Managing connectivity of a wireless device in a cellular communication network

1 . A computer implemented method for managing connectivity of a wireless device in a cellular communication network, wherein the wireless device is operable to connect to a cell of a first radio-access technology, RAT, and to a cell of a second RAT, the method, performed by a management node, comprising:

receiving, from a radio access node serving the wireless device, location information for the wireless device, wherein the wireless device is connected to a first RAT cell hosted by the serving radio access node, and is also connected to a second RAT cell, and wherein a signal strength of the second RAT cell, received at the wireless device, has fallen below a trigger threshold;

identifying first RAT cells in the communication network that are neighbours of the first RAT cell to which the wireless device is connected, and identifying second RAT cells in the communication network that are neighbours of the identified first RAT cells;

assembling, from the identified first RAT cells and second RAT cells, a candidate set of cell combinations, wherein each cell combination in the candidate set comprises a pair of a first RAT cell and a second RAT cell to which the wireless device is operable to connect;

for cell combinations in the candidate set, using a Machine Learning, ML, model to predict a received signal strength at the wireless device of the first RAT cells and the second RAT cells of the cell combinations, based on the received location information for the wireless device;

identifying, from the candidate set of cell combinations, a cell combination for which the predicted received signal strengths of the first RAT cell and the second RAT cell of the combination satisfy an operational criterion, wherein a cell combination that satisfies the operational criterion comprises the cell combination for which a function of the predicted received signal strength of the second RAT cell of the combination has the highest value of all cell combinations in the candidate set for which a function of the predicted received signal strength of the first RAT cell of the combination is above a minimum threshold value; and

causing the radio access node serving the wireless device to initiate handover of the wireless device to the first RAT cell of the identified cell combination.

2 . A computer-implemented method according to claim 1 wherein identifying, from the candidate set of cell combinations, a cell combination for which the predicted received signal strengths of the first RAT cell and the second RAT cell of the combination satisfy the operational criterion comprises:

selecting cell combinations in the candidate set for which a function of the predicted received signal strength of the first RAT cell of the combination is above a minimum threshold value; and

identifying, from the selected cell combinations, the cell combination for which a function of the predicted received signal strength of the second RAT cell of the combination has the highest value.

3 . A computer implemented method according to claim 1 , wherein the function of the predicted received signal strength of the first RAT cell comprises:

a normalised value of the predicted received signal strength of the first RAT cell.

4 . A computer implemented method according to claim 3 , wherein the function of the predicted received signal strength of the first RAT cell comprises:

a normalized and weighted value of the predicted received signal strength of the first RAT cell, wherein the weighting factor comprises a function of the frequency priority of the first RAT cell.

5 . A computer implemented method according to claim 1 , wherein the function of the predicted received signal strength of the second RAT cell comprises:

a normalized value of the predicted received signal strength of the second RAT cell.

6 . A computer implemented method according to claim 5 , wherein the function of the predicted received signal strength of the second RAT cell comprises:

a normalized and weighted value of the predicted received signal strength of the second RAT cell, wherein the weighting factor comprises a function of the frequency priority of the second RAT cell.

7 . A computer implemented method according to claim 1 , wherein using an ML model to predict a received signal strength at the wireless device of the first RAT cells and the second RAT cells of the cell combinations, based on the received location information for the wireless device, further comprises calculating the functions of the predicted received signal strengths.

8 . A computer implemented method according claim 1 , wherein a pair of a first RAT cell and a second RAT cell to which the wireless device is operable to connect comprises a pair in which:

the first RAT cell is operating on a first frequency band;

the second RAT cell is operating on a second frequency band; and

the combination of the first frequency band and the second frequency band is a combination that is supported by the wireless device.

9 . A computer implemented method according to claim 1 , further comprising:

receiving, from the radio access node serving the wireless device, capability information for the wireless device, wherein the capability information comprises frequency band combinations supported by the wireless device; and

wherein assembling, from the identified first RAT cells and second RAT cells, a candidate set of cell combinations, wherein each cell combination in the candidate set comprises a pair of a first RAT cell and a second RAT cell to which the wireless device is operable to connect comprises:

populating the candidate set with only those combinations of first RAT cells and second RAT cells whose frequency band combinations are supported by the wireless device.

10 . A computer implemented method according to claim 1 wherein using an ML model to predict a received signal strength at the wireless device of the first RAT cells and the second RAT cells of the cell combinations, based on the received location information for the wireless device, comprises, for each of the first RAT cells and second RAT cells of the cell combinations:

generating from the location information a location of the wireless device with respect to the radio access node hosting the cell;

inputting, to a trained ML model for the cell, the generated location of the wireless device with respect to the radio access node hosting the cell, wherein the trained ML model for the cell is operable to process the input location in accordance with its trained parameters; and

obtaining from the trained ML model for the cell an output comprising the predicted signal strength of the cell that would be received by the wireless device in the represented location.

11 . A computer implemented method according to claim 10 , wherein, for each cell, the trained ML model for the cell is trained using historic received signal strength measurements and wireless device locations with respect to the radio access node hosting the cell.

12 . A computer implemented method according to claim 10 wherein the location of the wireless device with respect to the radio access node hosting the cell comprises a Timing Advance and Angle of Arrival of the wireless device with respect to the radio access node.

13 . A computer-implemented method according to claim 1 , further comprising:

receiving, from the radio access node serving the wireless device, transmission power information for the wireless device; and

wherein using an ML model to predict a received signal strength at the wireless device of the first RAT cells and the second RAT cells of the cell combinations, based on the received location information for the wireless device, further comprises using the ML model to predict a received signal strength at the wireless device based on the received transmission power information for the wireless device.

14 . A computer-implemented method according to claim 1 wherein the received location information for the wireless device provides a location of the wireless device with respect to the serving radio access node, the method further comprising:

Converting the received location information to a location of the wireless device in a global coordinate system using a location of the serving radio access node in the global coordinate system.

15 . A computer implemented method according to claim 14 , wherein generating from the location information a representation of the wireless device location with respect to the radio access node hosting the cell comprises:

converting the location of the wireless device in the global coordinate system to a location of the wireless device with respect to the radio access node hosting the cell, using a location of the radio access node hosting the cell in the global coordinate system.

16 . A computer-implemented method according to claim 1 , wherein the received location information for the wireless device comprises a Timing Advance and Angle of Arrival of the wireless device with respect to the radio access node serving the wireless device.

17 . A computer implemented method according to claim 1 , wherein causing the radio access node serving the wireless device to initiate handover of the wireless device to the first RAT cell of the identified cell combination comprises: sending to the radio access node serving the wireless device an identifier of the first RAT cell of the identified cell combination.

18 . A management node for managing connectivity of a wireless device in a cellular communication network, wherein the wireless device is operable to connect to a cell of a first radio-access technology, RAT, and to a cell of a second RAT, the management node comprising processing circuitry configured to cause the management node to:

receive, from a radio access node serving the wireless device, location information for the wireless device, wherein the wireless device is connected to a first RAT cell hosted by the serving radio access node, and is also connected to a second RAT cell, and wherein a signal strength of the second RAT cell, received at the wireless device, has fallen below a trigger threshold;

identify first RAT cells in the communication network that are neighbours of the first RAT cell to which the wireless device is connected, and identify second RAT cells in the communication network that are neighbours of the identified first RAT cells;

assemble, from the identified first RAT cells and second RAT cells, a candidate set of cell combinations, wherein each cell combination in the candidate set comprises a pair of a first RAT cell and a second RAT cell to which the wireless device is operable to connect;

for cell combinations in the candidate set, use a Machine Learning, ML, model to predict a received signal strength at the wireless device of the first RAT cells and the second RAT cells of the cell combinations, based on the received location information for the wireless device;

identify, from the candidate set of cell combinations, a cell combination for which the predicted received signal strengths of the first RAT cell and the second RAT cell of the combination satisfy an operational criterion, wherein a cell combination that satisfies the operational criterion comprises the cell combination for which a function of the predicted received signal strength of the second RAT cell of the combination has the highest value of all cell combinations in the candidate set for which a function of the predicted received signal strength of the first RAT cell of the combination is above a minimum threshold value; and

cause the radio access node serving the wireless device to initiate handover of the wireless device to the first RAT cell of the identified cell combination.

19 . A radio access node for managing connectivity of a wireless device in a cellular communication network, wherein the wireless device is operable to connect to a cell of a first radio-access technology, RAT, and to a cell of a second RAT, the radio access node comprising processing circuitry configured to cause the radio access node to:

receive, from the wireless device, an indication that the received signal strength of a second RAT cell, to which the wireless device is also connected, has fallen below a trigger threshold;

transmit, to a management node, location information for the wireless device, wherein the management node is configured to identify a cell combination comprising a pair of a first RAT cell and a second RAT cell to which the wireless device is operable to connect and which combination satisfies an operational criterion, wherein the operational criterion comprises: for all cell combinations in which a function of the predicted received signal strength of the first RAT cell of the combination is above a minimum threshold value, the cell combination for which a function of the predicted received signal strength of the second RAT cell of the combination has the highest value; and

responsive to a prompt from the management node, initiate handover of the wireless device to the first RAT cell of the identified cell combination.