The method of forced enumeration for nondeterministic automata
Acta Informatica
Nondeterministic space is closed under complementation
SIAM Journal on Computing
Nondeterministic computations in sublogarithmic space and space constructibility
SIAM Journal on Computing
Tally versions of the Savitch and Immerman-Szelepcse´nyi theorems for sublogarithmic space
SIAM Journal on Computing
The alternation hierarchy for sublogarithmic space is infinite
Computational Complexity
The Sublogarithmic Alternating Space World
SIAM Journal on Computing
Bridging across the log (n) space frontier
Information and Computation
Some Results on Tape-Bounded Turing Machines
Journal of the ACM (JACM)
A Note Concerning Nondeterministic Tape Complexities
Journal of the ACM (JACM)
Separating Nondeterministic Time Complexity Classes
Journal of the ACM (JACM)
Turing Machines with Sublogarithmic Space
Turing Machines with Sublogarithmic Space
Hierarchies of memory limited computations
FOCS '65 Proceedings of the 6th Annual Symposium on Switching Circuit Theory and Logical Design (SWCT 1965)
Relationships between nondeterministic and deterministic tape complexities
Journal of Computer and System Sciences
Space bounds for processing contentless inputs
Journal of Computer and System Sciences
Hi-index | 0.00 |
We show that, for an arbitrary function h(n) and each recursive function l(n), that are separated by a nondeterministically fully space constructible g(n), such that h(n)∈Ω(g(n)) but l(n) ∉Ω(g(n)), there exists a unary language L in NSPACE(h(n)) - NSPACE(l(n)). The same holds for the deterministic case. The main contribution to the well-known Space Hierarchy Theorem is that (i) the language L separating the two space classes is unary (tally), (ii) the hierarchy is independent of whether h(n) or l(n) are in Ω(log n) or in o(log n), (iii) the functions h(n) or l(n) themselves need not be space constructible nor monotone increasing. This allows us, using diagonalization, to present unary languages in such complexity classes as, for example, NSPACE(log log nċlog*n) - NSPACE(log log n).