Topics covered include:

• Introduction to issues
  
• Overview of Wireless Andrew
  
• Types of buildings and wireless technology
  
• Bandwidth and wireless technology
  
• Relationship between wired and wireless networks
  
• A "wired" wireless classroom
  
• The scope of Wireless Andrew
  
• Factors for determining access points
  
• Frequency of transmission and interference
  
• Off-campus access to wireless networks
  
• Facilitating student access
  
• Hand-held Andrew
  
• Security and wireless networks
  
• Wireless networks and network congestion
  
• Coexistense of wireless and switched networks
  
• Wireless at home
  
• The Lucent partnership
  
• Covering open spaces
  
• Status of Hand-held Andrew Project
  
• Small wireless projects
  
• Future wireless network speeds
  
• Keeping cards and software up-to-date
  
• Success for the Wireless Andrew Project
  

[Top of Page]

JB: Welcome to the CREN TechTalk series for the fall of 1999 and to this session on Wireless Networking Directions. You are here because it's time. This is Judith Boettcher, your CREN host for today, and I invite your institution to help support these TechTalks if you not already a member.

I am pleased to welcome the technology anchor for TechTalk, Howard Strauss of Princeton. Howard is a well-known Web and all-around information technology expert. Welcome, Howard.

HS: Thank you, Judith.

I'm Howard Strauss, the technology anchor for the TechTalk series of technology webcasts. As technology anchor, my job is to engage our guest experts -- oops, our guest expert today -- in a lively technical dialogue that will answer the questions you'd like answered, and to ask those very important follow-up questions. You can ask our guest expert, Chuck Bartell, your own questions by sending e-mail to expert@cren.net anytime during this webcast. If we don't get to your question during the webcast, we'll provide an answer in the webcast archives.

Does technology really change the way we do things? Not always. Do you replace your cordless phone on its cradle after every use and have to run across the room to get it when it rings? Does your microwave oven have features on it that you've never used, even though you bought it because it had those features? Do you still know how to set the radio stations with the electronic buttons on your car radio? If they become un-set, would you remember what stations they were set to? Is your electronic address book just as disorganized as the paper version you probably still keep handy?

Not every new technology has the effect its advocates expected. Carnegie-Mellon University is in the midst of a bold experiment to understand the effect of introducing wireless technology in all academic and administrative buildings and their adjacent areas. They are installing about 400 wireless access points that are the base stations for wireless devices.

What exactly does CMU expect their students, faculty and staff to do with this futuristic network called Wireless Andrew? Well, they have some ideas, but they're not exactly sure. It is a daring, Field-of-Dreams experiment in which they will build the network and hope great things will happen.

Many colleges and universities have wireless initiatives under way, but this one will saturate a large portion of the campus with seamless wireless access. Students will be able to wander about from building to building and across the open spaces of the campus and remain network-connected. Early portable computers were just smaller desktop machines with handles attached to them. In fact, they were often accurately referred to as "luggable computers." Laptops can be more readily carried about, but their need to be plugged into a hard-wired network connection has made them serially portable; easy to take from place to place, but not totally usable in transit.

Wireless Andrew seems to let laptops be the mobile devices they were really meant to be and Hand-Held Andrew, a concurrent project at CMU, should do the same for palmtop computers and other small devices. Will this technology change the way courses are taught, labs are run and meetings are conducted? Will CMU really be a different place once it goes wireless, or is this just another novel technology?

James Boswell said of a dog walking on its hind legs, "It is not done well, but you are surprised to find it done at all." Are we doing this because we are surprised that it can be done at all? Although getting the infrastructure for wireless networking in place is a major undertaking, as Chuck Bartell will be quick to tell you, perhaps the biggest challenge is yet to come: to make sure that this technology is used in ways that do make a real difference.

We've been keeping a close watch on Wireless Andrew, and so should you. It just might turn out to be the wireless model that all of us should follow. Today, Chuck Bartell will give us the very latest news on it from the trenches at CMU. If you already have a wireless connection, you might want to take your laptop outside to enjoy today's edition of TechTalk!

Judith?

JB: Well, thank you, Howard, and I'm just really excited about our topic today on wireless. And just let me make one comment about what you're saying about how this technology makes a difference in our lives. I think that, while the technology keeps coming along, we keep doing things the same way -- not because we necessarily have to but because we just do it that way out of habit. So I think as we listen to Chuck talk about what they're doing at CMU, thinking about how people will be changing their habits for dealing with technology I think will maybe give us a clue as to what we're going to see.

So with that, let me tell you a little more about our guest expert for today's TechTalk. Our guest is Charles Bartell, known generally as "Chuck". Chuck is the Director of Operations for the Division of Computing Services at Carnegie-Mellon. He is with us today in his role with the Telecommunications Projects at CMU, including the ADSL trial and both phases of the Wireless Andrew project.

Welcome, Chuck, and thanks for being here.

CB: Well, thank you for inviting me to participate in the TechTalk series.

Topics

HS: Chuck, the last time we talked about -- I guess we didn't talk to you, but the last time we discussed this topic, Wireless Andrew was not in place yet. The students and faculty and staff were not able to use it. As I understand it, it is in place now and I wonder if you could just give us kind of an overview, for folks who missed this last time.

CB: Sure, I'd be happy to. In terms of Wireless Andrew, what we have essentially done is taken a look at our campus and we have decided to deploy the wireless technology on a par with how we have deployed wired technology in the past, so that we in general want to make wireless network access as available in a building as you could normally have a wired network connection. So we have taken a look at the bulk of our campus -- our primary academic, research and administrative buildings -- and have started to install the wireless technology.

Now, this phase of Wireless Andrew is actually what we call Wireless Andrew Phase II. We had done a first series of Wireless Andrew with non-standardized type of technology from Lucent Technology -- 900 megahertz wireless about four to five years ago, and we had deployed that in five buildings, primarily as a research network in support of a number of our researchers in mobile computing efforts and robotics. And so this is now really just an opportunity to refresh that technology -- bring it up to standard space and more fully deploy it across the campus, not just leaving it for researchers but also making it available to all the campus faculty, staff and students.

Topics

HS: Okay. Actually, normally we have to wait until we get about halfway through here for the questions to start coming in, but we already have a couple questions here. We have one here from Alexandra Kim, and Alexandra Kim is the Director of Staff Technology.

She has two questions, I think, that we could start with here. One is, she wonders for what type of buildings is this technology most useful: academic buildings, administrative buildings, residence halls? I believe at CMU, you haven't done residence halls. Is that correct?

CB: Yes. At this point, the grant of the equipment that we have received is from Lucent Technologies, and from our own estimates of what it will take to do wireless, we have decided at this current phase not to do the residence halls. The residence halls are pretty much wired already on what we call a "network connection per pillow basis," so we don't think there's any issue in terms of capacity or need. We would like to look at residences as maybe a future project, but we're not making any commitments to that right now.

In terms of the styles of buildings that this is useful for, I would say it's useful for all. One of the first buildings -- first two buildings that we opted to install the original wireless network in was actually our own Computing Services building (which is primarily an administrative type building), and then another building, Hamburg Hall, which is a combination of some administration, a number of classrooms and a number of different academic departments and some research departments as well. And what we have found is that it's extremely useful to just about everyone that has a need or finds a use for having an untethered connection to the campus LAN.

HS: Okay, Alexandra's second question -- which perhaps you touched on -- was, she said, "If a variety of different types of buildings were using this technology, would they all use the same user support structure?"

CB: If I understand the question correctly, in terms of the end-user itself support structure, yes.

For the most part, our wireless LAN -- the connections we are doing is just like a standard Ethernet-style connection, so any of the control panel changes, any of the IP-related software, anything that you need to run over there, you support in a similar manner.

In terms of configurations of devices, there are some specifics to wireless -- NIC cards, network type cards vs. your standard wired cards -- but that's all the same. In terms of user support as far as design goes, that might be somewhat different and we have really found design not so much to be specific to a particular function, but more actually in general to the construction style or the building materials used in a particular building.

Topics

HS: Okay. We have another question from Phil Long at Seaton Hall University, and his question really is about "throughput" or, I think, bandwidths on a wireless network. He points out that teaching and learning places extraordinarily high demands on network throughput, and he wonders what the implications of the current wireless systems are. So maybe you could talk to use just a little bit about bandwidth here.

CB: Okay. In terms of bandwidth -- again, when you think about the continuum of connectivity that's available to folks, if you go all the way to what we sort of consider the low extremes of dial-in networking, I guess V.90 or roughly 53 kilobits, 56 kilobit is probably the fastest you can go.

HS: Those are the standard modems that people now have at home?

CB: Yeah, the standard modems that you'll connect to. And then there are some faster services, cable television type modems or DSL types of technologies that are getting you more into the high-kilobit, megabit type of connection speeds. And then you go into campus LANs, which are anywhere from sort of your standard ten megabit shared up to 100 megabit. And people are even looking at some of the gigabit to the desktop type scenarios.

Where wireless fits in currently, with the current state of the art, is in the one-to-two megabit per second shared media. So this is not a switched connection, this is not a point-to-point -- this is essentially like a ten megabit shared ten base T Ethernet.

However, it is lower in speed. It is one to two megabits shared. So in terms of its use, we see that wireless is not necessarily a replacement for a wired LAN -- in particular if you have some high bandwidth applications -- but it's really a complement.

It's in that continuum of connectivity. It's the ability to do some things untethered, and what you really do is you have to match your applications to the bandwidth that's available.

HS: So people aren't going to be connecting to Internet2 from the wireless laptop?

CB: Yeah, I wouldn't expect to try to run supercomputer computations or basically simulations over a wireless link.

Topics

JB: You mentioned that in our prep talk, too, Chuck, about the idea that the wireless networks are really best when -- that really they depend on solid wired networks in order to work.

CB: Yes, and in fact, one of the jokes that we say here at Carnegie-Mellon is the first four letters in wireless spell "wire" -- and in fact, that's one of the things you need to do when you're installing a wireless network is install both power and data to a number of locations.

But in particular, we basically at Carnegie-Mellon have a number of different initiatives to get networking into the classroom. We have our traditional computing labs, which were initially available as sort of public access for students to use. What we have seen over the years is that those labs are turning more into teaching arenas or teaching theaters during the day, and then are being used by students at night.

We have done another project that we have nicknamed Netbar. Originally Netbar was the idea of having docking locations of switched ten base T5 connections for students outside of computing labs that wanted to just dock up with their laptops so they could check e-mail very quickly or print out a document if they needed to. We've extended Netbar now into certain classrooms where there's a requirement of the faculty or the course to basically allow students to connect up laptops in larger teaching theaters. And Netbar is eventually a switched, authenticated type of network.

So what we're seeing is that wireless helps out in some of those arenas -- in particular if we have a classroom that is not covered by Netbar, and if it's in sort of a lower, 20-seat type of classroom size, wireless will probably work for some of the impromptu use of laptops or the need of computing in a class. However, if you're really doing some large computational types of work or something that's going to be fairly bandwidth intensive, you're probably better off going to a classroom or scheduling a computer lab.

Topics

HS: When the student using Wireless Andrew now walks into a classroom in a classroom building that's wired -- it sounds funny to say "wired" for wireless.

CB: I know!

HS: But I guess that's the case. What does a classroom look like? What's different about the classroom?

CB: In terms of wired or wireless?

HS: Well, I bring my laptop in. Do I have a place to plug it in at my desk?

CB: Okay, in the case of a wireless type environment, you see nothing different than you would in any other classroom. We have not run power, we have not run dedicated wiring to the particular seats. In the case of what we would call the Netbar-enabled type classrooms, we have run some level of power and some level of Category 5 cabling. But in terms of wireless, it really is the case that the room is covered by one or multiple of the access points -- the devices that actually connect the wireless network to the wired world.

HS: So I come in with my laptop and if my batteries hold out, I can do whatever I want to do in there?

CB: Absolutely.

HS: I have complete network access. I can print on any printer. It looks just like I was wired into your network.

CB: Absolutely, and an interesting piece is (and I think I related this at a previous time) that I recently attended the Internet2 conference out in Seattle and they had done a similar wireless network in the hotel in terms of the conference area. And I was actually able to take my wireless card, which I do now normally carry with me, plug it into my laptop and, with a few minor configuration changes, actually be able to access all my e-mail, my centralized calendar, web browsing -- everything I needed to do from Seattle back to Carnegie-Mellon in Pittsburgh.

JB: And this was just in -- which part of the hotel conference area was that set up?

CB: Yeah, mainly the main ballroom and some of the associated meeting rooms.

Topics

JB: Okay. Interesting. Chuck, it might be a good time just to step back for a moment and, you know, try and ask a question about just how big a project -- what is the scope of Wireless Andrew? I mean, are you really covering the whole campus?

CB: Yes. In essence, our campus is a fairly compact campus in terms of -- if people are not familiar with it, it's around 103 acres. We refer to it as roughly a one-kilometer radius as far as the campus proper goes. We have about 50-plus buildings, about half of those are resident halls and that type of associated buildings. But the other half are academic, research and administrative buildings. We are focusing on roughly 25 of the academic, administrative and research buildings to start with, which covers almost, I think, a bit over 2,000,000,000 square feet.

HS: That's about what percentage of the campus? Is that half the campus?

CB: That is going to end up being more like 60 to 70 percent of the campus, when you actually count the actual square footage of buildings and the footprint of buildings.

HS: Okay. And the reason it's not the whole thing is because you're not going to do the residential halls?

CB: Yes, not at this point.

JB: And your goal for the project, again, the scope is to get those 25 buildings -- when are you going to have those all done?

CB: Okay, we essentially now have roughly -- I believe it is now seven or eight of our 25 buildings completely set up. We will be bringing on the remaining -- I guess it would be seven, up to a total of 15 -- by the end of this calendar year. And then we were going to work on the remaining ten buildings into our spring semester.

JB: Okay. So your goal really is to get them all up and finished by June of the year 2000 for the 25 buildings?

CB: Yes.

Topics

HS: Chuck, I was surprised to hear that you had 400 access points in those 25 buildings. Could you tell us a little bit about why it takes so many?

CB: Well, it varies for a number of different reasons. (And excuse me, I have to turn off my cell phone that's ringing.)

JB: Aha! Technology comes again.

CB: Yes, yes, it does.

HS: It's a wireless phone.

CB: Yes, a wireless phone that I'm shutting off so I can talk about wireless data.

Yes, in essence it's a number of different factors that come into play: the size of the building, whether it's a multi-floor building, actually the construction techniques that were used in the building. That all varies in terms of what's going on.

While we have primarily decided to do designs based upon maximizing coverage, we have not completely ignored capacity either. So as we were doing our designs, if there is an area that we believe would benefit by the overlaying of multiple access points, we will try to plan that into the design.

So for example, if we have a large lecture hall in the 200-plus seat range, we will try to serve that from more than one access point if it's at all possible. So we take those kinds of things into account.

Topics

HS: Okay, we have a question from Rich Weaver just down the road from you at Duquesne University.

CB: Okay.

HS: And Rich asks, he says, "Why is the effect of transmission frequency a concern for certain areas such as chemistry, biology or physics departments? Would the frequency of the transmission affect any ongoing research projects in those areas?"

JB: Interesting question.

CB: Not that we know of. Currently the product we are using is IEEE spec, it is FCC licensed. It is really in what they call the ISM band -- Industrial, Scientific and Medical, no license required. So there is some low amount of radio transmission that is required. All of this stuff seems to be in safe levels. In terms of frequencies, we have not run into problems to date with any of the labs where we have installed the equipment. But that is still to be determined.

HS: It would seem strange if any of that stuff would be a problem, especially since folks like you and everybody else are walking around with cell phones and people are walking around with GPS receivers. And there's so many gadgets, actually, not to mention the microwave transmission of telephone calls and TV signals and there's all kinds of stuff going through the air.

CB: That's absolutely correct. So we believe that because this is an FCC unlicensed type of product that its potential interference with other devices is going to be minimal. And the interference with other devices as well can be sometimes problematic, but we try to take as much of that into account as we're doing our designs.

Topics

HS: What about students who are going home or going somewhere else or going off campus? I mean, your campus is compact and you don't have any branch campuses and things like that. But what do people do about being home? Can they get into this wireless thing?

CB: Well, they can get into the campus network, but they cannot necessarily get into wireless because again, we're trying to keep the wireless coverage to the campus proper.

Now, that is almost impossible because one of the things that we are finding is that we are getting an awful lot of between-building coverage just from the installation of access points in the buildings themselves. So there will be a certain amount of leakage around buildings. If you happen to live close enough to our campus that you can actually get covered that way and you are a member of our campus community, then you could access the campus LAN that way.

HS: How far away from an access point can you be and still use it?

CB: It varies. In terms of normal design, they are recommending that the coverage is in roughly the 30 foot range. Again, your mileage will vary based upon --

HS: Thirty feet?

CB: Thirty feet within a building.

HS: You've got to be that close to it?

CB: Well, I'm saying that is sort of the manufacturer recommendation. We have seen much better coverage than that, and again, it all depends on a number of different propagation issues.

Some buildings -- newer style construction buildings -- the radios penetrate walls much better than some of our close-to-100-year-old buildings. So your mileage will vary! We really don't worry so much about the actual coverage of an access point itself. We worry about actually the blending of multiple access points coverage to provide a seamless level of coverage for the building itself.

Topics

HS: Okay, we have more questions coming in here. This is generating lots of questions, which is great! Folks, keep sending those questions in!

This is a question from Mark Cather at UMBC and he has a couple basic questions. One, he's asking about what are you doing to help students purchase network cards or computers? And I guess a related question which he didn't ask but we should ask is whether you require students to have computers at all.

CB: Okay, in terms of the requirement for student computers, it varies depending upon the school. For all undergraduates on our campus, there is no requirement to buy a computer. That said, we are finding that many of our students are coming to campus, either buying a computer prior to coming to campus or buying a computer shortly after being on campus. So we are seeing in the high eighties and into the low nineties as far as percentage of incoming freshmen coming and getting those computers.

We do have a business school, our GSIA is the name of our Graduate Studies in Industrial Administration, and that school has required over the last three years their incoming Masters' degree students to buy a computer, and in fact, they buy laptops.

JB: And don't you have figures, also, on just how many of the students overall have laptops as opposed to desktops?

CB: Yeah, we have. We've determined that roughly about 20% of our students coming in now -- the incoming freshmen -- are buying laptops or coming in with laptops. And we believe that as Wireless Andrew becomes more and more of a factor on campus, we're going to see that percentage increase.

As far as the question of what we are doing, we're really not doing anything in terms of subsidizing the cost of the cards, but we have been tracking the cost of the cards ever since we began this project roughly four to five years ago. And we have seen the cards go down from an original high of around $700 to what is now almost a $260 price point.

JB: What is the price point now?

CB: The current price point is around $260, available through our campus computer store. The other interesting piece -- if anyone is following what's going on with Apple and the IBook introduction, Apple has announced a product called AirPort which essentially is a compatible wireless card, and that is priced in the $99 range.

JB: Yes, in fact, I'm at a conference this weekend and they announced it was $89, I think, for the card for the IBook.

CB: Right. So we believe that wireless is now becoming enough of a commodity and will continue to become a commodity that those price points are going to continue to push down.

Topics

JB: Well, you know, this links into a question we have from Richard Danielson who's asking about the cheapest. Let me quote from his question, saying, "What is the cheapest, lightest, most durable receiver unit currently available for wireless educational applications?" Even going down into palm-sized machines running on AA batteries. Is there some comment you might make on that?

CB: Okay, well, actually this sort of segues a little bit into something you mentioned early, being our Hand-held Andrew project, and that is really a project that is running in parallel with Wireless Andrew.

And the concept of Hand-held Andrew is to look at these smaller form factor computers -- the palmtops, for example, or the much smaller versions of laptops, things that are running the CE style operating systems. And these are devices that also can take advantage -- as long as they have a PCMCIA card slot in them and currently if they're running Windows CE, they are also compatible with the Wireless Andrew network. So these are devices that are now getting into the 800 and lower price point cost, and these are devices that will allow you to do full-fledged web browsing, e-mail -- they can be part of our centralized calendar system that we're beginning to deploy on campus. Overall, they're a very useful device.

JB: You might want to talk a little bit about that calendaring application, Chuck, because a lot of folks wonder what is wireless particularly good for. Is there a killer app, for example, for the wireless?

CB: Right, and in fact, we don't know yet if there is the killer app. You had talked a little bit earlier about the idea of the "field of dreams," and to some extent, that is what Wireless Andrew is. We have our own notions about how people will use it, but this campus is very aggressive and surprises us a great deal with some of the new and innovative ways that people use the resources or the functionality that we put out on campus.

In terms of the calendaring, one of the things that we have done is we have partnered with a company called CS&T.; Their product is called Corporate Time. And we are looking at making a centralized calendaring system available across the entire campus so that you have the capability of coming to centralized calendar servers. You can then either basically upload and download your calendar, synchronize your calendar as you need to with the server, whether it be a laptop or whether it be a hand-held over the wireless LAN. In fact, you could be anywhere on campus and either sync up with your desktop machine if that's the way you need to do it, or else sync up with the centralized calendar server.

JB: That sounds exciting. We've got a lot of questions coming in here. Howard, did you have one that's ready?

Topics

HS: Yeah, there's one more. Mark Cather has another question. He actually asked two. We only got to one of his, but his other one is something we were going to talk about, so let me take the question from him.

His question was about security. He had a pretty specific question about security, but I think we could get a more general answer. But his question was how do you secure your wireless network from student-sniffing traffic? And he also asked how easily can connections be hijacked? In general, I guess the question is does all this wireless stuff make the network less secure and what precautions do you have to take?

CB: And I would say -- I will not say that it necessarily makes it less secure, but it actually makes it easier to get to, and I guess that's the fundamental difference.

One of our network architecture philosophies has been all along with the campus net is that for the most part, you don't count on the campus net for security per se. You build the security into the applications. So in fact, you should use some level of encrypted telnet or you should -- we use Kerberos in terms of the way that we pass the tokens around for access to our Andrew file systems or the like. So we try to make as many Kerberized things -- functions -- as we can.

There again, the idea is if the information is important enough to you, you in some way should encrypt it and that should be done at the application level.

That said, we are looking at ways to try to make our wireless network relatively as secure as we can. There are some industry initiatives. One thing that Lucent offers is something which is called WEP, which is Wired Equivalent Privacy. That is a fairly recent enhancement they have offered to their product. We are looking at how we would want to deploy that on campus, but also using similar things such as Kerberos or Radius authentication and services are things that we're looking at, not unlike what we do for our own dial-in modem pools.

Topics

HS: Okay, as Judith said, we have questions just pouring in. This wireless thing seems to have generated lots and lots of questions, which is great. Keep those questions coming.

We have a question from David Brown, and David says, "When we tried to do wireless in the classroom, all students needed to do the same thing at the same time and this meant the system was very slow. Isn't this a problem for you?" He also asks -- no one asks one question, Chuck --

CB: Doesn't cost any more to send the bits, right?

HS: Right, it doesn't take much more to send a few more characters. He also asks, "Can you give some examples of how professors are successfully using wireless?"

CB: Okay, to answer the second question first, I really can't give any specific examples because again, the production deployment is still fairly new. We began this October 1 as far as production goes. We were doing our designs in buildings and leaving the designs up as functional as we could throughout the summer and into part of the fall. But the campus, as they've been aware of it, are only really now starting to take a more active and aggressive use for Wireless Andrew.

In terms of the first question, and I'm trying to remember it again. I'm sorry.

JB: It's all students trying to do something at the same time.

CB: Yes.

HS: Doesn't that just clog the whole network because you have everybody going out to the same spot at the same time? Probably everybody's hitting the return key within a fraction of a second of each other.

CB: Yes, and in fact, we had done some testing very early on with the Lucent product. We had done sort of what we called the classic test: everyone hit the enter key at the same time. We went to one of our larger classrooms, set it up with a single access point, and we had done some wired to wireless comparisons using FTP and ten megabit files and then also doing some other what I would call "anecdotal comparisons."

And what we found is that when you start going beyond 20 simultaneous FTPs, you really start seeing a major performance hit. But that's not unlike what you would see on Ethernet if you were coming up with a larger number of machines. (Again, this is a shared Ethernet segment. It is lower in speed, two megabit vs. ten megabit, but it's going to have the same kinds of characteristics that you see in any shared network.)

Now, that's sort of the bad news. The good news is that we also ran the same kinds of tests in a slight variation, and that was we had a number of users do some orchestrated, what I would say heterogeneous types of testing. So we had some people reading e-mail, some people web browsing, some people doing a web download and some people doing the standard FTP. And what we found there is that better than 30 simultaneous users were able to do all of that mixture of types of downloads and use without any major performance hit.

So again, what we try to tell the faculty that have intents of using Wireless Andrew is, it is a complement to the network, it is not a replacement. If you believe your class application is going to be one that's going to require a lot of simultaneous downloads or is going to require an awful lot of bandwidth-intensive work, wireless is probably not the right media to use and you may be better off going to a Netbar enabled classroom or into a computer lab.

JB: Do you want to mention (and I hope I didn't miss this, Chuck) about the next wave in terms of the speed that's coming down the pike?

CB: Yes, actually I believe it was somewhere in mid to late September. The IEEE had ratified what will be the next variation on the 802.11 specification, and that is actually getting into 11 megabit per second as the upper bandwidth for a wireless network.

HS: Okay, and that's about what folks right now have at their desktops.

CB: Yes, some people either have ten shared or ten switched. If you have ten shared, then the wireless, 11 meg wireless is going to be roughly the same speed.

Topics

HS: Okay, we have two questions from Ed Caray at UIC. His first question is can the wireless setup on a personal computer coexist with a switched network setup so that you could easily change to high bandwidth networking in the office then to wireless networking, without doing complicated configuration changes and things?

CB: I would say that depends entirely on how your network is actually architected. For the most part, the way we typically do things on campus is you make a choice. You do one or the other, and there may be some rebooting that's required. I'd be a little concerned if you may be getting into some level of routing loops or how your DHCP service may handle those kinds of changes. I guess in theory, depending upon your operating system, you could make those switches fairly seamlessly. I think it really depends on your device and some more specifics to how your network's architected.

Topics

HS: Okay, Ed's other question is, do the higher-end solutions that you have investigated have a compatible, less-expensive solution for home use? I'd like to seamlessly use my wireless connected laptop on campus or at home, again without having to make any configuration changes.

CB: Okay, actually again, going back to Apple, as part of the IBook announcement, Apple has also announced something that they're calling the AirPort base station. And in essence, while an access point which is a full-fledged device that we use on campus -- one of these 400 access points we described -- is in roughly around the $1,000 price point, the AirPort base station Apple is talking about is a device more in the, I believe, $300 price point. And it has a ten base T connection -- has, I believe, a V.90 modem built in. And that is probably something more along the lines of a device that you would want to use.

Another possibility is there are some product manufacturers out there -- I believe Zoom is one as well -- that basically sells the IEEE compliant cards, but they also sell software that enables you to put a card into your home PC and, with the software, run it as a small access point.

JB: And that computer, then, can multi-task in doing that?

CB: Right. It can essentially multi-task and it can go out through a modem port or a DSL port and connect back up to the main campus, but also then act as more or less the wireless bridge or access point for the laptop or the other devices in the home that want to be connected wireless.

The person that sent the e-mail actually brings up a fairly good point. As we're seeing more and more homes becoming more computing-intensive, and as more devices are becoming more network clients (whether they be printers or each of your children wants to have their own computer or whatever that might be), a number of homes are really not wired to take advantage of that. Not too many people have wired their homes with Category 5 cabling, so wireless may be an interesting alternative for homes that want to have high speed access.

As we think about DSL being a technology that's becoming more deployed in homes for that continuous high speed access to the home, having wireless as a reasonably cost effective distribution mechanism for the home computing environment is becoming more and more attractive.

JB: It certainly sounds like a great solution.

CB: Plus you can now sit out on your deck and commune with nature and still stay connected for stock quotes or whatever else is --

JB: They just have to fix the screen a little bit.

CB: That's right.

HS: If only the screen lets you do that! You have to have a very shady porch.

CB: Yes.

JB: We might want to just say something. I know that some of the folks out there listening have had some problems. I just got a note from Julia saying that the server hung up for a bit and was restarted so I'm hoping that everyone who's out there is in fact doing well. So if you're not, if you want to let us know, that can help us isolate some problems for the next go-round.

HS: And of course, everything, whether you hear it or not, will be in the archive.

JB: That's right.

HS: We have another question from Alexandra Kim, actually two questions again (as I said, nobody will ask one question here, which is fine). Alexandra wonders how many students have elected to use Wireless Andrew and purchase the network cards.

CB: Well, we don't have the detail as far as the number of students that have done that. The statistics I have gotten through our computer store indicate that in the less than a month that we have, approximately 400 cards have been purchased between faculty, staff and students. So again, I can't really point to a specific number of students that have purchased this, but 400 in about three weeks is a pretty nice number as far as folks wanting to jump onto the network.

JB: And Chuck, just so folks know the scope of your community, you've got a total of what in your total community?

CB: We have roughly, I think -- I'm looking now at our fact sheet, just so I have the numbers correct. We average about 4,900 undergraduates and about 2,800 graduate students, 1,000 full time faculty and I think we are probably in the 2,000 plus staff range as well. Almost a 10,000 member community.

Topics

HS: And Alexandra's other question is she understands that you have some kind of partnership with Lucent in this project. Could you tell us a bit about it?

CB: Yes, actually we are very proud of the partnership that we have with Lucent. It really started back, as I said, almost four to five years ago with the original Wireless Andrew project -- the Phase I.

We're in essence -- we have worked with Lucent and they have come to appreciate how far we are actually pushing the envelope in deploying wireless networking to the degree that we have. For the most part, heretofore, wireless was really viewed as either a cable replacement type of technology, or a way of getting more cost-effective point-to-point link between buildings.

What we have done is actually shown that it can be a complementary network; it can be a very large deployed network; that you can take advantage of some of the ideas of roaming between access points to have that seamless coverage; and some of the design criteria and the ways that you do these kinds of deployments that we're doing on our campus vs. the other ways of doing things.

Lucent found a great deal of value in working with Carnegie-Mellon in how we were doing these kinds of designs -- so much so that when we asked them and approached them about what the future was holding in terms of the IEEE 802.11 spec, they had offered to grant us 400 access points towards the idea of doing the campus comprehensively.

JB: And you mentioned you're not certain at this point whether those 400 access points are going to enable you to really do the reach that you want in all of your 25 buildings?

CB: I think we will pretty much be able to do the reach we want in the 25 buildings. Again, what we also want to do is at some point look at what we might be able to do in the residence halls. That, again, is not part of the scope of this project but it's always something that we're interested in doing.

JB: So it might be the scope next year or the year after that, then?

CB: Yes.

Topics

JB: Okay. Do you want to talk at all or describe any of your open spaces on campus that you have covered now with the Wireless Andrew?

CB: Okay, well, actually what we have done is we have looked -- for folks not very familiar with the campus, we have some major large green spaces in sort of an L-type shape. And in fact, what we have found through some very recent testing is that one area of that L, which we sort of refer to as the Mall, is actually getting enough coverage from bleed-through from the buildings around the perimeter of the Mall that we may not have to do anything special as far as that outdoor coverage goes.

So what we are going to do is continue to deploy within buildings around campus, do outdoor measurements, supplement with outdoor antenna type coverage as needed and then, you know, beam some wireless coverage into areas that are not covered by perimeter buildings. So for example, you know, we'll look at doing our intramural fields on our football field and things like that.

HS: So the football players are going to carry laptops. Or they could make due with hand-held.

CB: Or the other way of thinking about it is, you know, there is the concept of voice-over IP and some of the things you can do over wireless technologies. Maybe we'll be doing what the NFL does with radios in the helmets. Who knows?

Topics

HS: Could you talk a little bit more about wireless, the hand-held wireless, or I guess it's Hand-held Andrew project? How far along is that? Are people starting to use that? What kind of devices?

CB: Okay, currently we are using -- we haven't really standardized on a particular device at this point. What we are saying is that if your device supports PCMCIA slot (which is required right now because that is what the Lucent card slot, the Lucent card particularly is), and if it is supporting Windows CE (because that is what Lucent has been running the drivers for at this point) -- if you can run Windows CE and it has a PC card slot, then you can run on Wireless Andrew.

What Wireless Andrew is is a combination of -- I'm sorry, what Hand-held Andrew is is a combination of making this Wireless Andrew infrastructure available, and then doing whatever pieces we can to work with corporate partners to make the hand-held device as much of a full peer on the Andrew computing environment as we can.

And let me describe what I mean by "full peer." So, for example, we all know that if we use a word processing package like Microsoft Word, there's probably about ten percent of the features that we use on any regular basis.

JB: Um-hum.

CB: And on occasions, we may dig through the manual or the Help function and find that one or two extra type functions. But there's an awful lot of functionality that's been built in that's specialized for some folks to really use, but in general, you really don't need it for your basic word processing or spreadsheets or whatever that might be.

What we have in mind is -- and the current problem with the current hand-held operating stuff is that, although you can run Microsoft Word, as an example, on a Windows CE platform and it has some of those limitations in functionality, being able to print your document is a little bit more difficult. You typically have to, you know, dock that to your desktop or your laptop (if it's acting as your desktop) and then do your printing out your desktop or laptop port. You really can't print from the hand-held device.

What we want to try to do is work with as many corporate partners as we can in trying to get the hand-held to be as much of a full peer on the network as possible, even if we realize that some of the applications may be somewhat stripped down. We still want you to be able to get network connectivity, be able to print if that's what you want to do. As I mentioned, in terms of the calendar server, we want to make it so that you can link up directly over the wireless connection and get your calendar updates directly from your hand-held to the server and not need to go through some intermediate piece.

JB: Let me just, since we're getting awfully close to the end here, let me just remind folks that this would be a last chance perhaps to get in a question for Chuck.

Topics

And Chuck, maybe I'll just ask a question, following up here quickly. We've talk a lot about this huge, you know, your comprehensive test that you're doing at Carnegie-Mellon. What if someone wanted to get started with a small project? What's the smallest size project that people might get started with?

CB: Well, I think what you really need to do is you really need to look first at what is it you would like to accomplish, and how you believe wireless may be able to help you accomplish that goal. You really do need to understand the application.

As I said before, some of the really bandwidth-intensive applications, wireless may not be the right solution with the current state of the art. But what I would advise people to do is to get a couple of access points, a handful of cards and just begin doing some radio coverage testing in the areas they would like to do. Find out, you know, start learning a little bit more about the technology.

This is a technology that we believed five years ago was eventually going to become a commodity and was going to be something that was very pervasive and around. And while we're going to be able to do that on our campus, another three or four years from now you may be able to go to your AirPort and be able to connect up over a wireless LAN, the card that you carry with you. You may actually find that wireless in your home is the way that you prefer to do computing.

So I think the idea is this technology is not as far-fetched as you might think it is. Now is a good time if you have any inkling in starting to get into wireless to get some devices, start becoming familiar with it. What are the similarities and what are the differences between wired computing?

JB: Well, it sounds like it's affordable now -- that people can do a small project and it doesn't break the bank.

CB: Yes.

JB: Howard, I think we have a couple other questions. Do you want to tell us those?

HS: Yeah. We have a question from Sahid Rahimi and one of the questions that Said asks is who makes the network cards? I think you said that was Lucent was making them.

CB: Um-hum.

HS: And also, Sahid asks what the cost is, and I think you said the students were buying them for about $260.

CB: Yes.

Topics

HS: But a question that you didn't answer that Sahid asks is, is there any plan to make cards higher, she says, than two megabits per second? But you talked about 11 megabits. Is there any plan to make them even higher than 11 megabits? Is that on the horizon?

CB: At one time (and again, this is many years ago) I had heard people talk about something they referred to as "wireless ATM" trying to get into the 25 megabit per second range. Now, when I say wireless ATM, it doesn't necessarily mean it would be doing cells like a standard ATM network does, but I think there's always going to be a Holy Grail of something higher.

HS: By ATM, we mean Asynchronous Transfer Mode (not the little places where you get money).

CB: Yes, I'm sorry. Yes.

HS: (At your banks.)

CB: I think the point is, there's always going to be some level of Holy Grail.

You probably recall way back in the early days of networking, one megabit networks were sort of the Holy Grail and then we go to ten and a hundred and now gigabit. I think the same thing you'll see in wireless. People will continue to try to push the envelope, realizing that the need for speed is great and that as you want to make wireless more of a commonplace network, you are going to have to make it faster or at least be able to have more bandwidth within it so you can support more people or more connections.

Topics

HS: Okay, Chuck Chamberlain has a question and I know the answer to it. I'm going to let you answer it because it has kind of an interesting answer, and it is -- he's talking about what happens as we get changes to the drivers and functionality of things? How do people keep these cards and software up to date?

CB: Okay, that actually is not much different in wireless than it is in the wired world.

In essence, what we are going to do is now that this has become a production network, we are going to be working as part of our network engineering and actually basically follow whatever normal release processes that we need to do. As new revisions come up, we will basically install firmware and appropriate software on the access points and levels of cards that supports the network in a production mode right now.

We'll be looking in the future as to what the next releases are, do adequate beta testing on campus to verify they work, and then roll out those changes to the campus at appropriate windows, whether they be sort of our holiday breaks in December or whether they be at the end of the academic year in the May-June time frame.

But we're going to just roll this out as we do standard support. Not much different than what you do right now in terms of how you may be upgrading new software or new drivers or IOS if you have Cisco boxes or the like.

HS: I understood, Chuck, that there was something about being able to upgrade the firmware right on the cards?

CB: Yeah, there are some capabilities where you can, actually what they've called "flashing the cards," that you can actually make the changes if you have -- you can install the software on your own machine and basically take the card and reflash it for whatever the appropriate firmware is.

We've set up similar stations in our own help center so that if you aren't familiar with that or you want some help, we have experts that can actually take you through that process. And actually, we've set up a similar station in our computer store so that our computer store sells the current supported revision of firmware on the cards.

HS: So if you have an old card, you don't necessarily have to go out and buy a new one to get the latest version?

CB: No.

HS: Right, you just get the card flashed.

CB: Yes.

HS: No, I think that's really an interesting kind of thing because we're used to the idea that you have some piece of hardware and then you have to throw it away. You have a modem and the modem is now too slow. You've got to go out and buy the faster one, you don't take it off and change the firmware.

JB: Well, that's interesting. I got my cell phone upgraded by calling a number on a phone.

CB: Ah, that is interesting!

JB: Yes, and you know, they encouraged me to go into a center, but since I was at least three hours away from any of their centers, they told me, "Oh, well, you can also do it this way."

HS: Yeah, well, in fact there's dishwashers now where you can download the latest stuff into the chip in the dishwasher and it'll do new things that they've decided dishwashers ought to be able to do.

JB: Well, and the wireless --

HS: We practically see it everywhere.

CB: And at some point you'll have wireless in your home and you'll put the card in your dishwasher and it'll be able to do that stuff itself.

JB: Ah, but I wonder if it will unload the dishwasher!

CB: Never!

HS: In fact, it probably won't even tell you when it should be unloaded.

JB: Well, all right, we're coming close -- I think we're probably at the point where we need to kind of wrap up for today. Howard, any final question or comment?

Topics

HS: Yeah, I do have a final question here. And that is, Chuck, it looks like by the summer you're going to have your 25 buildings wired and Wireless Andrew will be well on the way. I wonder how you're going to know the project is successful. What are you using to measure success in this project?

CB: Well, again, we're really using our end-users to measure success. We will be able to gather some network statistics. We will be able to determine how much the network is being used, where the hot spots are, things like that. But it's really going to be the practical applications that our end users are going to perform and that's really going to be the gauge of how successful Wireless Andrew is.

HS: Okay, thank you. Judith.

JB: Okay, let me thank all of our web participants for being with us here today for this time with Chuck, and you are invited to also send some follow-up questions to expert@cren.net and Chuck will answer those individually.

Be sure to mark your calendars for October 28, just one week from today -- which is a little bit off our schedule, but we're doing that because we're doing a live audience at Educause. For those of you who will be at Educause, you can join us for the live audience of TechTalk, and for those of you who will not be at Educause, we will welcome you on the web. We're going to feature two experts, John Booker from Oberlin College and also Barbara O'Keefe from the University of Michigan. Barbara will focus on a project with instructional objects and the newly launched Media Union at Michigan, and John Booker will bring the latest on hot issues and what he calls "hallway conversations" at Educause -- kind of getting a sense of what's happening there. Howard is going to wear two hats for this session, serving both as anchor and as expert, bringing his insights on what's happening with web and content convergence. So check the website for this in more detail, and always, we welcome suggestions and feedback on what and whom you would like to see and hear on TechTalk.

Thanks to all who helped make this event possible today: to the Board of CREN; to our guest expert, Chuck Bartell; to Howard Strauss, our technology anchor; to Terry Calhoun, event page producer; to David Smith and Patty Gaul of CREN; to Julia O'Brien, Jason Russell, Carol Wadsworth and the whole support team at the Merit Network; to Susie Berneis, our audio file transcriber; to Laurel Erickson, our transcript editor and indexer; and finally, our thanks to all of you for a great interchange and great questions. See you next week!

HS: Thank you, Judith, and see you at Educause.

JB: Bye, Howard and bye, Chuck. Take care.

HS: Bye-bye.

JB: Bye-bye.

CB: Good-bye.

Topics

[Top of Page]