Characterizing Noise in IBM-Q Devices Using Unitarity Randomized Benchmarking.

Date of Submission

December 2020

Date of Award

Winter 12-12-2021

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Master's Dissertation

Degree Name

Master of Technology

Subject Name

Computer Science


Cryptology and Security Research Unit (CSRU-Kolkata)


Paul, Goutam (CSRU-Kolkata; ISI)

Abstract (Summary of the Work)

Quantum computers have potential applications in cryptography, optimization problems and recently proposed quantum-enhanced machine learning. Theoretical research into this novel framework of computation has found ways to complement and enhance the abilities of classical computers. Practical large scale quantum computers with sufficient computing capacity suffer from various kinds of errors, which destroy any meaningful computation. Fault tolerant computation is an active area of research to realize quantum computers that are robust to noise. Quantum threshold theorem states that, if the physical error rates are below certain threshold (estimates give a figure around 1%) then by using error correction schemes, one can push the logical error rates to arbitrary low values. Benchmarking or quantifying the effect of noise induced in practically implementing quantum gates has been a topic of interest in recent years. Several protocols to effectively benchmark the physical error rates have been proposed that empirically calculate various metrics to quantify error rates of quantum gate operation irrespective of how the quantum computer works. This thesis is about one such protocol known as Unitarity Randomized Benchmarking (URB). In this paper, we have firstly, implemented the URB protocol using Qiskit which is an open source library to program and execute quantum programs in simulators or on actual IBM-Q devices. To the best of our knowledge, our work presents the first step in the direction of experimentally benchmarking the ‘coherence’ of noise induced during execution of quantum programs on any commercially available device. Secondly, we experimentally validate the effectiveness of our slightly modified URB protocol, to gauge the ‘coherence’ of noise channels. We perform URB experiments by simulating two well studied noise channels namely, the single qubit depolarising channel and the single qubit bit-flip channel and as a result, validate our implementation by comparing with the theoretical unitarity values. Thirdly, we propose yet another modification to the URB protocol, which we call the Native Gate URB, in order to study the noise in the native gates, i.e., the gates ultimately into which the quantum circuits are compiled and executed in the actual IBM-Q devices. Finally, We comment on the ability of Native Gate URB protocol to detect cross talk. We find experimentally and to the best of our knowledge, that both the IBM-Q processors have noticeable but considerably low cross talk when applying native CNOT gate, based on the unitarity values. However, there is considerable difference (~ 10−2 ) in the unitarity of CNOT gate for 5-qubit processor (burlington) as compared to 15-qubit one (melbourne).


ProQuest Collection ID:

Control Number


Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.


This document is currently not available here.