US 7,349,331 B2Granted

Procedure and system for scheduling a shared recourse between multiple information packet flows

AssigneeTelecom Italia S.p.A.

InventorsLuciano Lenzini, Enzo Mingozzi, Enrico Scarrone, Giovanni Stea

Life of this patent

Estimating the term…

Loading patent…

Abstract

Each synchronous flow (h=1, 2, Ns) is associated to a respective synchronous capacity value (Hh) indicative of the maximum amount of time for which a synchronous flow can be served before relinquishing the token. Each asynchronous flow (I=1, 2, NA) is, on the other hand, associated to a respective indicative value of the delay to be recovered so that the respective queue has the right to be served and to another value indicating the instant in which the server visited the respective queue in the pervious cycle. Each queue associated to a synchronous flow (h) is therefore served for a maximum amount of time that is equal to the aforesaid synchronous capacity value, while each queue associated to an asynchronous flow (i) is only served if the server's visit takes place with anticipation with respect to the expected instant. This anticipation is determined as the difference between the expected rotation time, needed by the server (10) to complete a visit cycle (T) of the queues associated to the aforesaid flows (h, i), and the time that has passed since the server's previous visit (10) and the delay accumulated. This difference, if positive, defines the maximum service time for the asynchronous queue. If the queue is empty when the server visits it, the server (10) moves on to the next queue even before the relative maximum service time has passed.

Claims15

Dependent claims

A method of scheduling a service resource shared between multiple information packet flows, said flows generating respective associated queues and being served by the attribution of a token, this plurality of flows including synchronous flows requiring a guaranteed minimum service rate and asynchronous flows destined to exploit the service capacity of said resource left unused by the synchronous flows,

comprising the steps of:
providing a server that visits the queues associated with said flows in successive cycles and and that determines a time value of expected rotation that in turn identifies an amount of time necessary for the server to complete a visit cycle to the respective queues;
associating with each synchronous flow a respective synchronous capacity value indicative of the maximum amount of time for which a synchronous flow can be served before relinquishing the token;
associating with each synchronous flow (i) a first respective delay value that identifies a value that must be made up for the respective queue to be served, and a second value that indicates an instant in which the server visited the respective queue in the previous cycle, determining for said respective queue an amount of time that has passed since the previous visit of the server,
serving each queue associated to a respective synchronous flow for a maximum service time equal to said respective value of synchronous capacity, and
serving each queue associated to a respective asynchronous flow only if the server's visit occurs before an expected instant, said advance being determined as the difference between said expected rotation time value and an amount of time that has passed since the server's previous visit and any accumulated delay;
wherein if positive, this difference defines a maximum service time for each said queue.

4 more elements

A system for the scheduling of a service resource shared between multiple information packet flows, said flows generating respective associated queues and being served by the attribution of a token;

this plurality of flows includes synchronous flows requiring a guaranteed minimum service rate and asynchronous flows destined to exploit the service capacity of said resource left unused by the synchronous flows, said system comprising
a server that is able to visit the queues associated, to said flows in successive cycles;
the system being configured to perform the following operations:
determine an expected rotation time value which identifies an amount of time necessary for the server to complete a visiting cycle of said, respective queues,
associate with each synchronous flow (h) a respective synchronous capacity value (Hh) indicative of the maximum amount of time for which a respective synchronous flow can be served before relinquishing the token,
associate with each asynchronous flow a first respective delay value that identifies the delay that must be made up for the respective queue to be served, and a second respective value that indicates an instant in which the server visited the respective queue in the previous cycle, determining for said respective queue, an amount of time that has passed since the previous visit of the server,
serve each queue associated to a respective synchronous flow for a maximum service time equal to said respective value of synchronous capacity, and
serve each queue associated to a respective asynchronous flow only if the server's visit occurs before the expected instant, said advance being determined as the difference between said expected rotation time value and an amount of time that has passed since the server's previous visit and any accumulated delay;
if positive, this difference defines a maximum service time for each said queue.

6 more elements

Dependencies are inferred from the claim text.

Facts

Application: 10/399,887Filed 19 Oct 2001
Priority: IT TO2000A100023 Oct 2000
Published: US 2004/0014470 A122 Jan 2004
Granted: 25 Mar 2008
Term ends: Not estimatedThe source records don’t support an estimate.

Classification10 codes

H04L 47/10
Traffic control in data switching networks: flow control; Congestion control
H04L 9/40
Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols: network security protocols
H04L 12/6418
Hybrid switching systems: hybrid transport
H04L 47/50
Traffic control in data switching networks: queue scheduling