Some Remarks on Real-Time Turing Machines

The power of real-time Turing machines using sublinear space is investigated. In contrast to a claim appearing in the literature, such machines can accept non-regular languages, even if working in deterministic mode. While maintaining a standard binary counter appears to be impossible in real-time, we present a guess and check approach that yields a binary representation of the input length. Based on this technique, we show that unary encodings of languages accepted in exponential time can be recognized by nondeterministic real-time Turing machines.