A five-qubit codeword stabilized quantum code is implemented in a seven-qubit system using nuclear magnetic resonance (NMR). Our experiment implements a good nonadditive quantum code which encodes a larger Hilbert space than any stabilizer code with the same length and capable of correcting the same kind of errors. The experimentally measured quantum coherence is shown to be robust against artificially introduced errors, benchmarking the success in implementing the quantum error correction code. Given the typical decoherence time of the system, our experiment illustrates the ability of coherent control to implement complex quantum circuits for demonstrating interesting results in spin qubits for quantum computing.PACS numbers: 03.67. Pp, 03.67.Lx, 03.65.Wj
INTRODUCTIONQuantum computers are promising to solve certain problems faster than classical computers [1]. The power of quantum computing relies on the coherence of quantum states. In implementation, however, the quantum devices are subject to errors, from inevitable cou-2 pling to the uncontrollable environment, or from other mechanisms such as imperfection in controlled operations. The errors damage the coherence, and consequently can reduce the computational ability of quantum computers. In order to protect quantum coherence, schemes of quantum error correction and fault-tolerant quantum computation have been developed [1][2][3][4][5][6][7]. Those schemes have greatly improved the long-term prospects for quantum computation technology.A quantum error correcting code (QECC) protects a K-dimensional Hilbert space (the code space) by encoding it into an n-qubit system, and is usually denoted by parameters ((n, K, d)), where d is called the distance of the code [8]. This n-qubit system is used in the process of quantum computing and hence subject to errors. A code of distance d is capable of correcting d − 1 erasure errors (i.e. loss of qubits at up to d − 1 known positions) or⌋ arbitrary errors (i.e. arbitrary errors on t qubits at unknown positions). At the end of the computation, the quantum code can be decoded to recover the quantum state in the original K-dimensional Hilbert space.In practice, one would always hope for a "good" QECC where more information is protected (larger K), while less physical resources are used (smaller n) plus more errors can be corrected (larger d). There are trade-offs among these three parameters, and one can readily develop upper bounds and lower bounds for the third parameter if two of them are fixed [7]. With increasing length n of the codes, in most of the cases there is a gap between these upper and lower bounds. Therefore, given two fixed parameters among the three, to find a code with the best possible value of the third parameter is one of the most important topics in studying the theory of QECC.Stabilizer codes, also known as additive codes, form an important class of QECCs, developed independently in [6,7] in the late 1990s. The construction of these codes is based on a simple method using Abelian groups, where the...