In computer networks, Controlled Access Protocols (CAPs) manage how multiple devices share a common communication medium. These protocols control the right to transmit in such a way that collisions are avoided, efficiency is maintained, and fair access is ensured.
Note: Unlike random access methods (e.g., ALOHA, CSMA), in controlled access, stations do not transmit at will. Instead, they follow a systematic method to determine which station has the right to send at a given time.
What is Controlled Access?
In Controlled Access, before transmitting data, a station must first obtain permission to access the channel. At any given time, only one node can send data, preventing collisions. This coordination is achieved using one of three methods:
Reservation
Polling
Token Passing
Reservation
In this method, stations reserve slots in advance before transmission. The timeline has two parts:
Reservation Interval (fixed length): divided into slots, one per station.
Data Transmission Period (variable length): reserved stations send data in order.
Reservation
Example: If there are M stations, the reservation interval is divided into M slots. Each station signals intent by inserting a bit into its slot. Then, only the reserved stations transmit in the data phase.
Advantages
Predictable access time and throughput.
Priorities can be set for faster access.
Supports QoS (Quality of Service) for applications like real-time audio/video.
Efficient bandwidth utilization.
Disadvantages
Performance decreases under light loads (wasted slots).
High dependency on controlled synchronization.
Polling
Here, a controller (primary station) polls each node (secondary station) in sequence, granting permission to send. The controller sends a message with the address of a station. The addressed station responds:
Sends data (if available)
Polling
Sends a NAK (poll reject) if no data is pending.
Polling
Advantages
Predictable access times and bandwidth.
High efficiency, since no slot is wasted.
Priorities can be assigned.
Disadvantages
Extra overhead due to polling messages.
High dependence on controller reliability.
Turnaround time increases under light loads.
Efficiency formula
\eta = \frac{T_t}{T_t + T_{poll}}
where:
T_t =data transmission time
T_{poll} = polling overhead time
Token Passing
In Token Passing, stations are logically connected (ring or bus). A special frame called a token circulates among stations:
The token represents permission to transmit.
A station can send data only when it holds the token.
If no data is ready, the token is simply passed on.
Token PassingCommon Implementations
Token Ring – token passed in a ring topology.
Token Bus – stations use a bus medium but pass the token in a predefined logical order.
Advantages
Excellent performance under high loads.
Built-in error recovery and debugging features.
Predictable throughput.
Disadvantages
Expensive setup and hardware.
Token management issues (lost/duplicate tokens, station failures).
Complex implementation.
Performance Parameters
Delay: Average delay = a/N, where a = T_p/T_t.
Throughput:
1. For a<1a < 1a<1: S = \frac{1}{1 + a/N} 2. For a>1a > 1a>1: S = \frac{1}{a(1 + 1/N)}