Probabilistic Model of Fault Detection in Quantum Circuits

Anindita Banerjee, Anirban Pathak

It is shown that the fault testing for quantum circuits does not follow conventional classical techniques. If probabilistic gate like Hadamard gate is included in a circuit then the classical notion of test vector is shown to fail. We have reported several new and distinguishing features of quantum fault and also presented a general methodology for detection of functional faults in a quantum circuit. The technique can generate test vectors for detection of di erent kinds of fault. Speci c examples are given and time complexity of the proposed quantum fault detection algorithm is reported. 1

Qing Zhong Gao, Daoyi Dong, Ian R. Petersen

Alexandru Paler, Ilia Polian, John P. Hayes

A new approach to detecting and diagnosing faults in quantum circuits is introduced. In order to account for the probabilistic nature of quantum circuits, collections of test experiments, called binary tomographic tests (BTTs), are generated. A BTT can identify a fault with respect to some user-defined confidence threshold τ. We present an algorithm to generate BTTs that either detect, or ensure the absence of, all modeled faults in a given circuit. We also present an adaptive diagnostic method to locate quantum faults. While classical circuits, even probabilistic ones, only handle ordinary probabilities, quantum circuits deal with quantum states, which have phase as an extra probabilistic parameter. The tomographic testing methods introduced previously for probabilistic circuits are unable to detect differences in phase, and therefore leave many quantum faults undetected. In contrast, we develop a design-for-test method which is specifically intended to detect faults that only affect the phase of a quantum state. We give experimental results for benchmark and random circuits which show high coverage of quantum faults by BTTs, and good resolution in the case of the adaptive diagnosis method.

Hafiz Md. Hasan Babu

Sowmya Aligala, Sreecharani Ratakonda, Kiran Narayan +4 more

It is believed that quantum computing will begin to have an impact around year 2010. Much work is done on physical realization and synthesis of quantum circuits, but nothing so far on the problem of generating tests and localization of faults for such circuits. Even fault models for quantum circuits have been not formulated yet. We propose an approach to test generation for a wide category of fault models of single and multiple faults. It uses deterministic and probabilistic tests to detect faults. A Fault Table is created that includes probabilistic information. If possible, deterministic tests are first selected, while covering faults with tests, in order to shorten the total length of the test sequence. The method is applicable to both binary and ternary quantum circuits. The system generates test sequences and adaptive trees for fault localization for small binary and ternary quantum circuits.

Alexandru Paler, Ilia Polian, Kae Nemoto +1 more