We consider the transmission of classical information over a quantum chan- nel. The channel is expressed by an alphabet of quantum states. For an example, we can express quantum states by photon polarization. For trans- mitting information, we use speci ed set of probabilities. If we nd that the receiver is unable to make separate measurement on the received letter then we have to use Holevo theorem. From this theorem we see that, in such case, the amount of information per letter we are sending cannot be larger than the Von Neumann entropy H of the letter ensemble. It happens most of the time that, the actual amount of information which will be transmitted is less than H. However if we use block coding scheme which has options to choose code words that respects the priori probabilities of the letter states then we nd a di erent situation. In this case the receiver distinguishes whole words rather than individual letters. In this way, the information transmitted per letter can be made arbitrarily close to H. Block coding scheme helps us to nd clear information of theoretical interpretation of Von Neumann entropy in quantum mechanics. We use this experiment in superdense coding and we consider this extension to noisy channels.
We consider the transmission of classical information over a quantum chan- nel. The channel is expressed by an alphabet of quantum states. For an example, we can express quantum states by photon polarization. For trans- mitting information, we use speci ed set of probabilities. If we nd that the receiver is unable to make separate measurement on the received letter then we have to use Holevo theorem. From this theorem we see that, in such case, the amount of information per letter we are sending cannot be larger than the Von Neumann entropy H of the letter ensemble. It happens most of the time that, the actual amount of information which will be transmitted is less than H. However if we use block coding scheme which has options to choose code words that respects the priori probabilities of the letter states then we nd a di erent situation. In this case the receiver distinguishes whole words rather than individual letters. In this way, the information transmitted per letter can be made arbitrarily close to H. Block coding scheme helps us to nd clear information of theoretical interpretation of Von Neumann entropy in quantum mechanics. We use this experiment in superdense coding and we consider this extension to noisy channels.
We consider the transmission of classical information over a quantum channel. The channel is defined by an ``alphabet'' of quantum states, e.g., certain photon polarizations, together with a specified set of probabilities with which these states must be sent. If the receiver is restricted to making separate measurements on the received ``letter'' states, then the Kholevo theorem implies that the amount of information transmitted per letter cannot be greater than the von Neumann entropy H of the letter ensemble. In fact the actual amount of transmitted information will usually be significantly less than H. We show, however, that if the sender uses a block coding scheme consisting of a choice of code words that respects the a priori probabilities of the letter states, and the receiver distinguishes whole words rather than individual letters, then the information transmitted per letter can be made arbitrarily close to H and never exceeds H. This provides a precise information-theoretic interpretation of von Neumann entropy in quantum mechanics. We apply this result to ``superdense'' coding, and we consider its extension to noisy channels. \textcopyright{} 1996 The American Physical Society.
We discuss the capacity of quantum channels for information transmission and storage. Quantum channels have dual uses: they can be used to transmit known quantum states which code for classical information, and they can be used in a purely quantum manner, for transmitting or storing quantum entanglement. We propose here a definition of the von Neumann capacity of quantum channels, which is a quantum-mechanical extension of the Shannon capacity and reverts to it in the classical limit. As such, the von Neumann capacity assumes the role of a classical or quantum capacity depending on the usage of the channel. In analogy to the classical construction, this capacity is defined as the maximum von Neumann mutual entropy processed by the channel, a measure which reduces to the capacity for classical information transmission through quantum channels (the ``Kholevo capacity'') when known quantum states are sent. The quantum mutual entropy fulfills all basic requirements for a measure of information, and observes quantum data-processing inequalities. We also derive a quantum Fano inequality relating the quantum loss of the channel to the fidelity of the quantum code. The quantities introduced are calculated explicitly for the quantum depolarizing channel. The von Neumann capacity is interpreted within the context of superdense coding, and an extended Hamming bound is derived that is consistent with that capacity.
Abstract This book has discussed the foundations of quantum information science as well as the relationship between physics and information theory in general. It has considered the quantum equivalents of the Shannon coding and channel capacity theorems. The von Neumann entropy plays a role analogous to the Shannon entropy, and the Holevo bound is the analogue of Shannon's mutual information used to quantify the capacity of a classical channel. Quantum systems can process information more efficiently than classical systems in a number of different ways. Quantum teleportation and quantum dense coding can be performed using quantum entanglement. Entanglement is an excess of correlations that can exist in quantum physics and is impossible to reproduce classically (with what is termed “separable” states). The book has also demonstrated how to discriminate entangled from separable states using entanglement witnesses, as well as how to quantify entanglement, and looked at quantum computation and quantum algorithms.
We consider a channel coding for sending classical information through a quantum channel with a given ensemble of quantum states (letter states). As is well known, it is generically possible in a quantum channel that the transmittable information in block coding of length n can exceed n times the maximum amount that can be sent without any coding scheme. This so-called superadditivity in classical capacity of a quantum channel is a distinct feature that cannot be found in a classical memoryless channel. In this paper, a practical model of channel coding that shows this property is presented. It consists of a simple code-word selection and the optimum decoding of the code words minimizing the average error probability. At first, optimization of decoding strategy is discussed. Then the channel coding that shows the superadditivity in classical capacity is demonstrated.
Information transmission via a quantum channel is one of the most important processes considered by the quantum information theory. Quantum channels obey nonequilibrium quantum dynamics. The task of the quantum information theory investigates properties of information transmission by means of quantum channels. The von Neumann mutual information and Holevo capacity are useful entropic quantities for this purpose. This short note discusses the relation between the von Neumann mutual information and the Holevo function and also provides a useful formula for calculating these quantities. The result is applied to the quantum dense coding under the influence of an external environment, which is the generalization of the results of ref. 4 Suppose that Alice sends a signal state carrying classical information to Bob through a quantum channel described by an arbitrary trace-preserving completely positive map L. The signal states are represented by orthonormal quantum states j ki (k 1⁄4 1; 2; . . .) which are assumed to be realized with prior probabilities k (k 1⁄4 1; 2; . . .). Then the quantum state of Alice’s system, i.e., the input system of the quantum channel, is given by in 1⁄4 P k k ðkÞ in , where ðkÞ in 1⁄4 j kih kj. In this case, we have the eigenvalue equation, inj ki 1⁄4 kj ki. After receiving the signal, Bob obtains the quantum state ðkÞ out 1⁄4 Lj kih kj with the probability k and thus the quantum state of the output system becomes out 1⁄4 Lin 1⁄4 P k k ðkÞ out. Then we consider the eigenvalue problems, outj ni 1⁄4 nj ni and ðkÞ outj ðkÞ n i 1⁄4 ðkÞ n j ðkÞ n i. To obtain the von Neumann mutual information between the input and output systems of the quantum channel L, it is necessary to introduce a composite state of the input and output systems, which is not uniquely determined. Since we now consider the transmission of classical information, we use a separable state in{out as a composite state, 5,6) in{out 1⁄4 X
Information arises out of a communication game played between a sender and receiver of signals. For physical systems, the players may be designated as Nature and the scientist. The average information obtained from a quantum system is given by the von Neumann measurement scheme, which represents a generalization of thermodynamic entropy that is in perfect accord with common sense when considering a mixed state. While testing multiple copies of such a state can reveal information about the choice made by the sender, entropy for an unknown pure state is zero. A related puzzle is how the total information in the universe appears to be increasing over time.1 Zero entropy for an unknown pure state is reasonable inasmuch as once that state has been identified, no further information may be gained from examining its copies. But such an idea will not be reasonable if the game between sender and receiver consists of the former choosing one of a certain number of polarization states (say, for a photon) and supplying several copies to the receiver. In this case, measurements made on the copies reveal information regarding the sender’s choices. If the set of choices is infinite, the ‘information’ generated by the source is unbounded. From the point of view of the preparer of the states, information in the pure state is limited by the ‘relationship’ between source and receiver, and by the precision of the receiver’s measurement apparatus. If the sender chose a polarization state with which the apparatus was synchronized, the receiver could readily recognize the state. In recent papers,2, 3 I have been investigating information obtainable from an unknown pure state within the framework of communication between source and receiver. I propose a measure of entropy that covers both pure and mixed states. In general, entropy has two components. One is informational, related to the pure components of the quantum state, which can vary by receiver. The other component is thermodynamic and independent of the receiver. The increase of information over time is a consequence of the interplay between unitary and non-unitary evolution, which makes it possible to transform one type of information into the other. Such complementarity indicates that a fundamental duality is essential for information. For a two-component elementary mixed state, the most information to be obtained from each measurement is one bit. Each further measurement of identically prepared states will also yield one bit. For an unknown pure state, information represents the choice made by the source from an infinity of choices related to the values of probability amplitudes with respect to the basis components of the receiver’s measurement apparatus. For a two-component pure state, each measurement also provides a maximum of one bit of information. If the source has made available an unlimited number of identically prepared states, the receiver can obtain additional information from each measurement until the probability amplitudes have been correctly estimated. Once that has occurred, unlike the case with a mixed state, no further information will be obtained from testing additional copies of this pure state. These ideas have potential application in the field of quantum computing. Estimates by the receiver can be made by adjusting the basis vectors to move closer to the unknown pure state. In the most general case, the information obtainable from such a state in repeated experiments is potentially infinite. But if the observer learns the value of the pure state, the information associated with the states vanishes. This suggests a fundamental divide between objective and subjective information. This approach is consistent with the positivist view that one cannot speak of information associated with a system except in relation to an experimental arrangement together with the protocol for measurement. The experimental arrangement is thus integral to the amount of information that can be obtained. The informational measure I have proposed resolves the puzzle of entropy increase. We can suppose that, in the begin-