@Mark Ollig
In the late 1960s, Professors Norman Abramson and Franklin Kuo at the college of engineering at the University of Hawaii created ALOHAnet, an early wireless data network using radio frequencies.
By June 1971, ALOHAnet had become operational, providing inter-island wireless access to the University of Hawaii’s central mainframe computer.
ALOHAnet’s use of randomized access allowed multiple users to share the same radio channel efficiently.
It also enhanced wireless data traffic management by enabling devices to transmit data immediately and resolving signal collisions through randomized retransmissions.
In 1972, Xerox Palo Alto Research Center (PARC) in California began developing the Alto computer, explicitly as part of their vision for “the office of the future,” which included wired networked personal computers and shared resources like printers.
That same year, Robert Metcalfe joined Xerox PARC to develop a wired network that linked Alto computers and shared devices such as printers.
Also in 1972, Metcalfe proposed his doctoral thesis at Harvard, focusing on connecting the Massachusetts Institute of Technology’s (MIT) mainframe computer to the Advanced Research Projects Agency’s (ARPANET) and analyzing its performance.
His initial thesis was rejected by his Harvard dissertation committee, which stated that “it wasn’t theoretical enough.”
To address this, Metcalfe studied Norman Abramson’s 1970 paper on the ALOHAnet system and incorporated its mathematical analysis of random access protocols into his revised thesis.
After visiting Hawaii to learn firsthand about ALOHAnet’s random access protocols, he constructed mathematical models to improve the academic accuracy of his work.
Metcalfe then revised his thesis, “Packet Communication,” which was accepted, and he earned his PhD in 1973.
While at Xerox PARC May 22, 1973, he wrote an internal memo officially titled “Alto Ethernet,” sometimes informally referred to as “Ether Acquisition” in later sources. In it, he proposed a shared 50-ohm coaxial cable to connect devices like the Alto and PDP-11 in a tree-structure topology.
The PARC internal memo begins: “The ether network. We plan to build a so-called broadcast computer communication network, not unlike the ALOHA system’s radio network, but specifically for in-building minicomputer communication.”
In November 1973, Xerox PARC created the first Ethernet prototype using a 50-ohm coaxial cable.
This prototype local area network (LAN) achieved a data transmission speed of 2.94 Mbps.
In 1973, Robert Metcalfe coined the term “Ethernet,” inspired by the “luminiferous ether,” which was believed to carry light waves.
He used it to describe the shared coaxial cable that transmits data between computers, likening it to how the ether carried light to all.
CSMA/CD (Carrier Sense Multiple Access with Collision Detection) is the protocol Metcalfe used in early Ethernet networks to manage access to a shared medium and detect data collisions.
Devices listen for traffic before transmitting, and if a collision occurs, the protocol uses a randomized backoff algorithm to retry transmission after a delay.
The transition from coaxial began in the late 1980s with the introduction of 10BASE-T in 1990.
This standard utilized Category 3 unshielded twisted-pair (UTP) cabling in a star topology, allowing each device to connect to a central hub or switch, which provided more flexibility and cost-effectiveness.
Vampire taps are physical connectors that attach computers and printers to Ethernet cables without interrupting the network.
They work by piercing (biting) into the coaxial cable’s insulation to connect directly to the copper conductor without cutting the main cable.
Vampire taps remind me of my days splicing telephone wires with 3M™ Scotchlok™ connectors at the local telephone company.
These connectors allowed telephone wires to be inserted with their insulation intact, speeding up the splicing process.
I mostly used the UR (red) connector, which has three ports and is used for splices joining two or three cut solid copper wires ranging from 19 to 26 AWG. It is a gel-filled connector designed to be durable and moisture-resistant for long-term reliability.
The UG (green) connector is specifically designed as a tap splice; it allows a new telephone wire to be connected to a continuous, uncut line, making it ideal for tapping into existing circuits.
For thicker wires, the UO (orange) connector, model U1O, is a gel-filled, moisture-resistant butt splice for two wires ranging from 18 to 14 AWG.
It’s been more than 30 years since I last spliced telephone wires using 3M Scotchlok UR connectors, and many of those splices are still in service.
First, I’d prepare the “joint,” (the specific point where the wire ends meet to be connected) by twisting the wires together one full turn.
Then, I would cut the wire ends evenly to about one inch and not strip the insulation, as the connector is designed for insulated wires.
Holding the UR connector with its red button facing down, I would insert the unstripped wires all the way into the individual ports: two ports for two wires, and three ports for three wires.
To complete the splice, I’d firmly crimp the red button using a Scotchlok E-9 series tool – I often called it the “Scotchlocker.” This action caused the sharp metal plate inside the UR to “bite” through the insulation and into the copper wires, creating a secure electrical connection.
The splicing procedure would be much the same for the UG, UY, and UO connectors.
In 1975, Xerox filed a patent application titled “Multipoint data communication system with collision detection” (US Patent 4,063,220, granted in 1977).
In a 2019 United States Patent and Trademark Office (USPTO) “Journeys of Innovation” interview, Robert Metcalfe credited ALOHAnet by stating, “And the key idea was to use randomized retransmissions.”
And so, ALOHAnet assisted in the birth of Ethernet’s LAN.
