NK(q)-Landscape Instances

Download the problem instances used in our GECCO'2012 paper entitled Constant Time Steepest Ascent Local Search with Statistical Lookahead for NK-Landscapes.

Data for JSP

Download the problem instances and code generator used in our INFORMS Journal on Computing paper entitled Contrasting Structured and Random Permutation Flow-Shop Scheduling Problems: Search-Space Topology and Algorithm Performance.

AFSCN Scheduling Problems

The U.S. Air Force Satellite Control Network (AFSCN) coordinates communications between civilian and military organizations and more than 100 satellites. There are 16 antennas located at nine ground stations that communicate with satellites around the globe. Customers submit requests to reserve an antenna for a specific time period, based on visibility windows between ground stations and target satellites and subject to specific task requirements. Customers may also specify alternate time windows and ground stations for a task. Problems may consist of over 500 task requests for a typical day. As a result, communication antennas are over-subscribed: there are many more requests than 16 antennas can accomodate.

AFIT Problem Data
This data distribution includes 14 files from the Air Force Institute of Technology (AFIT). The files represent seven multi-resource satellite scheduling problems, with two files for each problem: one file for high-altitude requests and one for low-altitude requests. These files contain AFIT benchmark problems for the dates 10/12/1992 through 10/18/1992 that have appeared in several papers (see Publications for additional references):

• Leap Before You Look: An Effective Strategy in an Oversubscribed Scheduling Problem (AAAI '04) PS, PDF
• Trading Places: How to Schedule More in a Multi-Resource Oversubscribed Scheduling Problem (ICAPS '04) PS, PDF
• Scheduling Space-Ground Communications for the Air Force Satellite Control Network (JOS '04) PS, PDF

We do not have approval to release our more recent AFSCN data; however, we are willing to run other's methods on them at our site. We also have a configurable problem generator that can create synthetic data similar to the AFIT problems. To request a copy of the problem generator or arrange to run your code on the more recent problems, please send e-mail to rogersma@cs.colostate.edu.