Title :
Statistical Property and Subclass Structure of Sequence Triple Code Space for Repeated Placements
Author_Institution :
Japan Adv. Inst. of Sci. & Technol., Nomi
Abstract :
Sequence triple code scheme was proposed for representing a "repeated placement" which arose from a layout of regular array structure, such as an array of homogeneous sub-circuits, FPGA, multiprocessor on a chip, etc. Each code consists of three permutations of module names, and represents left-right relation and below-above relation between modules. First in this paper, a probability imbalance between left-right relation and below-above relation in the sequence triple solution space is proven mathematically, and its relevance to a tall placement as a result of SA based placement optimization is discussed. Secondary, a subclass structure of the sequence triple solution space is proposed, which provides us a variety of solution sub-spaces having different sizes and different degrees of the probability imbalance between left-right relation and below-above relation.
Keywords :
circuit layout; probability; sequential codes; simulated annealing; statistical analysis; FPGA; array structure layout; below-above relation; homogeneous subcircuits; left-right relation; multiprocessor; placement optimization; probability imbalance; repeated placements; sequence triple code space; simulated annealing; statistical property; subclass structure; tall placement; Circuits; Extraterrestrial phenomena; Field programmable gate arrays; Information science; Probability; Space technology;
Conference_Titel :
Circuits and Systems, 2006. MWSCAS '06. 49th IEEE International Midwest Symposium on
Conference_Location :
San Juan
Print_ISBN :
1-4244-0172-0
Electronic_ISBN :
1548-3746
DOI :
10.1109/MWSCAS.2006.381797