Not only cost issues with broadband wireless repeaters (#2891)
I am back from a trip sans e-mail and enjoyed catching up on this thread. It can be fun to start from where I left off, and read a post or especially a post of a published article that seems to have logic or factual weaknesses, form a mental response, and then go on to read most of the my criticisms already leveled at the original. This is a very astute assembledge of readers/posters.
A very astute Bernard Levy pointed out some obstacles to wide-scale deployment of broadband wireless, and highlighted the cost of repeaters that would be used to overcome line of sight difficulties. Cost is certainly an issue, but after some deployment experience, my more fundamental concern has been topological. Even one-way repeaters (i.e., downstream only, as with the original CellularVision deployment) are a deployment, operational, and maintenance nightmare. They present a vivid, tangible reminder of the vagaries of SHF propagation and blockage. There may never be any practical way to effectively model fortuitous blockages that limit self-interference, or the season-variant foliage blockages that may make repeaters necessary for coverage. The real intellectual struggle, though, arises when these gadgets are two-way, as they obviously would be in any forward-looking implementation. The self-interference problem just get far uglier. In the TI/Bosch design (with their power levels, dynamic power control, and 90 degree sector topology),capacity is self-interference limited. Any system overlay that increases self-interference will noticibly impact system operation and capacity, so it is more than an academic curiosity. When you are in the field with active repeaters figuring out how to place, point, power, project specific beam-widths into shadow areas, you get really thankful if you only have to worry about a one-way broadcast architecture. Passive repeaters seem interesting until you stop to consider the problems of knowing what coverage you have really added or improved, absent a lot of wide-ranging and expensive reception testing. Whatever propagation tricks are being employed to discriminate adjacent sector signals get very complicated with any proliferation of what the MDS industry calls 'beam-benders', i.e., broadband repeaters.
I would be very interested to see theoretical descriptions of ways that additional, low-power two-way repeaters can be planned and deployed within an area illuminated by high-power sector transceivers such that the designed signal discrimination (say, c/i) is effectively and predictably maintained. This truly seems real-world difficult, given the propagation and blockage effects from structures and foliage, on top of the customary topology with general, averaged allowances for foliage. Averaging trees seems to no longer work at SHF, when the task is illuminating scattered signal-blocked areas.
Bosch, back when they were TI, did a lot of testing in Richardson and South America on the problems surrounding residential coverage. If that problem is defined as reaching multiple dwelling units, I think we are playing games. A ten story apartment building from a broadband wireless coverage perspective is a effectively business building -- its roof is above the tree line. There is probably a big roof area that presents substantial options for antenna placement. I put that case aside from discussion of residential coverage; yes, those sites can certainly be served by a broadband wireless system optimized for data transport to medium businesses. The tougher issue is serving suburban residential homes, either single family of attached housing. My recollection of the TI/Bosch coverage studies showed extreme difficulty in reaching over 60% of such addresses unless the subscriber antenna was located ~20 feet above the roof line. The Bosch 1-foot aperture reflector antenna, as well as most others that have been proposed, have too narrow a beam-width to be give reliable service if mounted on a guyed mast -- too much wind movement. What I believe TI concluded, and some of my independent experience suggested, is that the suburban residential market, where foliage is a factor, does not look like fruit likely to ripen at any height on the tree any time soon. The notion that the blockage issues can just be finessed with repeater placements seems to me to overlook the impracticalities of operating what would become a sort of wireless mesh network. I could imagine building it; I could even imagine making it work, but I can not imagine keeping it working at service levels that customers have come to expect. This would be a network necessarily in flux, with continual additions and changes. As such, it would need to be architecturally more robust than it could be with even a few two-way repeaters per sector.
As an investment observation, if some company really breaks the code on this, perhaps with some enhanced intelligence or smart antenna technology packaged so that the repeater is still economical and reliable, their market for such a proprietary unit could be substantial. It will be interesting to see if some of the smart antenna specialists currently aiming for the PCS/cellular base station market might connect with SHF specialists and figure out how to make a high-performance SHF model economically. I wonder if such companies or if such a hypothetical alliance would see moving to price points (say, <$1500) that might change the industry deployment picture. |
