We study the complexity of representing an array on a Pure Pointer Machine (PPM) or a pointer machine without arithmetic capabilities. In particular, we show that lower bounds in access time for information retrieval on a PPM arise from two different factors: the amount of information being stored and the limited number of unique neighborhoods with a fixed radius that can be constructed on a PPM. This result contrasts with earlier work by Ben Amram and Galil which showed that for pointer machines with arithmetic capabilities, look-up time only depends on the amount of information stored. We then apply these bounds to show optimal look-up times for comparing elements in partial and total orders on a PPM.
Proceedings Title: Proceedings of the International Conference on Foundations of Computer Science
Conference Dates: July 13-16, 2009
Conference Location: Las Vegas, NV
Pub Type: Conferences
incompressibility, pointer machine, random access memory