TechTalks Transcript
Wiring for a Wireless Campus
 Judith Boettcher [JB] |
 Howard Strauss [HS] |
 Ron Eaglin [RE] |
 Joel Hartman [JH] |
September 23, 1999
Audio
• Streaming
MP3
• Download
MP3 (Download
Tips)
JB: Welcome to the CREN TechTalk Series for Fall of 1999, and to this session for "Wiring for a Wireless Campus." You are here because it's time to discuss the core technologies in your future.
This is Judith Boettcher, your CREN host for today. 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. And welcome back to another season of the TechTalks!
HS: Thank you Judith. It's great to be back for another year of TechTalk Webcasts.
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 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 experts, Joel Hartman and Ron Eaglin, your own questions by sending email to expert@cren.net any time during this Web cast. If we don't get to your question during the Webcast, we'll provide an answer in the Webcast archive.
According to Yahoo, the University of Central Florida (where our two guest experts keep busy reshaping technology) is the 97th most wired college campus. My university, Princeton, is reported to be the 22nd most wired. And the most wired university of all is Case Western Reserve, with 100% of its dorm rooms wired. Princeton, Yahoo says, has 75% wired, and UCF has a tiny 9% of its dorm rooms wired.
All of us, even Case Western with only 41.2 shared computers per 100 students, still has lots of work to do to get their campuses completely wired. Of course, all of us have been spending prodigious amounts of time and money trying to wire our campuses with whatever technology -- twisted pair, coax cable, fiber, whatever -- whatever seems like the latest and best thing.
UCF is also wiring up its campus, but much of the wiring done lately has been to implement a wireless campus. That's just one of the very strange facts that we're going to explore today -- that it takes an extensive and expensive wired network infrastructure to support wireless computing on campus. Before the ink has even dried on our checks for zillions of dollars of copper and fiber, we are being asked to spend more on wireless networks that may obsolete the hard-wired connections that we just put in place.
We'd better have some solid reasons for embracing yet another new, expensive technology.
There are many obvious reasons to move to wireless computing. In some existing buildings, it's just too expensive to alter the structure to accommodate network connections, especially if altering the building exposes asbestos or requires the building to be brought up to new building codes. Laptop computers that have freed users from the dependence on 120 volt AC outlets have left them tied to hard-wired network connections. Wireless would free them, and a wireless infrastructure would support wireless palm-size computers and the host of wireless information appliances that will no doubt appear any day now.
But new technologies often produce unexpected results. Mainframe computers, for example, made it possible to solve otherwise intractable scientific problems, as expected. But they also led to the Sony Playstation, supermarket UPC scanners and the World Wide Web -- all quite unexpected directions.
What can we expect from the wireless revolution on campus? Will it change the way we do instruction, research and administration? Where is wireless computing going anyway? Which uses are silly and which make sense? And how do universities get wireless computing started on their campuses and keep it going?
Joel and Ron are facing those issues right now, in the face of the UCF student population that will more than double in the next ten years. We'll share their insights on these issues and others on today's edition of TechTalk.
Judith?
JB: Thank you so much, Howard. And I think we will continue to see -- to be surprised by technology as we move forward here into the world of wireless.
Let me tell our audience a little bit more about our guest experts for today. Joel Hartman is the Vice Provost for Information Technologies and Resources at the University of Central Florida, and in effect, is the CIO there for UCF. Joel came to UCF after decades of work in IT management positions at Bradley University. Our second expert today is Ron Eaglin, and Ron wears many hats at UCF and is often called "Dr. Engineering" (or Dr. E for short). Ron is the Director of Distance Learning at the UCF College of Engineering, and he's also in charge of new applications of technology as the Director of the Alternative Instructional Methodologies lab. And there's lots more information about our guests at today's Website.
So welcome, Joel and Ron, and thanks for being here.
JH: Thanks, Judith. Good to be here.
RE: Thank you.
JB: Great. And do you want to say anything about what's happening on your campus this very moment, Joel?
JH: Well, yeah, one of the things (talking about technology surprises) is that BellSouth, our telecom provider, about a half hour before the TechTalk, experienced a major regional fiber cut, and our networking capacity is pretty well out of filter right now. So if anyone wanted to send us an email, you might hold off a little bit until the network's back up.
HS: Okay, we'd like to remind our users, however, that expert@cren.net is working just fine. So you can send us questions at any time, even if UCF is not working.
JH: And then Howard can fax it to us!
JB: Oh, we'll do it by wireless!
HS: I think we'll just read them to you. That's all planned here. That's the kind of technology we're going to use right now.
JB: All right, but coming back for just a moment, for those of you who are up on the Website and would like to get a little bit larger picture of the campus and all of that, do come back to that as the network gets reconnected. You'll have a chance to take a look at all of those --
JH: If it happens during our conversation here, I'll let you know.
JB: All right, that'd be great, Joel.
HS: Okay, Ron, Joel, I understand that UCF is in quite a big elaborate program, going toward wireless. Perhaps you could start off by telling us why you're doing it.
JH: Yeah. I think the answer is, multiple things brought us to it, and we'll describe in a little bit just exactly what we're doing and to what extent we've made the campus or parts of it wireless.
One of the things that's happening is that the University is in a growth phase. You alluded to a doubling in size, which is not quite the case. We're at 32,000 students now and by 2010, we're projected to have a student enrollment of 52,000.
HS: Oh, I said more than double.
JH: Yeah.
HS: That sounds like more than double.
JH: Yeah. No, 32 to 52 is not quite double.
HS: Oh, okay, I'm sorry.
JH: But it's significant nonetheless, and it's going to increase our student population by roughly the head count of an entirely additional university. So it's significant, and we're in a major building phase.
That means we have opportunities, as we build buildings, to do some things with technology. And one of them, we've decided, is to couple both our wired and a wireless infrastructure in something of a hybrid fashion to take best advantage of both, given the needs of the buildings. So that's one opportunity.
The second, as Ron will talk about, is with the College of Engineering. It has begun a laptop mandate, and their model for that is wireless based.
And I guess the third is recognition that both students and faculty are essentially nomadic individuals and that their need for access to all these campus services and utilities that we're building extends to times when they're not at a desktop or a connection of some kind. And essentially, wireless gives you ubiquity on top of utility. So what we hope to provide is the ability to have both wired (for fixed work stations and connections) and wireless (for notebook, PDA and other kinds of devices) coexisting and ultimately saturating the entire campus so you can have both anywhere.
HS: Does that mean that if somebody has a laptop that's equipped for wireless, that your plan would also be to make sure that it's equipped so that it can tie into a fixed network connection?
JH: I think that's dependent on the user's needs. I mean, mine happen to be, Ron's happen to be, but we see the time when users will have perhaps a desktop machine -- a high-speed wired connection -- and then a notebook with a good but not quite as high a speed wireless connection they can take anywhere. And then we find places that people don't really use computers today very much (which would be meeting rooms, you know, lunch rooms, coffee counters and things of that kind) that they could take the computer with them and not just do computing, but have access to the entire range of information resources.
So it opens up possibilities and places -- where information exchange, communication, all this can take place -- that weren't simply thought of as locations you could do this before.
HS: So looking out as far as you can see, whatever your horizon is, you still see lots of fixed network connections?
JH: Yeah, I do. I see a gradual replacement of desktop machines with good notebooks, and I think as the speed of notebooks increases and the speed of wireless increases, we'll see probably a shift from wired to wireless connections.
For example, if someone were to have replaced their desktop machine with a laptop, and if they had wireless in the laptop that was accessible both in their office and other places, I know of a number of people today that have made the shift entirely to notebooks with wireless and have given up both the desktop and the wired connection. But there may be others who still must have 100 megabit or even ATM speed connection at the desktop, where wireless today won't provide, and so their only solution today is to go with a wired connection.
HS: Ron, I understand that there's some new engineering buildings that are going up that are designed for wireless?
RE: Exactly, and in fact --
HS: If you could tell us what that means, to have a building designed for wireless? How does one go about doing that?
RE: Actually, Joel's a good one on that one, but I'll answer that for you.
HS: Oh, okay. I'm not particular about who answers.
RE: He had a great answer I'll give you. Designing and building for wireless: well, there are some design components in designing a building for wireless, like don't make all your walls out of steel, as it does block the wireless.
JB: Okay!
HS: How about, should you make any of it out of steel?
RE: Well, if you don't want people to get -- you can block a signal going through the external walls. In fact, the Admin building is unique in that you can actually put a wireless access point on one side of a window, walk to the other side of the window and you'll get absolutely zero signal because it was built with lead windows and --
JH: Tinted glass, yeah.
RE: Tinted glass and it effectively blocks the signal 100%. And in some cases, maybe you want to do that.
JB: It may give whole new meaning to firewalls and things, right?
JH: Firewalls, love that!
RE: Now you have fire windows.
JB: Right.
JH: But Ron, the College of Engineering is about to build a major expansion of their building and we're also this fall going to be cutting over a new major classroom building. And both of these are among the first structures the University has built that are designed with wireless in mind, combined with wired networking.
In the case of the classroom building, we have wired connections all the places you would normally assume to have them -- desks, conference rooms.
HS: Yeah, I think a lot of people wouldn't assume to have them at every desk, so --
JH: Perhaps that's too big. It's an obvious place to put it, if not universal. The teaching console and multimedia classrooms, places of that kind. But what we are not doing is running a wired connection to every seat in every classroom. We're set up to do it if we need to, but we decided not to actually run the wire and instead will provide the entire building with universal wireless access.
And our hope is -- our assumption is -- that wireless will be the way the students will access the network in their classroom seats, whether they're in an auditorium, sitting in the hallway, in a classroom. We have docking ports and little desks in the hallways that will have both wired and wireless connections, should they want to use them. Now, the building is taking advantage of wireless in a way that will, we think, preclude the need to run wired connections to every seat in the classroom, for example.
JB: Joel, one question I think that people like to ask (particularly when you're talking about building a classroom building -- particularly in this way) is just what are the sizes of the actual classrooms. And also, what size is the auditorium in which this type of technology is possible?
JH: This particular building -- and of course, this may differ from campus to campus (it even differs on our own campus) -- this particular building is the first structure we've built that's entirely classrooms. Up until now, it's been a college of this or a department of that or a research institute. And it's had a variety of spaces in it, some of which are teaching spaces and many of which are labs and offices.
HS: But is this building going to be dedicated to the College of Engineering?
JH: No, the building I'm referring to, the classroom building, is really a campus-wide resource that will be used by all colleges. And what we've attempted to do is to --
HS: And when you say all classrooms, you mean all classrooms, no labs?
JH: Yes. Well, there's actually a PC lab in it for students and a PC lab for faculty, but again, it's a computing resource as opposed to a biology lab or such a thing.
JB: Okay, and how big is the auditorium?
JH: Oh, you were asking about the size, I'm sorry. This particular building has rooms that seat from 20, a seminar room, to around 30, 45, 75, then a 300 and a 400 seat auditorium.
JB: And it's the 300 and 400 seat auditoriums that provide special challenges for the wireless, isn't it?
JH: Well, two kinds -- one of which is power and the other is networking.
JB: Okay.
JH: And we're providing power -- not at every seat, but at many of the seats adjacent to the aisles, so it will be within reach with an extension cord, a power cord. And then there will be sufficient numbers of access points that a high percentage, if not all, of the students gathered in the room could be carrying on wireless communications at the same time.
HS: But if you had these 120 volt AC connections at every seat, I mean, the students are kind of pinned down. You're imagining the students are plugged in there. I mean, assuming there are batteries in their laptops --
JH: That's true, but --
HS: They can wander around, so why do they need the 120 volt AC outlet?
JH: Well, the thought is, you're going to be sitting some of the time, and that's charging time. And then, when you're ready to move around, you run off your battery, and you can simply unplug the computer and go. You don't need to shut down or do anything special, so you could maintain a permanent connection, using the times when you're not mobile to recharge your battery, so to speak. And then when you need to move, unplug from AC and go.
HS: But I take it you imagine the students are going to be mobile most of the time, so --
JH: I expect that. In fact, what we're hoping to be able to do is to allow them to move within the auditorium. We have a number of faculty who are going to be leveraging the capabilities of wireless to have students break up into teams, go off in the corners of the room and access information, do projects, report back and so on. So there's some active projects underway to look for ways how we can leverage the capabilities of wireless to actually do things in the classroom that couldn't be done before.
HS: Are you using any special software so that the professor can take any student's machine and display it at the front of the room, or so that students can see what's on each other's machines or that kind of thing?
JS: We have not actually made a decision on that yet, but there's been some discussions about using software such as PC Anywhere (or that kind of thing) to do that. But we've not made a decision, and we're not now doing that at this point in time.
HS: But do you have any plans to do that? I mean, is that going to be part of the rollout of this?
JH: I think it will happen, but there are no firm plans to do it at this time.
RE: We do, actually, run a number of computerized classrooms over here. And the situation where you do need to take something off of a student's PC and be able to display it up onto the screen or actually be able -- really, it's the exchange of information in a live fashion that enhances a classroom experience.
And what we will typically do in those cases is -- and this sounds funny, but you'll actually email people in the classroom. You'll email the instructor and send something over.
Or what we do with Web CT (and I'll let Joel talk a little bit about Web CT) is use the actual Web CT bulletin board to send files and have students that are sitting at their classroom seats post something to the bulletin board. They can be working on anything. It can be an Excel spreadsheet, they can be a Math CAD type of a thing, and attach it to a bulletin board posting. And then our computers are set up very easily so that the instructor, who is also connected to Web CT, just clicks on it and then displays whatever the file was that was sent over.
So we've been fairly successful. And working within our existing pieces of software that we have still, we're able to do that.
HS: As part of your plan to go wireless, obviously this works with laptops, not with desktops. People aren't going to carry around desktops with them.
RE: Actually, you can do it with desktops.
HS: I know you can put a wireless card in a desktop.
RE: We do have wireless desktop applications, actually. An example is that during registration periods, the Registrar's Office and the Finance Office set up service desks out in the foyer and commons of various buildings, particularly Administration Building.
HS: Sure.
RE: And we're now serving them with wireless connections so they can run out, plug in, do their service and then close down and we don't have to have network cabling out there, and it's working fine.
HS: Yeah, actually, we're looking at doing the same thing here. And we're also looking at doing that kind of thing for reunions. More and more folks are coming to reunions. We set them up in tents outside, and in addition to the beer they'd like to drink, they seem to want computer connections now.
RE: Um-hum.
HS: And wireless --
JB: Interesting. Wow!
HS: Wireless connections would be easier outside than running cables.
HS: But what I was getting to -- I think you raised a good point, though, that's kind of interesting: that it does make sense in some applications to have wireless connections on desktop machines.
But for students, obviously they're going to want laptops. And I wondered what plans you had to make sure that every student had one. Are laptops, for example, required at UCF?
JH: Not in the way you see it at some institutions, where there's a campus-wide mandate.
The approach we've chosen to take here is not to have a campus-wide PC mandate, but rather to approach the question from the opposite end, which is, when is it appropriate for a student to own a computer? And the answer is, when it makes sense academically to have ownership required (because we provide students with access to machines all over campus, so it's not like they're denied access). So when the faculty are prepared to use computers in the classroom in a way that fundamentally requires ownership or dedicated access is the time that a student should own one. So the mandates are coming not from the institution top down, but from the colleges, sort of bottom up. Ron's college, the College of Engineering, is the first on campus to do this, and I'll ask Ron to kind of explain what they're doing and where we'll go.
RE: What we're doing is looking at it on a pure per-class basis. We're taking individual classes and we're calling them the N classes, or the Notebook classes. And to actually enroll in a section that is an N section, you have to own a laptop computer and have a wireless connection.
Those classes are typically scheduled in what we would call a special room, but it's not really that special. It just means that it has enhanced power outlets and it has a computer projection system in the classroom that the instructors can use to display things. And that's actually a fairly strong model of classrooms throughout the entire campus.
And we make those decisions at a curricular level because some classes really naturally lend themselves to this approach and some classes don't. So it doesn't make a lot of sense, at least to us, it did not make a lot of sense to just flat say, "All students must own a laptop computer," especially being a state school where we have students that might -- it might really be a tremendous financial burden. But --
HS: But are you saying that students who can't afford one can't take these classes?
RE: That is something that we're grappling with.
HS: It seems like a really difficult problem, especially in a community college or place where students might not have the financial resources.
JH: We're putting in place both leasing and discounted purchase arrangements through our computer store on campus. We have yet to address the financial aid implications, however. Because it's not a mandate, the financial aid thing has not kicked in yet.
HS: Um-hum.
JH: And your point's well-taken. It's one of the issues that we're still working on with the particular model we've chosen.
JB: But if you are in this growth phase, then a student may have a choice of different sections -- one being an N section, then, and one being a non-N section. Is that right?
JH: That's correct, yes.
RE: They're not denied access to the class, yeah.
JH: And the other flip side of the coin is, in universities where you actually have a full laptop mandate, we're saying that some students will be able to get into sections because they have a laptop vs. other students that can't get in. But at the same time, if a university has a laptop mandate which would cause hardship and actually limits the student from even being able to enroll in the university as a whole. There's not a real quick and easy answer to that as an issue.
HS: Yeah, I mean, clearly, somebody has to pay for the laptop if a student wants to have one. And either it's going to be the university or the student or the state or some combination of those folk --
JH: Correct.
HS: -- are going to pay for the thing. And I think it's a bit of a problem to figure out who it's going to be.
JH: Yeah. And we're not saying that a PC mandate is a bad thing. On the contrary, I think many of the institutions that have done it are finding some exciting results. We have just taken, for our own reasons, a different approach based on what happens in the classroom and the faculty's ability and willingness and eagerness to do it and voluntary participation on the part of the students. We're going to try that for a couple of years and see where it goes, and if it doesn't produce any meaningful results, we may come back and revisit the mandate question again at that time.
HS: Yeah, you've said that one of the reasons you're going to wireless now is because the university is growing and because you're building a bunch of buildings and things, you have the opportunity --
HS: -- to do it. But I wonder what else has changed to make it sound -- I mean, you obviously built some buildings a couple years ago. What has changed now to make wireless attractive that wasn't true a couple years ago?
JH: Well, all the wireless we have in place today, in fact, is in older buildings that have been retro-fitted. So the new buildings are basically a future opportunity.
I think what's changed is that the technology has matured to the point where it's now practical and cost-effective to do it, and it works well. The pieces are in place to make a reasonable, supportable wireless infrastructure for a campus that has almost all of the attributes of wired networking in terms of robustness, manageability and so on and so forth.
The 802.11 standard -- the IEEE standard which was recently adopted that brought all of the disparate technologies into one common architecture where now Brand A, Brand B and Brand C can all work together pretty well seamlessly -- now makes it even more practical than it would have been before. Before, you'd have to choose a brand, hope that company stayed in business, and then stick with the one brand. Today, you can mix and match.
JB: So in your implementations, Joel, let's say the new classroom building that you have, that you're just finishing, is that a multi-vendor implementation in that building?
JH: Well, we could be influenced by partnerships, as anybody could, in terms of a singular opportunity, but the plan is to stick with a single vendor for the access points but to allow students to buy multiple vendors of wireless cards.
JB: Okay.
JH: Which, when you think about it, is similar to how you do a wired network. You probably stick with a single company for the hubs and routers, but a student or faculty member with a laptop could buy any of a dozen brands of NIC cards for their laptop with no problem.
HS: And who are you using for your access points?
JH: We're currently using a combination of Aeronet and Lucent on campus.
HS: Okay, I thought you said single vendor and now I just heard --
JH: Well, I was talking about a particular project, but in different buildings, we do have some of both and we've been trying them out.
JB: And are you finding that they do interoperate well, or have you found some leftover nagging problems in that area?
JH: During and shortly after the standard went into effect, there were still some nagging problems with software revs about once a week.
JB: Ooh!
JH: That has pretty well flushed its way through the system, and I would say today what we're finding is that we can put different brands of access points in the same building and have multiple different kinds of cards roaming throughout the building and it all works seamlessly.
HS: What's the difference in speed that you're seeing between what people get on desktops commonly and what's available today on wireless?
JH: Well, it's interesting. There's a difference between rated speed and observed speed. And let's take, for example, a typical ten megabit wired connection. And I guess you'd have to distinguish between switched and shared. But let's not get that fine a grain detail.
A ten megabit Ethernet connection could give you roughly maybe two megabits of useable bandwidth and a 100 megabit connection could give you -- I'm not sure I could give you an equivalent number, but certainly much more than two, but not a hundred.
HS: One would hope so!
JH: But not a hundred.
HS: No, it won't give you 100. But if it gave you only two, you'd have a problem.
JH: Yeah. The two flavors of wireless are two megabit and 11 megabit. Those are the two standards. And the two megabit wireless next to a ten megabit connection (so the hub itself is talking into ten megabits and transmitting two out to you) and you can get about 200 K out of that kind of a connection. An 11 megabit access point we're seeing 300, 350, maybe 400, so it's one order of magnitude, not five orders of magnitude. Follow what I'm saying?
JB: Interesting.
JH: The access points, one's a two, one's an 11, but you don't actually get five times the throughput. You get roughly twice, maybe half again to twice.
JB: Let me perhaps remind our listeners that they can send questions in to the email address at expert@cren.net. And while I'm already interrupting here, Joel, maybe it would be good to just re-emphasize.
Then you're saying that there's two different rated speeds, then -- two megabits and 11, but even the difference between observed speed then is only like a one to two ratio.
JH: Yeah, two to one ratio.
JB: Two to one.
JH: And that may vary under selected conditions, but typically, on average, that's what I see anyway.
HS: Can we talk for a few minutes about some of the costs here? If you're a student and you're buying a laptop -- obviously, any laptop that has a PC card capability (which is every laptop I can think of) can participate.
JH: Yes.
HS: So the only added cost, really, is the cost of the card. Is that correct?
JH: That's correct.
HS: Okay. And these two different speed cards, what's the street price of these things now? Or what are people on your campus paying for these two megabit or 11 megabit cards.
JH: Ron keeps up with that more than I do. I'll let him answer that one.
RE: I do keep up with that. Right now, the two lower priced cards that we have are the Zoom Air and the Lucent card which are both retailing at $265. Right now, in fact --
HS: That's a two megabit card?
RE: Two megabit card. However, these prices are fairly variable, and they're variable in one direction -- which is down, which is very nice. In fact, street price for the cards will probably halve within the next couple of months and there's a lot of competition, once the standard came out. Now anybody can make these, and there are a number of chips that are out there and a number of vendors that are out there and they're in a fairly cutthroat competition situation right now. So all we've seen is -- and when I mean price down, it went from $500 to $265 and we're looking at $130 range now for the cards. So we like that!
JH: It's also worth mentioning that Apple's decision to develop the AirPort card in the IMAC or the IBook --
HS: No, the IBook.
JH: -- is going to have some impact on pricing as well. Apple's technology is developed in partnership with Lucent and it's 802.11 compliant. And Apple's own Website is talking about a card with a retail price of $99, and the access point base station having a suggested retail of under $300. So if that's the price around the corner, we can see that the market is again going to collapse and that's probably even suggested retail. You're going to buy when the street price is lower than that.
HS: Again, we're talking about a two megabit card.
JH: Two.
RE: No, we're talking about eleven.
JH: It's eleven.
HS: What?
JH: That technology's 11 megabits.
JB: You mean the IBook tech stuff is?
JH: Yeah, um-hum.
JB: Okay, all right.
RE: On a manufacturing decision at the level of Lucent, they can pretty much choose to make the two meg or make the 11 meg cards, and that's not a very difficult decision to make if you're mass producing.
JH: Yeah, AirPort's eleven.
HS: Okay, so you expect in the near future that the additional cost to the student will be about $100.
JH: Yes.
HS: That's what a student will pay.
JH: That's a reasonable expectation, I think.
HS: For --
JH: Which puts it a little more than a wired card would be, but still not out of sight.
HS: Right. I mean, the major expense is obviously the laptop, not the card.
JH: That's right.
HS: If you can afford the laptop, obviously, you can afford another $100 for the card.
JH: Yes.
JB: Although this might be a good time to mention, do students need to have both Ethernet as well as the card for their notebooks, as well as the wireless card?
JH: The truth is that today, it probably would be, in that we don't have wireless everywhere yet.
JB: Um-hum.
JH: But when we reach that goal -- and I do mean everywhere, including outdoors and dorm rooms and the entire campus -- the answer would be no.
However, many of our students live off-campus and pretty much all of our faculty and staff live off-campus, despite what they may say at times. And they want this at home, so having a high-speed connection in the office still means they have to find some solution at home. And I think maybe for another discussion, Howard, the question of what do people do at home? And now that cable modems and DSL are coming around, it's likely that people who really are into networking may have an Ethernet speed connection at home and do they need a wired connection at home and a wireless in the office? How do you do that? And I think that the emergence of wireless home networking now would become a new phenomenon.
HS: Oh, I think that at least in the immediate future, even with DSL and cable modems, I think that most people at home still are using just regular telephone modems.
JH: That's true, but it seems to me that, in fact, my own future plans are that when I'm able to get the service, I plan to have a cable modem at home with a wireless access point. And that means wireless computing throughout the house, without having to worry about what room you're in and so on. And I think that'd be a delightful way to go.
HS: Is any of this wireless stuff going to use the cellular telephone network?
JH: Yes. When I'm off campus, I use a CDPD (Cellular Digital Packet Data), which is a relatively low speed service that runs piggybacked on the voice cellular network. And again, you plug a PC card into your computer, you bring up a piece of software -- which is essentially the same as dialing in to a campus modem pool.
HS: So it turns your laptop into a cell phone.
JH: Well, it's purely digital and it runs IP.
HS: Okay.
JH: So it's actually very much like a modem, but the technology today is fixed at 19.2 kilobits, and that's awfully slow for a lot of applications, but it works well for email and limited Web surfing. So when I'm traveling, I can open the notebook in an airport in certain cities, bring up a CDPD connection and have slower but still wireless access -- just as I could have in my office with a high speed 802.11 connection.
HS: Is that service available over a wide region? I mean, is that available wherever cell phone service is available?
JH: Yeah, there are a number of major vendors that have it. I happen to subscribe to AT&T Wireless's service and they have it in many cities nationwide, although not every city. So it's still not what you'd consider a fully ubiquitous service, but it's very --
HS: And the cost of that? Since you have it, you can tell us what you're paying.
JH: Well, their particular service is approximately $50 a month, in that range -- and that's all the data you can eat for that price, so you don't pay by the packet. And so that means for a flat fee, you can use it as much or as little as you want anywhere in the country, no roaming charges. That's a fairly convenient arrangement.
HS: Okay, we've talked about the cost to the user pretty much. What about the costs for the campus in putting these wireless hubs in and whatever other infrastructure you need?
JH: We're setting standards campus-wide. And the folks in the College of Engineering who are both doing research as well as deployment have been among the leaders in working through some of the technical issues.
From my position as CIO, once we've settled on brands and models and so on, that what we'll do is the central computing organization will provide wireless service for many of the shared buildings like the Student Union, the Administration Building and so on. And we're partnering with colleges who may want to foot the bill to do wireless in their buildings. In fact, we have a couple of colleges lined up now wanting to do that. And the cost is relatively nominal.
RE: In fact, you have very broad coverage with a single access point. And if you're looking at an access point even at the current prices today (which are the $1,200 to $1,300 range -- and of course, those costs are coming way down), you've got about a 300 foot radius in all directions -- not just side to side, but also up and down. So in multi-floor buildings, you basically put all your access points on the second floor and your first floor is completely covered. At that, I have a 63,000 square foot usable. It's about 100,000 square foot building that's completely covered with eight access points. Four stories.
HS: But we also have a bandwidth issue -- whether it can cover a big volume. If we put enough people in that volume, it just won't be useful.
JH: Well, what you do then, it's like taking a box and filling it full of ping pong balls basically. That gives you base coverage. Then, to handle volume of data or number of users, you double up access points in those locations where the heavy data flow is coming from, such as an auditorium or a lab or a commons area, and they do load sharing which allows --
JB: Load sharing between the access points?
JH: Between the access points, yeah.
JB: Okay.
JH: So if you run out of bandwidth, you could add access points, being kind of careful about frequencies and so on, and accommodate increased usage.
HS: This new classroom building you're building, can we talk about how many access points are in it?
JH: The building's not finished yet. And I'm trying to think about how many actually --
HS: I'm just trying to get some sense of how many access points do you need for how many students? You're going to have so many seats in that building.
JH: Well, the building's fully loaded capacity is 1,200 individuals. That is, you have 1,200 seats in the building for occupants, including office workers and students, at a given time.
JB: How many? 1,200?
JH: 1,200, plus the possibility that another 1,200 are on their way in or on their way out from preceding or following classes.
JB: Okay.
JH: We will have probably somewhere in the vicinity of ten or 11 access points for base coverage, and then additional units in the auditoria and other areas that turn out to be heavy use. So probably less than 15.
JB: Okay.
JH: And that may be an understatement of the auditoria, depending on usage as it builds up, but initially, that should cover it.
HS: So there's going to be on the order of 20 to 25 access points in that building?
JH: No, I'm thinking 15 will cover it initially, maybe 15 to 20.
HS: Okay, so about 20 in that building for about 1,200 students attending class and things. And these access points are about $1,200 apiece, did you say?
JH: That's currently the case, yeah. But keep in mind that not every student is carrying a notebook with them at all times, nor if they were, would they be using it at all times. So even though there are 1,200 bodies in the building, the number of devices using wireless service at a given time is some fraction of that yet to be determined.
RE: In fact, it becomes a little bit more complex than formerly because you also have to deal with you have students that may be using it, but are not actually transferring data. So there can be quite a few students connected at the same time. And, you know, the worst-case scenario is you have an auditorium filled with 400 students on three access points who, all at the same time, the professor says, "Okay, now, everybody download this file."
HS: Yeah, it's a large file and they all hit the RETURN key at the same moment.
RE: Yeah.
JB: Listen, I do have a question coming in. Actually, two questions coming in from one person, Sylvia Elliott at the University of Georgia. And she's mentioning that the video cam is not responding, so I think we need to remind everyone about the network, larger than life network problems that you're having down at the campus.
JH: Yeah, there's been a Bell South fiber cut in a location, a major hub near the university, which just coincidentally happened to sever all network connectivity to our campus about a half hour before our TechTalk, so any of our URL's will not be responding. And I just was handed a note that says the fiber will be restored by 9:30 tonight, so it's going to be awhile.
HS: Okay, so we'll just keep this TechTalk going until about 9:30 tonight!
JH: Yeah, then we could see the campus.
JB: Then they could see the video cam, except that since Ron had to move since they were backhoeing out of his window, right?
RE: In fact, the video camera -- which I can see -- is very interesting right now because you can see where I work because it's the one with the large bulldozer tearing down a piece of my wall!
HS: Okay, we hope that's not the bulldozer that cut the cable!
JB: But anyway, let me -- Sylvia had a couple other good questions. She's asking about just how the students and users access your network. Do you distribute IP networks or do you use DHCP or just what do you do with that?
JH: We use DHCP.
JB: Okay.
JH: And that allows a couple of things to happen, but one of them is full roaming throughout the campus. Our backbone is ATM, and our networking folks working with Ron and his colleagues in the College of Engineering have set up all the access points so that they're on a V-LAN (virtual LAN) which treats all the wireless space on campus essentially as a flat network. And that means that if you have a laptop with a wireless card, you can boot it in any wireless-equipped space on campus and not have to fiddle with any of the settings.
HS: Okay, that's very nice.
JB: All right, and then there was one other question, and that is, I think, really about how users connect and reconnect the network as they move across, say, from access point to access point. Do they notice that, or do they need to do anything different, or just what happens? Maybe you could answer that.
JH: Well, take, for example, a building that has full in-building coverage, engineered properly with no dead spots. As an example, I brought up RealAudio, put on a pair of headphones and tuned into my favorite Internet radio station.
JB: Um-hum.
JH: And I walked into every room through every corridor of the building with the laptop.
JB: Yeah?
JH: And the music didn't skip a beat. So it gives you an idea that you do, in fact, have fully seamless roaming coverage just as you would in a properly engineered cellular system, cellular radio.
HS: What's really happening with those access points? Because obviously, at some point, you're transmitting to both of them.
JH: Maybe you ought to explain that one, Ron.
RE: Well, the access points work primarily the same way a cellular data system does. Right now, in fact, I'm looking at my computer right now, which I've brought up the access point manager, and I'm currently seeing three access points. My computer has visibility to the three, and it's chosen to connect to the access point which actually is giving me the largest power of the signal.
If I started walking away from the access point (which I won't do right now because I'm on the phone and the phone cord isn't that long) but what would happen is, I can actually see the signal strength of the one that I'm walking away from decreasing and the signal strengths of the ones that I'm walking towards increasing. The software itself makes a decision when to actually switch you from one access point and then connect to the other one.
The switch is not an infinitesimally quick switch, but it's fast enough that, like, if you're listening to something on RealAudio which does have a little bit of caching capability, it takes about a second to a second and a half to actually make the connection. So if you have a mission critical type of application that losing the connection for a second will kill it, you'll actually see the connection die. And I've seen that, but I don't run that many types of applications either.
JH: But if you're doing a Web browsing situation where you're downloading a page, an FTP, transmitting or receiving email, receiving Real Audio or Video, all that seems to work pretty much without a glitch.
RE: That's right. In our building, it's been working without a glitch for a good six months.
In fact, here's something that you might have to worry about. We set up all the access points, it was roughly a year ago, and we wrote down on a piece of paper a map of our building where all the access points were. We happen to have lost the piece of paper that had all that information, so even though the access points have been working fine and we can try to kind of figure out where they are based on their signal strengths, we don't quite remember where they are, and we didn't put them in places where they are right out in the open, so they're actually hidden in ceilings and places like that. So we had to do a search and find on one of the access points recently. Somebody had unplugged it, and we didn't know where it was! We wanted to bring it back up, and we were trying to find it.
JB: That calls to mind all these new little tools that they have for spying, that the cameras are so small that they just meld into the walls. It sounds like you have achieved that.
RE: Let me get you an example here. You can't see our cams right now because we're offline, but if you could see our cams, the one in my office (which I did have pointing out the window right now so that you can watch the people destroying the loading dock outside my window, causing my ground to shake -- which is what's going on right now) we have cams all over the building and the cams are typically hooked to a computer on a wired connection.
But we did a broadcast, an actual floating broadcast when we did the groundbreaking. We took a wireless laptop and walked around the groundbreaking talking to people and broadcast that, not only remotely, but with a remote video over a wireless connection. The cam was hooked to the laptop which was then hooked to the wireless network. So we were able to have a basically floating live video feed. Which is kind of interesting.
JB: That is very interesting! I think, you know, let me just remind everyone, we've got time for perhaps one or two more questions for people who send them in. And we're also getting close to the time where we usually try and ask the most practical kind of question that we have. I think, Joel, you have shared a couple instances in which it's really good to use wireless, and one is where it's difficult to reach areas or perhaps temporary areas, and then other places where you're really using the highbred model. Do you have any practical suggestion for people as they're perhaps planning their first wireless project, where they might best try that?
JH: Well, I guess my answer is the circumstances will maybe dictate where it will be of most use. We've only thought of a few here. I'm sure others have others that would go beyond what we've done.
But the beauty of wireless is that you can operate untethered -- and if you have broad coverage, you can take it with you wherever you want to go with it. I mean, literally anywhere you want to go. And one of the things we've observed is that one of the places that's now become an active hub of computing is our coffee shop in the bookstore, which is not a place you would necessarily have -- I mean, the cybercafes are all over the place, but you would never have thought of that as a core computing location. We have both wired and wireless connections there, and it's now not uncommon to find three or four people with wireless machines there at the various tables or places that you wouldn't have had connections before, actively communicating on email or on the Web.
JB: And then you had mentioned before that you also have a library project going, so a place where people have tables and chairs we haven't --
JH: That's another one, that this fall, our library will be fully-equipped with wireless. In fact, it's being installed currently, and the library will have both laptops with wireless cards in them or cards that could be installed and used by students that bring in their own notebooks, which gives them full roaming throughout the building with a wireless connection.
JB: And they would just borrow those for the time they are there?
JH: They would just borrow those for the time they were there and then return them.
JB: Okay. Interesting.
HS: To follow up on Judith's question, if one was at a university that had been thinking about wireless but really hadn't started, where should they start? How do they get going?
JH: It depends on the scale with which you want to take your first step. My recommendation is to have the appropriate parties -- which means not only the technical staff, the networking staff, but also the faculty or staff who will be using it -- get together and discuss the application and to build a pilot project of some scope. It could be a building, it could be part of a building, it could be a single access point with a few machines in the vicinity. But set something up and get some experience with it
And I think one of the things that most people find is that it's a very captivating technology. Once you've had the ability to use wireless to roam, to have access almost anywhere you care to sit, it's difficult to go back.
HS: I think one of you mentioned some conference or something that was going to be discussing some of these issues.
RE: That was me. On February 3rd and 4th in 2000, we're having a conference called Laptops in the Classroom. It's a conference that'll be here at the University of Central Florida, and what we're trying to do is, as we started getting our laptop program together and working on it, we discovered that even though there's a lot of universities that have a laptop mandate -- there's a very large number -- there are few universities that are doing it well.
There are a lot of universities that are not doing it well. Basically, they've come up with a mandate and just throw people to the wolves and the faculty try their best to figure out how to deal with a situation of all the students having computers. But it's not really being approached from the -- the approach that's taken is not "Let's figure out the best way to use these laptops in a classroom setting and get the most value out of the laptops and train the faculty and prepare the infrastructure for what needs to be done."
And I found this just by going around and talking to the people at different universities and sending out emails and then talking to people who are representatives of Dell who have a very large, broad scope and have met all these people. And so I thought it would be timely to actually put together a conference.
Well, we invited experts -- an expert is somebody who's been doing this for a little while, and in this case, a little while was not very long because you can only say one to two years. But people who have had practical experience in the setting of dealing with a classroom where you have a laptop in there. So we've put together a large list of topics, many of them pedagogical issues, of literally what you do in the classroom once you've got these students sitting in front of you with all the laptops open and staring at you. Also to the infrastructures and the wireless vs. wired and some of the other issues that you have to deal with.
HS: I assume there's a Website that describes this.
RE: Actually, I've given that one to the -- you can get to it through the CREN (hopefully) Website, except that you probably can't get to it now through our Website since we're currently down.
HS: Okay, I urge our listeners to go take a look at that.
JH: You have the URL, Ron?
RE: Yeah, actually, it's www.engr.ucf.edu/students/laptops.
HS: And it's linked from the CREN TechTalk series Website, so as soon as we get that cable repaired somewhere in Florida --
RE: Sometime after 9:30 tonight you can actually go out and take a look at the information that's on there. And it does have a list of the topical issues we're planning on covering with the conference itself. It's also a good opportunity for networking because, as I've found by going to other conferences, every now and then I'll run into somebody who has a little pool of expertise that I need to tap on these types of issues. But it makes a lot of sense to have those pools of expertise handy, right there where everybody's together.
JB: Okay, and actually, I'm looking at our Website right now and on the Web Event page and I don't see it there, but we'll get it there and it will also be in the archives, so the dates again, Ron, are February --
RE: February 3rd and 4th, 2000.
HS: Right, and that's far enough away from the Y2K things so you'll be able to travel again and things like that.
RE: Yeah, you should be able, yeah.
JB: And it's in Orlando, so it would be a good time to do everything, right?
RE: That's right.
I also have a situation that I had to deal with today where wireless turned out to be a good solution. We had talked earlier about how can you incorporate wireless into distance learning, and it doesn't really incorporate that tremendously well because you have students that are far away and access points that are local.
But right now, we have an interactive compressed video system that is running to a number of remote sites. And a problem occurred at one of the campuses that they were saying that the notes that were sent over the compressed video just weren't coming in well enough that they could read them.
So what we were doing was we were having the professor fax the notes down to the students, which of course, meant that you had to have somebody in a room that can get them and make copies and then give them to the students. However, in the classroom itself, it's not a classroom that we could easily go and put a computer in, a permanent computer, but we have set up a laptop checkout so that a student can come and get a laptop, walk into the classroom itself, open the laptop up and actually look at the notes which we have scanned into PDF format and on a Web page.
Well, it solved all the different problems that we had because we couldn't put a laptop in the classroom and just leave it there because it would sprout legs. It wasn't even a situation where we could put a wired desktop unit in there because we had no ability to have security there. But we did have a location where we could actually have the students go and check the computer out to bring into the classroom.
And then the next problem was, how do you deal with the network connection? And the people that were down there were saying, "We have no network connection to hook this thing up to." So there was another problem that had arisen, but we did have network connections that were nearby, and nearby means within 300 feet. But we had one in an adjacent room.
So the solution was to use a wireless laptop that the students could check out and they could walk into the room and open it up. So while our students are actually sitting there watching an ITV class, they could open up their laptops and access the notes in nice high clarity that they weren't able to see very well on the screen in many cases. So we solved a problem.
HS: So that's a very interesting situation there.
JB: Interesting. Okay, I think our time has come.
HS: More than!
JB: More than!
HS: This is very easy to keep discussing this with you folks until 9:30 tonight so folks can use your Website, but I think we're going to have to sign off. Judith?
JB: Oh, yes, and just as we're signing off, one final question came in, and maybe we'll just go ahead and take that -- and that is because I think it's one that folks would be interested in.
And that is from C from Ft. Lauderdale, who asks have there been any major improvements in wireless systems that would assist selling the idea to administrations?
RE: C in Ft. Lauderdale?
JB: Yes.
RE: Hi, C, in Ft. Lauderdale! Joel, you want me to field this one?
JH: Sure, go ahead, Ron.
RE: Well, there was the major improvement of the 802.11 specification which gives you 11 megabit access, which is fairly fast access. I mean, you are approaching desktop capabilities. It's the major improvement that we've had over the last few years.
That doesn't mean that, of course, we're going to stop right there. In fact, there is another technology that's going to be down -- it's about four years down the road right now -- that is built on a different frequency range, the 50 gigaHertz frequency range, which has 25 megabit access. Currently, you're going to have a hard time buying products for that one because there's nobody selling them, but of course, there will always be improvements. Somebody's going to find a better and faster way to do everything.
The big thing that makes her able to sell it to the administration is the fact that within the next couple of months, the prices are going to be to the point where you can go in and do a wireless application for less than you can do a wired application, which is unique.
JH: Yeah. I would say that's music to an administrator's ears.
JB: Um-hum. All right, well, thanks very much.
And I'd like to thank all of our Web participants for being here today for this time with Ron and Joel. And if you do have follow-up questions, you may continue to send them for the next 24 hours to expert@cren.net and Joel and Ron will answer them individually or on the Web.
And be sure to mark your calendars for October 7, just two weeks from today. And on that day we'll have a topic that came up number two on our member survey list, and that was the Digital Millennium Copyright Act. Our expert for that session will be Laura -- who is know by most as Lolly Gassoway -- from the University of North Carolina. So please plan on joining that session and inviting your friends and colleagues. So check the Website for upcoming events and, as always, we welcome your suggestions and feedback.
Thanks to all who helped make this possible today: the Board of CREN; our guest experts, Joel Hartman and Ron Eaglin; technology anchor, Howard Strauss; Event Webpage creator, Terry Calhoun; to David Smith and Patty Gall of CREN; Julie O'Brien, Jason Russell and the whole support team at the MERIT Network; and transcriber, Susie Berneis; and editor, Laurel Erickson, and all of you for being here. You were here because it's time.
Good-bye, now, Joel and Ron. I hope your network's up soon.
RE: So do we! Thanks a lot, Judith.
JB: Okay, bye, Howard, and bye, everyone online.
HS: Bye-bye.
JB: Bye-bye, now.
JH: Bye.
