Self-tuning

In control theory a self-tuning system is capable of optimizing its own internal running parameters in order to maximize or minimize the fulfillment of an objective function; typically efficiency or error.

Self-tuning systems typically exhibit non-linear adaptive control. Self-tuning systems have been a hallmark of the aerospace industry for decades, as this sort of feedback is necessary to generate optimal multivariable control for nonlinear processes. In the telecommunications industry, adaptive communications are often used to dynamically modify operational system parameters to maximize efficiency and robustness.

Examples

Examples of self-tuning systems in computing include:
*TCP (Transfer Control Protocol)
*SQL Server (Newer implementations only)
*FFTW (Fastest Fourier Transform in the West)
*ATLAS (Automatically Tuned Linear Algebra Software)
*libtune (Tunables library for Linux)
*PhiPAC (Self Tuning Linear Algebra Software for RISC)

Performance benefits can be substantial. Professor Jack Dongarra, American computer scientist, claims self-tuning boosts performance often on the order of 300%.

Digital Self-tuning Controllers are an example of self-tuning systems at the hardware level.

Architecture

Self-tuning systems are typically composed of four components: expectations, measurement, analysis, and actions. The expectations describe how the system should behave given exogenous conditions.

Measurements gather data about the conditions and behavior. Analysis helps determine whether the expectations are being met- and which subsequent actions should be performed. Common actions are gathering more data and performing dynamic reconfiguration of the system.

References

External links

* [http://www.eecs.harvard.edu/~sullivan/thesis.html Using Probabilistic Reasoning to Automate Software Tuning]
* [http://www.fftw.org/fftw-paper-ieee.pdf Frigo, M. and Johnson, S. G., "The design and implementation of FFTW3", Proceedings of the IEEE, 93(2), February 2005, 216 - 231. DOI:10.1109/JPROC.2004.840301.]
* [http://math-atlas.sourceforge.net/devel/atlas_devel/ A Collaborative guide to ATLAS Development]
* [http://www.cs.berkeley.edu/~richie/phipac/paper/paper.html Optimizing Matrix Multiply using PHiPAC: a Portable, High-Performance, ANSI C Coding Methodology]
* [http://www.siam.org/pdf/news/781.pdf Faster than a Speeding Algorithm]
* [http://www.vldb.org/conf/2000/P001.pdf Rethinking Database System Architecture: Towards a Self-tuning RISC-style Database System]
* [http://www.barrera.org/selftune/selftune.htm Self-Tuning Systems Software]
* [http://www.microsoft.com/presspass/features/2000/oct00/10-18sql2000.mspx Microsoft Research Adds Data Mining and Self-tuning Technology to SQL Server 2000]
* [http://www.lanl.gov/radiant/pubs/hptcp/hpdc02-drs.pdf A Comparison of TCP Automatic Tuning Techniques for Distributed Computing]
* [http://libtune.sourceforge.net/index.html Tunables library for Linux]
* [http://www.springer.com/dal/home/instructors/textbook+engeneering?SGWID=1-40190-22-45362381-0&teaserId=48695&CENTER_ID=95210 Digital Self-tuning Controllers]