The interbank financial obligations routing problem

Abstract

The problem of speeding up cross-currency transfers and mutual settlements between banks with the increasing amount of national electronic currencies, cryptocurrencies, and the overall volume of financial transactions can be considered as a routing problem. The solving time of this class of combinatorial  optimization  problems  can  be  significantly  reduced  by  using  hybrid  quantum-classical cloud services and heuristic algorithms, which provide approximate solutions for such problems. Numerical experiments were conducted on a hybrid service provided by the company D-Wave  using  the  quantum  annealing  algorithm  and  algorithms  from  the  networkx  library  to compare the speed of solving the traveling salesman problem for a fully connected graph with 105 vertices.  The  developed  routing  algorithms  provide  an  approximate  solution  to  TSP  or  VRP problems on D-Wave quantum computers in approximately 30 milliseconds.