UE-MEASUREMENT ASSISTED CLOSED LOOP LEARNING APPROACH FOR REAL-TIME OPTIMIZATION OF SYSTEM METRICS

1 . A method for assessing an impact of a design choice on a system level performance metric of a radio access network (RAN) deployed in an environment, comprising:

receiving messages from a plurality of UEs over time by a plurality of RNs in the RAN;

selecting a design choice for a set of operating parameters of the RAN;

processing one or more of measurement values in each of the received messages and the selected design choice to compute a set of derivatives; and

determining a system level performance metric as a function of the computed set of derivatives.

2 . The method of claim 1 , wherein the set of derivatives comprises a signal-to-interference-and-noise ratio (SINR) at a plurality of spatial locations in the environment at which the UEs are located when the measurement values in the messages are measured.

3 . The method of claim 1 , wherein the set of derivatives comprises a logarithmic operation of signal-to-interference-and-noise ratio (SINR) at a plurality of spatial locations in the environment at which the UEs are located when measurement values in the messages are measured.

4 . The method of claim 1 , wherein the function is a weighted summation of the derivatives, where each derivative is associated with a unique weight.

5 . The method of claim 4 , wherein all weights are identical.

6 . The method of claim 1 , wherein the function selects the least derivative from among the set of derivatives.

7 . The method of claim 1 , wherein the messages are Radio Resource Control (RRC) measurement reports.

8 . The method of claim 7 , wherein the measurement values include a RSRP received from a specified cell.

9 . The method of claim 7 , wherein the measurement values are obtained from a report selected from the group including RSRQ for LTE, RSCP, RSSI, Echo for UMTS and CQI reports.

10 . The method of claim 1 , wherein the measurement values include a radio resource management (RRM) measurement value.

11 . The method of claim 1 , wherein processing the measurement values, selecting a design choice, and determining a system level performance metric are performed by an access controller operatively coupled to the RNs, the access controller being configured to receive the measurement reports from the RNs.

12 . The method of claim 1 , wherein the RAN is selected from the group including a small cell RAN, a macro network or a combination of a small cell RAN and a macro network.

13 . The method of claim 6 , wherein derivatives based on measurement values in more recent messages are given more weight than derivatives based on measurement values in less recent messages.

14 . The method of claim 6 , wherein derivatives based on measurement values in messages received from certain UEs are given more weight than derivatives based on measurement values received from other UEs.

15 . The method of claim 14 , wherein derivatives based on messages received from cell-edge UEs are given more weight than derivatives based on measurement values received from other UEs.

16 . The method of claim 1 , further comprising:

assessing an impact of each of a plurality of different design choices for the set of operating parameters; and

selecting for use in operation of the RAN the design choice that optimizes the system level performance metric.

17 . The method of claim 1 , wherein the design choice is a downlink transmit power from the RNs.

18 . The method of claim 1 , wherein the design choice is an operating frequency of each RN

19 . The method of claim 2 , further comprising a gateway that includes a plurality of access controllers each configured to control and coordinate a different RAN, the plurality of access controllers including said access controller that determines the SINR at a plurality of spatial locations in the environment.