Token Ring Networks and IEEE 802.5

Every networking protocol is ultimately defined by its frame format—the precise structure of bits that travel across the physical medium. Token Ring's frame format reflects its design philosophy: careful engineering for reliability, explicit support for priority and reservation, and sophisticated error handling baked into every transmission.

Unlike Ethernet's relatively simple frame structure, Token Ring frames contain multiple control fields that support its deterministic operation. Each field serves a specific purpose, from the special delimiter patterns that use code violations to the dual-redundant status bits that survive CRC verification.

In this page, we'll dissect the Token Ring frame byte by byte—often bit by bit—revealing how each component contributes to the protocol's reliability and performance. This deep understanding is essential for troubleshooting, analysis, and appreciating why Token Ring was chosen for mission-critical environments.

Token Ring defines two primary frame types: the minimal token frame and the full data/command frame. Both share common delimiters but differ in content and purpose.

Token Frame (3 bytes):

The token is the control frame that circulates when no station is transmitting data:

Data/Command Frame (variable length):

When a station captures the token and transmits data, the frame expands to include addressing, data, and error checking:

Key Structural Differences from Ethernet:

Comparing Token Ring to Ethernet frames reveals important design differences:

1. No Preamble: Token Ring uses code violations in delimiters instead of a preamble for synchronization
2. Access Control Field: Carries priority and reservation bits (Ethernet has no equivalent)
3. Frame Control Field: Explicitly distinguishes MAC and LLC frames
4. Frame Status Field: Provides immediate acknowledgment (Ethernet ACKs are at higher layers)
5. Variable Maximum Size: MTU depends on Token Holding Time, not collision detection

These differences reflect Token Ring's deterministic, prioritized design versus Ethernet's simpler, best-effort approach.

The Starting Delimiter (SD) and Ending Delimiter (ED) are remarkable engineering achievements—they mark frame boundaries using patterns that cannot appear in normal data.

The Code Violation Technique:

Differential Manchester encoding defines two normal symbols:

• Data 0: Transition at the start of the bit period
• Data 1: No transition at the start of the bit period

Both always have a transition in the middle of the bit period (for clock recovery).

Code violations are symbols that intentionally break this rule:

• J (Violation): High for entire bit period (no mid-bit transition, holds high)
• K (Violation): Low for entire bit period (no mid-bit transition, holds low)

Since J and K never appear in normal data, the following sequences are unambiguously frame delimiters—no escape characters or bit stuffing needed!

The I Bit (Intermediate Frame):

The I bit supports frame fragmentation at the MAC layer. If a station's data exceeds the maximum frame size, or if the Token Holding Time is about to expire:

• First frame: I=1 (more frames coming)
• Middle frames: I=1 (more frames coming)
• Last frame: I=0 (sequence complete)

Receivers use the I bit to reassemble fragmented data or to know when transmission is complete.

The E Bit (Error Detected):

The E bit provides hop-by-hop error signaling:

1. Transmitter sends frame with E=0
2. Each forwarding station checks frame integrity (FCS)
3. If any station detects an error, it sets E=1 before forwarding
4. By the time the frame returns to the sender, E indicates whether any station found corruption

This is faster than waiting for FCS verification at the destination—errors are flagged as early as the first detecting station.

Why Code Violations Are Superior to Bit Stuffing:

HDLC and similar protocols use bit stuffing: insert a 0 after five consecutive 1s to prevent the flag pattern (01111110) from appearing in data. This has overhead and complexity.

Token Ring's code violations:

• Add zero overhead (no extra bits inserted in data)
• Are unambiguous (impossible to confuse with data)
• Simplify receiver logic (no unstuffing required)
• Provide reliable synchronization (receiver re-syncs at each delimiter)

The Access Control (AC) field is the most important byte in Token Ring operation. It controls who may transmit, at what priority, and provides critical ring maintenance functionality—all in just 8 bits.

AC Field Structure: PPP T M RRR

The T Bit: Token to Frame Transformation:

The T bit is the central mechanism of token capture. When a station with data receives a free token:

1. Station examines incoming AC byte: T=0, Priority ≤ station's priority
2. As the AC byte is forwarded, station flips T from 0 to 1
3. Instead of forwarding rest of token (ED), station transmits its frame
4. The 3-byte token has become the header of a data frame

This single bit change—performed in real-time as bits stream through—is the essence of Token Ring access control.

The M Bit: Monitor Cleanup:

The Monitor bit prevents orphan frames from circulating forever:

1. Every frame/token passing the Active Monitor has M set to 0
2. After a complete ring circulation, a frame returns to the AM with M=0
3. If the AM sees M=1, the frame has already passed once without being stripped
4. This happens when the originating station failed before stripping
5. AM strips the orphan and generates a new token

Priority and Reservation Interaction (Preview):

The P and R fields implement Token Ring's priority system, which we'll explore in depth in the next section. Briefly:

• P (Priority) controls current token access rights
• R (Reservation) allows stations to request future high-priority access
• Together, they ensure time-critical traffic gets precedence while preventing starvation

The Frame Control (FC) field immediately follows the Access Control field in data/command frames. It identifies the frame type and provides additional control information.

FC Field Structure: FF ZZZZZZ

MAC Frames (FF = 00):

MAC frames carry ring management traffic—they're generated and processed by the Token Ring adapter itself, not by upper-layer protocols. The Z bits specify the MAC frame type:

LLC Frames (FF = 01):

LLC (Logical Link Control) frames carry user data—IP packets, NetBIOS traffic, and other higher-layer protocols. The 802.2 LLC sublayer provides:

• Service Access Point (SAP) addressing
• Connectionless and connection-oriented services
• SNAP (Sub-Network Access Protocol) for protocol identification

For LLC frames, the Z bits are typically 000000, and the LLC header follows immediately after the source address.

Frame Type Significance:

The clear separation between MAC and LLC frames allows:

1. Adapter-level processing: MAC frames can be handled entirely in hardware/firmware
2. Efficient delivery: LLC frames are passed to higher layers without adapter interpretation
3. Priority differentiation: Some implementations give MAC frames higher implicit priority
4. Protocol isolation: Ring management doesn't interfere with user data

Token Ring uses 48-bit (6-byte) MAC addresses, sharing the same addressing architecture as Ethernet. However, important differences exist in bit ordering and special address handling.

Address Format:

Destination Address (DA):

The 6-byte Destination Address identifies the intended recipient(s):

• Individual Address: Unicast to a single station (I/G=0)
• Group Address: Multicast to a group of stations (I/G=1)
• Broadcast Address: FF-FF-FF-FF-FF-FF—all stations receive
• Functional Address: Group addresses with specific meanings

Source Address (SA):

The 6-byte Source Address identifies the transmitting station:

• Always an individual address (I/G bit is 0 for SA)
• The U/L bit in SA has special meaning: Routing Information Indicator (RII)
  • RII=0: No routing information field present
  • RII=1: Routing information field follows SA (for source routing bridges)

Functional Addresses (Special Group Addresses):

Token Ring defines Functional Addresses for ring management entities. These are group addresses (I/G=1) with specific bit patterns in bytes 4-6:

The Information field carries the actual payload, and the Frame Check Sequence (FCS) provides error detection. Both are critical but often overlooked in frame analysis.

Information Field:

The Information field (also called the data field or INFO field) contains:

For MAC frames: The specific MAC frame content varies by frame type. For example:

• Beacon frames include the sender's NAUN address and beacon type
• Active Monitor Present frames include the sender's Stored Upstream Address
• Error reporting frames include counter values

For LLC frames: The LLC header followed by user data:

+------+------+------+------+------+...
| DSAP | SSAP | Ctrl | Data | Data |...
+------+------+------+------+------+...
|  1   |  1   | 1-2  | variable    | bytes

Where:

• DSAP: Destination Service Access Point (identifies receiving protocol)
• SSAP: Source Service Access Point (identifies sending protocol)
• Control: LLC control field (1 or 2 bytes depending on frame type)
• Data: Upper-layer protocol data (IP packet, NetBIOS data, etc.)

Maximum Information Field Size:

Frame Check Sequence (FCS):

The 4-byte FCS provides robust error detection using CRC-32 (Cyclic Redundancy Check).

CRC-32 Polynomial: x³² + x²⁶ + x²³ + x²² + x¹⁶ + x¹² + x¹¹ + x¹⁰ + x⁸ + x⁷ + x⁵ + x⁴ + x² + x + 1

FCS Coverage: The FCS is calculated over:

• Frame Control (FC)
• Destination Address (DA)
• Source Address (SA)
• Information field

NOT covered by FCS:

• Starting Delimiter (SD) — Contains code violations
• Access Control (AC) — May be modified in transit (reservation bits)
• Ending Delimiter (ED) — Contains code violations and status bits
• Frame Status (FS) — Set by receiver after FCS calculated

CRC Verification Process:

1. Transmitter calculates CRC over covered fields
2. CRC value is appended as FCS (transmitted MSB first)
3. Each forwarding station may verify CRC
4. If any station detects CRC error, it sets E=1 in ED
5. Receiving station verifies CRC before accepting data
6. Returning frame reaches sender with E bit indicating any errors

The Frame Status (FS) field is Token Ring's built-in acknowledgment mechanism—it provides immediate feedback about frame delivery without requiring a separate response frame.

FS Field Structure: A C r r A C r r

Interpreting Frame Status:

When the transmitting station receives its frame back, it examines FS to determine delivery outcome:

Immediate Acknowledgment Benefits:

1. No Separate ACK Frame: Reduces ring overhead
2. Guaranteed Feedback: Every frame transmission gets status
3. Rapid Error Detection: Know immediately if destination unreachable
4. Application Awareness: Higher layers can react to delivery failure

FS in Broadcast/Multicast:

For broadcast and multicast frames (multiple destinations):

• Each destination that recognizes address sets its local A bit
• Each destination that copies sets its local C bit
• The FS bits combine (OR logic) as frame circulates
• Final FS indicates whether at least one station copied

This doesn't tell you which stations copied (or how many), just that at least one did. For detailed multicast acknowledgment, higher-layer protocols are needed.

We've dissected every field in the Token Ring frame format, from code violation delimiters to the duplicated status bits. Let's consolidate the essential concepts:

What's Next:

With the frame format mastered, we'll explore Token Ring's priority mechanism—how the P and R bits in the Access Control field enable differentiated service, ensuring time-critical traffic gets precedence while maintaining fairness for lower-priority stations.