On Quantum Turing Machine Halting Deterministically

We define a subclass of quantum Turing machine (QTM) named SR-QTM, which halts deterministically and has deterministic tape head position. A quantum state transition diagram (QSTD) is proposed to describe SR-QTM. With the help of QSTD, we construct a SR-QTM which is universal for all near-trivial transformations. This means there exists a QTM which is universal for the above subclass. Finally we prove that SR-QTM is computational equivalent with ordinary QTM in the bounded error setting. It can be seen that, because SR-QTM has the same time steps for different branches of computation, the halting scheme problem will not exist when considering SR-QTM as a model of quantum computing.

💡 Research Summary

The paper introduces a restricted class of quantum Turing machines (QTMs) called SR‑QTM (Stationary‑head Reversible QTM). The defining feature of SR‑QTM is that every computational branch halts after exactly the same number of steps, and the tape head’s position is a deterministic function of the step count. This eliminates the notorious “halting‑scheme problem” of ordinary QTMs, where different branches may stop at different times and require an external measurement to detect termination. To describe SR‑QTMs compactly, the authors propose the Quantum State Transition Diagram (QSTD). In a QSTD, each node encodes a pair (internal state, head position) and each directed edge carries a unitary transition together with its complex amplitude, providing a visual, circuit‑like representation of the machine’s evolution.

Using QSTD, the authors construct an SR‑QTM that can implement any near‑trivial transformation – a unitary that acts non‑trivially only on two basis states, rotating them by an angle θ and adding a phase φ. Since any unitary matrix can be decomposed into a product of such near‑trivial operations, this SR‑QTM is universal for the whole class of quantum transformations. The construction shows that the machine can be built with a polynomial number of steps relative to the size of the input, and that its transition rules can be expressed as a finite set of elementary gates.

The core theoretical contribution is the proof of computational equivalence between SR‑QTMs and ordinary QTMs in the bounded‑error setting (BQP). The authors give two simulation directions. First, any SR‑QTM can be simulated by a standard QTM by expanding each QSTD edge into a sequence of elementary quantum gates; this incurs only polynomial overhead. Second, any standard QTM can be transformed into an SR‑QTM without increasing error beyond a prescribed bound. The transformation proceeds by (i) “time‑normalising” the original machine so that all branches run for the same number of steps, (ii) introducing auxiliary tracks to keep the head stationary while logically moving data, and (iii) decomposing the original unitary transitions into a series of near‑trivial transformations that the universal SR‑QTM can execute. Consequently, SR‑QTMs recognise exactly the same language class as ordinary QTMs under the usual bounded‑error criteria.

Because SR‑QTMs have a fixed halting time and deterministic head movement, the halting‑scheme paradox disappears: one can run the machine for the known number of steps and then perform a single final measurement, without worrying that some branches have already terminated earlier. This property greatly simplifies both theoretical analyses (e.g., complexity proofs) and practical implementations (e.g., hardware scheduling, error‑correction timing). The paper concludes by suggesting that SR‑QTMs provide a more physically realistic model of quantum computation while retaining full computational power, and that the QSTD formalism could serve as a useful design tool for future quantum processors.