Carl Malamud: Internet Talk Radio, flame of the Internet.
Malamud: This is Geek or the Week, and we’re proud to have the ultimate geek here today. This is Jun Murai from Keio University. Jun, welcome to Geek or the Week. I’d like to ask you about some of the work you’ve been doing in Japan. Maybe you can start by telling us about the work you do at Keio university. You have students there that need to compute, right?
Murai: Right. Okay. Our university, Keio University has eight campuses. And this is a pretty new campus, just four years old. So we’re having the first graduation in next March. So each of the grades has about 1,000 students. So we have 4,000 students now. And we’re going to have a graduate school. But this university is very special for the Japanese students and Japanese universities, and maybe for universities worldwide. When they get into the university, they have to learn how to use the Internet. Every single student, right. And then they have to use email and other networking functions on Unix workstations for their entire campus life.
Malamud: So these are freshmen, they come in, they need to learn Unix, they need to learn Internet skills, that’s a part of their education.
Murai: Right. That’s the first thing they have to do.
Malamud: And how do they do that? Do you have workstations all over campus that they sit down and use?
Murai: Yeah, that’s the point. Yeah, we wanted a lot of workstation for them, but you know, it was impossible because of the financial problem we had, and also we didn’t have the space. So then we were asking all the Japanese computer makers to provide us with a very light Unix laptop computer. And they worked very hard but they couldn’t make it at that time, four years ago. And then Sony produced the Soney NEWS workstation, which is a Unix workstation, and also there was a SpArc laptop, the first one, from Toshiba. But those are very expensive and very heavy, and it was impossible for students.
So, what we thought was to provide very light laptops to the students, so that they can use their workstation but if the number of workstations is not enough then they can use the laptop in exactly the same manner. That’s was what we wanted. So, what did gave up on the Unix workstation laptop solution. So instead we chose a DOS laptop up Toshiba for the first year, and then we started all the students having a very similar environment on a DOS laptop. And then in the universities, probably one third of the students could use Unix workstations. So that was how we started.
Malamud: And are they still running DOS laptops and Unix workstations now?
Murai: Yes. That’s a very good question, actually. Okay. So now we’ve been working on realizing our dream. So we have been working on making the laptops run Unix. And then finally we got that kind of environment. So we actually…probably last year, we got a copy of BSD running on the 386 architecture. And then we put the all the environments we’re using at the universities, which is mostly hand-made and public domain Japanese character-handling environment… So we put all those environments on that, so we are now running machines.
Malamud: Now how do these things get into the network then? I can understand with the workstation, but you just have a bunch of docking ports that people get into? Do you have wires all over campus?
Murai: Yeah, that’s on the problems, too. So of course we are working with an Ethernet solution so that laptops can connect to the university campus network via normal Ethernet like 10BASE‑T and 10BASE2, because our campus was designed that way from the first point. So every single room is equipped with Ethernet anyway.
But then we hit the problem that know, because all 4,000 students are carrying their laptop and trying to connect to the campus network somewhere in some room or…then we need the wires and the cables everywhere, right. We can’t do that, so we are now working on a wire solution as well. But now that we have Unix running on it, that’s a normal Unix, so we can develop any kind things over it.
And also we are now trying to put— Okay. Since we have buildings, right, and each of the buildings are on a separate subnet. So if a student is carrying a laptop from one building to the next, then the subnet address would change. So the WIDE Project, our group, is working on a mobile IP protocol. So we are now up putting that protocol onto the laptop Unix so that they can move around the buildings, and hopefully the wireless solution will be completed by March.
Malamud: So there’s two aspects. One is the wireless communications, the other is the mobile IP.
Murai: Right. Yeah. [crosstalk] Mobile—
Malamud: How do you do wireless?
Murai: Mobile came first, and we’re going to work on the wireless.
Malamud: Okay, wireless is just going to be some infrared or microwave, or—
Murai: Yeah— Well… Yeah, currently we are thinking about using the telecom wireless.
Malamud: Now what about the mobile IP problem? I know a lot of people have been working on it. It sounds like you have solution for mobile like IP that’s in the works.
Murai: Well actually the one solution proposed for IETF is from a Sony [indistinct phrase] working on that. He’s very actively working within the WIDE Project.
Malamud: Now how does that work? What happens when I move from— I’m in one subnet and I’m computing, and I pick up my laptop and I walk into another building. What happens? Is there some dynamic readdressing going on? Uh…
Murai: Uh…yeah. Actually the… Yeah, there are solutions for the various cases, but the solution we are employing is called VIP—Virtual IP. So basically one machine’s got two types of IP addresses. And then an intermediate node, either…kind of switching the other for the, you know, to assure the…you know.
Malamud: So it’s a level of indirection, is that what we’re doing here?
Malamud: Okay. And that means that as I walk from building to building, things like my TCP connections are not broken, right?
Murai: That’s true.
Malamud: I just keep on going.
Murai: Yeah, that’s true. And even the— You know, we have a lot of experiments of the variations on the implementation of the protocol. So even the TCP connection…while using the TCP connection and moving, or cutting, and you know, switching from Ethernet to a SLIP connection, you still keep the TCP connection, so…
Malamud: That’s amazing. So that’s… Walking is fairly slow. Now, is this gonna work if I’m driving in my car? Is this robust enough to handle a car situation?
Murai: Uh…well, yeah. Yeah, I’m sure it… Okay the VIP solution is independent from the lower-level solution like wireless things. So that’s an interesting point we are doing in our campus because we’re going to use the wireless thing andthe mobile VIP solution at the same time, right. So you know, carrying the laptop from classroom to classroom we’re using the lower-level solution of the wireless networking technology. And then if you jump from building to another, then the mobile IP solution would help you.
Malamud: Interesting. Now, you run…in addition to doing work at Keio, you’re the founder of WIDE, which is the Japanese Internet. Give us some idea of how big this network is.
Malamud: Are talkin’ a dozen computers? A couple dozen?
Murai: Okay. We have seven…a little bit more than a dozen, network operation centers all over Japan, from the northern island Hokkaido to the southern islands. So seven locations, and we have a so-called backbone. It’s not very big bandwidth, though. So that entire network is used for the experimental work of the development of the protocols and you know, that kind of thing. But at the same time we are providing the Internet services to each of the members of the WIDE Project. The Wide Project is to realize the widely-integrated distributed environment, that is what “WIDE” stands for. To realize a global [indistinct] environment which is very similar to—
Malamud: How many people are part of the WIDE Project?
Murai: Okay, the registered researchers working for the WIDE Project is 130. No, 136 mem… [crosstalk] researchers and students.
Malamud: So 136 researchers. I mean we’re talkin…—
Murai: Yeah, researchers and students. But active members are about eighty-six or…yeah.
Malamud: And how many computers do you estimate are in area served by WIDE? [crosstalk] How big is the Japanese Internet?
Murai: Well… Oh how big is the Japanese Internet. Okay. About… Let’s see, about 500 organizations connected? Network numbers connected, sorry.
Malamud: Mm hm.
Murai: Yeah. So—
Malamud: So that’s 500 networks, and presumably many thousands of computers, or tens of thousands of computers.
Murai: Well…yeah, not. Yeah, there are a lot of a small entities. But yeah, there are very big entities as well.
Malamud: Now, WIDE originally started out as a research network, as a research experiment in fact.
Malamud: And now it looks like the Internet in Japan is starting to enter a new phase. You’ve been instrumental in helping to give birth to IIJ, which is a…service provider?
Murai: Yeah. Okay. So the problem in Japan was… We’ve been working on promoting this kind of environment like Internet in Japan. And that has been very successful and everybody understands that okay, the Internet is a good thing and that this kind of computer network is very useful. And now everybody wants to get on the Internet. That was what I wanted.
Malamud: Jun Murai, let’s talk a bit about your research. You actually are a researcher in addition to running a large network. There’s a couple projects you’ve been involved in that I find I absolutely fascinating. Maybe we could start by talking about satellite-based multicasting. Could you tell us what that project is?
Murai: Ah yes. Okay. When we are running WIDE Internet, we found that there are a lot of applications over the WIDE Internet were basically the multicast communication. So we were interested in multicast communication technology. And so we worked on a Class D IP transmission thing.
Now, at the same time we are are a research network and so we are funded by various companies. But for the research, we are not rich at all. So we are always looking for more bandwidth. And then one file was shared by everybody, and the network sometimes congested. So then when you look up the [indistinct] there is a satellite broadcasting started. And I talked with them and there was certain bandwidth that they weren’t using, and they wanted that boundaries for data communication but they didn’t know how. So we’ve been working with them.
Malamud: Now, we’re talking about the K band satellite system, right, in Japan?
Malamud: Now, K band for listeners who don’t know, it consists of if you walk around Tokyo you will see little satellite dishes in almost every home. And these dishes cost a hundred bucks, right?
Murai: Right. Yeah. That’s very cheap. And that is a point we’ve been working for. And then okay, satellites. We’re gonna have bandwidth. And we can design the multicast transmission. But if there’s just one too many just using the satellite, that is not helping the Internet world at all. So what I was working on was integrating the satellite communication paths together with the existing terrestrial communication path, which is to use the satellite communication path as part of the normal Internet communication path, right.
So then we designed a special kind of routing mechanism so that if you issue a multicast datagram, then it would be automatically routed to use a satellite. The mechanism is basically very similar to the [force?] routing type of thing, but it’s a mixture of the existing technology plus some new ideas.
But anyway. So then as you mentioned, there is a satellite market around in Japan now. And so you can get a receiver in a very… Yeah, it’s kind of very cheap, right.
Malamud: But it connects to a TV normally, right? [chuckles]
Murai: Right. Yeah. So what we did was we designed a small hardware adapter which can retrieve information from the normal receiver, and then provide that data to the computer, right. It’s an adapter; we call it a satellite adapter. We designed the hardware. And the interesting portion of the hardware is then— Okay. Actually, potentially the satellite could carry 8 megabits per second, right. Then if you think about the interface board, right, to receive 8 megabits of serial line communication, then usually you have to buy a special kind of serial line board, right. Which is very expensive usually. So if we do start this technology for the actual communication, probably the universities and people who can buy that board are you know, probably limited. And also the products are limited. So for your computer receive the satellite communication would be limited. And that is not interesting at all. So that is not very good.
So what we designed was an 8 megabit [indistinct], and then a cheap interface. Then the only download communication would be used for that path. Okay. Then why can’t we use Ethernet? Then if that is emulating the Ethernet interface, then the interface board is very cheap and we can get— Okay, that Ethernet interface just providing the satellite datastream, so nothing else. So in that case we can get probably very good performance from that board, that adapter. So we decided to emulate the Ethernet interface from the satellite adapter. And we about thirty pilot hardwares now, and we are running, and we are testing, and it’s running pretty well.
Malamud: So for a few hundred dollars for a K band satellite dish, which is cheap, an Ethernet card, which is cheap, and this satellite adapter that sits in the middle, you give your users 8 megabits of wide-area communications.
Murai: That’s true, yeah.
Murai: Class D. Currently Class D is routed to the satellite communications. And then on the way back, like an ACK type of communication, is now using terrestrial. So it’s a combination of satellite one way, and then terrestrial on the other way. So the other way it’s normal IP routing. It’s something with an assumption that your machine has a receiver, then you get the benefit from the satellite communication which is you know, a broadcast and multicast type of thing. And at the same time you have to a terrestrial IP link the WIDE Internet right now.
Malamud: And is that ISDN? Is that a [crosstalk] fast modem?
Murai: Any. Any kind of— We have developed IP over ISDN thing for the first stages of our project. So and any kind of—
Malamud: Now, you make extensive use of ISDN in WIDE, don’t you?
Malamud: That’s an important technology for you.
Murai: Yeah, right. In our country. And you know, the Japanese government is turning to start up the BISDN, FTTH in a pretty— Probably by the end of the century.
Malamud: But regular ISDN is out there, it’s just available. You go any place in Japan and you [crosstalk] can buy ISDN lines?
Murai: Yeah, any place, any place. And you know—
Malamud: Pay phones, and…
Murai: Yeah. Pay phones, yea. That’s an interesting portion. So if you visit Japan— Actually you did. But there you’ll see phones with an RJ-45 jack. So if you have a laptop equipped with an ISDN port, you can get probably 128 Kbps on a street corner. So you can [indistinct] from a street corner.
Malamud: Which you do on a regular basis.
Malamud: Well that’s fascinating. One of the other projects you’ve been working on is something that you call XITV. What’s XITV?
Murai: Yeah that’s a new one, and actually we are preparing for that, and then we have some demonstrations. It’s TV communications, or video communication, over the Internet [fame?]. But it’s currently a very simple thing. The only protocol the software is using is FTP. And—
Malamud: Now wait a minute. Let me just make sure I understand this right. Your transport layer for your television-over-the-Internet is…FTP.
Murai: FTP. Because we wanted to use anonymous FTP. Because— Okay. That is implemented and coded by Professor Yamamoto of Hokkaido University. So the basic idea of the software is very easy implementation and handling on the client side. So basically if you have X Window running, then you need nothing else, right. So what the software is doing is showing the video images to the X Window and passing the voice data to the normal—.au file, actually, but you know, if you have a Sun workstation then you can hear the voice continuously. And then of course if it’s over the Internet, and especially the Japanese Internet is not that high-bandwidth, then the software is decreasing the number of frames per second.
Malamud: But we’re talkin’ full-color, full-motion, if the bandwidth is there.
Murai: Well, full-color, full-motion, but the number of frames is small. It depends on the condition of end-to-end link. And also there are certain servers around, and they’re capturing live video images and storing that to disk, and then using the FTP protocol retrieving both voice data and TV images.
So that’s one. And then sometimes the [indistinct] conditions the link, but sometimes only one frame per five seconds or that kind of condition would happen. But then found out that that is good. I mean, normally good for the normal broadcasting of TV, like if you are watching the TV news, and if you can hear the voice and if you can get one frame per three seconds, then you can probably get most of the information out of the TV broadcast.
Malamud: Well that’s fascinating. What other things are you working on now?
Murai: A very important is handling global multilingual messages on email messages or Internet messages in general. We have a long history from the start of the network of handling of local characters, like know, kanji characters, Japanese characters. And that was a very important thing. And so we’ve just started the translation of the RFC documentation into Japanese. And that will be very important for promoting the technology of the Internet in our country, right. So when we started the email culture in our country, nobody was using it until we had developed handling of local characters on email and the related environments—operating systems and the email software, and everything.
Malamud: This is Internet Talk Radio, flame of the Internet. You’ve been listening to Geek of the Week. You may copy this program to any medium, and change the encoding, but may not alter the data or sell the contents. To purchase an audio cassette of this program, send mail to email@example.com.
Support for Geek of the Week comes from Sun Microsystems. Sun, The Network is the Computer. Support for Geek of the Week also comes from O’Reilly & Associates, publishers of the Global Network Navigator, your online hypertext magazine. For more information, send mail to firstname.lastname@example.org. Network connectivity for the Internet Multicasting Service is provided by MFS DataNet and by UUNET Technologies.
Executive Producer for Geek of the Week is Martin Lucas. Production Manager is James Roland. Rick Dunbar and Curtis Generous are the sysadmins. This is Carl Malamud for the Internet Multicasting Service, town crier to the global village.