I can tell you my experience from working on a multi-state trunking system for a very large petroleum company.
We had 6GHz DS3 backbone for most of the linking (radio, phone, computer, etc) but also had a few T1s and 3005s. I remember only one problem ever with one of the microwave links, and it was a hardware failure in one of the radios. Our Harris Constellation radios were set up with hot standby units so that the back took over as soon as the primary had a problem. All sites were also set up for vertical space diversity. It was just a rock-solid system; due mostly to the top-quality equipment and engineering.
Then we had the land-based telco circuits. I became very good friends with all the techs from Sprint and Valor Telecom. At least twice a month something would go wrong with the circuits. Downtimes of up to a day were not uncommon. This is on a system that cost upwards of $1,000/second for downtime. Needless to say, I cursed them all the time.
We also had a small VSAT system for some pipeilne SCADA data. Satellite is probably overkill for regional communications, however, it is a real blessing for super rural and remote locations. The VSAT was also very reliable.
To sum it up, I would definitely push for the microwave system. A well-engineered system will be extremely reliable, and cheaper in the long run. The initial capital investment can be very high depending on how many paths there are and the distances. The only major recurring costs would be tower rental if you needed it, plus any maintenance. There is also the factor of how much does downtime costs...both in dollars and human lives.
Jeff Walsh - KC5IBY
jeff@waltel.com
SmartZone: Microwave linking, as opposed to Fiber - costs
Moderator: Queue Moderator
In general, I agree with Wayne, but add a couple of observations:
1. The engineering and licensing costs for a DS3 microwave system can be very high.
2. You will get good performance out of a DS3 microwave system only if you buy top of the line equipment (Harris) and spend what it takes for hot standby, UPS, and the like.
3. Depending on terrain, hop length and frequency, microwave can be subject to precipitation fading when it is raining or snowing out. Be sure to quiz your engineer hard on his frequency selections. Two or more short hops are better on this point than one long one, though even short hops can experience fade.
In doing your research, consider that the telcos are not the only source for "wirelines." There are a number of 3d-party vendors with their own fiber backbone systems, and most of these are quite reliable. What you have to work on is getting a drop path from the nearest fiber "POP," and if this is done with owned fiber, it, too, can be quite reliable.
1. The engineering and licensing costs for a DS3 microwave system can be very high.
2. You will get good performance out of a DS3 microwave system only if you buy top of the line equipment (Harris) and spend what it takes for hot standby, UPS, and the like.
3. Depending on terrain, hop length and frequency, microwave can be subject to precipitation fading when it is raining or snowing out. Be sure to quiz your engineer hard on his frequency selections. Two or more short hops are better on this point than one long one, though even short hops can experience fade.
In doing your research, consider that the telcos are not the only source for "wirelines." There are a number of 3d-party vendors with their own fiber backbone systems, and most of these are quite reliable. What you have to work on is getting a drop path from the nearest fiber "POP," and if this is done with owned fiber, it, too, can be quite reliable.
Just to clarify, I don't think they need a DS3 backbone for their PS communications system. I was only using it as a real-world example from a large system. You can buy excellent T1 radios from WMux and such for a fraction of the cost of of the Harris stuff, and they work just as well 99% of the time.
I will reiterate what RKG said about the engineering. The planning, site survey/selection, path analysis, etc is what really makes or breaks a microwave system. If you did decide to spend the big bucks on the Harris stuff, they have an excellent engineering team who will design the entire system. There also alot of independant microwave engineers out there, too, and it may be worth talking to one to see what they think before you even start the planning. They may see something that would make the microwave system not feasible that we on the internet can't see.
About the third-party fiber companies, it can be hit and miss in your area. Also, if you wanted to put in your own fiber from their POP to your site, it can get very expensive very fast. Not just for the actual laying of the fiber, but for all the permits and right-of-ways and such.
One thing that you could consider is a mixed fiber and microwave system. You could intermix fiber and microwave where each would be most practical. Expanding on what RKG said, you could use fiber to go between the carrier's POPs and then use microwave for the "last mile" hop between their facility and yours. This could end up saving you alot of money in recurring charges because the most expensive part of circuit is from the carrier's POP to your location, not the intra-POP link. I don't know where you are located, so I don't know who the major fiber and telco players are, but most would be willing to make special arrangements like above to accomodate public service and government agencies.
Jeff Walsh - KC5IBY
jeff@waltel.com
I will reiterate what RKG said about the engineering. The planning, site survey/selection, path analysis, etc is what really makes or breaks a microwave system. If you did decide to spend the big bucks on the Harris stuff, they have an excellent engineering team who will design the entire system. There also alot of independant microwave engineers out there, too, and it may be worth talking to one to see what they think before you even start the planning. They may see something that would make the microwave system not feasible that we on the internet can't see.
About the third-party fiber companies, it can be hit and miss in your area. Also, if you wanted to put in your own fiber from their POP to your site, it can get very expensive very fast. Not just for the actual laying of the fiber, but for all the permits and right-of-ways and such.
One thing that you could consider is a mixed fiber and microwave system. You could intermix fiber and microwave where each would be most practical. Expanding on what RKG said, you could use fiber to go between the carrier's POPs and then use microwave for the "last mile" hop between their facility and yours. This could end up saving you alot of money in recurring charges because the most expensive part of circuit is from the carrier's POP to your location, not the intra-POP link. I don't know where you are located, so I don't know who the major fiber and telco players are, but most would be willing to make special arrangements like above to accomodate public service and government agencies.
Jeff Walsh - KC5IBY
jeff@waltel.com
There are other very good microwave vendors beside Harris. I have not been very happy with their over priced, poorly sold and designed gear. In the past, they would try to sell you a hop and give you a bare bones radio. When you asked what happen to the order wire, or what ever, you got the answer that you didn't ask for that. Some of those features, you needed to make the radio play. They just kept adding and adding and adding the dollars to it.
Do yourself a favor and start looking some place else. Don't forget to ask about what the primary power is. Not every company has the ability to run their microwave from + 24 volts. If not, you need to have some form of - 48 volts available.
Try looking at Aclatel or DMC. Some of the radios have the ability to grow into more capacity. Like buying a single DS3 radio that can be expanded into a 3 DS3 radio. Hot standby may almost double the cost for a radio, BUT IT WILL STAY UP AND RUNNING.
The cost trade off comes in some place between 3 and 4 T1 circuits to compare microwave with. You can get some real low cost microwave equipment in non redundant that will walk circles around the cost of the 3 to 4 T1 cable costs.
If it's a true Public Safety operation, they should own the towers too. That way, there is no monthly rental for tower space. They have total control over what happens on the towers. If they were smart, (doesn't happen very often) they build extra capacity into the tower loading specs when they ordered the tower. If the tower is a Pirod, there is no extrra loading in it. You got only what you asked for. Not a pound extra.
Jim
Do yourself a favor and start looking some place else. Don't forget to ask about what the primary power is. Not every company has the ability to run their microwave from + 24 volts. If not, you need to have some form of - 48 volts available.
Try looking at Aclatel or DMC. Some of the radios have the ability to grow into more capacity. Like buying a single DS3 radio that can be expanded into a 3 DS3 radio. Hot standby may almost double the cost for a radio, BUT IT WILL STAY UP AND RUNNING.
The cost trade off comes in some place between 3 and 4 T1 circuits to compare microwave with. You can get some real low cost microwave equipment in non redundant that will walk circles around the cost of the 3 to 4 T1 cable costs.
If it's a true Public Safety operation, they should own the towers too. That way, there is no monthly rental for tower space. They have total control over what happens on the towers. If they were smart, (doesn't happen very often) they build extra capacity into the tower loading specs when they ordered the tower. If the tower is a Pirod, there is no extrra loading in it. You got only what you asked for. Not a pound extra.
Jim
Microwave vs Fibre or HDSL for SmartZone
Great points! But someone mentioned the cheaper unlicensed uwave radios "work 99% of the time".... sorry.. public safety won't buy unlicensed to begin with, and 99% won't cut it either. Will have to be 99.995 at least for T1 or even 99.999% for DS3.
Long 7 GHz Microwave hops (Hot standby radios, but only one antenna & feedline) cost about US$130k each... tower analysis, equipment, rigging, commissioning... etc. And that is when you have control of the process and costs and haven't outshopped it to a third party.
Then you add the cumulative behaviour of multiple hops and your outage potential (seconds per year) increases.... although adding the worst case scenario for multiple hops seems like doomsday.... if the K factor changes that much wouldn't it affect most hops at the same time? So why add them up as if each hop outage occurs at separate times.
SmartZone drops into site trunking if you have a uwave hit that is more than 1.5 seconds or so. The uwave radio will have to resync, which takes a dozen seconds to get stable, followed by a couple of minutes for the controller to recycle the site and bring it back into wide area. OUCH! The cops in their cars won't like that at all. Then your numbers go really bad... the 99.999% figure is for the uwave outage... add in the time the SZ site stays in site trunking and it gets ugly fast. Seems you almost need simplex radios to get back in touch as fast as possible without waiting for the SZ controller to do its thing. If you loose a DS3 with lots of sites riding, it will take much longer than 2 mintues to bring them back.
HDSL... can be problematic. This is what you will get for most rural sites if you request a T1 to the tower. Who has responsibility when an outage hits... is it in the CO...wire cut, repeater, or CPE? They usually have different union people that respond... can take 12 hours to fix.
Does your service provider with the fibre have decent backup power everywhere? What is the weakest link in the event of a large blackout (like August 14th)?
Lastly... the fibre rings you will be on... they are redundant, but can you guarantee that if a fibre cut happens and your T1 goes the long way around an OC-48 loop.... will your trip latency stay below the 5 ms required by SZ?
No easy answer.... still trying to figure it out.
RFDude.
Long 7 GHz Microwave hops (Hot standby radios, but only one antenna & feedline) cost about US$130k each... tower analysis, equipment, rigging, commissioning... etc. And that is when you have control of the process and costs and haven't outshopped it to a third party.
Then you add the cumulative behaviour of multiple hops and your outage potential (seconds per year) increases.... although adding the worst case scenario for multiple hops seems like doomsday.... if the K factor changes that much wouldn't it affect most hops at the same time? So why add them up as if each hop outage occurs at separate times.
SmartZone drops into site trunking if you have a uwave hit that is more than 1.5 seconds or so. The uwave radio will have to resync, which takes a dozen seconds to get stable, followed by a couple of minutes for the controller to recycle the site and bring it back into wide area. OUCH! The cops in their cars won't like that at all. Then your numbers go really bad... the 99.999% figure is for the uwave outage... add in the time the SZ site stays in site trunking and it gets ugly fast. Seems you almost need simplex radios to get back in touch as fast as possible without waiting for the SZ controller to do its thing. If you loose a DS3 with lots of sites riding, it will take much longer than 2 mintues to bring them back.
HDSL... can be problematic. This is what you will get for most rural sites if you request a T1 to the tower. Who has responsibility when an outage hits... is it in the CO...wire cut, repeater, or CPE? They usually have different union people that respond... can take 12 hours to fix.
Does your service provider with the fibre have decent backup power everywhere? What is the weakest link in the event of a large blackout (like August 14th)?
Lastly... the fibre rings you will be on... they are redundant, but can you guarantee that if a fibre cut happens and your T1 goes the long way around an OC-48 loop.... will your trip latency stay below the 5 ms required by SZ?
No easy answer.... still trying to figure it out.
RFDude.
Re: Microwave vs Fibre or HDSL for SmartZone
Sounds like RFdude has the red ass over microwave. I will say that a pooly enfineered system will work just that way. A well engineered system will work like it should. If microwave was so unreliable and costly, do you think that the major cellular carriers would use it.
There are many microwave systems around that handle the backbone system for many cellular carriers. They have millions of dollars riding on customer satisfaction and use of their cell systems. They can't afford to have a hop go down.
As for the fiber, it has this problem called the back hoe outage that can't be controlled. A ring operation is a must. As RFdude pointed out, the distance length of the loop in of some concern.
With microwave, once you get over about 4 T1 circuits, it becomes practicle on the money side. The exact number will depend on what the T1 circuits will cost you from the local service provider.
On the negative side of microwave, unless you own the towers, you have the monthly rent that has to be put into the cost. You need a generator to keep the system up when the power goes out.
Take and make a spread sheet of the cost involved. Spread it out say over 5 years and then you can see just what the true cost will be. The initial cost to install the microwave will be high, but over a 5 year period, you should see a payback and better reliability.
Just make sure you pick some good equipment and have good engineering done on the paths. It doesn't take a PHD to do the path studies, but it does take someone with the background and track record.
Harris gear would not be my first choice. They have a nasty habbit of forgetting to include needed parts in their price to the end user. Their answer is, "OH you didn't tell us you wanted to do that".
Jim
There are many microwave systems around that handle the backbone system for many cellular carriers. They have millions of dollars riding on customer satisfaction and use of their cell systems. They can't afford to have a hop go down.
As for the fiber, it has this problem called the back hoe outage that can't be controlled. A ring operation is a must. As RFdude pointed out, the distance length of the loop in of some concern.
With microwave, once you get over about 4 T1 circuits, it becomes practicle on the money side. The exact number will depend on what the T1 circuits will cost you from the local service provider.
On the negative side of microwave, unless you own the towers, you have the monthly rent that has to be put into the cost. You need a generator to keep the system up when the power goes out.
Take and make a spread sheet of the cost involved. Spread it out say over 5 years and then you can see just what the true cost will be. The initial cost to install the microwave will be high, but over a 5 year period, you should see a payback and better reliability.
Just make sure you pick some good equipment and have good engineering done on the paths. It doesn't take a PHD to do the path studies, but it does take someone with the background and track record.
Harris gear would not be my first choice. They have a nasty habbit of forgetting to include needed parts in their price to the end user. Their answer is, "OH you didn't tell us you wanted to do that".
Jim
-
nc5p
- Posts: 293
- Joined: Tue Jul 01, 2003 6:52 pm
- What radios do you own?: XPR5550 XTL5000 XTS2500(V&U)
I've had a few bad experiences with unlicensed spread spectrum radios. I could see them for emergencies but not for permanent installation.
If you are doing any simulcast you will want microwave. The telcos tend to hose things up there. They will route your bit stream all over the place. Harris does make some nice medium capacity equipment, you don't need to go DS3. For protection you can go with a ring topology instead of monitored hot standby, just make sure your equipment can re-align itself when it turns around. I've seen it done both ways, just depends on the geography of your service area and your budget. From a homeland security point the ring is better, if somebody blows up a site MHS does no good! Be careful in big cities, I know an electric utility that lost their loop protection because a new tall building went up in their path.
Sometimes you can get access to free fiber, just depends on who you are and if your entity (government, utility, railroad) control any rightaways. I've seen that done a few times. Somebody wants to put in a fiber and in exchange for letting them you get a few fibers for your own use. That is a good way to go, at least you control it instead of some phone company who decides to route you through the next state. I even knew of a tv station in west texas that had a free fiber to their transmitter, don't know how they managed that. Of course, fiber is subject to backhoe, earthquake, and mudslide outages. Microwave can get rained out, especially the 18 and 23 GHz stuff.
Doug
If you are doing any simulcast you will want microwave. The telcos tend to hose things up there. They will route your bit stream all over the place. Harris does make some nice medium capacity equipment, you don't need to go DS3. For protection you can go with a ring topology instead of monitored hot standby, just make sure your equipment can re-align itself when it turns around. I've seen it done both ways, just depends on the geography of your service area and your budget. From a homeland security point the ring is better, if somebody blows up a site MHS does no good! Be careful in big cities, I know an electric utility that lost their loop protection because a new tall building went up in their path.
Sometimes you can get access to free fiber, just depends on who you are and if your entity (government, utility, railroad) control any rightaways. I've seen that done a few times. Somebody wants to put in a fiber and in exchange for letting them you get a few fibers for your own use. That is a good way to go, at least you control it instead of some phone company who decides to route you through the next state. I even knew of a tv station in west texas that had a free fiber to their transmitter, don't know how they managed that. Of course, fiber is subject to backhoe, earthquake, and mudslide outages. Microwave can get rained out, especially the 18 and 23 GHz stuff.
Doug
".... sorry.. public safety won't buy unlicensed to begin with, and 99% won't cut it either. Will have to be 99.995 at least for T1 or even 99.999% for DS3."
Most public safety funding comes from politicians who will buy (licensed or not) equipment from the vendor who has the best lobbying technique. And while all those 9's look impressive, there are many users of microwave links today that cope with a reliability factor of only slightly more than 99 percent, due to budget constraints.
It is hard to get the money for the last .499%, in either microwave or VHF/UHF infrastructure.
Most public safety funding comes from politicians who will buy (licensed or not) equipment from the vendor who has the best lobbying technique. And while all those 9's look impressive, there are many users of microwave links today that cope with a reliability factor of only slightly more than 99 percent, due to budget constraints.
It is hard to get the money for the last .499%, in either microwave or VHF/UHF infrastructure.
-
ASTROMODAT
- Posts: 1825
- Joined: Tue Nov 05, 2002 12:32 am
rOf, the issue of microwave vs. fiber is not something that can be easily answered in a few paragraphs here. For instance, to determine the true long run comparative economic costs of these two technologies, typically Cellular companies, Public Safety Agencies, etc. run Net Present Value (NPV) studies with very sophisticated economic tools, such as the Bellcore CUCRIT NPV Evaluator program.
A properly designed and engineered digital microwave radio system, with proper fade margins and an acceptable Minutes of Outrage/Year threshold, will do you just fine. You need to run the economics for your particular circumstances. You certainly can not summarily dismiss one technology over the other.
However, when you evaluate the overall NPV of both alternatives, make sure you don't forget to put in ALL of the relevant cost items. For instance, don't forget to load some costs in for paying your Tec triple time to go up to the remote site New Years Eve in a snowcat to restore the system due to a Murphy's Law parameter.
Don't forget that there are also a whole host of non-pecuniary factors that you will also need to broadly consider. For instance, what happens if the Telco goes on strike, and their fiber goes down during such a period. Will the Telco managers restore service as quickly as otherwise? If not, what is the impact to your business if this happens? How often might this happen, and what is the probability, etc.?
ALL of this can be modelled in the Bellcore NPV CUCRIT economic evaluator engine, and you can even generate and evaluagte risk analysis scenarios that allow you to do very sophisticated Monte Carlo Risk Analysis for competing alternatives. One important parameter to decide on up front is the length of your study period. Is it 3, 5, 7 , 10 or even more years?
There are many very good economic evaluator tools that can help you pecunarize and quantify as many of the pertinent factors as possible, so you can make a true business decsion, as opposed to gut level instincts and emotion based decision making.
Good Luck!
Larry
A properly designed and engineered digital microwave radio system, with proper fade margins and an acceptable Minutes of Outrage/Year threshold, will do you just fine. You need to run the economics for your particular circumstances. You certainly can not summarily dismiss one technology over the other.
However, when you evaluate the overall NPV of both alternatives, make sure you don't forget to put in ALL of the relevant cost items. For instance, don't forget to load some costs in for paying your Tec triple time to go up to the remote site New Years Eve in a snowcat to restore the system due to a Murphy's Law parameter.
Don't forget that there are also a whole host of non-pecuniary factors that you will also need to broadly consider. For instance, what happens if the Telco goes on strike, and their fiber goes down during such a period. Will the Telco managers restore service as quickly as otherwise? If not, what is the impact to your business if this happens? How often might this happen, and what is the probability, etc.?
ALL of this can be modelled in the Bellcore NPV CUCRIT economic evaluator engine, and you can even generate and evaluagte risk analysis scenarios that allow you to do very sophisticated Monte Carlo Risk Analysis for competing alternatives. One important parameter to decide on up front is the length of your study period. Is it 3, 5, 7 , 10 or even more years?
There are many very good economic evaluator tools that can help you pecunarize and quantify as many of the pertinent factors as possible, so you can make a true business decsion, as opposed to gut level instincts and emotion based decision making.
Good Luck!
Larry
Thanks Larry... Sometimes one forgets that someone else has "BTDT" and the best place to look is IEEE or Telcordia. I think I do enough of this to warrant the tool.
I checked google and Telcordia for NPV CUCRIT and came up with version 5.4 but no current info on the product. Any other ideas about where to find this?
When dealing with public safety, cost isn't always an issue, but it does help to quantify risk management and put various $ vs risk choices.
Another place where one encounters various shades of grey is battery plant backup time... how many hours is enough? And when a resident generator is in place... how much battery in case the machine throws a rod? etc.
There is always a solution. All it takes is money.
Regards,
RFDude
I checked google and Telcordia for NPV CUCRIT and came up with version 5.4 but no current info on the product. Any other ideas about where to find this?
When dealing with public safety, cost isn't always an issue, but it does help to quantify risk management and put various $ vs risk choices.
Another place where one encounters various shades of grey is battery plant backup time... how many hours is enough? And when a resident generator is in place... how much battery in case the machine throws a rod? etc.
There is always a solution. All it takes is money.
Regards,
RFDude
In the installations for which I have some responsibility, we use the battery backup primarily to avoid station/computer reboots while the genset is starting, warming, and transferring load. We assume that if the genset will fail to accept load, we'll know that in a couple of minutes, and that it should take no more than 2 hours to get the Highway up with a skid mount and an extension cord. We therefore set our objective to have the batteries carry the radios for 4 hours without discharging beyond the 50% point. I real life, some experimentation shows that, depending in part on duty cycle (which might be on the high end, given that something must be going on for the power to have failed), we can get at least twice the target 4 hours.