Date: Mon, 26 Aug 1996 14:41:12 -0700 From: Greg Wiley <greg@loop.com> To: freebsd-isp@freebsd.org, isp-telco@bsd1.sytex.net Subject: Heads Up Message-ID: <1.5.4.32.19960826214112.00726058@pop.loop.com>
next in thread | raw e-mail | index | archive | help
Dialup Providers- This is one from a number of comments on this subject being discussed in the TELECOM list. This one points out the need for our voices to be heard by the regu- lators. Whether you're for or against the usage charges, you need to be heard or the issue will be resolved with- out you. -greg >To: greg >Subject: more from telcom group re: isps and co capacities > >------------------------------ > >Date: Sat, 24 Aug 1996 15:27:53 -0700 >From: lars@anchor.RNS.COM (Lars Poulsen) >Subject: Can ISP Dial-ins Really Cause Blocking in the CO? > > >Both in TELECOM Digest and on the COM-PRIV mailing list, the issue has >been raised about telephone companies complaining to the regulatory >authority that home access to the Internet through modem dial-in to >a local ISP places an undue burden on the local exchange facilities, >and the telco wants a regulatory change to put an end to this "misuse" >which is caused by the availability of flat rate local calling. > >As a suggested remedy, at least in the US, the telco wants ISPs to >be subject to the same two cent per minute access charges as the long- >distance telephone carriers. > >Many outside of the telco management are sceptical of these claims of >blocking, and observe that this request for tariff relief (which ISPs >claim will drive them out of business) comes just as the telcos >themselves are getting ready to roll out internet access services. > >The following quote from a knowledgeable journalist is illustrative: > > The [telcos] aren't provisioning their switches in suburban > areas to provide access rates at anything near the blocking > rates. > > In the city areas I am told that the modern switches are > usually provisioned to something near the 70-75% blocking rate > limit, while in suburban areas it is down around the 30%. So, > suburban calls are more likely to come up against the busy > barrier - however, no one I know has ever seen this happen at > times when people are surfing the Internet - which, is 9 to 11 > at night. > > The problem is that the carrier is claiming that the Internet > users are pushing the 'technical limits' of their exchange, and > leaving everyone with the suggestion that it would be a very > expensive thing to fix. In fact, we see no limit, and it is a > financial decision rather than a technical one. > >The carrier argument is cogently expressed by Bell Atlantic: > > Bell Atlantic did a study of the impact of the Internet explosion > during February and March, 1996. We submitted the study to the >Federal > Communications Commission. ... > > We have posted both the article and the original study on the >Bell > Atlantic Internet site (http://ba.com/ea/fcc). > >The study shows that there are in fact some exchanges where Internet >access traffic has exceeded the traditional busy hour, creating a new >busy hour during the evening hours, around 9 PM. In particular, this >seems to happen where a suburban area with no concentration of >business subscribers acquires an ISP. Such a bedroom community may for >years have been served by a Remote Switching Unit (RSU) using a >minimum amount of connection paths within the RSU and an even smaller >amount of trunkage to connect the RSU to the main switch, located in >another community. > >The typical subscriber line in this area may have had 20 minutes of >local calls per day and 10 minutes of calls outside of the RSU. As the >ISP moves in, some percentage of the residences now have 120 minutes >of internet access per day (i.e. five to six times to previous >traffic), and if the ISP is on the main switch, this new traffic >requires interoffice trunkage (so that the need for interoffice >trunkage for these subscribers is 12-15 times the previous >traffic). If the ISP is located on the same RSU, they take up a number >of line groups which will be fully loaded during the busy >hour. According to the Bell Atlantic report, each line group module >can accept 512 station ports, but provides only 64 channels into the >switching fabric. Thus, it is much more expensive to equip the RSU to >accommodate the ISP lines. > >After pondering these facts for a while, I realize that this doesn't >have to be a problem. If the telco spreads the ISP lines evenly across >the switch (putting no more than 16 on any one 512-port module) the >switch as described can easily take the load. > >Friends, if we are to survive this assault, we need to educate the >commissioners, so that they will be able to understand that the >"dangerous overload caused by ISP traffic" is just another >manifestation of a total failure of the telcos to understand the >nature of Internet access, leading to a failure to construct a working >network out of the perfectly good building blocks that they have on >hand. > >In the long run, I think we are nearing the end of flat-rate local >calls for residential subscribers. Back in February, I wrote an >article about the issues, it is still available on: > http://www.silcom.com/~lars/editorial/telecom.html > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1.5.4.32.19960826214112.00726058>