the simplex. A current best point in the simplex can be resampled to determine if the environment has changed. If a sufficiently different utility value is obtained from a previously sampled utility value for the point in the simplex, the simplex is expanded. If the difference is not sufficient enough, then contraction of the simplex is performed.

Dynamic online multi-parameter optimization system and method for autonomic computing systems

A method for detecting intrusions that employ messages of two or more protocols is disclosed. Such intrusions might occur in Voice over Internet Protocol (VoIP) systems, as well as in systems in which two or more protocols support some service other than VoIP. In the illustrative embodiment of the present invention, a stateful intrusion-detection system is capable of employing rules that have cross-protocol pre-conditions. The illustrative embodiment can use such rules to recognize a variety of VoIP-based intrusion attempts, such as call hijacking, BYE attacks, etc. In addition, the illustrative embodiment is capable of using such rules to recognize other kinds of intrusion attempts in which two or more protocols support a service other than VoIP. The illustrative embodiment also comprises a stateful firewall that is capable of employing rules with cross-protocol pre-conditions.

Growing community of inventors

West Lafayette, IN, United States of America

Saurabh Bagchi