Emc IP Holding Company LLC

Flavia Coimbra Delicato

Ana Cristina Bernardo De Oliveira

Tiago Salviano Calmon

Igor Leão Dos Santos

Paulo F Pires

Paulo De Figueiredo Pires

Ana C Oliveira

Tiago Cariolano De Souza Xavier

Marcelo Pitanga

Market-based distributed resource allocation techniques are provided for edge-cloud systems. One method comprises obtaining an application request at a given edge node in a multi-tier environment comprising cloud resources and multiple edge nodes. The edge nodes host a plurality of virtual nodes to process the application request. The application request is assigned to at least one of the virtual nodes based on a utility value of each virtual node. The utility value of each virtual node is based on a cost value representing a total cost incurred by each virtual node to process the application request. The utility value of each virtual node is optionally further based on a priority value of the application request. Master nodes from different edge node groups can collaborate to identify a given edge node group that can serve the application request when local edge nodes are unable to process the at least one application request.

Market-based distributed resource allocation for edge-cloud systems

A resource management framework is provided for edge-cloud systems that supports heterogeneity of devices and application requirements. One method comprises obtaining an application request in a multi-tier environment comprising cloud resources and edge nodes, wherein the edge nodes host virtual nodes to process the application requests, wherein each of the plurality of virtual nodes corresponds to a given predefined virtual node type; providing data from a given data source to at least two virtual nodes based on a data type of the given data source and the virtual node type of the at least two virtual nodes; and providing a given application request to at least one virtual node based on a data type of the given application request and the virtual node type of the at least one virtual node, wherein the at least one virtual node provides data in response to the application requests to corresponding applications and/or additional virtual nodes.

Data centric resource management for edge cloud systems

Techniques are provided for resource allocation and provisioning in a multi-tier edge-cloud virtualization environment. An exemplary method comprises: obtaining an application request for processing a given data type in a multi-tier environment; processing application requests received within a decision window to allocate resources for virtual nodes to process the application requests received within the decision window, wherein the allocated resources for each virtual node is on a corresponding one of cloud resources and a given edge node; instantiating the allocated virtual nodes to process the application requests; and providing the application requests to the instantiated virtual nodes, wherein the instantiated virtual nodes obtain the data of the given data type from a data repository. The virtual node waits to process a given application request for output data of any predecessor requests and sends the output data of the given application request to any additional virtual nodes holding successor requests to the given application request.

Resource allocation and provisioning in a multi-tier edge-cloud virtualization environment

An architecture, methods and apparatus are provided for managing sensor data. Sensor networks comprised of a plurality of sensors are managed by obtaining measurement data and context data from the plurality of sensors; storing the obtained measurement data and context data using a Massively Parallel Processing Database Management System (MPP DBMS); and managing the sensor network from outside of the sensor network using the MPP DBMS. Context-aware adaptation of sensors is based on context regarding a state of the sensor network and context regarding a state of one or more applications. The sensor nodes are optionally clustered based on semantic similarities among sensor readings from different sensor nodes and a distance among the sensor nodes. A subset of the sensor nodes is optionally selected to be active based on a residual energy of the sensor nodes and a relevance of the sensor nodes to an application. Data prediction models are generated and employed for data sensing and analytics.

Growing community of inventors

Rio de Janeiro, Brazil

Flavia Coimbra Delicato