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.

Strategies & Market Trends : Gorilla and King Portfolio Candidates -- Ignore unavailable to you. Want to Upgrade?

To: 100cfm who wrote (35211)	11/23/2000 8:32:57 AM
From: tbancroft	Read Replies (1) \| Respond to of 54805

What exactly is an erlang? "erlangs" were originally units of telephone traffic, associated with some relatively complex (to me) formulae for calculating the number of trunk lines needed in a network based on the acceptable number of "blocked" calls. In this age of wireless and data, erlang calculations remain an important means of quantifying the necessary capacity for a system. In the case of CDMA, EDGE, etc. it is a significant indicator of the efficiency of an approach, which translates directly into infrastructure cost advantages. The numbers you see are a major reason for believing in Qualcomm's continuing gorillahood. Also, for a little history lesson, see: pass.maths.org.uk tim@gottagocookaturkey.eat

To: 100cfm who wrote (35211)	11/23/2000 4:31:14 PM
From: voop	Respond to of 54805

erlang a unit of telephone traffic Also see Erlang, a programming language. In telecommunications, an erlang is a number between 0 and 1 that indicates how busy a telephone facility is over a period of time. An erlang of 1 applied to a particular telephone circuit would indicate Busy 100% of the time. An erlang can be applied to the group of lines in a telephone trunk line or to the traffic in a telephone call center. The term is named after the Danish telephone engineer, A. K. Erlang, the originator of queueing theory. Erlang B is a calculation for any one of these three factors if you know or predict the other two: Busy Hour Traffic (BHT), or the number of hours of call traffic during the busiest hour of operation Blocking, or the percentage of calls that are blocked because not enough lines are available Lines, or the number of lines in a trunk group An extended version of Erlang B lets you add the factor of how many people who are blocked retry their calls immediately. Erlang C is a calculation for how many call agents (answerers) you'll need in a call center that has a given number of calls per hour, a given average duration of call, and an acceptable level of delay in answering the call. Selected Links Westbay Engineers in the U.K. lets you use the Erlang formulas in their Online Traffic Calculators.

To: 100cfm who wrote (35211)	11/26/2000 12:39:59 PM
From: Eric L	Read Replies (1) \| Respond to of 54805

100, Re: QCOM slide set featuring "3G CDMA for Wireless Internet Access" qualcomm.com Notice that Q actually confirms Docomo's 5/01 release of their WCDMA system in this slide. They label it as initial deployment of 2nd qtr 01, then they show Q's commercial availability for their wcdma chip as q3 01. Then they show an estimate of commercial availability for docomo's wcdma system as 2003. Don't get that >> When I first studied that slide, I considered it to be a bit "flaky" for several reasons. The good Doctor seems to be saying that (in reality) 1xMC & 1xEV will be commercially available before WCDMA is. It appears to me he may be applying "commercial availability" a little inconsistently in each of the 3 line items. As for WCDMA he has the NTT DoCoMo 1st Phase Deployment plugged in as Q2 2001. He then has WCDMA "Commercial Availability" (I read this as not exclusively DoCoMo) as 2003 (ESTIMATED). This is Qualcomm's estimate obviously, and not 3GPP's (which estimates 2002. I guess we will have to wait till 2002 to see whether this is correct in this estimate, or Dr. Jacobs. The key point is we are talking about "Commercial Availability" as opposed to "Limited Deployment" or "General Deployment" (terms used by Andrew Seybold in his "Wireless Roadmap" which has recently been updated: wirelessroadmap.com Andrew Seybold can hardly be considered "technology neutral" as he has a decided bias (as I do, and I'm sure you do) towards CDMA for any number of valid reasons. Despite this, I think he attempts to be objective in his mapping. I somewhat question his start dates on 3GDS-WCDMA "Trials Start". It seems to me that these actually proceeded any 1x-EV "Trials Start", but I won't nitpick. It appears he has 1xMC IS-2000 plugged in right for limited deployment end of Q1 2001 (MSM5105 IS-2000 compliant chipset rather than MSM5100) and he is showing IMT-2000 WCDMA commencing end 1st quarter 2002, where it is supposed to (which dictates "commercial availability", one would think. At the very least, Seybolds roadmap is a good supplement to Dr. Jacobs, to my way of thinking. Seybold explains his terms this way: >> The table above summarizes the main wireless data technologies and key deployment dates. When Infrastructure manufacturers provide dates that their technology will be available, such dates indicate when the technology will be available to service providers. Typically, trials begin three to six months after a technology is "available," and about a year later there is limited deployment- the technology is made available commercially, but in limited geographic locations. General deployment has been achieved when the data technology is available in nearly the same coverage area as the existing voice technology . At this stage, products and solutions enter the market. << [EL - the tornado begins] >> For example, a computer manufacturer would most likely deploy a handheld device with an internal wireless modem at the general deployment stage. We wouldn't expect to see name-brand palm-size computers with GPRS wireless modems until the end of 2001. << << If docomo launches in q2 and we have commercial availability of our chips in q3 why would it be 2003 until docomo has commercial availability for their system. >> This is where I think the term "commercial availability" may be being misapplied or applied spuriously. Here is what I think I have been able to piece together about DoCoMo's plans. * DoCoMo plans to (hopes to) commence a VERY limited commercial trial of WCDMA on an Evolved GSM/MAPP core end of May 2001. Necessary debugging will take place in this stage and will impact on future dates. The trial will use infra and handsets manufactured to a stripped down version of the DRAFT 3GPP 'R99' standard. One of the reason that the trial will be very limited (infra. handsets, subs or users) is that components are NOT to IMT-2000 'R99' standards, and no question but there will be bugs. The network is probably in place now, operating in test mode, but not commercially "On Air". * DoCoMo plans to (hopes to) take delivery of standardized 'R99' infrastructure about 6 or 7 months before commencing a VERY limited commercial launch of WCDMA. This commercial launch is scheduled for late 2001 but could stretch into early 2002. Focus will be on corporate customers. There will be some introduction and testing of new services during this period. * DoCoMo plans to (hopes to) build out and begin commercial general deployment by end of Q1 2003 at which time they have conservatively targeted 150,000 WCDMA subs on their 3G 2GHz network. << What exactly is an erlang? >> Voop and Tim ("tbancroft") nailed this one nicely. One interesting aspect of WCDMA is that it (like 1xMC) CLAIMS to double VOICE capacity as compared to CDMA IS-95 and is 7x as spectrally efficient as 2G GSM). One of these days we'll see some slides out of the WCDMA proponents to compare to Qualcomm's widely used example. I'll let the folks that are into "erlangs" debate them. << is this the reason for the recent docomo confession re 3g shortfalls >> ROFLMAO! I guess you are talking about what Ben Garrett called "THE CONFESSION". Cracked me up when I saw it at 5 AM on Turkey Day" before motoring up the turnpike. I must say that his post has encouraged some very good posts on the "Moderated Qualcomm" thread. Let me direct your attention to the "1xEV Supporters" slide in Dr. Jacobs presentation: qualcomm.com "1xEV (Operator) Supporters" in the slide include: * KTF (SKT curiously lacking) * KDDI * Verizon * Sprint PCS And most importantly: * Vodafone I have long stated that Vodafone (perhaps more than any other GSM operator) is the potential key to unlocking GSM land for cdma2000 - either straight up or as an overlay). Vodafone is every bit as much a proponent and driver of the UMTS UTRA (WCDMA) specifications as DoCoMo and is also a proponent and driver of the GPRS specifications. They are a part owner of Globalstar, owner of one of the largest cdmaOne networks in the world, and managed the first cdma overlay of a GSM network (trial basis). If 1xEV has as much potential as we think it has why not WCDMA/1xEV (as well as 1xMC/3xMC/1XEV)? Regulatory barriers? Chris Gent would have something to say about that. This might require 3GPP2 disbanding and CDG & Qualcomm moving into 3GPP. I have conjectured before that when Qualcomm management thinks that this is appropriate, that it will happen. - Eric -