Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : LAST MILE TECHNOLOGIES - Let's Discuss Them Here -- Ignore unavailable to you. Want to Upgrade?

To: DenverTechie who wrote (1648)	7/23/1998 2:34:00 PM
From: Hiram Walker	Respond to of 12823

Mike et all, a good article about the "Killer App is killing us". internettelephony.com It should first be noted that the entire network is not built using remote terminals (RTs) as illustrated here. The older approach was to run individual paths (cable pairs) directly from the central office to each telephone. In the 1980s, telcos (led by the Bells but followed by virtually all the telcos) began introducing remote switches (RTs) as intermediate points in the local exchange network. At first, these RTs were fed by "t-lines" (copper pairs with repeaters operating as digital lines at 1.54 Mb/s providing 24 equivalent voice paths--known as the DS-1 level of digital multiplexing), but now are almost universally fed by fiber optics. These RTs reduced the investment needed to keep up with new growth and offered the possibility of extending fiber into the exchange distribution plant. Now, the RT-based design is the standard approach for meeting new growth requirements with the exception of growth very close to COs. We have developed a scenario that looks at the trunking required from the RT to the central office, currently and as Internet traffic continues to grow over the next few years (see sidebar, Remote Terminal Impact Calculations). Table 2 represents the results of evaluating this traffic on the central offices today and in the next few years. Even understanding that there are many assumptions behind the calculations, this table represents a very important result in terms of degrading the capacities of the central offices. The last column--% Reduction--shows how much will be lost in the installed central office capacities (while retaining the existing quality of service). The bottom line is, if Internet traffic continues to grow at current rates, i.e., at 60%, then we are going to face a very serious reduction (about 50%) in our CO capacity. This reduction would require an investment of approximately $25 billion to replace that capacity.19 Even in the stratospheric investment environment of the major telcos, that is a lot of money. Perhaps even more important, there is very little obvious revenue attached to this huge investment, and no indication that the telcos are moving to make this investment. This lack of activity suggests that the CO is going to be the first serious victim of the Internet. This result implies that we are going to see very substantial reductions in the quality (i.e., many more blocked calls) of our overall telephone service in the next few years. How Do We Save the Network? Everybody Has an Opinion The obvious answer to stopping this killer is to stop using the Web. If we agree that this is not a reasonable solution, then we must look elsewhere. Some possibilities are: Revise the tariffs. There is little doubt that a usage-based tariff at some level would reduce the load on the access networks.20 The question remains "Is this a realistic solution from a political and policy point of view?" The answer must be--"No, this is not acceptable; we want to broaden access to information, not limit it."21 Upgrade the traffic capacity of public networks.While this has a lot of attraction, the issue of paying for the upgrades must be addressed. How would the LECs ever be paid for this enormous upgrade? Without making substantial changes in the tariffs to pay for the upgrades, there is very little likelihood of this ever being a serious consideration. Use the remote terminals to deflect the traffic. This has the potential to be fairly effective. One vendor has recently offered a remote terminal that will do exactly this.22 Their terminal recognizes a call to an ISP and takes it off the network to a true data network (a packet network like the Internet itself). This approach would certainly resolve many of the problems identified in this paper. It would still leave the traffic on the remote terminal itself, but it could be designed to handle a very high incoming traffic load. The problems with this approach are: A large part--perhaps as many as 75%--of the existing CO lines are not served via RTs. The existing RTs would all have to be replaced. This replacement is close in cost to the entire network upgrade discussed above. (In fact, it would certainly be a candidate approach to achieve that upgrade.) Increase Internet network trunking. This has the potential for helping some, particularly in cases where the problem is a slowdown in processing (caused by the Internet itself, rather than its accesses). However, it would have no impact (except perhaps the negative impact of making the Web work better, therefore attracting more usage) on the real killer in the access networks.23 IP switching. This is just a special case of the above discussed "increase Internet trunking." It consists of introducing IP-based switches in the Internet network itself. It will have no impact on the fundamental problem of access traffic.24 Broadband. Here, we are addressing the simple approach of increasing the available bandwidth to the home, e.g., by a hybrid fiber/coax architecture. Yes, this may help some, in that it would allow data to come down the network (and up, for that matter) much faster. However, would it really help the long holding times? There has to be a real question as to its impact. If the tariffs were still the same, would the fact that data comes fast enough to have real motion significantly cut the holding times? My feeling is that it would not. There is no real research on this issue, but I think we would have to severely question the proposition that making the Web work much better (by making it faster) would cut down on usage. I think that as good an argument could be made for exactly the opposite.25 Take the load off the public network. This will be a very unpalatable answer as far as the LECs are concerned, but it is really just about the only answer. This would require the development of a separate network to the home. While, in some ways, this may sound as expensive as the upgrade of the network discussed above, it should be noted that there are many players today that want to do exactly this. Let us look a little more closely at this solution. Hiram

To: DenverTechie who wrote (1648)	7/25/1998 1:17:00 AM
From: Hal Barnett	Read Replies (1) \| Respond to of 12823

DenverTechie and Thread, Just read the last 75 posts. Excellent contributions. To put the current "Last Mile Technologies" discussion into a historical context consider my grandfather's first job as a "runner" for Illinois Bell in the 1890s. He worked in an office building in the Chicago Loop with a few telephones in the basement. His job was to answer the phone when it rang and then run to the office of the person being called and bring him back to the proper telephone set in the basement. Hal

To: DenverTechie who wrote (1648)	8/3/1998 5:12:00 PM
From: MikeM54321	Read Replies (1) \| Respond to of 12823

Question: Is xDSL solution a circuit-switched solution? In other words, if you have xDLS service, is this a fixed point to point wire that the CLEC comes out and runs a wire that operates like a typical voice line that isn't shared? Answer: DSL is a circuit switched solution, but even more accurately, it is a point to point solution that only operates between the customer premises and the central office. The high speed is accomplished by incorporating frequency division multiplexing (FDM) at frequencies well above those needed by traditional voice traffic. Typical voice circuits only require about 4 kHz of bandwidth while the potential on the copper pair is as high as 1 mHz. The voice is carried at the bottom, there is a guardband above it, then the data is multiplexed on as many as 250 seperate frequencies above the guardband. ________________________________ DenverTechie and Frank, I received this answer from the networking monitor on America Online. Just thought I would repeat his answer for others. You and Frank already replied to my same question and it agrees with his. Thanks. So even if every single neighbor of mine, connected to my local CO, decided to subscribe to 24x7, xDSL service, there would be no problem with the local CO not being able to place voice calls? Right? And it's all because of the splitter, right? Thanks, MikeM(From Florida)