sequence-toolbox/SeQUeNCe#253 described a use case where each QNode may have different QuantumMemory and sources/detectors, which leads to different entanglement generation fidelities between different nodes.
According to that issue, it is limiting if init_fidelity is a per-node property; instead, it should be defined per-qchannel.
Currently, the LinkLayer constructor accepts init_fidelity and several other parameters:
|
attempt_rate: max entanglement attempts per second (default: 1e6) (currently ineffective). |
|
alpha_db_per_km: fiber attenuation loss in dB/km (default: 0.2). |
|
eta_s: source efficiency (default: 1.0). |
|
eta_d: detector efficiency (default: 1.0). |
|
frequency: entanglement source frequency in Hz (default: 80e6). |
|
tau_0: local operation delay in seconds for emitting and absorbing photon (default: 0.0). |
|
init_fidelity: fidelity of generated entangled pairs (default: 0.99). |
It seems that some/all of these parameters should be moved to
QuantumChannel or
LinkArch type instead.
sequence-toolbox/SeQUeNCe#253 described a use case where each QNode may have different QuantumMemory and sources/detectors, which leads to different entanglement generation fidelities between different nodes.
According to that issue, it is limiting if
init_fidelityis a per-node property; instead, it should be defined per-qchannel.Currently, the LinkLayer constructor accepts
init_fidelityand several other parameters:mqns/mqns/network/protocol/link_layer.py
Lines 94 to 100 in 51853ae
It seems that some/all of these parameters should be moved to QuantumChannel or LinkArch type instead.