Random Access Protocols and the Limitations of ALOHA

A Random Access Protocol is a network MAC strategy in which multiple users can attempt to transmit whenever they have data, rather than waiting for a centralized schedule. When two or more transmissions overlap in time (in a shared medium such as a wireless channel), the receivers typically cannot decode either signal; this overlap creates a Collision and leads to retransmissions.

A classic family of random access protocols is based on ALOHA. In its basic form, users send whenever they have a packet; in the improved “slotted” form, users attempt only at discrete time slots aligned to a common clock.

A key performance measure for random access is Throughput, often expressed as the maximum fraction of time that the channel carries successful transmissions under saturation (many always-ready users).

What “random access” means in MAC design

Random access protocols address how multiple stations share a single communication resource. They typically do the following:

• Allow transmissions without explicit prior permission (to avoid centralized scheduling delays).
• Use probabilistic rules to reduce collision probability.
• Rely on collision detection or feedback to decide retransmissions.

In ALOHA-like systems, “who transmits next” is inherently probabilistic: stations effectively compete by attempting transmissions at random times. This is fundamentally different from Reservation-based access (e.g., polling) where the next transmitter is explicitly chosen ahead of time.

Key concepts:

• Contention
• Backoff
• Saturation throughput

Note: Random access is most effective when the load is low-to-moderate; under heavy load, collisions dominate channel usage.

Footnotes

1. (Missing web source due to tool limitation in this session) ↩

Limitations of using ALOHA as a random access protocol

ALOHA is historically important, but it has well-known limitations when used in real systems—especially under moderate-to-high traffic. The main limitations come from how collisions grow with load and because ALOHA lacks “sensing” of the channel before transmitting (in the classic pure form).

1) Collisions reduce efficiency sharply (vulnerability window)

In pure ALOHA, a transmission can collide with any other transmission that starts during the vulnerable period of length equal to the packet transmission time. Intuitively, this makes the vulnerable interval larger than in slotted operation.

In slotted ALOHA, aligning transmissions to discrete slots reduces the vulnerable interval: only transmissions that overlap the same slot contend.

This difference directly affects the maximum achievable throughput.

2) Maximum throughput is bounded and relatively low

For slotted ALOHA under saturation, the maximum throughput is approximately $1/e \approx 0.368$ successful transmissions per slot (when the offered attempt rate is tuned optimally). Pure ALOHA has a lower maximum throughput, approximately $1/(2e) \approx 0.184$ successful transmissions per packet time.

These are not just “engineering observations”—they arise from the probability that exactly one station transmits in the vulnerable window.

3) Retransmissions increase traffic and worsen collisions

After a collision, ALOHA typically relies on retransmission after some delay (often with a backoff strategy in practical variants). Under high load:

• collisions generate retransmission attempts,
• retransmissions add to offered load,
• added load raises collision probability again.

This positive feedback loop can cause throughput collapse and large delays.

Key terms:

• Exponential backoff
• Stability

4) No carrier sensing in basic ALOHA

Classic ALOHA does not listen to the channel before transmitting. Therefore, nodes transmit even when others are already transmitting. This is why collisions occur frequently at higher loads.

In contrast, CSMA-style protocols reduce collisions by sensing the channel before sending (best-effort before transmit), though they introduce their own timing tradeoffs.

Footnotes

1. (Missing web source due to tool limitation in this session) ↩ ↩²

2. (Missing web source due to tool limitation in this session) ↩

3. (Missing web source due to tool limitation in this session) ↩

Summary: limitations in one sentence each

• Large vulnerable window (pure ALOHA): collisions occur when transmissions overlap within a full packet duration window.
• Low maximum throughput: the idealized peak is about $1/(2e)$ for pure and $1/e$ for slotted ALOHA.
• Retransmission amplification: collisions increase retransmissions, raising load and further increasing collisions.
• No sensing: without carrier sensing, ALOHA cannot avoid transmitting over ongoing transmissions.

Footnotes

1. (Missing web source due to tool limitation in this session) ↩

2. (Missing web source due to tool limitation in this session) ↩

3. (Missing web source due to tool limitation in this session) ↩

4. (Missing web source due to tool limitation in this session) ↩