Thrust IV: Designing a quantum network architecture suitable for quantum repeaters.

Vision: In this thrust we will design an allocation quantum network stack applied to our particular physical implementation using entanglement sources and quantum memories. We will define services, case requirements and design considerations based upon the link layer presented above and we study how to implement them in the physical network configuration described above.

Design considerations for our quantum repeater architecture. Our quantum network possesses two classes of quantum nodes. On one hand we have controllable quantum nodes, that will perform controllable quantum operations, such as storing qubits and entanglement. Specifically, these nodes enable decision making, for example, which nodes are connected by entanglement swapping. On the other hand, we will have automated quantum nodes which will be only time- controlled, performing the same pre-programmed action at all times. Our target will be to define mechanisms and protocols to control these quantum devices, for example by individual FPGA’s programmed by online services.

There are four “use cases” that can be defined in a quantum network:

  • Measure Directly (MD): Application protocols that produce many pairs of entangled qubits sequentially, where both qubits are immediately measured to produce classical correlations. Examples of such applications are Quantum Key distribution, and device-independent protocols.
  • Create and Keep (CK): Protocols that require genuine entanglement, even multiple entangled pairs existing simultaneously and operations on multiple qubits while keeping the qubits in local quantum storage. Examples of such protocols are sensing, metrology, and distributed systems.
  • Send Qubit (SQ): Distributed quantum computing applications require the transmission of (unknown) qubits. This can be realized using teleportation.
  • Network Layer (NL): The network layer is responsible for producing entanglement between nodes separated by larger distances. The key is to produce pairwise entanglement effectively, enabling simultaneous entanglement swapping along the entire path with minimal delay.

Based on these considerations, we will develop an elementary quantum network stack (see Figure 9) describing our physical quantum network. In analogy to classical networking, the lowest element of the stack will be the physical layer, including all quantum hardware devices and optical fibers. The hardware at the physical layer will be responsible for timing and synchronization. The task of the data link layer will be to run the physical layer, making entanglement attempts to produce entanglement between controllable quantum nodes. Requests can be made by higher layers to the link layer to produce entanglement. The network layer is responsible for producing long-distance entanglement between nodes that are not connected directly. This is achieved by means of entanglement swapping, using the link layer to generate entanglement between neighboring controllable nodes. The transport layer takes responsibility for transmitting qubits deterministically using teleportation. The application layer enables applications to utilize both the classical and quantum capabilities of the network. Examples of such applications include entangling qubits of distant quantum computers, data transfers at bandwidths potentially multiple times those of classical networks, etc. We will investigate the range of functionalities necessary for each layer and the interactions between the layers.

Figure 9. Comparison of a TCP/IP 5-layer reference model with a quantum repeater network architecture: The physical layer includes all quantum hardware devices and optical fibers. The hardware at the physical layer is responsible for timing and synchronization. The data link layer produces entanglement between controllable quantum nodes based on requests made by higher layers. The network layer produces long-distance entanglement between nodes that are not connected directly through entanglement swapping, with the link layer generating entanglement between neighboring controllable nodes. The transport layer transmits qubits deterministically using teleportation. The application layer enables applications to utilize both the classical and quantum capabilities of the network.

Control plane for a quantum repeater network. As indicated in figure 9, he control plane of the quantum repeater network is the connection between the classical communication stack and the quantum physical layer. In our envisioned quantum repeater prototype this control plane will take the form of interconnected, field-programmable gate arrays (FPGAs) controlling the functionality of all quantum devices in the network. We will have several controlled groups of devices. Device group 1 (FPGA I in figure 10) will control the timing and qubit sequences to be teleported, establishing a common reference frame for all other layers. A second device group (FPGAs II and III in figure 10) will control the emission of pulses containing entangled photons to other elements in the network.  Device group 3 (FPGAs IV, V, VI and VII) will control the portable quantum memories, establishing the adequate time delays in order to prepare the memory in advance to receive the entangled photons. Heralding signals obtained by quantum non-demolition measurements of the state of the memory will control the FPGAs to synchronize the retrieval of entangled photons, once all the memories are ready to initiate the entanglement swapping procedure. Device group 4, (FPGA VIII in figure 10) will control and establish the success of the entanglement swapping procedures.  Device group 5 (FPGA IX in figure 10) will control the emission of the qubit to be teleported from another quantum memory, controlled by a successful entanglement swapping operation. Device group 6 (FPGA X in figure 10) will control and determine the outcome of a second Bell state measurement between one long-distant entangled photon and the qubit to be teleported. Finally, device group 7 (FPGA XI in figure 10) will control the measurements that finalize the quantum teleportation protocol.

Figure 10. Control plane for a basic quantum repeater network: Interconnected, field-programmable gate arrays (FPGAs) control the functionality of all quantum devices in the network. Device group 1 (FPGA I) controls the timing and qubit sequences to be teleported, establishing a common reference frame for all other layers. Device group 2 (FPGAs II and III) controls the emission of pulses containing entangled photons to other elements in the network. Device group 3 (FPGAs IV, V, VI and VII) controls the portable quantum memories, establishing the adequate time delays in order to prepare the memory \textit{in advance} to receive the entangled photons. Heralding signals obtained by quantum non-demolition measurements of the state of the memory control the FPGAs to synchronize the retrieval of entangled photons, once all the memories are ready to initiate the entanglement swapping procedure. Device group 4, (FPGA VIII) controls and establishes the success of the entanglement swapping procedures. Device group 5 (FPGA IX) controls the emission of the qubit to be teleported from another quantum memory, controlled by a successful entanglement swapping operation. Device group 6 (FPGA X) controls and determines the outcome of a second Bell state measurement between one long-distant entangled photon and the qubit to be teleported. Finally, device group 7 (FPGA XI) controls the measurements that finalize the quantum teleportation protocol.

Metrics of success: In this final part of the proposal, the primary goal is the development and deployment of a functioning control plane that can accurately and effectively coordinate the quantum devices and sequence the operations to achieve quantum teleportation of qubits. A properly operating control plane will further enable the linking of multiple quantum repeater “hops” to achieve teleportation over greater distances.