ML model based power management in a wireless communication network
A computer-implemented method for optimization of downlink (DL) transmit powers in a wireless communication network includes acquiring deployment data describing a deployment of a cluster of cells of the wireless communication network. Further, the method includes acquiring measurement data representing measurements performed with respect to a plurality of connections established between wireless devices and the wireless communication network. Further, the method includes, based on the deployment data and the measurement data, emulating effects of applying different DL transmit powers in at least some cells on the plurality of connections. Further, the method includes estimating state information for each of the cells based on the emulated effects. Further, the method includes, based on the estimated state information, training a machine learning (ML) model for optimization of the DL transmit powers of the cells.
1 . A computer-implemented method of controlling transmissions in a wireless communication network, the method comprising:
acquiring deployment data describing a deployment of a cluster of cells of the wireless communication network;
acquiring measurement data representing measurements performed with respect to a plurality of connections established between wireless devices and the wireless communication network;
based on the deployment data and the measurement data, emulating effects of applying different downlink transmit power values in at least some cells on the plurality of connections, wherein emulating the effects comprises, for each of the plurality of connections, determining to which cell of the cluster the connection would be established under the different downlink transmit power values;
based on the emulated effects, estimating state information for each of the cells;
based on the estimated state information, training a machine learning model for optimization of downlink transmit power values of the cells;
applying the trained machine learning model to determine optimized downlink transmit power values of the cells; and
providing control signaling for implementing the optimized downlink transmit power values.
2 . The method according to claim 1 , wherein the state information is based on deriving from the emulated effects, for each of the connections, a downlink signal strength of the connection.
3 . The method according to claim 1 , wherein the state information is based on deriving from the emulated effects, for each of the connections, a respective downlink signal strength from one or more neighboring cells in the cluster.
4 . The method according to claim 1 ,
wherein the state information is based on deriving from the emulated effects, for each of the connections, a level of interference affecting the connection.
5 . The method according to claim 1 ,
wherein the state information is based on deriving from the emulated effects, for each of the connections, an uplink transmit power level applied by the wireless device.
6 . The method according to claim 1 ,
wherein the state information is based on deriving from the emulated effects, for each of the connections, a serving cell of the cluster resulting from a virtual handover performed according to handover rules of the wireless communication network.
7 . The method according to claim 1 ,
wherein the machine learning model is based on maximizing a reward function.
8 . The method according to claim 7 , wherein the reward function depends on an amount of time in which, for one or more of the cells of the cluster, utilization of available radio resources exceeds a threshold.
9 . The method according to claim 7 , wherein the reward function depends on a number of the connections for which uplink interference created in one or more neighboring cells in the cluster exceeds a threshold.
10 . The method according to claim 7 , wherein the reward function depends on a number of the connections for which downlink signal strength level is below a threshold.
11 . The method according to claim 7 , wherein the reward function depends on a number of the connections for which a ratio of downlink signal strength to interference level is below a threshold.
12 . The method according to claim 7 , wherein the reward function depends on a sum of the downlink transmit power values of the cells of the cluster.
13 . The method according to claim 1 , wherein the measurement data comprise, for each of the connections, a measured downlink signal strength of the connection.
14 . The method according to claim 1 ,
wherein the measurement data comprise, for each of the connections, a respective downlink signal strength from one or more neighboring cells in the cluster.
15 . The method according to claim 1 ,
wherein the measurement data comprise, for each of the connections, identifiers of one or more neighboring cells in the cluster.
16 . The method according to claim 1 ,
wherein the measurement data comprise, for each of the connections, an amount of data transmitted via the connection.
17 . The method according to claim 1 ,
wherein for each of the cells of the cluster the deployment data indicate neighbor relationships to other cells in the cluster.
18 . The method according to claim 1 ,
wherein for each of the cells of the cluster the deployment data comprise: one or more parameters describing directivity of an antenna arrangement provided for serving the cell and/or a number of antenna ports available for downlink transmission.
19 . The method according to claim 1 ,
wherein for each of the cells of the cluster the deployment data comprise: a downlink bandwidth, a utilized downlink frequency channel, and/or a maximum downlink transmit power value.
20 . The method according to claim 1 ,
wherein for each of the cells of the cluster the deployment data comprise at least one parameter for uplink power control.