Machine learning-based, inter-cell neighbor relation management in a cellular network

A method comprising:

• determining, by a computing system for a cell of a cellular network, a boundary that depends on a direction and a type of coverage associated with the cell;
• determining a plurality of engineering parameters associated with neighbor relations between the cell and a plurality of neighbor cells within the boundary;
• inputting the plurality of engineering parameters to a machine learning model trained to estimate a probability that each neighbor cell of the plurality of neighbor cells is a related neighbor cell for purposes of cellular device handoffs;
• and
• storing, by the computing system, in a neighbor relation database in memory, an identity of each neighbor cell having a probability that satisfies a threshold probability for the cell.

A non-transitory computer-readable storage medium storing instructions, which when executed by one or more processing devices of a computing system, causes the one or more processing devices to perform operations comprising:

• determining, for a cell of a cellular network, a boundary that depends on a direction and a type of coverage associated with the cell;
• determining a plurality of engineering parameters associated with neighbor relations between the cell and a plurality of neighbor cells within the boundary;
• inputting the plurality of engineering parameters to a machine learning model trained to estimate a probability that each neighbor cell of the plurality of neighbor cells is a related neighbor cell for purposes of cellular device handoffs;
• and
• storing, in a neighbor relation database in memory, an identity of each neighbor cell having a probability that satisfies a threshold probability for the cell.