I have written much about this, however most recently want to dispell marketing fluff, but also want to lend some valid technical considerations if any of this is being thought about.  As I continually tell people, it is physics (yes you cannot change E=Mc2, and the pesking link budget. The link budget is the accounting of all the gains and losses from transmitter, through the medium (free space, cable, waveguide, fiber), to the receiver in a telecommunications  system. It accounts for the attenuation of the transmitted signal due to propagation, as well as the antenna gains, feedline and miscellaneous losses.

Received Power (dBm) = Transmitted Power (dBm) + Gains(db)-Losses (db)  or logarithmically            Prx = Ptx+Gtx-Ltx-Lfs-LM+Grx-Lrx

There are three salient issues about WLAN AP's and the WLAN Controller as a part of or integrated with a  DAS. 

1. Interference between adjacent AP's

2. LBS (location based solutions) beacon location

3. Wireless voice over IP compatibility. 

4. Cost

5. Support

1. AP Interference  – The interference can be somewhat solved by putting the AP's in metal bags or partioned in racks with metal dividers.  It is never a perfect solution. If there is interference then the software power controls from the WLAN controller through the AP do not work correctly and the bandwidth of the system drops by 30-40% or often it can just shut down.  This a is major issue for some manufactures of APs and WLAN controllers.  AP(s), need to beacon each other by design when out in the enterprise, but this is a really bad idea if stuck together with coaxial infrastructure in the IDF.

2. LBS Compatibility – Customers using Location Based Services (LBS) running on the DAS WLAN have a whole other set of problems most them relating to design changes.  The system can work on a DAS, but the problem is that when designing a WLAN that is LBS capable you have two rules that need to be followed.  1) APs positioned on the perimeter and 2) the ability to move AP(s) AFTER installation to optimize location performance.  Both 1. and 2. are very expensive and difficult to execute on when you have to run 1/2" coaxial cable to each and every antenna and they are at fixed locations for the antenna element.  Hospitals do not want to rip out their ceiling tiles a second time!  Understand that the AP has diversity of which is combined, so you also have to contend with this loss as well.

3. Wireless VoIP compatibility – RSSI or "received signal strength (indicator)" is the issue here.  In the early days of wireless VoIP, -72dbm was acceptable.  As hospitals started to deploy wireless VoIP, they quickly learned that -67dbm and some cases -65dbm was what was needed.  This increased the density of AP(s) in a dramatic fashion.  Any area where you had RSS to drop, you would experience handoff issues as well as dropped calls. Some DAS vendorshave have such a layered approach, but this changes the coverage issue to a capacity issue as APs are limited to 3 per layer. 

4. Cost – What is the cost delta between a separate WLAN and separate broadband services, (two separate quotes, versus trying to combine all of this.  Heck of a lot cheaper being separate. The potential client simply needs to compare these costs in detail.

5. Support – Will the manufacture of the WLAN infrastructure provide in writing that they will support all warranty and after warranty support (on site and calls), if they WLAN AP(s) and controller is combined with a DAS?

POSITIVES – Relocation of AP assests, some aesthetic improvement by putting AP not in the ceiling or in the IDF.  However, with LBS, and Voice over IP, going into the ceiling after the fact is a major issue.

NEGATIVES- 802.11n, no MIMO, or not true MIMO because antennas not within 2-3', high cost of cabling and additional combiners splitters, not LBS capable, lack of future compatibility with future WLAN or cellular vendor.  For instance support of LTE at 700MHz for Verizon 4G rollout?  Ask existing installations.  Will true 802.11n MIMO performance be guaranteed?  Will the WLAN provider of infrastructure put in writing that they will provide technical after installation support?  if any issue comes up, then the hospital will point the finger at the AP vendor and DAS vendor, and then they both say it is not our problem and it ends's up the customer and and end user problem. 

 Better to keep the WLAN separate, a lot less cost, can work right, and big time less issues. Who will assume the potential liability and eat the technology change costs if the applications do not perform the way they should?