Back to results
US 8,429,111 B1Granted
Encoding and compression of statistical data
AssigneeZscaler, Inc.
InventorsKailash Kailash, Arcady V. Schekochikhin, Srikanth Devarajan, Narinder Paul
Life of this patent
Estimating the term…
Abstract
Methods, systems, and apparatus, including computer program products, for generating or using augmentation queries. In one aspect, statistical model of statistical data is used to support lossless predictive compression. Data instances are identified in statistical data and classified into one of a plurality of data types. Each data type is associated with a corresponding compression process that is used to compress data instances of that type.
Claims24
A system, comprising:
- a data type classifier configured to identity data instances in statistical data associated with system monitoring and classify each data instance into one of a plurality of data types, each data type being associated with a corresponding compression process and being indicative of one of a first order redundancy, a second order redundancy, or a third order redundancy;
- a first field encoder configured to remove first order redundancies in the data instances associated with a data type indicative of the first order redundancy;
- a second field encoder configured to remove second order redundancies in the data instances associated with a data type indicative of the second order redundancy;
- and
- a frame encoder configured to remove third order redundancies in the data instances associated with a data type indicative of the third order redundancy, wherein the corresponding compression process used by each of the first field encoder, the second field encoder, and the frame encoder is different and all compression is lossless;
- wherein the plurality of data types are based on the statistical data and comprise any of a constant data type, a predictable time varying type, an occasional occurrence type, a time varying type, and a unit change type;
- and
- wherein the system monitoring is performed by a distributed security system and the statistical data is from an external network from the distributed security system, wherein the distributed security system is configured to interface between a plurality of users of the external network and the Internet and to monitor data communications therein to form the statistical data.
5 more elements
A method, comprising:
- performing system monitoring with a distributed security system that is connected to an external network, wherein the distributed security system is configured to interface between a plurality of users of the external network and the Internet and to monitor data communications therein to form statistical data;
- identifying data instances in the statistical data associated with the system monitoring;
- classifying each data instance into one of a plurality of data types, each data type being associated with a corresponding compression process and being indicative of one of a first order redundancy, a second order redundancy, or a third order redundancy;
- removing first order redundancies in the data instances associated with a data type indicative of the first order redundancy;
- removing second order redundancies in the data instances associated with a data type indicative of the second order redundancy;
- and
- removing third order redundancies in the data instances associated with a data type indicative of the third order redundancy, wherein the corresponding compression process used for removing each of the first order redundancies, the second order redundancies, and the third order redundancies is different and all compression is lossless;
- wherein the plurality of data types are based on the statistical data and comprise any of a constant data type, a predictable time varying type, an occasional occurrence type, a time varying type, and a unit change type.
5 more elements
A computer-implemented method, comprising:
- performing system monitoring with a distributed security system that is connected to an external network, wherein the distributed security system is configured to interface between a plurality of users of the external network and the Internet and to monitor data communications therein to form statistical data;
- identifying data instances in the statistical data associated with the system monitoring;
- classifying each data instance into one of a plurality of data types, each data type being associated with a corresponding compression process, wherein the plurality of data types are based on the statistical data and comprise any of a constant data type, a predictable time varying type, an occasional occurrence type, a time varying type, and a unit change type;
- and
- compressing each data instance according to the compression process of the associated data type of the data instance, wherein the compressing is different based on the associated data type and lossless for all of the plurality of data types.
2 more elements
Dependencies are inferred from the claim text.