Nortel Networks Corporation

Don W Bennett

Biswajit Nandy

Nabil Seddigh

Carl F Cao

Yajun Liu

Dabin Wang

Nausheen Naz

Richard A Houle

Satheesa Nadaragh

Fozia A Zaidi

A system and method for data flow classification based on a configurable rule-engine, is presented herein. In accordance with an embodiment of the invention, the system includes a data flow managing mechanism configured to identify, track, and manage the data flows and a rule set, which includes a plurality of rules for comparing information contained within data flow with pre-specified values. The system also includes a configurable classification rule engine for classifying the data flows into one of a plurality of traffic classes based on results of the comparisons. The configurable classification rule engine is configured via a configuration file that specifies and allows for the modification and reconfiguration of the pre-specified values and information regarding the data flows, the rule set, and the traffic classes.

Configurable rule-engine for layer-7 and traffic characteristic-based classification

Currently there are two techniques used to classify data packet flows, those techniques being static and dynamic. A proper combination of these two techniques can allow for the advantages of both techniques to be gained while reducing their respective disadvantages. With this combined flow classification method a flow is classified initially with the use of a static classification method while, at the same time, a dynamic classification method is initiated. If at any point during the dynamic classification method, the classifications derived from the static and dynamic algorithms converge, the flow's classification is henceforth assigned by the dynamic classification method. If a predetermined number of packets or length of time that would be required for the dynamic and static classification algorithms to converge is exceeded, the flow's classification, once again, is henceforth assigned by the dynamic classification method. This allows the initial classification of the flow to occur quickly with the static classification method and the accuracy of dynamic flow classification techniques to take over once such a classification is deemed accurate.

Apparatus and method for classifying data packet flows

Systems and methods applicable in the context of the TCP/IP suite of protocols explicitly are provided which allow for separate and preferential treatment of L -L control packets compared to the treatment of L data packets. This provides a performance gain at the cost of providing such preferential treatment for high error and long delay channels such as wireless links. Advantageously, channel utilization may be increased, for example by 5-50% depending on link conditions, on high error and long delay transmission links. More generally the invention distinguishes between control packets such as transport layer (for example, TCP, OSI Layer or L ) and application layer (for example, OSI Layer ) control packets and data packets, and transmits the control packets with higher reliability and lower delays. The benefit may be significant for IP applications where data objects transferred are small and are bursty, e.g., Web browsing and e-mail access.

Method and apparatus for improving channel utilization

A method and apparatus for dynamically classifying TCP-based traffic, is presented herein. In accordance with an embodiment of the invention, the method includes defining a first class to classify TCP data streams requiring interactive data transactions, a second class to classify TCP data streams requiring small data transactions, and a third class to classify TCP data streams requiring large data transactions. A rule set containing rules to define transitions between the defined classes, is established and an incoming TCP data stream as pertaining to the first class and detects attributes of the TCP data stream. The TCP data stream may include a single data transaction or may include a plurality of data transactions. Based on the detected attributes and the rules, the method dynamically classifies the TCP data stream into one of the defined classes. The method further includes forcing the classified TCP data stream to transition back to the first class in response to detecting that a data transaction has terminated or commenced.

Growing community of inventors

Ottawa, Canada

Don W Bennett