We demonstrate a previously unknown two-photon effect in a discrete-time quantum walk. Two identical bosons with no mutual interactions nonetheless can remain clustered together as they walk on a lattice of directionally reversible optical four-ports acting as Grover coins; both photons move in the same direction at each step due to a two-photon quantum interference phenomenon reminiscent of the Hong-Ou-Mandel effect. The clustered two-photon amplitude splits into two localized parts, one oscillating near the initial point and the other moving ballistically without spatial spread, in solitonlike fashion. But the two photons are always clustered in the same part of the superposition, leading to potential applications for transport of entanglement and opportunities for novel two-photon interferometry experiments.
All existing optical quantum walk approaches are based on the use of beamsplitters and multiple paths to explore the multitude of unitary transformations of quantum amplitudes in a Hilbert space. The beamsplitter is naturally a directionally biased device: the photon cannot travel in the reverse direction. This causes rapid increases in the optical hardware resources required for complex quantum walk applications, since the number of options for the walking particle grows with each step. Here we present the experimental demonstration of a directionally-unbiased linear-optical multiport, which allows reversibility of the photon direction. An amplitude-controllable probability distribution matrix for a unitary three-edge vertex is reconstructed with only linear-optical devices. Such directionally-unbiased multiports allow direct execution of quantum walks over a multitude of complex graphs and in tensor networks. This approach would enable simulation of complex Hamiltonians of physical systems and quantum walk applications in a more efficient and compact setup, substantially reducing the required hardware resources.
Recently, a generalization of the standard optical multiport was proposed [ Phys. Rev. A 93, 043845 (2016)]. These directionally unbiased multiports allow photons to reverse direction and exit backwards from the input port, providing a realistic linear optical scattering vertex for quantum walks on arbitrary graph structures. Here, it is shown that arrays of these multiports allow the simulation of a range of discrete-time Hamiltonian systems. Examples are described, including a case where both spatial and internal degrees of freedom are simulated. Because input ports also double as output ports, there is substantial savings of resources compared to feed-forward networks carrying out the same functions. The simulation is implemented in a scalable manner using only linear optics, and can be generalized to higher dimensional systems in a straightforward fashion, thus offering a concrete experimentally achievable implementation of graphical models of discrete-time quantum systems.
Reconfigurable distribution of entangled states is essential for operation of quantum networks connecting multiple devices such as quantum memories and quantum computers. We introduce a quantum distribution network architecture enabling control of the entangled-state propagation direction using linear-optical devices and phase shifters and offering reconfigurable connections between multiple quantum nodes. The basic two-photon entanglement distribution scheme is first introduced to illustrate the principle of operation. The scheme is then extended to a network structure with an increased number of spatial modes connecting potential end-users. We present several examples of controllable network configuration modifications using time-dependent phase shifters.
Linear-optical systems can implement photonic quantum walks that simulate systems with nontrivial topological properties. Here, such photonic walks are used to jointly entangle polarization and winding number. This joint entanglement allows information processing tasks to be performed with interactive access to a wide variety of topological features. Topological considerations are used to suppress errors, with polarization allowing easy measurement and manipulation of qubits. We provide three examples of this approach: production of two-photon systems with entangled winding number (including topological analogs of Bell states), a topologically error-protected optical memory register, and production of entangled topologically-protected boundary states. In particular it is shown that a pair of quantum memory registers, entangled in polarization and winding number, with topologically-assisted error suppression can be made with qubits stored in superpositions of winding numbers; as a result, information processing with winding number-based qubits is a viable possibility.
It is shown that quantum walks on one-dimensional arrays of special linear-optical units allow the simulation of discrete-time Hamiltonian systems with distinct topological phases. In particular, a slightly modified version of the Su-Schrieffer-Heeger (SSH) system can be simulated, which exhibits states of nonzero winding number and has topologically protected boundary states. In the large-system limit this approach uses quadratically fewer resources to carry out quantum simulations than previous linear-optical approaches and can be readily generalized to higher-dimensional systems. The basic optical units that implement this simulation consist of combinations of optical multiports that allow photons to reverse direction.