Document contributor contact points Revision history Guidelines for the reader Overview results of the study

Copyright © 2011 Open Geospatial Consortium 9 Recipients of this document are requested to submit, with their comments, notification of any relevant patent claims or other intellectual property rights of which they may be aware that might be infringed by any implementation of the standard set forth in this document, and to provide supporting documentation.

1.2 Document contributor contact points

All questions regarding this document should be directed to the editors: Name Organization J. Jansou AtoS T. Dacla Atmosphere gbmh We also want to thank the following contributors to this benchmarking campaign, or report writing: Name Organization J. Arquey AtoS M. Aguidi AtoS R. Jarzmik AtoS S Vingataramin Atmosphere gbmh

1.3 Revision history

Date Release Editor Primary clauses modified Description 2011-06-24 V0R1 J. Jansou First draft 2011-08-26 V0R2 J. Jansou Second draft 2011-09-30 V1R0 J. Jansou First release

1.4 Guidelines for the reader

Chapter §2 presents the reference material used to write this report Chapter §3 lists frequently used terms and definitions Chapter §4 gives conventions used along this document, in 4.1 you will find abbreviated terms Copyright © 2011 Open Geospatial Consortium 10 Chapter §5 introduces the platform used for the benchmarking, as well as the various candidates under test with their possible configurations Chapter §6 enumerates the AIXM files used as input data for the benchmark Chapter §7 treats results obtain and proposes explanation Chapter §8 shows how whose results could be used for actual applications Chapter §9 opens ways to further work on the AIXM compression topic Chapter §10 finally concludes, summing up achievements and facts from this study

1.5 Overview results of the study

The following lines resume the work done in the study and browse conclusions for the hurried reader. 3 families of AIXM files were identified and populated with AIXM files AIXM selected files were analyzed to show general aspects of AIXM and specificities due to each feature available. Measurements were made on compaction performance, CPU consumption and memory footprint for both setup, encoding and decoding phases for various algorithms deflate alone, FI, FI with deflate, CWXML, deflate with dictionary, deflate with several levels of compression, EXI without schema knowledge neither deflate, EXI without schema knowledge but with deflate, EXI with schema knowledge but without deflate, EXI with both schema knowledge and deflate using the exi-ttfms platform, modified for the purpose. Main conclusions regarding results obtain are: EXI is very efficient to compress small AIXM files without too much coordinates inside. Exificient with both AIXM schema knowledge and deflate post- compression allow to produce compressed D-NOTAMs around 700 bytes only 13 of the original file size. The dark spot of using exificient for such data is the huge memory footprint necessary ~100MB and the time needed to perform deflate post compression. This is not much a problem for a 2011 server, but will certainly raise some issue for an EFB. For compaction of bigger AIXM files 100KB, the difference of performance between EXI and other compression algorithms decrease, and thus as you can reach a comparable level of compaction with Fast Info Set with a CPU and memory consumption so much lower, we strongly suggest to stick on Fast Info Set with deflate post-compression. Copyright © 2011 Open Geospatial Consortium 11

1.6 Future work