6 - LTE Massive MIMO Radio Optimization Solution
6 - LTE Massive MIMO Radio Optimization Solution
6 - LTE Massive MIMO Radio Optimization Solution
For example, PA/PB parameters, SRS resource allocation optimization, and TM mode
selection
Traditional beamforming enables the eNodeB to adapt narrow beams only on the
horizontal plane to the UEs moving at low speed and cannot adjust the beams in the
vertical plane. 3D beamforming enables the eNodeB to adjust narrow beams on both the
horizontal and vertical planes based on the varying UE locations so that the beams are
always directed towards target UEs, increasing the SINR of target UEs.
Based on the MR and neighboring cell NI data, tool analyzes the distribution of UEs in
the local cell and neighboring cells, and adjust beam weights to direct beams toward
UEs to improve spectral efficiency.
1. For in-cell users: based on the strongest SRS signal received from each beam,
locate the UE location home beam in the local cell, and measure the total duration
of services on each beam, which will be reported in the MR.
2. For potential users: UEs that access the neighboring cells generate interference
with the highest NI to beam 1 (figure in slides) in the massive MIMO cell. Estimate
the beam serving the most majority of potential UEs based on the NI statistics of
different beams. The average and maximum NI values will be reported in the MR.
3. Analyze the MR data, and adjust the beam weight so that the beams are most likely
to direct to UEs in the local cell and more potential UEs access the massive MIMO
cell to maximize the capacity of the massive MIMO cell.
Parameter Description
Loss RSRP If the RSRP value of an MR is less than this threshold, it is regarded that the
Threshold MR will be discarded after weight adjustment and the throughput will be
affected after weight adjustment.
If the threshold is increased, more users are regarded as lost users. If the
threshold is decreased, fewer users are regarded as lost users.
RSRP Gap If the RSRP difference between two cells is greater than this threshold, it is
regarded that switchover between the primary and neighboring cell occurs.
If the threshold is increased, fewer users are regarded as users absorbed by
Massive MIMO cells. If the threshold is decreased, more users are regarded
as users absorbed by Massive MIMO cells. (You are recommended to set
this threshold based on handover hysteresis parameters of the live
network.)
Main to
Based on this range, the result of the RSRP value of the primary cell minus
Neighbor
the RSRP value of the neighboring cell is calculated at a step of 1 dB.
Cell RSRP
Difference A greater range results in more accurate calculation but lower tool
Range efficiency.
Neighbor to
Based on this range, the result of the RSRP value of the neighboring cell
Main Cell
minus the RSRP value of the primary cell is calculated at a step of 1 dB.
RSRP
Difference A greater range results in more accurate calculation but lower tool
Range efficiency.
The higher load of a massive MIMO cell, the more layers of multi-user multiplexing hence
better spectral efficiency. In general, for network with mixed MM sites and traditional sites,
most of the load could be shifted to MM sites considerably for cell throughput and user
speed enhancement.
When traffic load increases and saturated, due to the fact that users with poor SINR could
not participate in user pairing, which affects MM cell throughput, hence
If the spectral efficiency observation shows that it increases with the increment of number
of pairing layers, then an easier pairing criteria can be set, including lower threshold for UE
correlation, CQI and SRS SINR.
MaxLayerMMVMIMO is used to determine maximum number of layers of UEs paired for
MU-MIMO in the uplink in a cell with massive MIMO
Reference value for high load: PRB utilization > 60%, connected UE > 300