International Business Machines Corporation

Ting He

Chatschik Bisdikian

Liang Ma

Dakshi Agrawal

Kang-Won Lee

Ramya Raghavendra

Mudhakar Srivatsa

Yang Song

Erich Miles Nahum

Jian Tan

Murtaza Zafer

Joel W Branch

Bonnie Kathryn Ray

Yuting Ji

Dinesh C Verma

Qing Fa Wang

Vasileios Pappas

Yonghua Lin

Petros Zerfos

Nirmit V Desai

Thai Franck Le

Hai Zhan

Shoel D Perelman

Ali Munir

Animashree Anandkumar

A computer-implemented method, computer program product, and computer processing system are provided for client-side monitoring in a cloud provider environment including a network having a plurality of nodes. The method includes estimating a logical topology of the network by clustering source-destination node pairs while excluding intermediate nodes from among the plurality of nodes based on end-to-end performance measurements between various ones of the plurality of nodes forming the source-destination node pairs. The method also includes at least one of scheduling and customizing a workload within the network, and executing the workload within the network, based on the estimated logical topology.

Client-space network monitoring

Techniques to adaptively launch/replace applications and services on edge devices in a cellular infrastructure and/or adaptively place content and computation at the edge devices based on logic at the network core are provided. In one aspect, a method for dynamic placement of applications in a cellular network mobile cloud is provided which includes the steps of: (a) obtaining: (i) a model of the cellular network, (ii) a model of user mobility patterns in the cellular network, and (iii) a model of a profile of the applications; (b) upon receipt of requests from users for the applications, obtaining runtime states at edge servers in the mobile cloud; and (c) upon obtaining the runtime states at the edge servers, placing the requests among the edge servers and a core server in the mobile cloud based on the models obtained in step (a) and the runtime states obtained in step (b).

Techniques for mobility-aware dynamic service placement in mobile clouds

A method is provided for client-side monitoring in a cloud provider environment including a network having a plurality of nodes. The method includes estimating a logical topology of the network by clustering source-destination node pairs from among the plurality of nodes based on end-to-end performance measurements between various ones of the plurality of nodes. The method further includes estimating a performance state of the network for a client application based on end-to-end performance measurements. The method also includes at least one of scheduling and customizing a workload within the network based on the estimated logical topology and the estimated performance state of the network.

Techniques for network state-aware and network scheduling policy-aware task scheduling are provided. In one aspect, a method for scheduling tasks in a distributed computing network includes the steps of: collecting a pool of candidate resources in the distributed computing network for performing a given one of the tasks; predicting a performance of each of the candidate resources in performing the given task based on both i) a state and ii) a scheduling policy of the distributed computing network; and selecting a best candidate resource for the given task based on the performance. A system for scheduling tasks in a distributed computing network is also provided which includes a task scheduler; and a network scheduler, wherein the task scheduler is configured to schedule the tasks in the distributed computing network based on both i) the state and ii) the scheduling policy of the distributed computing network.

Joint network and task scheduling

There is provided a method, a system and a computer program product for improving performance and fairness in sharing a cluster of dynamically available computing resources among multiple jobs. The system collects at least one parameter associated with availability of a plurality of computing resources. The system calculates, based on the collected parameter, an effective processing time each computing resource can provide to each job. The system allocates, based on the calculated effective processing time, the computing resources to the multiple jobs, whereby the multiple jobs are completed at a same time or an approximate time.

Scheduling mapreduce jobs in a cluster of dynamically available servers

Monitoring a network is provided. A set of predicted network topology snapshot graphs of nodes in the network are generated based on a set of network topology prediction models for the network. A subset of the nodes is selected as monitors in the set of predicted network topology snapshot graphs of the network so that each link in a set of selected network links is identifiable using end-to-end measurements between the monitors. Network states of interest are determined in the set of selected network links within a selected time window based on the end-to-end measurements between the monitors.

Monitoring dynamic networks

A method, computer storage medium, computer system of performing monitor deployment under budget constraints includes obtaining a network topology and a deployment budget, decomposing the network topology into a plurality of components having a predetermined property, computing a benefit and a cost for identifying each of the components, selecting at least one targeted component based on a ratio of the benefit to the cost for identifying each of the components, and deploying monitors to identify links in the at least one targeted component within a remaining portion of the deployment budget.

Bounded-budget monitor deployment in monitoring networks via end-to-end probes

A method (and corresponding computer readable storage medium) of selective monitor deployment under budget constraints, wherein the mechanism includes iterating through each layer of an overlay network multiple times and wherein monitor deployment in a lower layer results in recommending adjustments of monitor deployment in an upper layer so as to reduce the overall deployment cost. Also, a method (and corresponding computer readable storage medium) of selective monitor deployment under budget constraints, wherein a deployment budget is a residual budget after deployment at higher layers of an overlay network.

According to exemplary embodiments, a method for resource management of network systems includes sampling channel states of a first set of channels from at least one base station associated with a radio network controller providing an application and estimating channel states of a second set of channels from the at least one base station, wherein the estimated channel states are based on previously sampled channel states and currently sampled channel states. The method further includes adapting at least one runtime parameter of the application based on the sampled channel states of the first set of channels and the estimated channel states of the second set of channels.

Enhanced resource management for a network system

A method for reducing bandwidth used in multiple transmissions on a cellular network includes mapping at least one request of at least one end-device to a unicast channel between server and an intermediary in the network, creating a correlated multicast communication channel between the intermediary and the at least one end-device, and transferring data from the unicast channel to a multicast channel at the intermediary.

Multi-stream communication

A system and method for adaptive remote decision making includes steps of: receiving from an application layer a target range for a level of reporting quality for processed data; setting data collection parameters to meet the target range; collecting the data from a plurality of remote data collecting devices deployed in the distributed computing system, a portion of said data being compromised during the collecting process; processing the collected data to produce the processed data; evaluating the processed data based on observable metrics of current collected data and reported data losses; forecasting an expected reporting quality while continuing to collect the data; comparing the expected reporting quality with the target range; and reporting the processed data when the expected reporting quality falls within the target range for the level of reporting quality.

Adaptive remote decision making under quality of information requirements

A mechanism is provided for reassigning virtual machines to resources in a computing environment. Monitoring data is collected about virtual machine request history and resource usage in a form of time series for measured indicators. Forecasted time series are computed for the measured indicators over a time window by utilizing time series forecasting. A new assignment is computed by executing a snapshot based assignment algorithm on the forecasted time series. The new assignment is used for reassignment of virtual machines on the resources in the computing environment.

Forecasting based service for virtual machine reassignment in computing environment

Spectrum access for Internet of things (IOT) applications including receiving information about expected use by a primary user of a spectrum band in a radio frequency spectrum. The spectrum band is classified into at least two working modes based on the expected use. The spectrum band is sensed to determine a current access pattern of the primary user. Based on the classifying and the sensing, one of the working modes is selected as a current working mode of the primary user. Transmissions are scheduled on the spectrum band using a current schedule that is responsive to the current working mode of the primary user. If the current working mode of the primary user changes, the sensing, selecting, scheduling, and determining whether the current working mode has changed are re-performed. Otherwise, the transmission scheduling and determining if the current working mode has changed are re-performed.

Growing community of inventors

Piermont, NY, United States of America

Ting He