I never said the utilities were required to buy back excess power from cogeneration sites. Just that many will do it.
The gain isn't really in generating a profit by selling power back to the utility. Its just a way of dropping your bill if you have to pull power from the grid occasionally for high-load activities on your site (home/business).
I don't think having a single "grid" covering a larger area is itself a problem. The problem arises when you don't have an appropriate set of safeguards in place to protect that grid from itself and an effective SCADA (Supervisory Control and Data Acquisition) network in place to monitor it all.
The technology is there to do it safely and reliably. It just costs money. Right now the major utilities have no profit motive to deploy technologies to harden and protect the power distribution and transmission systems.
In most states you can now sell power back to the utilities. A local generation plant (solar, hydro, wind, etc) can be connected to the power system via a utility intertie rated device.
This can as simple as a utility intertie rated inverter as part of you home solar system. Unfortunately for everyone else on your block, as soon as commercial input fails your system will stop providing power out to the utility side. This is to protect power company personnel during line repairs.
The SMARTAPAT stuff is all part of the US Army's FCS stuff. It won't be fielded for several years yet.
I understood the sarcasm. That is exactly why they have been developed. If you look back at Desert Strom a great number of the US casualties were caused by unexploded ICM (Improved Conventional Munitions) bomblets. Killing/Injuring troops with your own unexploded munitions is a bad thing. Killing/injuring noncombatants is even worse. Believe it or not the military is actually very sensitive to that. The use of mines is very tightly controlled.
(Note that I say noncombatants. It doesn't matter if they are wearing a uniform or not. If they are a threat then they are a combatant. A cute little 12-yr old girl with a mobile phone calling in troop's positions and adjusting artillery fire on them is a combatant. Is it a good thing? - NO! Its the way the real world works however.)
Mines themselves are a movement obstacle. Their value lies in reducing an ememy's options for manuever. Thats why they are often used to protect a flank or deny travel on a road/trail network. If you want to really chew up an enemy supply convoy then let them get in the middle of a decently sized minefield and then activate it. That will keep them in the zone until you can range artillery or air strikes on them to make sure those supplies don't get where they were headed.
The use of RF based sensor networks goes back at least as far as Vietnam. Thousands of seismic and audible sensors were dropped along the HoChiMin (sp?) trail network and used to attempt to monitor traffic. There were also seismic sensor packages which could be deployed around firebases and bivouack sites to detect personnel moving in the area. (I don't recall the nomenclature).
Its already part of the plan. They are generally classified Unattended Ground Sensors (UGS). They then get subclassed into SLUGS/BUGS/MUGS. They can be deployed in various ways and then used to form a sensor fence in an area.
The networking fits (mostly) the P2P definition. Its actually a MANET (Mobile Ad-Hoc Networking) implementation. Take a look around for the DARPA GloMoSim work or a couple other projects.
When you get into actual explosive mines you are talking about SMARTAPAT (at least that used to be the name). Same concept except with a boom and less battery. The general idea is that they can be fired out, placed into sense-only mode, and then when the sensors trigger an operator can check the indications and decide whether or not to allow detonation of a single node or blow the whole array. There are several other modes we'll just skip for the moment. The great thing is that they can be deactivated to allow friendly units to manuever through an area and most implementations automatically go into sense-only mode if they lose all links with an element capable of monitoring and commanding them. Add that to GPS derived location data for each unit and you've got an obstacle that is much easier to cleanup up at the end of hostilities.
I started out with a little company out of Dallas, TX called Metro Access Networks (MAN). I loved the name and loved the company even more. Finally, I could saw I was "workin' for the MAN!"
After about a year they got bought by Brooks Fiber Communications (Burger, Fries, and a Coke). About 6-9 months after that they got bought by the Borg (Worldcom). Shortly after that Worldcom bought MCI.
I said it then and I'll say it now, If WorldCom would have gotten out of its own way they could have ruled telecom in the US and been the major foreign carrier in most of the developed world. They had local, LD, internet, etc. Unfortunately Bernie was a huge M&A guy. He didn't know or care about integrating those companies once he bought them. The big BMW (Brooks-MCI-WorldCom) project to move all possible services on-net that started before I left WorldCom (Jan99 - I volunteered for the first layoff) was still in its 3rd/4th round of "planning" in late 2000 when I left another CLEC to head overseas. They wouldn't even let the local markets do circuit and building inventories to see what could be done. It all had to come from Bernie's chosen.
You're absolutely correct. Now that you remind me I think the calling card operator was using PRIs. In that particular case they were running the PRIs back to another state to be terminated in their switch. Their initial install had been done before I took over the market in question. When they went to add another 28 PRIs I (under advice of counsel) asked the sales rep to get a signed statement from them that they were operating a network which qualified under the "enhanced services exemption" rule. That was not a happy sales rep!
I'm not a big fan of the origination/termination tariffs for LD calls. If you look at the pricing on LD plans you'll almost always see a lower rate for interstate than intrastate. Its the same LATA tandem terminating both calls on exactly the same hardware. The difference is what the PUC has granted the ILEC (RBOC) on intrastate calls ($.025-.05/min) vs. what the FCC has set as the rate for interstate calls (?? ~$.005/min). An average of $.03/min starts adding up rapidly when you've got millions of channel/minutes per day. There is no technical difference so the only difference is regulatory.
Take it another step and look at the traffic settlements of most CLEC/ILEC interconnection agreements. The ILECs screamed until they got them put into the interconnection agreements. They expected a huge profit center. When the CLECs started signing up all the ISPs the ILECs screamed about "abuse of their network" amazing the difference in "fairness" to the ILEC mind between depositing the checks and having to write them. Even with settlements, they are pretty low. IIRC, $.0002-.0005/min for direct EO term, higher plus mileage for access tandem termination. It varies for each interconnection agreement. If the IXC owns a CLEC and terminates the call either directly to the EO (end-office - also known as LSO) or even via the access tandem then why should the interconnection tariff be 5-10 times higher? Its all in the regulation folks. I'm not sure what the solution is, but I'm very sure its not more regulation.
First, how is it supposed to work. SS7 pointcodes are like the IP address of a telephone switch. Messages are routed through an SS7 network that runs between switches to route calls, identify the source and destination information, and generate billing data. There are rather simple ways to conceal the origin of these calls. The ILECs (who own the InterLATA tandems) have gotten their friends, the state PUCs, to continue with quite high orig/term interconnection tarriffs. This is a huge source of revenue for them. The original concept was to pay for the large upfront expenditures to install the interlata tandems with the breakup of AT&T and the entrance of the new (at the time) IXCs. Those switches (and the required capacity upgrades) have been paid for hundreds of times over. When you consider $.05/min long distance and the orig/term fees are $.03-.04/min for both ends you see the IXC isn't exactly making much. Its a little present to the ILECs from the PUCs.
Many companies are doing this today via what is known as the "enhanced service provider exemption". In short, this states that Inter-LATA traffic which is carried across an enhanced services network (VoIP, VoATM, VoFR, etc) is not subject to InterLATA termination fees at the distant end of the call. The rules are pretty vague here and there doesn't appear to be a minimum percentage in the quantity of calls which must be handled by the enhanced services network or a percentage of the overall call distance that must be handled by the enhanced services network. What you get is folks that buy some to handle perhaps a T1s worth of trunks, place them next to each other in the rack, and route a few calls through it within a single office. Under the current rules they now operate an "enhanced services network" and are thus exempt from paying the orig/term InterLATA tarriffs. There is at least one large calling card provider (especially catering to the Hispanic population in the US) that does exactly this. The company then finds a friendly CLEC to allow them to dump their calls into the local network via MF (tone signalled, non-SS7) trunking and the origin of the call will appear to be a local number.
In the old days (pre-1999) there were several companies doing this without bothering to claim the enhanced service provider exemption. I've personally seen companies locate in a CLEC colocation facility and house nothing but a patch panel in a closed cabinet. MF trunks from IXCs (long distance carriers) are brought in on one side, and MF local-access trunks head out the other. This is also known as "dump and term".
When you're MCI (WorldDom) this becomes trivially easy. MCI owns at least 2 CLECs. WorldCom bought Brooks (I ran local operations in 2 cities for Brooks) shortly before the MCI deal. They also bought MFS several years before that. It would be a very simple matter to use an intermediary in each LATA to launder the traffic via MF trunks back into their MFS/Brooks switches and then pass them off to the ILEC (incumbent local exchange carrier) as what appear to be local calls. There isn't any high-tech SS7 munging required here.
This could also be accomplished via some sexy work with SS7 on a switch. It would be like NAT and would rewrite the originating point code and phone number to a local one. The same SS7 hardware would take the messages coming back and rewrite them to go to the proper switch. We do NAT with IP addresses every day. Its not a large stretch to imagine doing it with SS7. I don't see much of a need to though. There are much simpler ways to accompish it.
Hell, if MCI/Worldcom doesn't mind the exposure just run the MF trunks between local and LD switches without the intermediary. It opens up a huge liability hole, but it may have been deemed acceptable.
Sounds like you're talking about mobile ad-hoc networking. With some judicious configuration of routing (TORA/AODV/ODMRP/etc) protocols and a good firewall you can do all this today over 802.11x links.
A good portion of the software currently runs on Linux as that is what the research/development folks are using to build prototype systems.
I checked a little bit ago and SCOs (SCOX) market cap is only +- $180M US. The value of their lawsuit against IBM is $1B US. I think their initial aim was that IBM would buy them (with all their $0.66/share options - read the latest SEC filings) and they would make out like bandits.
Unfortunately for SCO, IBM didn't bite and take the bait. IBM realized they have much more staying power and enough resources to bleed SCO dry with legal costs. I don't think you're going to find too many financial institutions willing to lend SCO money to finance their legal jihad against IBM and Linux. The outcome is too uncertain and they don't have near enough assets on their balance sheets to make it a good bet for the financiers.
I realize that is the case. Unfortunately with a typical cable modem/DSL system you'll be able to do a fair bit of damage before you cut yourself off. Every time the links come up you take them down again. Combine this with a little wardriving around a large metro area and you've got a very large problem for the engineers. If you've got a large botnet you can do even more.
I was presenting it more as food for thought than a suggestion of an attack method. I'm on the other side of the food-chain for such things (network engineer) so I didn't post a full and complete dissertation.
More than likely you'll see solutions coming in stages.
1 - Very fragile and formative stage with various networks deployed by US organizations (civil and military) and international NGOs
-- HF radio networks (Voice and PACTOR/SITO)
-- Data modem use in areas that have a semi-working telephone system and some surviving wirelines
This is the system that will be the springboard for further development
2 - Initial efforts at restoring the PSTN.
-- Installing/repairing satellite ground stations for international commercial connectivity
-- A few international cable routes coming back on-line
-- Kabul and a couple other cities get some primary telephone exchanges rebuilt
-- Some type of national microwave network get s started up. Probably digital
-- Initial wireline data at decent speeds
This will leave large areas of the country without service and you'll see the HF/VHF networks pushing out towards the fringes. The wireline dedicated data network that starts here will be conceived as a small project and be horrendously overloaded and more popular than ever imagined.
3 - The initial real national infrastructure
-- Fiber routes are laid along road routes connecting major cities as the roads are rebuilt. (hopefully they can do all the rip/tear at one time)
-- Microwave routes start heading out to the less populated areas.
-- Cellular networks start getting rebuilt and offer more than regional connectivity.
-- Someone realizes the data network needs to be rethought and rebuilt......................
It goes on from there. I hope we don't see any national level patchwork of consumer grade equipment. While an interesting hobby project these folks need a real national communications infrastructure. Once that is in place you'll start seeing the less adventurous NGOs move in.
Very interesting thought. As I understand the LPFM rules they cannot be commercial in nature, but that would not necessarily be a problem.
Unfortunately I expect the NAB (who ramrodded the initial house resolution) will still argue that possible degradation within a 1300m radius of the LPFM antenna site is unacceptable. The initial HR was run through using receiver performance figures from about 1950-1960 (IIRC). Although the NAB will find it has a tough time arguing the technical quality of the work done by MITRE.
Seems like a decent open project. If they do relax the rules (which certainly isn't guaranteed) when will you have the site up? (G)
Its a little known fact that over 50% of the data the CIA gathers is unclassified. In the business it is known as "open source intelligence". What makes the CIA analysis classified is the collection, collation, and sifting of that data. There are many examples where the individual pieces of source data are unclassified, but the fusion of many pieces of data combined with analysis and algorithms is classified.
All they need to do is write up a quick 2-3 page government contract with a SOW (scope of work) to produce what he's already done. It would be an FFP (firm fixed price). $100-250K/yr would be a bargain. Once the government owns it they classify it and make extracts available to the involved companies to allow them to fix the problems. This would not be something new. An abstract can remain unclassified that he can include in his resume. The best idea would be to get him a job in either NIMA or CIA doing exactly what he's doing now.
The comment attributed to Richard Clarke -- "burn it" seems very unwise. Mr. Gorman has done an excellent job of intelligence fusion and synthesized a very useful body of information from a large collection of seemingly unrelated data. This is the holy grail of the intelligence community.
There are a few basic kinds of CLECs (competitive local exchange carriers).
1 - Local Facilities based: The CLEC has an independent CO (switch site) in the local area and can either extend facilities to you (not likely unless you are buying 2-4 DS1s+ of service) or extend POTS/DSL/DS1 service via copper loops from the closest LEC (local exchange carrier - the old baby bell) exchange.
2 - Non-Local Facilities Based: The CLEC's switch is located somewhere else and simply trunked into the area. They CLEC may or may not have direct colocation in the LEC's COs.
3 - Reseller: The CLEC just takes your order and passes it to the LEC to fulfill. Its still the LECs lines, switch, numbers, etc.
There are also myriad variations on the above. In general if you can get service from a local facilities based CLEC go for it. Most of them aren't really setup for residential (not profitable), but you may get lucky in your area.
Not really that expensive. The last few networks I've built had a GPS 1pps output disciplining a pair of rubidiums. Plenty accurate for telco requirements (even with a couple days of hold-over).
I've never seen a hydrogen maser used in a telco network time sync setup. Just LORAN and GPS.
Based on what I see in the article they are talking about using a VOIP infrastructure to replace their TDM infrastructure between switches.
(I'll use the US network as an example since its the one I'm most familiar with)
Right now each state is broken up into LATAs (Local Access Telecommunication Area - IIRC). Within each LATA there is a LD tandem which lets IXCs (long-distance carriers) accept and terminate calls to the local phone providers within that LATA.
The IXCs in turn have their own switches which are connected to the LD (inter-lata) tandem. The IXC switches are interconnected with each other via a TDM network. The TDM network uses multiplexers to build T1 (E1) service up to high-speed carrier level for transport between physical switch sites. Each T-1 gets broken out again and plugged in individually to another switch somewhere. (I know some switches can handle DS3 directly - I'm simplifying a bit) For the duration of the call a full 56/64Kbps DS0 line is nailed end-to-end.
Since this gets a bit expensive the network isn't physically a full-mesh. Smaller switches (leaf nodes) connect to major switches to route calls across the network. To acheive the P.01 service grade (1 call in 100 blocked) a large percentage of this capacity sits idle most of the time.
By going to a VOIP infrastructure for inter-switch trunking you get several advantages:
1 - Full mesh network. A call from Frog Jump, Kentucky can go directly to Fort Stockton, Texas. This frees up a large number of switch ports at the major sites and can greatly simplify call routing.
2 - Efficient bandwidth utilization. The packets are only flowing when there is audio to send. During pauses in speech (Probably something like 40% of a typical call) no data is sent (silence detection). (NOTE: This is where the noise comes in. A small bit of noise is sent to the near-end receiver during the silence intervals) With compression in he codec even more savings can be realized. Most implementations can detect fax/modem tones and switch over to a full-bandwidth codec so they won't degrade/disable fax/modem transmissions.
3 - Efficient bandwidth loading. Bandwidth can be sized for the total utilization of a switch rather than attempting to discern "average" calling patterns. Having done trunking analysis within a CLEC intra-lata network I can only imagine its a nightmare on LD networks.
4 - Equipment consolidation. The softswitch device can likely be plugged directly into some number of OC-3/12/48 interfaces. This cuts out things like M13 muxes, a DACS, copper patch panels, etc.
5 - Network consolidation. From a single office the IXC probably has a TDM voice network, IP network, frame-relay network, and possibly (especially outside the US) an X.25 network. The more of these things you can run across a single network the more economies you get.
All in all, its a very good thing. Its totally transparent to the customer, but it saves the IXC a big chunk of change.
Buffy - Wrapping up the series? (don't watch it, but recall the wife mentioning it.)
Should it be seen as a sign of the times that the nominees are all either going or gone? Makes you wonder about the intelligence of the masses. Oh wait, we already know about the intelligence of the masses.
Agreed, Mr. Moore chose to register the legal address of his business as his home address. He made this choice with full knowledge that spammers are not typically loved by internet users. The public owes him absolutely nothing.
Now if someone could arrange to get a couple tons of manure delivered to his front lawn, that would be funny.
I never said the answer was easy or convenient. I only said there was an available set of solutions.
Laws only exist to keep honest people honest and punish those who break them when they are caught.
I would not depend on the rule of law to protect my sensitive communications from interception and exploitation by entities that may desire to do so. Its illegal (in most countries) to steal credit card information and use it to make fraudulent purchases. We don't, however, rely solely on the protection of law. We insist that vendors we do online business with use proper cryptographic protection of our card information during the transaction, and we hope they take steps to protect it once they've got it. Why should our voice communications be any different?
I find the likelihood of effectively constraining the evil and unfair world very low. Its good to work toward it, but in the meantime I'll keep running encryption where I feel it is warranted.
If you don't want to be monitored ENCRYPT!!
on
Snooping on VOIP
·
· Score: 2, Informative
The only protection against eavesdropping is strong end-to-end encryption. We got the ECPA (86 - US) shoved down our throats so cellular companies could claim their systems were "protected" from unauthorized monitorin without having to actually spend money on embedding crypto hardware in subscriber units.
CALEA was just a pitiful attempt to keep LE agencies from having to spend big bucks on upgrading their monitoring hardware.
If an individual, organization, government agency, or other entity wants to monitor your communications badly enough they will. If you don't like that then use stong end-to-end encryption.
--BEGIN RANT MODE-- Instead of wringing you hands over the evil and unfair world we live in just deal with it and work around it.
Its not exactly difficult to properly encypt just about anything you send. How many actually do it? Want to bet those same people that can't be bothered to use strong encryption are some of the first to whine about monitoring? --END RANT MODE--
The US mothballed its only active ASAT system in the late 80s/early 90s (IIRC). It was carried by the F-15 and had definite limits on the altitude of satellite it could engage and negate.
Effectively intercepting an orbiting platform is not a trivial task.
A more achievable solution would be some scattered jamming dishes broadcasting wideband noise on the uplink frequency. These would, of course, be a huge target for air strikes. With a combination of fixed and mobile units you might manage to jam reasonably effectively on civillian birds for about 12-18 hours before all your assets were located and negated.
1 - I wholeheartedly agree with you. I work with FCS at work and if enough folks can keep from tweaking it to uselessness we might just have a decently working system.
I think most of the posters here just don't quite get it. FCS is part of the transformation process that DOD is currently going through. The reality is that the current force structure is not well suited for what are likely to be the primary threats of the first half of the 21st century.
Throw in sub-national groups (terrorists) and you have an even larger gap in the current force structure.
The vehicles are depicted with the bulges at the base so you can't tell for sure whether they have wheels or tracks. At this point they haven't issued a spec that calls for either. If they showed wheels, the tracked proponents would bog the whole thing down with calls for studies and investigations into why they weren't using tracks. If they showed tracks it would be the opposite. By masking that area somewhat they can avoid the petty bickering and get a mission task oriented requirement done and then let the vendors submit test articles that have either (or do that part of the chassis work then.)
Slashdot Mirror
I never said the utilities were required to buy back excess power from cogeneration sites. Just that many will do it.
The gain isn't really in generating a profit by selling power back to the utility. Its just a way of dropping your bill if you have to pull power from the grid occasionally for high-load activities on your site (home/business).
I don't think having a single "grid" covering a larger area is itself a problem. The problem arises when you don't have an appropriate set of safeguards in place to protect that grid from itself and an effective SCADA (Supervisory Control and Data Acquisition) network in place to monitor it all.
The technology is there to do it safely and reliably. It just costs money. Right now the major utilities have no profit motive to deploy technologies to harden and protect the power distribution and transmission systems.
In most states you can now sell power back to the utilities. A local generation plant (solar, hydro, wind, etc) can be connected to the power system via a utility intertie rated device.
This can as simple as a utility intertie rated inverter as part of you home solar system. Unfortunately for everyone else on your block, as soon as commercial input fails your system will stop providing power out to the utility side. This is to protect power company personnel during line repairs.
The SMARTAPAT stuff is all part of the US Army's FCS stuff. It won't be fielded for several years yet.
I understood the sarcasm. That is exactly why they have been developed. If you look back at Desert Strom a great number of the US casualties were caused by unexploded ICM (Improved Conventional Munitions) bomblets. Killing/Injuring troops with your own unexploded munitions is a bad thing. Killing/injuring noncombatants is even worse. Believe it or not the military is actually very sensitive to that. The use of mines is very tightly controlled.
(Note that I say noncombatants. It doesn't matter if they are wearing a uniform or not. If they are a threat then they are a combatant. A cute little 12-yr old girl with a mobile phone calling in troop's positions and adjusting artillery fire on them is a combatant. Is it a good thing? - NO! Its the way the real world works however.)
Mines themselves are a movement obstacle. Their value lies in reducing an ememy's options for manuever. Thats why they are often used to protect a flank or deny travel on a road/trail network. If you want to really chew up an enemy supply convoy then let them get in the middle of a decently sized minefield and then activate it. That will keep them in the zone until you can range artillery or air strikes on them to make sure those supplies don't get where they were headed.
The use of RF based sensor networks goes back at least as far as Vietnam. Thousands of seismic and audible sensors were dropped along the HoChiMin (sp?) trail network and used to attempt to monitor traffic. There were also seismic sensor packages which could be deployed around firebases and bivouack sites to detect personnel moving in the area. (I don't recall the nomenclature).
Its already part of the plan. They are generally classified Unattended Ground Sensors (UGS). They then get subclassed into SLUGS/BUGS/MUGS. They can be deployed in various ways and then used to form a sensor fence in an area.
The networking fits (mostly) the P2P definition. Its actually a MANET (Mobile Ad-Hoc Networking) implementation. Take a look around for the DARPA GloMoSim work or a couple other projects.
When you get into actual explosive mines you are talking about SMARTAPAT (at least that used to be the name). Same concept except with a boom and less battery. The general idea is that they can be fired out, placed into sense-only mode, and then when the sensors trigger an operator can check the indications and decide whether or not to allow detonation of a single node or blow the whole array. There are several other modes we'll just skip for the moment. The great thing is that they can be deactivated to allow friendly units to manuever through an area and most implementations automatically go into sense-only mode if they lose all links with an element capable of monitoring and commanding them. Add that to GPS derived location data for each unit and you've got an obstacle that is much easier to cleanup up at the end of hostilities.
I started out with a little company out of Dallas, TX called Metro Access Networks (MAN). I loved the name and loved the company even more. Finally, I could saw I was "workin' for the MAN!"
After about a year they got bought by Brooks Fiber Communications (Burger, Fries, and a Coke). About 6-9 months after that they got bought by the Borg (Worldcom). Shortly after that Worldcom bought MCI.
I said it then and I'll say it now, If WorldCom would have gotten out of its own way they could have ruled telecom in the US and been the major foreign carrier in most of the developed world. They had local, LD, internet, etc. Unfortunately Bernie was a huge M&A guy. He didn't know or care about integrating those companies once he bought them. The big BMW (Brooks-MCI-WorldCom) project to move all possible services on-net that started before I left WorldCom (Jan99 - I volunteered for the first layoff) was still in its 3rd/4th round of "planning" in late 2000 when I left another CLEC to head overseas. They wouldn't even let the local markets do circuit and building inventories to see what could be done. It all had to come from Bernie's chosen.
You're absolutely correct. Now that you remind me I think the calling card operator was using PRIs. In that particular case they were running the PRIs back to another state to be terminated in their switch. Their initial install had been done before I took over the market in question. When they went to add another 28 PRIs I (under advice of counsel) asked the sales rep to get a signed statement from them that they were operating a network which qualified under the "enhanced services exemption" rule. That was not a happy sales rep!
I'm not a big fan of the origination/termination tariffs for LD calls. If you look at the pricing on LD plans you'll almost always see a lower rate for interstate than intrastate. Its the same LATA tandem terminating both calls on exactly the same hardware. The difference is what the PUC has granted the ILEC (RBOC) on intrastate calls ($.025-.05/min) vs. what the FCC has set as the rate for interstate calls (?? ~$.005/min). An average of $.03/min starts adding up rapidly when you've got millions of channel/minutes per day. There is no technical difference so the only difference is regulatory.
Take it another step and look at the traffic settlements of most CLEC/ILEC interconnection agreements. The ILECs screamed until they got them put into the interconnection agreements. They expected a huge profit center. When the CLECs started signing up all the ISPs the ILECs screamed about "abuse of their network" amazing the difference in "fairness" to the ILEC mind between depositing the checks and having to write them. Even with settlements, they are pretty low. IIRC, $.0002-.0005/min for direct EO term, higher plus mileage for access tandem termination. It varies for each interconnection agreement. If the IXC owns a CLEC and terminates the call either directly to the EO (end-office - also known as LSO) or even via the access tandem then why should the interconnection tariff be 5-10 times higher? Its all in the regulation folks. I'm not sure what the solution is, but I'm very sure its not more regulation.
Here is how something like this typically works.
First, how is it supposed to work. SS7 pointcodes are like the IP address of a telephone switch. Messages are routed through an SS7 network that runs between switches to route calls, identify the source and destination information, and generate billing data. There are rather simple ways to conceal the origin of these calls. The ILECs (who own the InterLATA tandems) have gotten their friends, the state PUCs, to continue with quite high orig/term interconnection tarriffs. This is a huge source of revenue for them. The original concept was to pay for the large upfront expenditures to install the interlata tandems with the breakup of AT&T and the entrance of the new (at the time) IXCs. Those switches (and the required capacity upgrades) have been paid for hundreds of times over. When you consider $.05/min long distance and the orig/term fees are $.03-.04/min for both ends you see the IXC isn't exactly making much. Its a little present to the ILECs from the PUCs.
Many companies are doing this today via what is known as the "enhanced service provider exemption". In short, this states that Inter-LATA traffic which is carried across an enhanced services network (VoIP, VoATM, VoFR, etc) is not subject to InterLATA termination fees at the distant end of the call. The rules are pretty vague here and there doesn't appear to be a minimum percentage in the quantity of calls which must be handled by the enhanced services network or a percentage of the overall call distance that must be handled by the enhanced services network. What you get is folks that buy some to handle perhaps a T1s worth of trunks, place them next to each other in the rack, and route a few calls through it within a single office. Under the current rules they now operate an "enhanced services network" and are thus exempt from paying the orig/term InterLATA tarriffs. There is at least one large calling card provider (especially catering to the Hispanic population in the US) that does exactly this. The company then finds a friendly CLEC to allow them to dump their calls into the local network via MF (tone signalled, non-SS7) trunking and the origin of the call will appear to be a local number.
In the old days (pre-1999) there were several companies doing this without bothering to claim the enhanced service provider exemption. I've personally seen companies locate in a CLEC colocation facility and house nothing but a patch panel in a closed cabinet. MF trunks from IXCs (long distance carriers) are brought in on one side, and MF local-access trunks head out the other. This is also known as "dump and term".
When you're MCI (WorldDom) this becomes trivially easy. MCI owns at least 2 CLECs. WorldCom bought Brooks (I ran local operations in 2 cities for Brooks) shortly before the MCI deal. They also bought MFS several years before that. It would be a very simple matter to use an intermediary in each LATA to launder the traffic via MF trunks back into their MFS/Brooks switches and then pass them off to the ILEC (incumbent local exchange carrier) as what appear to be local calls. There isn't any high-tech SS7 munging required here.
This could also be accomplished via some sexy work with SS7 on a switch. It would be like NAT and would rewrite the originating point code and phone number to a local one. The same SS7 hardware would take the messages coming back and rewrite them to go to the proper switch. We do NAT with IP addresses every day. Its not a large stretch to imagine doing it with SS7. I don't see much of a need to though. There are much simpler ways to accompish it.
Hell, if MCI/Worldcom doesn't mind the exposure just run the MF trunks between local and LD switches without the intermediary. It opens up a huge liability hole, but it may have been deemed acceptable.
Sounds like you're talking about mobile ad-hoc networking. With some judicious configuration of routing (TORA/AODV/ODMRP/etc) protocols and a good firewall you can do all this today over 802.11x links.
A good portion of the software currently runs on Linux as that is what the research/development folks are using to build prototype systems.
NRL
Assorted Papers
I checked a little bit ago and SCOs (SCOX) market cap is only +- $180M US. The value of their lawsuit against IBM is $1B US. I think their initial aim was that IBM would buy them (with all their $0.66/share options - read the latest SEC filings) and they would make out like bandits.
Unfortunately for SCO, IBM didn't bite and take the bait. IBM realized they have much more staying power and enough resources to bleed SCO dry with legal costs. I don't think you're going to find too many financial institutions willing to lend SCO money to finance their legal jihad against IBM and Linux. The outcome is too uncertain and they don't have near enough assets on their balance sheets to make it a good bet for the financiers.
I realize that is the case. Unfortunately with a typical cable modem/DSL system you'll be able to do a fair bit of damage before you cut yourself off. Every time the links come up you take them down again. Combine this with a little wardriving around a large metro area and you've got a very large problem for the engineers. If you've got a large botnet you can do even more.
I was presenting it more as food for thought than a suggestion of an attack method. I'm on the other side of the food-chain for such things (network engineer) so I didn't post a full and complete dissertation.
I don't think its status as a worm really matters much.
All an individual has to do is something like this:
1 - Generate a list of hosts on the net
2 - Grab the first host off the list
3 - traceroute to the host
4 - send 10-20 of the "specially crafted packets" to each hop in reverse order
5 - cache the IP addresses that have been used
6 - Loop around and start again on the next host. Skip addresses that have already been done.
Its not that hard folks. Get enough machines running something like the above and we will have some very serious problems.
I leave it up to your imagination on how to deniably inject the packets into the net.
More than likely you'll see solutions coming in stages.
......................
1 - Very fragile and formative stage with various networks deployed by US organizations (civil and military) and international NGOs
-- HF radio networks (Voice and PACTOR/SITO)
-- Data modem use in areas that have a semi-working telephone system and some surviving wirelines
This is the system that will be the springboard for further development
2 - Initial efforts at restoring the PSTN.
-- Installing/repairing satellite ground stations for international commercial connectivity
-- A few international cable routes coming back on-line
-- Kabul and a couple other cities get some primary telephone exchanges rebuilt
-- Some type of national microwave network get s started up. Probably digital
-- Initial wireline data at decent speeds
This will leave large areas of the country without service and you'll see the HF/VHF networks pushing out towards the fringes. The wireline dedicated data network that starts here will be conceived as a small project and be horrendously overloaded and more popular than ever imagined.
3 - The initial real national infrastructure
-- Fiber routes are laid along road routes connecting major cities as the roads are rebuilt. (hopefully they can do all the rip/tear at one time)
-- Microwave routes start heading out to the less populated areas.
-- Cellular networks start getting rebuilt and offer more than regional connectivity.
-- Someone realizes the data network needs to be rethought and rebuilt
It goes on from there. I hope we don't see any national level patchwork of consumer grade equipment. While an interesting hobby project these folks need a real national communications infrastructure. Once that is in place you'll start seeing the less adventurous NGOs move in.
Very interesting thought. As I understand the LPFM rules they cannot be commercial in nature, but that would not necessarily be a problem.
Unfortunately I expect the NAB (who ramrodded the initial house resolution) will still argue that possible degradation within a 1300m radius of the LPFM antenna site is unacceptable. The initial HR was run through using receiver performance figures from about 1950-1960 (IIRC). Although the NAB will find it has a tough time arguing the technical quality of the work done by MITRE.
Seems like a decent open project. If they do relax the rules (which certainly isn't guaranteed) when will you have the site up? (G)
Its a little known fact that over 50% of the data the CIA gathers is unclassified. In the business it is known as "open source intelligence". What makes the CIA analysis classified is the collection, collation, and sifting of that data. There are many examples where the individual pieces of source data are unclassified, but the fusion of many pieces of data combined with analysis and algorithms is classified.
All they need to do is write up a quick 2-3 page government contract with a SOW (scope of work) to produce what he's already done. It would be an FFP (firm fixed price). $100-250K/yr would be a bargain. Once the government owns it they classify it and make extracts available to the involved companies to allow them to fix the problems. This would not be something new. An abstract can remain unclassified that he can include in his resume. The best idea would be to get him a job in either NIMA or CIA doing exactly what he's doing now.
The comment attributed to Richard Clarke -- "burn it" seems very unwise. Mr. Gorman has done an excellent job of intelligence fusion and synthesized a very useful body of information from a large collection of seemingly unrelated data. This is the holy grail of the intelligence community.
There are a few basic kinds of CLECs (competitive local exchange carriers).
1 - Local Facilities based: The CLEC has an independent CO (switch site) in the local area and can either extend facilities to you (not likely unless you are buying 2-4 DS1s+ of service) or extend POTS/DSL/DS1 service via copper loops from the closest LEC (local exchange carrier - the old baby bell) exchange.
2 - Non-Local Facilities Based: The CLEC's switch is located somewhere else and simply trunked into the area. They CLEC may or may not have direct colocation in the LEC's COs.
3 - Reseller: The CLEC just takes your order and passes it to the LEC to fulfill. Its still the LECs lines, switch, numbers, etc.
There are also myriad variations on the above. In general if you can get service from a local facilities based CLEC go for it. Most of them aren't really setup for residential (not profitable), but you may get lucky in your area.
Not really that expensive. The last few networks I've built had a GPS 1pps output disciplining a pair of rubidiums. Plenty accurate for telco requirements (even with a couple days of hold-over).
I've never seen a hydrogen maser used in a telco network time sync setup. Just LORAN and GPS.
Based on what I see in the article they are talking about using a VOIP infrastructure to replace their TDM infrastructure between switches.
(I'll use the US network as an example since its the one I'm most familiar with)
Right now each state is broken up into LATAs (Local Access Telecommunication Area - IIRC). Within each LATA there is a LD tandem which lets IXCs (long-distance carriers) accept and terminate calls to the local phone providers within that LATA.
The IXCs in turn have their own switches which are connected to the LD (inter-lata) tandem. The IXC switches are interconnected with each other via a TDM network. The TDM network uses multiplexers to build T1 (E1) service up to high-speed carrier level for transport between physical switch sites. Each T-1 gets broken out again and plugged in individually to another switch somewhere. (I know some switches can handle DS3 directly - I'm simplifying a bit) For the duration of the call a full 56/64Kbps DS0 line is nailed end-to-end.
Since this gets a bit expensive the network isn't physically a full-mesh. Smaller switches (leaf nodes) connect to major switches to route calls across the network. To acheive the P.01 service grade (1 call in 100 blocked) a large percentage of this capacity sits idle most of the time.
By going to a VOIP infrastructure for inter-switch trunking you get several advantages:
1 - Full mesh network. A call from Frog Jump, Kentucky can go directly to Fort Stockton, Texas. This frees up a large number of switch ports at the major sites and can greatly simplify call routing.
2 - Efficient bandwidth utilization. The packets are only flowing when there is audio to send. During pauses in speech (Probably something like 40% of a typical call) no data is sent (silence detection). (NOTE: This is where the noise comes in. A small bit of noise is sent to the near-end receiver during the silence intervals) With compression in he codec even more savings can be realized. Most implementations can detect fax/modem tones and switch over to a full-bandwidth codec so they won't degrade/disable fax/modem transmissions.
3 - Efficient bandwidth loading. Bandwidth can be sized for the total utilization of a switch rather than attempting to discern "average" calling patterns. Having done trunking analysis within a CLEC intra-lata network I can only imagine its a nightmare on LD networks.
4 - Equipment consolidation. The softswitch device can likely be plugged directly into some number of OC-3/12/48 interfaces. This cuts out things like M13 muxes, a DACS, copper patch panels, etc.
5 - Network consolidation. From a single office the IXC probably has a TDM voice network, IP network, frame-relay network, and possibly (especially outside the US) an X.25 network. The more of these things you can run across a single network the more economies you get.
All in all, its a very good thing. Its totally transparent to the customer, but it saves the IXC a big chunk of change.
Hmmm,
Firefly - Cancelled (and it was just getting fun)
Enterprise - Might be cancelled soon
Buffy - Wrapping up the series? (don't watch it, but recall the wife mentioning it.)
Should it be seen as a sign of the times that the nominees are all either going or gone? Makes you wonder about the intelligence of the masses. Oh wait, we already know about the intelligence of the masses.
Agreed, Mr. Moore chose to register the legal address of his business as his home address. He made this choice with full knowledge that spammers are not typically loved by internet users. The public owes him absolutely nothing.
Now if someone could arrange to get a couple tons of manure delivered to his front lawn, that would be funny.
FWIW, most of the CATV trunk lines have an aluminum jacket. I don't recall for sure, but I believe the center conductor may also be aluminum.
I never said the answer was easy or convenient. I only said there was an available set of solutions.
Laws only exist to keep honest people honest and punish those who break them when they are caught.
I would not depend on the rule of law to protect my sensitive communications from interception and exploitation by entities that may desire to do so. Its illegal (in most countries) to steal credit card information and use it to make fraudulent purchases. We don't, however, rely solely on the protection of law. We insist that vendors we do online business with use proper cryptographic protection of our card information during the transaction, and we hope they take steps to protect it once they've got it. Why should our voice communications be any different?
I find the likelihood of effectively constraining the evil and unfair world very low. Its good to work toward it, but in the meantime I'll keep running encryption where I feel it is warranted.
The only protection against eavesdropping is strong end-to-end encryption. We got the ECPA (86 - US) shoved down our throats so cellular companies could claim their systems were "protected" from unauthorized monitorin without having to actually spend money on embedding crypto hardware in subscriber units.
CALEA was just a pitiful attempt to keep LE agencies from having to spend big bucks on upgrading their monitoring hardware.
If an individual, organization, government agency, or other entity wants to monitor your communications badly enough they will. If you don't like that then use stong end-to-end encryption.
--BEGIN RANT MODE--
Instead of wringing you hands over the evil and unfair world we live in just deal with it and work around it.
Its not exactly difficult to properly encypt just about anything you send. How many actually do it? Want to bet those same people that can't be bothered to use strong encryption are some of the first to whine about monitoring?
--END RANT MODE--
In God we trust -- All others we monitor
The US mothballed its only active ASAT system in the late 80s/early 90s (IIRC). It was carried by the F-15 and had definite limits on the altitude of satellite it could engage and negate.
Effectively intercepting an orbiting platform is not a trivial task.
A more achievable solution would be some scattered jamming dishes broadcasting wideband noise on the uplink frequency. These would, of course, be a huge target for air strikes. With a combination of fixed and mobile units you might manage to jam reasonably effectively on civillian birds for about 12-18 hours before all your assets were located and negated.
1 - I wholeheartedly agree with you. I work with FCS at work and if enough folks can keep from tweaking it to uselessness we might just have a decently working system.
I think most of the posters here just don't quite get it. FCS is part of the transformation process that DOD is currently going through. The reality is that the current force structure is not well suited for what are likely to be the primary threats of the first half of the 21st century.
Throw in sub-national groups (terrorists) and you have an even larger gap in the current force structure.
The vehicles are depicted with the bulges at the base so you can't tell for sure whether they have wheels or tracks. At this point they haven't issued a spec that calls for either. If they showed wheels, the tracked proponents would bog the whole thing down with calls for studies and investigations into why they weren't using tracks. If they showed tracks it would be the opposite. By masking that area somewhat they can avoid the petty bickering and get a mission task oriented requirement done and then let the vendors submit test articles that have either (or do that part of the chassis work then.)