Multiple Access Protocols

Multiple access protocols coordinate how many nodes share a single broadcast communication channel so that their transmissions are delivered efficiently and fairly without permanently destroying one another in collisions.

Definition

A multiple access protocol is a medium access control mechanism that determines how nodes sharing a single broadcast channel decide when to transmit, so as to use the channel efficiently while resolving or avoiding collisions among simultaneous transmissions.

Scope

This topic covers the three families of medium access control protocols: channel-partitioning protocols (time-, frequency-, and code-division multiplexing) that statically divide the channel; random-access protocols (ALOHA, slotted ALOHA, CSMA, CSMA/CD, CSMA/CA) that let nodes contend and recover from collisions; and taking-turns protocols (polling and token passing) that grant access in turn. It analyzes their efficiency, fairness, and behavior under load. It excludes the higher-layer switching and forwarding that operate above link access.

Core questions

• What problem do multiple access protocols solve on a shared broadcast channel?
• How do channel-partitioning, random-access, and taking-turns approaches differ?
• How do ALOHA and CSMA variants detect or avoid and recover from collisions?
• What determines the maximum efficiency of a random-access protocol?
• How do these protocols balance efficiency against fairness and simplicity?

Key concepts

• broadcast channel
• collisions
• channel partitioning (TDMA, FDMA, CDMA)
• ALOHA and slotted ALOHA
• carrier-sense multiple access (CSMA)
• CSMA/CD and CSMA/CA
• binary exponential backoff
• polling and token passing
• channel efficiency

Key theories

Random access and collision recovery

Random-access protocols let nodes transmit without coordination and recover from the resulting collisions; ALOHA and slotted ALOHA bound efficiency, while carrier sensing (CSMA) and collision detection or avoidance reduce wasted time on busy channels.

Channel partitioning

Time-, frequency-, and code-division multiple access divide a shared channel into non-overlapping portions assigned to nodes, giving collision-free, predictable access that is efficient under heavy, steady load but wasteful when only a few nodes are active.

Taking-turns protocols

Polling and token-passing schemes coordinate access by granting the channel to one node at a time, avoiding collisions and bounding delay, at the cost of polling overhead or vulnerability to a lost token or failed coordinator.

Clinical relevance

Multiple access protocols govern how shared media work in practice: CSMA/CD shaped classic Ethernet, CSMA/CA underlies every Wi-Fi network, and channel-partitioning schemes such as TDMA and CDMA are central to cellular systems. Their efficiency and fairness properties determine throughput and latency on any medium where devices contend for a shared channel.

History

The ALOHA system, built by Norman Abramson at the University of Hawaii around 1970, introduced random access over a shared radio channel and inspired the analysis of slotted ALOHA. Metcalfe and Boggs adapted carrier sensing and collision detection into Ethernet in 1976. Token-passing and polling schemes were developed in parallel, and modern wireless systems combine partitioning and contention-based access.

Key figures

• Norman Abramson
• Robert Metcalfe
• Leonard Kleinrock

Seminal works

• abramson1970
• metcalfe1976
• kurose2021

Frequently asked questions

Why do shared channels need a multiple access protocol at all?

On a broadcast channel, if two nodes transmit at the same time their signals interfere and both frames are typically lost (a collision). A multiple access protocol decides who may transmit and when, so the channel is used productively instead of being wasted on repeated collisions or left idle.

What is the difference between CSMA/CD and CSMA/CA?

Both sense the channel before transmitting. CSMA/CD, used in wired Ethernet, also detects collisions while transmitting and aborts immediately. CSMA/CA, used in Wi-Fi, instead tries to avoid collisions in advance using backoff and acknowledgments, because a wireless node generally cannot reliably detect a collision while sending.