Abstract
Quantum computing exploits quantum-mechanical principles such as entanglement and superposition to offer significant computational advantages over conventional classical computing. Many complex and computationally challenging problems are expected to be solved by quantum computing in a number of fields, such as data science, industrial chemistry, smart energy, finance, secure communications, and many others. In order to understand the current status of quantum computing and identify its challenges, a systematic review of the existing literature will be valuable. An overview of quantum computing literature and its taxonomy is presented in this chapter. Further, the proposed taxonomy aims to identify research gaps by mapping various related studies. There is a detailed analysis of quantum technologies with the most current state of the art. Finally, the chapter presents a highlight of open challenges and future research directions.
Top1. Introduction
Quantum Computing will rescript the definition of impossible, making a world of all possibilities. To begin with we will go back to a popular example of Schrödinger's cat of Quantum Mechanics that demonstrates the paradox of superposition. A cat is left in room with assumption that no one is watching. A hammer is present connected to pulley above a green coloured poison. If the cat tamper with the pulley, it falls on the glass tube containing poison and the cat will die, otherwise, it is alive. So, an observer, is in a superposition of state that whether the cat is dead or alive. The probability of being alive and dead is half. Similarly, if we take an un-biased coin toss it to make a decision, it ends up with a head or a tail. But, if it is a quantum-based coin, it will spin and spin and never come back to make a decision, once it takes a decision, the problem becomes classical and not quantum mechanics based. Why Quantum? The universe can be thought of a quantum computer, with a massive amount of quantum computation going on regular basis. Quantum Computing is one of the most hyped topics in today’s world due to the capability to compute task in real-time and used in numerous fields like Machine Learning, Artificial Intelligence, Big Data, Neural Networks, etc. The beginning of this concept had already originated in 1980 by the pioneering proposal of Benioff and Feynman, who proposed the idea that systems using quantum concept rather than classical ones are capable of dealing with more complex problems (Benioff, 1980).
Quantum Turing Machine was proposed and universal models were developed based on quantum mechanics (Deutsch, 1985) and proved. The computational capabilities of classical computers are enhanced by quantum models, showing advantage over classical algorithms to solve complex problems. Factorization of large prime numbers is considered as one of the NP-Hard problem. A quantum algorithm for solving this was proposed by Shor in 1994 (Shor, 1994). He proved that only in polynomial time, the RSA encryption used in factorization problem can be deciphered by using quantum mechanism. Further, Groover research mechanism for finding data from an un-ordered database by use of quantum searching. He implemented quantum parallelism, processing all data at the same time thereby reducing complexity and time consumption. By these and many more quantum algorithms have attracted researchers in the area of algorithm design.
Since ten researchers and scientist have study various aspects of quantum algorithms and its applications and further have experimented and designed many efficient algorithms in various fields. For this they have implemented and brought forth the concept of qubits which are quantum bits for computation in quantum environment. A conventional classical digital computing has the information stored in bits taking definite binary values as ‘0’ and ‘1’. In contrast, Quantum computers, has the capability of accessing large Hilbert space i.e computational space, where there are n number of bits called qubits. These are in 2n number of outcomes at a given time called superposition state, this overcomes large space problem. Although these qubits are more reliable for computation than classical bits but are prone to environmental noise. Physicist John Preskill proposed the concept of Noisy Intermediate Scale Quantum (NISQ) technology for upcoming generation quantum systems which will be resistant to external noise factor (Preskill, 2018). The major challenge faced by qubits are decoherence, thereby degrading the mechanism of information processing capabilities. This paved a path for the future research in quantum computing arena, by minimizing the noise effect without comprising on information processing by various quantum algorithms. Next, major problem arises for connectivity of qubits with inter and inter qubits connectivity and couplings for a better quantum device functionality. Another, major challenge is to achieve real-time applications and problem not traceable by classical computing, which has been started by many companies like Google long back (Ball, 2016) and also is adapted globally for development.
Key Terms in this Chapter
Quantum Mechanics: Fundamental to physics, quantum mechanics describes the physical properties of matter on an atomic level and at a subatomic scale. Among its many applications are quantum chemistry, quantum field theory, quantum technology, and quantum information science.
Quantum Machine Learning: Quantum machine learning is the integration of quantum algorithms into machine learning programs. Quantum algorithms can be used to analyze quantum states instead of analyzing classical data. Furthermore, quantum algorithms can be used to analyze quantum states instead of classical data. These routines can be more complex in nature and executed more quickly on a quantum computer.
Noisy Intermediate Scale Quantum: A noisy intermediate-scale quantum (NISQ) processor contains 50 to a few hundred qubits, but are neither sufficiently advanced nor large enough to profit sustainably from quantum supremacy. This term describes the current state of the art in quantum processor fabrication. The term 'noisy' refers to quantum processors that are very sensitive to their environment and may lose their quantum state due to quantum decoherence. During the era of NISQ, quantum processors are not advanced enough to continuously use quantum error correction.
Quantum Optimization: A quantum optimization algorithm is used to solve a problem of optimization. From a set of possible solutions, mathematical optimization aims to find the most optimal solution. Most optimization problems are presented as minimization problems, where one attempts to minimize an error which is a function of the solution: the optimal solution has the smallest error. There are many optimization techniques used in fields like mechanics, economics, and engineering, and as the complexity and amount of data involved rise, it becomes increasingly important to find more efficient solutions to optimization problems. There is a possibility that quantum computing may allow problems which are beyond the capabilities of classical computers to be resolved, or may suggest a significant speedup over the fastest known classical algorithm.
Quantum Computing: A quantum computing system is a computer technology based on quantum theory (which explains how energy and matter behave at atomic and subatomic levels). Modern computers can only encode information in bits that have a value of 1 or 0, which limits their capabilities. Qubits, on the other hand, are an essential part of quantum computing. It takes advantage of the unique property of subatomic particles, which allows them to exist in more than one state at once (i.e., as a 1 and a 0).
Quantum Neural Network: The quantum neural network is a computational neural network that is based on quantum mechanics. Machine learning algorithms and quantum neural networks (QNN) combine concepts from quantum computing and artificial neural networks. In the past decade, the term has been applied to describe a variety of ideas, ranging from quantum computers that emulate the exact functions of neural nets to general trainable quantum circuits that bear little resemblance to the multilayer structure of perceptron.
Evolutionary Algorithm: The evolutionary algorithm (EA) emulates the behavior of living organisms by using mechanisms inspired by nature to solve problems. Both evolutionary computing and bio-inspired computing incorporate EA. Evolutionary algorithms are modeled after Darwin's concepts.
Quantum Clustering: Quantum Clustering refers to a class of algorithms that use concepts and mathematical tools from quantum mechanics to cluster data. Clusters are defined by higher densities of data points, and QC belongs to the family of density-based clustering algorithms.