Plaintree Systems Inc C.NPT

Pros and Cons of FSO in Small Cells Discussion

[Painter4]

Lengthy discussion on Pros and Cons on using FSO for small cells.  Obviously these are individuals that are close to the business.

https://www.linkedin.com/groups/Why-is-no-one-promoting-1065157.S.145039947

Dave

Why is no one promoting FSO as a inexpensive solution for small cell inter-connectivity?

I have seen a lot of discussion about 60GHz and E-band RF for small cell interconnectivity and mobile backhaul but nothing mentioned related to FSO. While a big proponent of MMW RF, I also recognise that FSO could satisfy some short range applications, interference free, for extremly short ranges or in areas where atmospherics are not an issue. Circuit availability could also be improved through redundancy or with a hybrid apporach using lower capacity, sub 6GHz back-up. In my opinion, FSO has always been a 9-5's vs. 5-9's tecnology but in the right situation it may be an economic solution. It certainly best resembles physical fiber, shares many of the same already high volume components and provides a nice dumb, low latency, physical layer, pipe.
Please weigh-in on why FSO proponents are not more vocal in promoting this technology.

7 months ago

• Close viewer

• Like
• Comment
• Follow
• Flag
  • Flag as Inappropriate
• More

Michael M., Mohamed A. and 2 others like this

You, Michael M., Mohamed A. and 2 others like this

22 comments • Jump to most recent comments

Steve

Steve H. • Agreed. When FSO is used appropriately it is a good technology at a good price. However, to achieve the level of 9's mentioned there are many places you cannot deploy it and over far less distances the equivalent mmw. When I came into the millimetric wireless business, I was struck by how successful 60GHz was due to it being needed to replace FSO that fell short in terms of reliability. The morale is that if the location is suitable, the range appropriate and the skills of the installation people up to it, then it makes sense to use FSO. However, bear in mind though that 60GHz prices are set to fall as new silicon becomes available, and volumes increase. Even more FSO vendors will be squeezed out and their pricing will not save them.

7 months ago

Dave

Dave R. • The promise of low cost 60GHz systems based on SiGE transceivers is a classic chicken and egg situation. Any monlithic semiconductor technology requires high volume to recover investment for low cost. Conversely, low cost is required to drive high volumes. Many people are investing now so hopefully we will see the economic benefits. I am more concerned about the effects of low power SiGE based systems and lower gain/wider beam width antennas on overall relaibility at 60Ghz. Anyone who has done a range calculation knows that when you reduce TX power to only 0-10dBm and use a lower gain 2 degree beamwidth antenna, reliable ranges will drop. These low cost, low power, easy to align 60GHz systems may not be a significant reliability improvement over FSO. Also, how will these new wider-beam lower power 60GHz system co-exist in the same urban canyons now being used for LOS by existing higher powered 60GHz systems?
I initiated this discussion because I knew I would get some good responses. I was hoping someone from LightePointe or another FSO company would chime-in.

7 months ago

Steve

Steve O. • I’m not sure I agree with the comment about range. Most SC sites are likely to be deployed in ranges in the order of 50-300m, well within what 60GHz is capable of, even when you drop down to the minimum antenna gains prescribed by ETSI (I haven’t looked at FCC or others closely).
Realistically, all the things which have precluded FSO in the past are still real issues, and you need to add in any potential Health & Safety implications of lasers down in street clutter, which I don’t profess to have much knowledge of .
The main driver for SCB systems has to be total cost of deployment, and once you start adding in expensive tracking systems, then FSO costs are likely to rise significantly. Also remember that mounting arrangements for SCB are going to be small and structurally less rigid than traditional point to point, so antenna's will either need to be as a wide a beamwidth as possible (bad for interference, especially from ITS) or self aligning. There are some quite novel self aligning schemes coming into the V band space at present.
Lastly don’t lose sight of some of the PtMP or MPtMP systems coming onto the market in 28/32/42Ghz bands. One in particular has been designed specifically for SCB, and it very impressive conceptually. SCB is likely to be the biggest contribution to new and innovative Microwave/Millimetric technologies that we have seen for many years, and its an exciting time to be part of this ecosystem.

7 months ago

Rory

Rory C. • Come up with a sub-$3000 300Mbps or faster FSO that can go 1/2 mile in Phoenix or a 10Gbps FSO for $40,000 that can go a mile and I'll be your first customer. Hey, throw in a 2Gbps 1 mile device for $10K while you are at it. Unfortunately, I don't see any of these coming to pass in the near future.

7 months ago

Jim

Jim J. • Great subject Dave. FSO or Free Space Optics has been around for quite some time. In the Broadcast Television space the HD-SDI Canon Cano-Beam has been a viable solution for many years. We have done some projects with MRV and some of their FSO data devices. Perhaps 60GHz is new and more exciting. FSO is susceptible to rain and heavy fog signal interruptions while 60GHz is more immune. FSO solutions are typically more expensive than our 60GHz solutions. FSO can be interrupted by the flight of a bird due to the narrow beam of light while a 60GHz beam is much wider making it less susceptible to irruption. The wider beam of 60GHz requires less accuracy when aiming and can permit some movement of the signal targets. FSO would not have worked for our NHL hockey in-net goal camera project as the hockey goal can move slightly during game action.

7 months ago

Nikolaos Μ.

Nikolaos Μ. T. • I would agree with all that FSO is overpriced in comparison with 60GHz technologies coming out in the market. Mind you the price is bound to fall even further when LTE related deployments increase. Another aspect is related with the installers and their radio background... Overall FSO has its place in the market though I would say not when it comes to SCB..that to me is stretching it. Those with radio planning experience are also correct in pointing out the 5-9s. Optics generally, in my opinion at least, would be more suited in fiber-to-the-antenna applications, especially with the density that would be required and the related capacity, though that would entail infrastructure concerns.

7 months ago

Dave

Dave R. • I am extremely pleased with the educated responses to my proposition. A check of my profile will clearly show my allegiance firmly within the MMW camp. The FSO/MMW debate remains as contentious as Chevy vs. Ford or Sox vs.Yankees. To compare apples to apples, consider a new generation of small FSO systems that do not necessarily produce a pencil beam transmission defined in Milliradians. A very small optical aperture might produce a 1-2 degree beamwidth that would project a 3-7 meter spot at 200 meters distance. A lens assembly similar to a rifle scope could double as a simple visual alignment tool and transmission lens. The receive lens could be larger to produce a good visual target and provide more receive gain. This system could also benefit from WDM to segregate individual mobile carrier services or increase capacity. Using a wider beam transmission would require higher transmission power but the eye safe regions from 1300 - 1550nm could mitigate safety issues. The benefits are that high power laser sources in the WDM wavelengths are already in mass production to support FiOS and similar deployments. Inexpensive single mode fiber with GBIC plugins could insert into any switch or router port and form the tranmitter and optical transmission line requiring only a small passive lens assembly on the light pole. More importantly beyond eye safety issues, this spectrum is unregulated. Finally, the low power yielded from SiGE 60Ghz systems using easily aligned wide beam antennas may not yield greater relaibility to justify the cost or delay in development.

7 months ago

Nikolaos Μ.

Nikolaos Μ. T. • Great to see more insight in the technology. So what you are suggesting is a novel use of FOS properties, or rather a novel adaptation of technology. This would resemble the fiber-to-the-antenna logic, but both techniques remain to be proven commercially... Do you have any such application in mind?

7 months ago

Pinhas

Pinhas R. • I am reading a lot of reference to 5-9's vs 9-5's here....From what I have read of millmetrewave datasheets, the distances specified there are typically for 3, maybe 4 9's - never 5-9's- and that's in Crane Rain Zone D - which is hardly extreme conditions either. You have to read the specs carefully. The mention of 99.999% is typically in a claim of "Up to 99.999% availability" - no mention of what conditions. So 5-9's availability can really only be achieved at shorter distances than the max specified - and the same can be said of FSO - just under different weather limitations.
When it comes down to it, both mmW and FSO are good technologies for SCB and virtual fibre-to-antenna - that each have their place and limitations.
In fact, a combination of both in parallel would offer 5-9's and at a longer distance than either can achieve separately - if 60GHz prices ever come down (still waiting.....) - that may become cost-effective.
I think FSO suffered from too much hype before the bubble burst in 2001, and people are wary of the technology.
In reality, FSO, when specified properly, offers a viable high bandwidth communication line for short distances, at lower pricing than 60GHz, and often better availability in rainy areas. I would also contend that in the urban canyon where much small cell backhaul will be installed, rain is far more common than fog, which is practically non-existent betweeen the buildings. Add to that questions of multipath and crosstalk and security of the radio waves, and I think FSO will be a better solution in many cases and parts of the world.

7 months ago

Nikolaos Μ.

Nikolaos Μ. T. • I will have to partly agree with Pinhas. If FSO is treated individually it could and in some cases I know it does constitute a better solution. But when one talks about MMW one has to take into acount the E-Band products as well which are not subject to the same attenuation scenarios (see ITU-R P.676-6, "Attenuation by atmospheric gases"). Most companies offering 60GHz radio links will offer 70-80GHz links as well. At this point technology is not the issue anymore...economy of scale is however.

7 months ago

Pinhas

Pinhas R. • Please note that while 70/80GHz is not prone to oxygen absorption, its susceptibility to rain attenuation is basically the same as 60GHz. So as far as availability in wet weather, its not a panacea.
As far as economies of scale, for short range a ~$3K gigabit FSO link is achievable at moderate volumes, and that's based on today's component costs (not the hope of SiGe prices dropping)........

7 months ago

Rory

Rory C. • Ubiquiti is selling a 700Mbps full-duplex 24GHz radio for $3K that can go up to 1.5 miles in pretty much most conditions and farther in more arid clients. Going to be hard to justify FSO with these types of radio products coming out..

7 months ago

Nikolaos Μ.

Nikolaos Μ. T. • I second Rory's response although, in a dense urban environment 24GHz might not be the frequency of choice for very short distance last mile deployments (I am more inclined to see these technologies in the underserved suburban and rural regions). As Pinhas suggested a ~$3K FSO could be achievable at moderate volumes..but the industry is commonly building up on acquired knowledge. So based on the fact that the major telecommunication manufacturers have the know-how on ODUs, moving forward to an all-outdoor all-IP unit makes economic and R&D sense. Subsequently volume production leans towards radio links rather than FSO links...

BTW how cool is to see a cool refreshing design like Ubiquity airFiber :)

7 months ago

Drew

Drew L. • Only problem is that you still can't get AirFiber. I haven't heard of anyone who has received one yet.

7 months ago

Rory

Rory C. • Quite a few people have them but they are back on back-order status. They are just shipping slowly.

7 months ago

Pinhas

Pinhas R. • That is a nice looking product. I am curious how the bandwidth scales down vs. distance with the different modulation formats. 700Mbps is clearly not at maximum distance.

7 months ago

Rory

Rory C. • Speed, distance, quality, pick 2. 700Mbps is good out to at least 4.5 miles with no rain but 1-1.5 miles in heavy rain.

7 months ago

Greg

Greg A. • just so someone is promoting.......taking orders for NOVEMBER 2012.

https://www.microcom.us/af24.html

7 months ago

Obinna

Obinna E. • Deployed a number of them and it never worked as advertised even in very short hauls between 1 and 2 miles. There were a lot of support calls and outages especially when there was fog, rain, or snow. I will only recommend when deployed together with a redundant OFDM RF backhaul.

7 months ago

Pinhas

Pinhas R. • deployed what? if you mean FSO - 1 to 2 miles is not "very short haul" for the technology.

7 months ago

Obinna

Obinna E. • Pardon the use of the term "short haul" - It was used relative to distances other technologies would typically handle. My point is that decisions on the choice of equipment should be done based on proven field-tested results and not on sales pitch by the OEM vendors or theoretical's'.

Personally speaking I would not touch FSO except for last mile solution between two adjacent premises, and would do so with a redundant RF link. There are better technologies, that deliver similar speeds without the baggage, and that is why there is a decline in its use.

7 months ago

Pinhas

Pinhas R. • my point was that indeed FSO is a "last-mile" solution, and using it at 1 to 2 miles, availability will not be good without backup.
And it is a problem when salespeople do not help the customer specify needs and performance properly.
But for true gigabit solutions up to a kilometer, there are many places (anywhere there is a good amount of rain, etc.) where it gives the best availability, and with an RF backup, all the better.
not to mention upwards of gigabit......

7 months ago