AlgoPlusAlgoPlus
Learn/Networking
Lesson

ALOHA (Pure & Slotted)

The simplest way to share one channel: transmit whenever you like, detect collisions, and retry after a random pause.

8 min read Watch it move Build it

ALOHA is the original random-access scheme — born in 1970s Hawaii to link island terminals over one radio channel. The rule is almost reckless: transmit whenever you have a frame, with no checking first. If two frames overlap in time they collide and both are lost, so each sender waits a random delay and tries again. Crude as it sounds, it's the ancestor of every Wi-Fi and Ethernet access method.

Pure ALOHA and the vulnerable period

In Pure ALOHA a station may start at *any* instant. That's the weakness. Your frame is destroyed not only by a frame that starts *during* yours, but also by one that started just *before* and is still on the air. So your frame is vulnerable for two full frame times — the frame before you and the frame after you can each clobber it.

  1. 1A station with data to send transmits immediately — no listening.
  2. 2It then waits for an acknowledgement (or a timeout).
  3. 3If no ack arrives, it assumes a collision and schedules a retry.
  4. 4It waits a random backoff so the two collided stations don't retry in lockstep.
The throughput ceiling
Model frame arrivals as Poisson with load G. Pure ALOHA's success rate is S = G · e^(−2G), which peaks at G = 0.5 giving S = 1/(2e) ≈ 0.184. So Pure ALOHA can use at most about 18% of the channel — the rest is burned on collisions and idle time.

Slotted ALOHA cuts the window in half

Slotted ALOHA adds one rule: every frame must begin only at a fixed slot boundary, where one slot equals one frame time. Now two frames either start in the *same* slot and collide completely, or land in different slots and never touch — there is no partial overlap. The vulnerable period shrinks from two frame times to one.

Halving the window doubles throughput
With a one-slot window the success rate becomes S = G · e^(−G), peaking at G = 1 with S = 1/e ≈ 0.368. Synchronising to slots roughly doubles the best-case usage to about 37%.
Slots need synchronisation
Slotted ALOHA assumes every station shares a common clock to agree on slot boundaries. That coordination cost is the price of the doubled throughput — and neither scheme listens before sending, which is exactly what CSMA fixes next.
OperationTimeSpace
Pure ALOHA · peak at G = 0.5, window = 2 frame timesS = G·e^(−2G)≈ 18.4% max
Slotted ALOHA · peak at G = 1, window = 1 frame timeS = G·e^(−G)≈ 36.8% max
Check yourself
Why does Slotted ALOHA achieve about double the throughput of Pure ALOHA?