mechgam@tness1.UUCP (08/18/87)
Could anyone give me a brief history of how LANs came about, or direct me to some reading material? Also, is there really such a thing as Aloha Net? Gary Mink -- Gary A. Mink Southwestern Bell Telephone Co. (713) 521-6958 tness1!mechgam
socha@drivax.UUCP (Henri J. Socha (x6251)) (08/21/87)
Yes there was (is?) Alohanet. It was one of the first networks that ended up growing into Ethernet. (i.e. one of the conceptual bases'). It was developed by the U of Hawaii? to link campuses on different islands. The system was similar to Ethernet but worked over radio (ether) waves. Others can give more details I am sure. I was not involved but knew of/read of it back when "lans" were in their infancy. Also, you can't say "when they started" because they evolved. Let me give you an example: I worked at Datapoint Corp. in San Antonio in the R&D deptartment as they developed ARCnet. I was there as ARCnet was developed. In 75 Datapoint had a networking system (some would call it that) but I wouldn't. They took their modems (MSI design) and replaced the DAA (phone) interface with a simple driver board (easy to do the way it was designed). This driver had negligible load when disabled and therefore you could have many of these multidropped. So, they did and used telephone wires from each machine to a patch panel and then connected them all together (in parallel) right there). It was a single user multi-drop system. If you want the "Wire" listen and then place carrier on it and listen some more to be sure nobody else wanted it. Then, transmit and you have the complete wire until you were done marked by dropping carrier. This system was used for sharing printers and for file transfer between machines. The data rate was 76,000 baud the max rate of the modems. This was not too shabby considering the time/hardware/cost involved to set it up, etc. It was never sold (that I know of) just a local hardware/software hack to provide internal facilities at Datapoint. At Datapoint, the concept of Local Area Networking grow out of this, research literature, and other dark and mysterious sources. At the time ARCnet was being developed, we were reading the XEROX PARC reports on the original Ethernet. We analysed their math, etc. and decided that their system (CSMA/CD) just was not up-to par. The idea that a system would slow down and fail to send packets if heavily loaded seemed preposterous. Anyway, ARCnet developed. ARCnet uses isochronous transmission (re-sync per character) at 2.5 meg-baud on the wire for an effective rate of 2.5*(8/11)=~1.8 mega(data)baud in a message. It uses 93 ohm coax but has been implemented using open-air! high-power LED transmitters, micro-wave, fiber cable, and other systems. It is a Token bus system with nodes connected to hubs in a star configuration with hubs linked non-cyclically. But, the data (token) transfer is through a logical ring formed by the sequencing of addresses on the bus. i.e. if node A has the token it will give it to B at address higher than A (with full wrap from highest address to lowest). It was designed and implemented by a small group with input/discussion among a few others of us. The only major mistake Datapoint made was to keep it proprietary from '77 until the SMC 9026 chip was developed. By the way ARCnet still has a lot of followers especially in Europe, performs as well as 10Mbaud Ethernet (please no flames unless you tested it!), and its interface is simpler and easier to programme. So there!-) -- UUCP:...!amdahl!drivax!socha WAT Iron'75 "Everything should be made as simple as possible but not simpler." A. Einstein
bob@uhccux.UUCP (08/30/87)
The Aloha System project started in September 1968, the brainchild of Norm Abramson here at the University of Hawaii. The original idea was to test the idea of remote computer access via satellite, using instead a much simpler to implement UHF radio network. The starting point in the design was the question: "Given the availability of a fixed amount of communications capacity, how does one employ this capacity to provide effective communication from remote users to a central machine". The project was primarily funded by DARPA, through NASA and the U.S. Air Force Office of Aerospace Research. By June 1971 the central UHF station was in place and tested; around the summer of 1972 the complete system was in operation. DARPA funding ceased fairly shortly thereafter, since the original concept had been proven. The Aloha System setup continued in operation until the late 1970s when it essentially died from lack of interest and lack of funding to maintain the thing. It used two 24,000 baud channels in the UHF band (Norm has said that in retrospect he should have used only one). One channel was used by the remote stations (usually terminals) talking two the central MENEHUNE (a packet-switched multiplexor based upon early ARPANET IMP designs); the other channel carried traffic from the MENEHUNE. A considerable amount of theoretical analysis was done before, during, and after the system was operational...which lead to some classic papers. One of the visiting scientists went on to develop a variant of the Aloha System using carrier sense on a cable, which he then called Ethernet.