Browse Prior Art Database

A Method for accessing context information of mis-ordered event streams

IP.com Disclosure Number: IPCOM000214841D
Publication Date: 2012-Feb-07
Document File: 2 page(s) / 85K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for refering a mis-ordered event stream to implement various context-dependent processes. This method requires much less computer resources and less latency than "sort" as a well-known generic solution.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 2

A Method for accessing context information of mis -ordered event streams
Disclosed is a method for refering a mis-ordered event stream to implement various context-dependent processes. This method requires much less computer resources and less latency than "sort" as a well-known generic solution.

Real time processing of time series event sequence is getting more popular because of readiness of ubiquitous sensor network or increase of prove devices , such as car navigation system, smart phone, or energy sensor. Sensor data is usually collected and achieved in a central data center through network and processed for collective analysis or reaction to situations.

Time series event sequence over multiple network routes may arrive in mis -ordered manner, because of network bandwidth, status, physical network distance and so on. Hence, central event processing sometimes refers successive multiple events of a specific event source in time order and process event set continuously , which is called context dependent process. For example, tons of moving devices send GPS information to a server once in a second and a server calculates direction of moving devices continuously. In this case, the server refers two successive events from each moving device and calculate direction from two GPS points . If time stamp of each event sequence is out of order, the server has to correct time order. Disclosed method works for this case.

For example, let assume 7 time series events are assigned like 1,2,3,4,5,6 and 7 in chronological order and arriving order to the server is like 1,2,4,5,6,7 and 3. If one of context dependent processes requires successive two events to calculate something , 'sort' as general solution requires memory buffers for five events and has to wait five event arrivals to correct sequences. Considering longest delay of missing events to define the buffer size , it cannot be determined in many real cases. Hence disclosed never waits missing event delivery and aggregates all required successive event immediately series once they get together . Disclosed also solves unknown delay issue using simple time-out technique. Disclosed consumes an input event sequence, outputs continuous sets of successive event series and issues time-out delayed events from another port. Fig.1 shows overview of disclosed. Disclosed has an input port, an output por...