Near-optimal QoS-based resource allocation for hybrid-medium communication networks

Patent metadata

Field	Value
Publication number	US-9565689-B2
Application number	US-201414522153-A
Country	US
Kind code	B2
Filing date	Oct 23, 2014
Priority date	Oct 23, 2013
Publication date	Feb 7, 2017
Grant date	Feb 7, 2017

Abstract

Embodiments of the invention provide a method for performing network resource allocation over hybrid networks so that application QoS requirements are met. The hybrid networks are represented as an undirected weighted graph with the communication links as edges irrespective of the communication medium. A QoS-based resource allocation model. Another embodiment further includes utility functions that capture the QoS attributes such as reliability, timeliness, fault-tolerance and lifetime allows determining routes and bandwidth allocation such that the total system utility across the entire network is maximized.

First claim

We claim:

1. A method for resource allocation in a network, the method comprising:
representing the network as an undirected graph with vertices of the graph representing a plurality of nodes of the network for routing data packets and with edges of the graph representing communication links of the network;
for each of a plurality of tasks assigned to the network, using a quality of service (QoS)-based resource allocation model (Q-RAM) to:
determine one or more QoS points for the task based on at least one of a timeliness factor, a reliability factor, a fault tolerance factor, and a lifetime factor;
determine a set of available bandwidth settings;
determine a set of available routes;
determine a set of available timeliness settings;
determine set-points for the task, each set-point including a selected one of the bandwidth settings, a selected one of the available routes, and a selected one of the available timeliness settings; and
for each set-point, determine a utility level based on the one or more QoS points, determine a resource requirement value for each edge associated with the selected route and determine a resource allocation penalty metric for each edge; and

based upon on the utility levels and resource allocation penalty metrics corresponding to the set-points associated with the plurality of tasks, determining route and bandwidth allocations for the network.

2. The method of claim 1 , wherein the network comprises a hybrid-medium network and the plurality of nodes includes at least a first node and a second node, the first and second nodes being connected in the graph by two or more edges, each edge representing a different communication medium.

3. The method of claim 1 , wherein the edges of the graph are weighted based on the resource requirement values for each edge associated with the selected route for a given set-point of the set-points.

4. The method of claim 1 , wherein each task is defined by a message size parameter, a time factor parameter indicative of when a message is sent, an end-to-end deadline parameter, a source node, and a destination node.

5. The method of claim 4 , wherein the set of available routes comprises all sets of edges in the graph that define a route from the source node and the destination node.

6. The method of claim 1 , wherein the resource allocation penalty metric for each edge is based upon an overall demand for resources associated with that edge.

7. The method of claim 1 , wherein determining the route and bandwidth allocations for the network based upon on the utility levels and resource allocation penalty metrics corresponding to the set-points associated with the plurality of tasks comprises:
determining, from a curve of the utility levels for each set-point, a concave majorant subset of the set-points;
ordering the set-points of the concave majorant subset according to a gradient value associated with each of the set-points; and
determining the route and bandwidth allocations for the network beginning with the set-point having the highest gradient value.

8. The method of claim 1 , wherein the Q-RAM model is defined by a QoS space having a plurality of dimensions, an operational space having a plurality of dimensions, and a resource space.

9. The method of claim 8 , wherein the dimensions of the QoS space include the timeliness factor, the reliability factor, the fault tolerance factor, and the lifetime factor.

10. The method of claim 9 , wherein determining the one or more QoS points for each task comprises determining each of a timeliness factor, a reliability factor, a fault tolerance factor, and a lifetime factor for the task.

11. The method of claim 8 , where determining the utility level for each set-point comprises applying a respective weight factor to each of the QoS space dimensions to obtain utility function curves for each of timeliness, reliability, fault tolerance, and lifetime, and determining the utility level in accordance with the utility function curves.

12. The method of claim 11 , wherein the utility function curves for timeliness and lifetime are substantially linear, and the utility function curves for reliability and fault tolerance are substantially concave.

13. The method of claim 8 , wherein the dimensions of the operational space include the set of available bandwidth settings, the set of available routes, and the set of available timeliness settings.

14. The method of claim 1 , wherein the resource requirement values of the edges associated with the selected route are represented as a first vector and the resource allocation penalty metrics of the edges associated with the selected route is represented as a second vector, wherein a scalar metric describing resource allocation cost is determined using the second vector for each set-point, and wherein determining route and bandwidth allocations for the network based upon the utility levels and resource allocation penalty metrics includes using the scalar metric for resource allocation cost for each set-point.

15. The method of claim 14 , wherein the resource allocation cost is a cost expressed in terms of utility and quality of service (QoS).

16. The method of claim 1 , wherein determining route and bandwidth allocations for the network based on the utility levels and resource allocation penalty metrics comprises selecting route and bandwidth allocations in a manner that prioritizes highest utility and lowest resource allocation penalty.

17. The method of claim 1 , wherein the resource requirement value for each comprises a bandwidth associated with the edge.

18. A computer program product embodied in a non-transitory computer-readable medium, the non-transitory computer-readable medium storing instructions that, when executed by a processor, causes the processor to:
represent a network as an undirected graph with vertices of the graph representing a plurality of nodes of the network for routing data packets and with edges of the graph representing communication links of the network;
provide a quality of service (QoS)-based resource allocation model (Q-RAM) and, for each of a plurality of tasks assigned to the network, use the Q-RAM to:
determine one or more QoS points for the task based on at least one of a timeliness factor, a reliability factor, a fault tolerance factor, and a lifetime factor;
determine a set of available bandwidth settings;
determine a set of available routes;
determine a set of available timeliness settings;
determine set-points for the task, each set-point including a selected one of the bandwidth settings, a selected one of the available routes, and a selected one of the available timeliness settings; and
for each set-point, determine a utility level based on the one or more QoS points, determine a resource requirement value for each edge associated with the selected route and determine a resource allocation penalty metric for each edge; and

based upon on the utility levels and resource allocation penalty metrics corresponding to the set-points associated with the plurality of tasks, determining route and bandwidth allocations for the network.

19. The computer program product of claim 18 , wherein the instructions causing the processor to determine route and bandwidth allocations for the network based upon on the utility metrics and resource allocation penalty metrics corresponding to the set-points associated with the plurality of tasks include instructions causing the processor to:
determine a curve of the utility levels for each set-point;
determine, from the

Assignees

• Texas Instruments Inc

Inventors

• Kandhalu Raghu Arvind
• Xhafa Ariton E

Classifications

• H04W72/543 Primary

  based on requested quality, e.g. QoS · CPC title

• H04W40/12

  based on transmission quality or channel quality · CPC title

• H04L47/805 Primary

  QOS or priority aware · CPC title

• H04W72/087 Primary

  Electricity · mapped topic

• H04W28/24

  Negotiating SLA [Service Level Agreement]; Negotiating QoS [Quality of Service] · CPC title