Make-before-break mobility of machine learning context
A method comprising: storing a received first and machine learning model instance and a received second machine learning model instance in a cache of a terminal, wherein the first machine learning model instance is associated to a first cell and configured to make, if activated, a first prediction for the terminal, and the second machine learning model instance is associated to a second cell different from the first cell and configured to make, if activated, a second prediction for the terminal; checking if a predefined first requirement is fulfilled; activating the first machine learning model instance to make the first prediction if the predefined first requirement is fulfilled; inferring a decision involving the terminal based on the first prediction if the predefined first requirement is fulfilled; inhibiting to infer the decision involving the terminal based on the second prediction if the predefined first requirement is fulfilled.
1 . Apparatus comprising:
one or more processors, and
memory storing instructions that, when executed by the one or more processors, cause the apparatus to:
store, in a cache of a terminal, a first machine learning model instance and a second machine learning model instance that are received, wherein the first machine learning model instance is associated to a first cell and configured to make, if activated, a first prediction for the terminal, and wherein the second machine learning model instance is associated to a second cell different from the first cell and configured to make, if activated, a second prediction for the terminal;
check if a predefined first requirement is fulfilled;
activate the first machine learning model instance to make the first prediction if the predefined first requirement is fulfilled;
infer a decision involving the terminal based on the first prediction if the predefined first requirement is fulfilled;
inhibit to infer the decision involving the terminal based on the second prediction if the predefined first requirement is fulfilled,
wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
if the terminal was handed over from the second cell to the first cell:
check if the terminal is in a predefined cache area related to the second cell and if a predefined cache time has elapsed since the terminal was handed over from the second cell to the first cell;
remove the second machine learning model instance from the cache if the terminal is outside the predefined cache area and the predefined cache time has elapsed.
2 . The apparatus according to claim 1 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to: monitor if an instruction is received from the first cell, wherein the instruction instructs the apparatus to determine that the predefined first requirement is fulfilled.
3 . The apparatus according to claim 1 , wherein the predefined first requirement comprises at least one of the following:
the terminal is served by the first cell; or a criterion is not fulfilled, wherein the criterion is received from the first cell.
4 . The apparatus according to claim 1 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
inform the first cell that the terminal is storing the first machine learning model instance if the terminal is storing the first machine learning model instance and that the terminal stores the second machine learning model instance if the terminal is storing the second machine learning model instance.
5 . The apparatus according to claim 4 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to at least one of the following:
monitor if an inquiry is received and inform the first cell in reply to the inquiry, wherein the inquiry asks if the terminal is storing at least one of the first machine learning model instance and the second machine learning model instance; or
monitor if the terminal performs a handover to the first cell and inform the first cell when the terminal performs the handover.
6 . The apparatus according to claim 1 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
activate the second machine learning model instance to make the second prediction if the predefined first requirement is fulfilled.
7 . The apparatus according to claim 6 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
prohibit the activating the second machine learning model instance if a predefined second requirement is not fulfilled.
8 . The apparatus according to claim 1 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
monitor if the terminal performs a handover from the first cell or reestablishes a radio link after a handover failure in a handover from the first cell or a radio link failure in the first cell;
check if the terminal is served by the second cell if the terminal performs the handover or reestablishes the radio link;
activate the second machine learning model instance to make the second prediction if the terminal is served by the second cell;
infer the decision involving the terminal based on the second prediction; inhibit to infer the decision involving the terminal based on the first prediction if the terminal is served by the second cell.
9 . The apparatus according to claim 1 , wherein
the decision involving the terminal is related to at least one of a mobility of the terminal and a positioning of the terminal.
10 . The apparatus according to claim 1 , wherein at least one of the first machine learning model instance and the second machine learning model instance is received from the first cell or the second cell.
11 . A non-transitory computer readable medium comprising instructions stored thereon: which, when executed by an apparatus, cause the apparatus at least to:
store, in a cache of a terminal, a first machine learning model instance and a second machine learning model instance that are received, wherein the first machine learning model instance is associated to a first cell and configured to make, if activated, a first prediction for the terminal, and wherein the second machine learning model instance is associated to a second cell different from the first cell and configured to make, if activated, a second prediction for the terminal;
check if a predefined first requirement is fulfilled;
activate the first machine learning model instance to make the first prediction if the predefined first requirement is fulfilled;
infer a decision involving the terminal based on the first prediction if the predefined first requirement is fulfilled;
inhibit to infer the decision involving the terminal based on the second prediction if the predefined first requirement is fulfilled,
wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
if the terminal was handed over from the second cell to the first cell:
check if the terminal is in a predefined cache area related to the second cell and if a predefined cache time has elapsed since the terminal was handed over from the second cell to the first cell;
remove the second machine learning model instance from the cache if the terminal is outside the predefined cache area and the predefined cache time has elapsed.
12 . The non-transitory computer readable medium according to claim 11 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
monitor if an instruction is received from the first cell, wherein the instruction instructs the apparatus to determine that the predefined first requirement is fulfilled.
13 . The non-transitory computer readable medium according to claim 11 , wherein the predefined first requirement comprises at least one of the following:
the terminal is served by the first cell; or
a criterion is not fulfilled, wherein the criterion is received from the first cell.
14 . The non-transitory computer readable medium according to claim 11 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
inform the first cell that the terminal is storing the first machine learning model instance if the terminal is storing the first machine learning model instance and that the terminal stores the second machine learning model instance if the terminal is storing the second machine learning model instance.
15 . The non-transitory computer readable medium according to claim 14 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to at least one of the following:
monitor if an inquiry is received and inform the first cell in reply to the inquiry, wherein the inquiry asks if the terminal is storing at least one of the first machine learning model instance and the second machine learning model instance; or
monitor if the terminal performs a handover to the first cell and inform the first cell when the terminal performs the handover.
16 . The non-transitory computer readable medium according to claim 11 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
activate the second machine learning model instance to make the second prediction if the predefined first requirement is fulfilled.
17 . The non-transitory computer readable medium according to claim 16 , wherein the instructions, when executed by the one or more processors, further cause the apparatus to:
prohibit the activating the second machine learning model instance if a predefined second requirement is not fulfilled.
18 . A method comprising:
storing, in a cache of a terminal, a first machine learning model instance and a second machine learning model instance that are received, wherein the first machine learning model instance is associated to a first cell and configured to make, if activated, a first prediction for the terminal, and wherein the second machine learning model instance is associated to a second cell different from the first cell and configured to make, if activated, a second prediction for the terminal;
checking if a predefined first requirement is fulfilled;
activating the first machine learning model instance to make the first prediction if the predefined first requirement is fulfilled;
inferring a decision involving the terminal based on the first prediction if the predefined first requirement is fulfilled;
inhibiting to infer the decision involving the terminal based on the second prediction if the predefined first requirement is fulfilled; and
if the terminal was handed over from the second cell to the first cell:
checking if the terminal is in a predefined cache area related to the second cell and if a predefined cache time has elapsed since the terminal was handed over from the second cell to the first cell;
removing the second machine learning model instance from the cache if the terminal is outside the predefined cache area and the predefined cache time has elapsed.