RRM Is a tricky thing for any system to get right, some of the automatic systems out there are better than others, some of them just outright stink.
One thing that is common between all of them though in my experience is the need to tune them for a given style of deployment. Left to its own devices any automatic system can be detrimental and make poor decisions – this has a knock on effect when you take dynamic channel planning into account too and can cause a cascade effect through your network with constant channel and power changes. This behaviour in an RRM system can also indicate a deficiency in network design – but that’s a different subject all together!
I often find with Cisco gear it is sensible to tweak the min / max TX power levels that TPC is allowed to use and often this is one of the most underused features of the Cisco RRM suite. I’ve long since moved to using TPCv2 for all of our networks and these settings are based on using that algorithm, although would suggest using the “run once” mode of operation as it seems unable to converge to a steady state otherwise.
High Density TPCv2 Settings:
|TX Min||TX Max||Neighbour|
Coverage Optimised TPCv2 Settings:
|TX Min||TX Max||Neighbour|
I have found the settings above to deliver predictable results with most of the AP’s in a high-density deployment settling at around power level 3/4 for 2.4GHz and power level 1/2 for 5GHz (more channel space there so DCA can move around CCI/ACI easier without needing TPC to reduce TX Power). They also allow you to drive the 5GHz band a bit “hotter” to help encourage some devices to move away from 2.4GHz.
Another way that I also manipulate TPC to do my bidding is the use of RF-Profiles for different areas / densities of deployment that may require a more coverage optimised approach than outright density and small cells. The settings required for a large auditorium are very different for offices or open space deployments for instance.
Some people argue that we should manually plan channels and power for a network, this does work and can certainly yield very good results.
However in my opinion for a dynamic RF environment or large network a better understanding of how an RRM system works under the hood enables the engineer to tune it for their network and deployment requirements.
The radio environment wireless networks operate in is not static, and these automated systems were developed to help us and to allow the infrastructure to react in real-time to changes in the RF landscape. Some do it better than others, but I’d rather not go back to the days of manually channel and power planning a large network! We have these tools to help us do our jobs and with the correct understanding of how they operate and how to leverage them to get the most out of any vendors gear is key and at the end of the day what makes someone a wireless engineer.
Vendor RRM Literature: