To: rharris who wrote (341 ) 4/7/1998 3:54:00 PM From: david waitt Read Replies (3) | Respond to of 638
OK.Great.We are getting some technical information.. And I have some questions. <Dave displayed his 2nd generation ISA card which is basically his wireless modem attached to a board that fits into any PC or laptop (these will be used in Santa Fe -the next city). It is a 2 Meg card with an ethernet card and wireless modem all as one unit which fits into any available slot.> So, the user has to install a card in his computer AND have an external modem? OK. This makes it a little difficult for laptop users. I guess MEEM will be in the business of manufacturing PCMCIA cards as well as ISA cards AND wireless modems and a network infastructure. < Because of RF utilization, MEME's system is capable of high-speed, large bandwidth access. Using a "Direct Frequency Spectrum" (as opposed to a frequency hopping system which doesn't provide enough bandwidth) MEME can offer:> WHAT ON EARTH IS "DIRECT FREQUENCY SPECTRUM"??? I must be falling behind. I have been in the field of wireless communications from Military SATCOM to commercial wireless data for 12 years and I have never heard of this. Could he have meant "Direct Sequence Spread Spectrum?" As far as Frequency hopping Spread Spectrum not providing enough bandwidth, this is not exactly true. The bandwidth limitation is not due to the fact that it is a frequency hopping system, it is due to the fact that in the 900MHz License free band ( where it sounds as if this system operates) the FCC states that you have to have at least 50 channels. The actual band is 902 - 928 MHz, thus each channel could be a maximum of about .5 MHz wide. If the FCC did not have that rule in place, then the bandwidth of a channel could be much wider. I have to add that it is these same rules (Part 15) that limit the transmit power ( of the modem and the radios that make up the network) to 1 Watt. There are many other restrictions that will force a network topology similar to that of Metricom's. I'm sorry to be beating this up people, but when I put on my RF Engineer hat, there are lot of things here that do not make sense. Regardless of what company I work for.. <1. 64K - 10 Megabits connectivity (residential to corporate applications).> Lets keep in mind here that there is a thing called "Bits per Hertz", A pretty high number for a wireless data system would be 4 bits per hertz. What this means is that you would need at leats 2.5MHz to transmit @ 10 Mbps and physics says that as you increase bandwidth, you lower the distance you can transmit that signal for a given transmit power ( which is limited by the FCC remember to 1 watt) Bottom line is that 10 Mbps is not going to go far with 1 watt of power, even with the highest allowed antenna gain ( 6 dB ).and a 6 dB antenna @ 900 MHz is to large to put on the wireless modem. < One main unit linked to different cells which could overlap can easily reach a 50 mile radius (actually Dave said it is essentially an "infinity" system).> I assume a MAIN UNIT is a unit connected to a T1 ( or a T3 ) line somewhere. ( It has to connect to the "Wired" internet somewhere!) If they are claiming that they will need just one of these for a 50 mile radius circle, they will need MANY MANY repeater radios to get the data from the user on the outside of that circle to the center of the circle. This is quite a infrastructure challenge. ( the sam one we face here at Metricom). Again, 1 watt, 2.5MHz Bandwidth, 6 dB of antenna gain, 25 miles to the center of the circle. Physics is going to get in the way! HHHmm. with Ricochet, I can connect from my office in Los Gatos CA to a modem in Washington DC with no problem! I Guess we have a "infinity" system also! < It operates on a 900 or 2.4 Mhz signal. Evidently the 900 Mhz band was developed by the military (The 900 Mhz uses a "spread" spectrum (zig-zag wave) and is harder to intercept and decipher).> Well, which is it, 900 MHz or 2.4GHz? I would guess it is 900 MHz between the modem and a radio in the network ( you get virtually NO in building penetration at 2.4GHz) and 2.4 GHz from radio to radio over the network. (Exactly what we are developing here at Metricom. I have a few of the radios here in my office). AND the 900 MHz band was not "developed" by the military. It has always been there. The military used spread spectrum methods to make it hard for an enemy to detect receive the information. Frequency Hopping was actually patented by the movie actress Hedy Lamar years ago. And the military has NOT abandoned spread spectrum at all! And yes. I know you were wondering, the user of the Ricochet network has the option of turning encryption on or off already! < The system is now operational in Baton Rouge, Louisiana. Santa Fe, New Mexico is scheduled next with a target date of June 1st. It would be 100% wireless. Santa Fe was picked because it has little foliage with few trees,> Why do you think this was? I can tell you. It is because trees and foliage absorb RF energy, meaning the signal cannot be transmitted as far. The current network topology may work fine in Santa Fe, but I guarantee that the same topology has trouble in a place with a lot of foliage. You would need more repeater radios closer together. < By the way, MEME holds a lot of spectrum in different markets.> OK.now here is where I am lost. To say they HOLD spectrum implies that they OWN it. If so, how much did they pay for the spectrum and what spectrum do they own. (BTW.Metricom owns a lot of spectrum also) If they are using the Part 15 bands (as Metricom is also) they do not OWN it. They are shared bands. And, if this is the case, will be sharing the band with us if we deploy in the same areas. Do they own spectrum or not? Why does nobody have an answer to this? This is a VERY VERY important question! < Weather will not be a significant factor. Under normal weather conditions rain and storms will not affect connectivity> ABSOULUTE NONSENSE! Rain and snow will affect connectivity! People, do some homework! Even right here it say connectivity WILL BE AFFECTED. Here----- < IF there is a momentary halt in data flow there is no re-dialing. It just temporarily stalls then automatically re-connects itself.> This does not happen in ZERO time. Every packet that has to be retransmitted will slow the system down DRAMATICALLY. < As far as I know, Ricochet (speeds only as high as 19,200) and WaveRider (which is evidentally not ready for delivery till the 3rd Qtr.) are the only existing competitors.> FALSE--- I get connectivity at the rate of 35 to 40 kbps from inside my apartment using Ricochet. and Waverider will never have a viable system until they hire a few engineers that understand RF. Then they will find out that physics prevents them from doing what they claim they will and they will go away. < With their proven technology a fact,> Where have they proven it? In a hotel room? And why did they use a ISDN line? They are clai9ming 10 Mbps! Why did they not use at least a T1? I am sorry people, but sending data across a hotel room does not impress me and more. Send it wirelessly from CA to Wash DC in a few miliseconds, and then you have something! I am looking forward to the demo in San Francisco. Rest assured that I will be there, with my modem, -since we have SF well covered, and I will be asking a lot of questions! This will be fun! Sorry, for the length, but hopefully, this will cause a few people to do thier TECHNICAL DD! Fell free to forward any quiestions regarding the MCOM Ricochet system to me, and stop by the Metricom home page, it may be a good source of questions for MEME! take care david david@metricom.com