Hello Experts,
Excuse me for asking but how idle mode UEs could contribute to the load of a cell, they don’t consume PRBs at all.
I don’t understand the relevance of balancing number of users in idle mode between different cells.
Has anyone tried Idle Mode Moobility Load Balancing in 4G Huawei?
2 Likes
RRC connected user can shift if accessibility goes less in that band automatically.
PRB utilization will be less, but coverage overlapping should be Ok.
Otherwise coverage hole can be the issue.
That’s from Connected Mode UEs that are load balanced by PRB utilisation.
I am Ok with that.
I still don’t understand the Idle Mode load balancing purpose…
Snon intra fast measurement means inter freq will be fast, but there should be proper coverage in that neighbor’s frequency thresxlow or high should satisy.
Or else continuous measurement of interfreq.
It works in overlapping timing advances.
What do you try to achieve by shifting users in Idle Mode to different frequencies?
What is the relevance of one cell having 20 users in idle mode and another overlapped cells having 50 users in idle mode?
What is wrong with that?
50 users in Idle Mode means when connected to Accessibility will automatically take more PRB.
For me, Load Balancing on Connected Mode based on PRB utilisation is the only one that has relevance.
It means high PRB band have more users, it’s good.
For e.g. 20 MHz should have more user than 5 MHz BW.
As we have more PRB for that.
Nope.
MLB is evaluated every 5 seconds.
So those Idle UE will be shifted in Connected Mode immediately.
They don’t load the cell.
No.
It’s not this way.
MLB is most effective.
Idle is second option.
And interfreqcoverage the last option.
Priority is MLB setting.
Or tilts azimuth wherever required for RF overshooters.
Idle mode LB is complementary to MLB.
IMLB can make MLB job easy. 
For example cell 1 is loaded and cell 2 is less loaded. Surely MLB will try to balance by LB Handover from cell 1 to cell 2.
But in addition to that, RRC release UEs from cell 1 can be asked to camp on cell2. Because its less loaded.
That is Connected Mode MLB based on PRB.
I am ok with that.
But Idle Mode Load Balancing, what are you achieving by it?
Idle mode UEs don’t consume any kind fo resources.
Idle LB take action before connected MLB can take.
If you distribute Idle UE from loaded cell to less loaded cells, more RRC attempt will happen to less loaded cell.
And RRC attempt will be reduced in loaded cells. This in turn will reduce RRC users and PRB utilization of high loaded cells.
This feature gives very good improvement in load balancing, specially when coverage of 2 carriers is nearly the same.
Morever, Connected Mode LB involve lot of signalling which affect RRC drop, reconfig, fail, etc…
This is not an issue at all with Idle LB.
Idle goes into Connected eg minus 114 inter measurement it is delayed to go another freq so it fast measurement it accessibility user will shift.
After thresxlow or high achieved.
But overlapping and reselection satisfy condition required.
And if CCE utilization high, Idle can play very good role.
If CCE more than 50 percent.
Ok, but I still think that PRB Load Balancing is better than Idle Mode Load Balancing.
Reselection doesn’t matter here here because it si ruled by IMMCI in rrc connection release.
That lasts for 3 hours those priorities.
Yes Connected will lead to poor throughput mainly HO.
Sorry guys, but I am still not convinced: a UE goes from Idle to Connected in less than 100msec.
So each UE after 100 msec is subject to MLB PRB based in connected mode.
So for those 100 msec (per UE) you are trying to activate the Idle Mode MLB?
I think Idle Mode UEs can be well managed by reselection priorities and threshservlow and ThreshHhigh and threshxlow.
No need to makle a storm by idle mode MLB so no one really knows which UE willl sit in which layer.
Yes it is always there.
That’s why carriers with similar coverage can be balanced well with IMLB.
Static reselection threshold and prioritues cannot handle dynamic loading scenarios.
How many times can we keep on changing Idle Mode parameters?
Idle LB distributes more UEs in less loaded cells and less UEs in loaded cells.
Of course I am talking about Samsung implementation, but more or less it is generic.