Frame transmission and Reception

 The frame transmission sequence is summarized in Figure 8.4(a). When a frame is to

be transmitted, the frame contents are first encapsulated by the MAC unit into the
format.
 To avoid contention when other transmissions on the medium, the MAC unit first
monitors the carrier sense signal and, if necessary, defers to any passing frame. After a
short additional delay (known as the interframe gap) to allow the passing frame to be
received and processed by the addressed station(s), transmission of the frame is
initiated.
 As the bitstream is transmitted, the transmitter simultaneously monitors the received
signal to detect whether a collision is occurring. Assuming a collision has not been
detected, the complete frame is transmitted and, after the FCS field has been sent, the
MAC unit awaits the arrival of a new frame, either from the cable or from the LC
sublayer within the station.
 If a collision is detected, the transmitter immediately turns on the collision detect signal
and enforces the collision by transmitting the jam sequence to ensure that the collision
is detected by all other stations involved in the collision. After the jam sequence has
been sent, the MAC unit terminates the transmission of the frame and schedules a
retransmission attempt after a short randomly-computed interval.

 Figure 8.4(b) summarizes the frame reception sequence. The MAC unit first detects the
presence of an incoming signal and switches on the carrier sense signal to inhibit any
new transmissions from this station.
 The incoming preamble is used to achieve bit synchronization and, when the start-of-
frame delimiter has been detected, with an IEEE802.3 LAN, the length indicator is read
and used to determine the number of bytes that follow.
 The frame contents including the destination and source addresses are then received
and loaded into a frame buffer to await further processing.
 The received FCS field is first compared with the computed FCS and, if they are equal,
the frame content is further checked to ensure it contains an integral number of bytes
and that it is neither too short nor too long.
 If any of these checks fail then the frame is discarded. If all checks pass, then the
destination address is read from the head of the frame and, if the frame is intended for
this station — that is, the address of the station is the same as that in the frame or, if it
is a group address, the station is a member of the specified group — the frame contents
are passed to the LC sublayer for processing.
 Multimedia Information Networks

8.6 Token ring

 Token ring networks are also used in technical and office environments and, in addition,
in industrial environments. As we can deduce from the previous section, with an
Ethernet LAN the time to transmit a frame is nondeterministic since, during heavy load
conditions when collisions are likely to be frequent, the transmission of a frame may be
delayed and, in the limit, unsuccessful.
 Although in an office environment this is a degree acceptable, in industrial
environments such as manufacturing and process control, it is unacceptable.
 Hence although token ring LANs also use a high bit rate shared/broadcast transmission
medium, in order to provide a deterministic service, they utilize a completely different
MAC method.
 Also, because the type of information — and hence frame contents — to be transmitted
in an industrial application may have different levels of importance, frames can be
assigned different priorities.
 The MAC method, therefore, also contains a priority control algorithm to ensure higher
priority frames — for example those containing alarm messages — are transmitted
before lower priority frames. We shall discuss this and other issues separately.

Control token

 In a token ring LAN, all the stations are connected together by a set of unidirectional
links in the form of a ring and all frame transmissions between any of the stations take
place over it by circulating the frame around the ring.
 In its basic form, only one frame transfer can be in progress over the ring at a time.
When the ring is first initialized, a single control token (frame) is generated and, in the
absence of any frames to transmit, this continuously circulates around the ring.
 Then, when a station generates a frame to send, it intercepts the MAC unit within each
station area as illustrated in Figure 8.5.
 Whenever a station wishes to send a frame, it first waits for the token. Upon receipt of
the token, it initiates transmission of the frame, which includes the address of the
intended recipient at its head.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

 The frame is repeated (that is, each bit is received and then retransmitted) by all the
stations in the ring until it circulates back to the initiating station, where it is removed.
In addition to repeating the frame, each station reads and stores the frame contents.
 The intended recipient — indicated by the destination address in the frame header —
retains a copy of the frame and indicates that it has done this by setting the response
bits at the tail of the frame.
 A station releases the token in one of two ways depending on the bit rate (speed) of the
ring. With slower rings (4 Mbps), the token is released only after the response bits have
been received.
 With higher speed rings (16 Mbps), it is released immediately after transmitting the last
bit of the frame. This is known as early (token) release and, as we shall explain later,
this is done to improve the level of utilization of the ring.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Wiring configurations

 A small token ring network is shown in Figure 8.6(a). As we can see, in this
configuration is a single hub — also known as a concentrator and located in, say, the
wiring closet of an office — to which all the stations are attached by a cable containing
two twisted-pairs, one for each direction of transmission.
 Within the hub, associated with each station is a station coupling unit (SCU) and, as we
show in the figure, these are interconnected so that all the stations are interconnected
together to form a unidirectional ring.

Figure 8.6(a) Token ring wiring configuration- Single hub

 An SCU is shown in Figure 8.6(b) and, as we can see, it contains a set of relays and
additional electronics to drive and receive signals to and from the cable. The relays are
so arranged that whenever a station is switched off, the SCU is in the bypassed state
and a continuous transmission path through the SCU is maintained.
 The insertion of a station into the ring is initiated when the station is switched on. A
separate pair of wires in the drop cable is used to pass power from the station to the
SCU and, when activated, the relays change position so that the station becomes
inserted.
 When the SCU is in the inserted state, all signals from the ring are routed through the
MAC unit of the station. The receive/transmit electronics in the MAC unit then either
simply read and relay (repeat) the received signal to the transmit side, if this station is

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

not the originator of the frame, or remove the received signal from the ring, if it initiated
the transmission.

Figure 8.6(b) Token ring wiring configuration

 The use of two pairs of relays connected in this way means that the MAC unit can detect
certain open-circuit and short-circuit faults in either the transmit or the receive pair of
signal wires.
 Also, in the bypassed state, the MAC unit can conduct self-test functions, since any
data output on the transmit pair is looped back on the receive pair.
 Each station is connected to the SCU by a shielded twisted-pair (STP) cable containing
three twisted-pair wires; one for transmission, one second for reception, and the third
to supply power to the SCU.
 As we show in Figure 8.6(c), larger configurations are formed by interconnecting
multiple nodes/concentrators together by means of either an STP or optical fibre trunk
cable.
 In this case, associated with each hub is a second relay unit known as a trunk coupling
unit (TCU). This has the same function as an SCU except that the power to activate the
relays — and hence insert the TCU and its attached stations into the ring — is from the
power supply of the hub/concentrator.
 In this way, should a hub fail, the remainder of the ring will continue functioning.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Figure 8.6(c) Multiple hubs/concentrators

Frame formats

In addition to the control token and information frames, additional frames are used for various
ring management functions including the selection of an active ring monitor. Collectively,
these are known as MAC frames.

Two basic frame formats in use. One is for the control token and the other for normal frames.
The control token is the means by which the right to transmit is (as opposed to the normal
process of repeating) is passed from one station to another, whereas a normal frame is used by
a station to send either data or MAC information around the ring. The format of the two types
of frame are given in Figure 8.7 together with the bit sequence used for each field.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Figure 8.7: Token ring network frame formats and field descriptions(a) token format; (b) frame
format; (c) field descriptions.

 The start delimiter (SD) and end delimiter (ED) fields are special sequences used to achieve
data transparency. They exploit the symbol encoding method used on the cable medium: all
information bits transmitted on the medium are differential Manchester encoded, except for the
bits in the SD and ED fields.
 In contrast, the J and K symbols depart from normal encoding rules, being used instead to
represent constant levels the complete bit cell period. The J symbol has the same polarity as the
preceding symbol, whereas a K symbol has the opposite polarity to the preceding symbol.
 In this way the receiver can reliably detect the start and end of each transmitted token or frame
irrespective of its contents and length. Note, however, that only the first six symbols (JK1JK1

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

in Figure 8.7(c)) are used to indicate a valid end of frame. The other two bits, I and E, have
other functions:

 In a token, both the I- and E-bits are 0.

 In a normal frame, the I-bit is used to indicate whether the frame is the first (or an
intermediate) frame in a sequence (I = 1) or the last (only) frame (I = 0).
 The E-bit is used for error detection. It is set to 0 by the originating station but, if any station
detects an error while receiving or repeating the frame (for example, FCS error), it sets the
E-bit to 1 to signal to the originating station that an error has been detected.
 The access control (AC) field comprises the priority bits, the token and monitor bits, and
the reservation bits. As its name implies, the AC field is used to control access to the ring.
 When it is part of the token, the priority bits (P) indicate the priority of the token and
hence which frames a station may transmit on receipt of the token. The token bit (T)
discriminates between a token and an ordinary frame (0 indicates a token, 1 a frame).
 The monitor bit (M) is used by the active monitor to prevent a frame from circulating
around the ring continuously. Finally, the reservation bits (R) allow stations holding high-
priority frames to request (in either repeated frames or tokens) that the next token to be issued
is of the requisite priority.
 The frame control (FC) field defines the type of the frame (MAC or information) and
certain control functions. If the frame type bits (F) indicate a MAC frame, all stations on the
ring interpret and, if necessary, act on the control bits (Z).
 If it is an I-frame, the control bits are interpreted only by the stations identified in the
destination address field. Both the source and destination addresses (SA and DA) are standard
16/48 bit MAC addresses.
 The information (INFO) field is used to carry either user data or additional control
information when included in a MAC frame. Although no maximum length is specified for
the information field, it is limited in practice by the maximum time which a station is allowed
to transmit a frame when holding the control token.
 A typical maximum length is 5000 [Link] frame check sequence (FCS) field is
derived from a 32-bit CRC. Finally, the frame status (FS) field is made up of two fields: the
address-recognized bits (A) and the frame-copied bits (C). Both the A- and C-bits are set to
0 by the station originating the frame.
 If the frame is recognized by one or more stations on the ring, the station(s) sets the A-
bits to 1. Also, if it copies the frame, it sets the C-bits to 1. In this way, the originating station

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

can determine whether the addressed station(s) is non-existent or switched off, is active but
did not copy the frame, or is active and copied the frame.

Frame transmission

 On receipt of a service request to transmit a block of data (which includes the

destination address and the priority of the data as a parameter), the data is first
encapsulated by the MAC unit into the standard format shown in Figure 8.7.
 The MAC unit awaits the reception of a token with a priority less than or equal to
the priority of the assembled frame. Clearly, in a system that employs multiple
priorities, a procedure must be followed to ensure that all stations have an
opportunity to transmit frames in the correct order of priority. This procedure works
as follows.
 After formatting a frame and prior to receiving an appropriate token (that is, one
with a priority less than or equal to the priority of the waiting frame), each time a
frame or a token with a higher priority is repeated at the ring interface, the MAC
unit reads the value of the reservation bits contained within the AC field.
 If these are equal to or higher than the priority of the waiting frame, the reservation
bits are simply repeated unchanged. If they are lower, the MAC unit replaces the
current value with the priority of the waiting frame.
 Then, assuming there are no other higher priority frames awaiting transmission on
the ring, the token is passed on by the current owner (user) with this priority.
 On receipt of the token, the waiting MAC unit detects that the priority of the token
is equal to the priority of the frame it is waiting to be transmitted.
 It therefore accepts the token by changing the token bit in the AC field to 1, prior
to repeating this bit, which effectively converts the token to a start-of-frame
sequence for a normal frame.
 The MAC unit then stops repeating the incoming signal and follows the converted
start-of-frame sequence with the preformatted frame contents.
 While the frame contents are being transmitted, the FCS is computed and
subsequently appended after the frame contents, before transmitting the end-of-
frame sequence.
 Once transmission of the waiting frame(s) has been started, the MAC unit stops
repeating, thus removing the transmitted frame(s) after it has circulated the ring. In

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

addition, the MAC unit notes the state of the A- and C-bits in the FS field at the tail
of the frame(s) to determine whether the frame(s) has(have) been copied or ignored.
 It then generates a new token and forwards this on the ring to allow another waiting
station to gain access to the ring.
 More than one frame may be sent on receipt of a usable token providing, firstly,
that the priority of the other waiting frame(s) is greater than or equal to the priority
of the token and, secondly, that the total time taken to transmit the other frame(s) is
within a defined limit known as the token holding time.
 The default setting for the latter is 10 ms. The frame transmission operation is
illustrated in Figure 8.8(a).

Figure 8.8 Token-ring MAC sublayer operation: (a) transmit; (b) receive.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Frame reception

 In addition to repeating the incoming signal/bitstream, the MAC unit within each
active station on the ring detects the start of each frame by recognizing the special
start-of-frame bit sequence. It then determines whether the frame should simply be
repeated or copied.
 If the F-bits indicate that it is a MAC frame, the frame is copied and the C-bits are
interpreted and, if necessary, acted upon.
 However, if the frame is a normal data-carrying frame and the DA matches either
the station’s individual address or relevant group address, the frame contents are
copied into a frame buffer and passed to the LLC sublayer for further processing.
 In either case, the A- and C-bits in the frame status field at the tail of the frame are
set accordingly prior to being repeated. The reception operation is shown in Figure
8.8(b).

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Routing algorithm

The routing information field contained within each frame immediately follows the source
address field at the head of the normal information frame. The modified frame format is thus
as shown in Figure 8.16(a).

Figure 8.16: Token ring frame format: (a) position of routing information field; (b) structure of
routing information field.

 Since a routing information field is not always required — for example, if the source
and destination stations are on the same segment — the first bit of the source address
— the individual/group (I/G) address bit — is used to indicate whether routing
information is present in the frame (1) or not (0).
 This can be done since the source address in a frame must always be an individual
address, so the I/G bit is not needed for this purpose.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

 If routing information is present, its format is as shown in Figure 8.16(b). The routing
information field consists of a routing control field and one or more route designator
fields.
 The routing control field itself comprises three subfields: frame type, maximum frame
size, and routing field length. In addition to normal information frames, two other frame
types are associated with the routing algorithm. The frame type indicates the type of
the frame.
 Source routing bridges can be used for the interconnection of different types of LAN
segments in addition to token rings.
 Since there is a different maximum frame size associated with each segment type, the
maximum frame size field determines the largest frame size that can be used when
transmitting a frame between any two stations connected to the LAN.
 Prior to transmitting a route finding frame, a station sets the maximum frame size field
to the (known) largest frame size that can be used in the total LAN.
 Before a bridge forwards the frame on a segment, the bridge checks this field with the
(known) maximum frame size of the new segment.
 If the latter is smaller, the bridge reduces the frame size field to the lower value. In this
way, the source station, on receipt of the corresponding route reply frame, can use this
information when preparing frames for transmission to that destination.
 Since the number of segments (and bridges) traversed by a frame when going from a
source to a destination may vary, the routing field length indicates the number of route
designators present in the rest of the routing information field.
 Each route designator comprises a pair of segment and bridge identifiers. The two
additional frame types associated with the route finding algorithm are the single-route
broadcast frame and the all-routes broadcast frame.
 To find a route, a station first creates and transmits a single-route broadcast frame with
a zero routing field length and the maximum frame size set to the known largest value
for the total LAN.
 As with spanning tree bridges, source routing bridges operate in the promiscuous mode
and hence receive and buffer all frames at each of their ports.
 On receipt of a single-route broadcast frame, a bridge simply broadcasts a copy of the
frame on each of the segments connected to its other ports.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

 Since this procedure is repeated by each bridge in the LAN, a copy of the frame
propagates throughout the LAN and is thus received by the intended destination station
irrespective of the segment on which it is attached.
 As we indicated earlier in Figure 8.13, if there are redundant paths (and hence loops) in
the LAN topology, multiple copies of the frame propagate around the LAN.
 To prevent this, before any route finding frames are sent, the bridge ports are configured
to give a spanning tree active topology.
 On the surface, this may appear to be the same procedure used with transparent bridges.
With source routing bridges, however, the resulting spanning tree active topology is
used only for routing the initial single-route broadcast frames.
 This ensures that only a single copy of the frame propagates through the network. The
spanning tree active topology is not used for routing either normal information frames
or the all-routes broadcast frames.
 On receipt of a single-route broadcast frame, the required destination station returns an
all-routes broadcast frame to the originating station.
 Unlike the single-route broadcast, this frame is not constrained to follow the spanning
tree active topology at each intermediate bridge.
 Instead, on receipt of such frames, the bridge simply adds a new route designator
(comprising the segment identifier on which the frame was received and its own bridge
identifier), increments the routing field length, and then broadcasts a copy of the frame
on each of its other port segments.
 In this way, one or more copies of the frame will be received by the originating source
station via all the possible routes between the two stations. By examining the route
designators in their routing control fields, the source station can select the best route for
transmitting a frame to the destination.
 This route is then entered into its routing table and is subsequently used when
transmitting any frames to that station.
 Since the all-routes broadcast frame is not constrained to follow the spanning tree active
topology, on receipt of such frames additional steps must be taken by each bridge to
ensure that no frames are simply circulating in loops.
 Before transmitting a copy of the all-routes broadcast frame on an outgoing segment,
each bridge first searches the existing routing information in the frame to determine if
the segment identifiers associated with the incoming and outgoing ports are already

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

present together with its own bridge identifier. If they are, a copy of the frame has
already been along the route, so a copy of the frame is not transmitted on the segment.
 Note that it is not necessary to perform the route finding operation for each frame
transmitted. Once a route to an intended destination has been determined and entered
(cached) into the routing table of a station, this can be used for the transmission of all
subsequent frames to that destination.
 Moreover, since most stations transmit the majority of their frames to a limited number
of destinations, the number of route finding frames is relatively small compared with
normal information frames for modest sized LANs.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Frame transmission and reception

 The short latency time (and hence number of bits in circulation) of the basic token ring
means that a station, after initiating the transmission of an information frame, can wait
until the FS field at the tail of the frame has been received before transmitting a new
token without any noticeable loss in ring utilization.
 However, from Example 8.4 we can deduce that with an FDDI ring the loss in ring
utilization will be significant if this mode of operation is adopted. With an FDDI ring,
therefore, the early token release method is utilized; that is, a new token is transmitted
immediately after the station has transmitted the FS symbols at the tail of a frame. The
station then follows the token with IDLE symbols until it receives the SD symbols
indicating the start of a new frame or token. The basic scheme is shown in Figure 8.24.

Figure 8.24: FDDI Transmission examples

 As we can see, as with the basic token ring, the source station removes a frame after it
has circulated the ring. However, because of the long latency time of an FDDI ring,
more than one frame may be circulating around the ring at one time.
 Although not shown in Figure 8.24, the ring interface must repeat the SD, FC, and DA
fields (symbols) of any received frames before it can determine its own address in the

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

SA field. This can result in one or more frame fragments — comprising SD, FC, and
DA fields — circulating around the ring.
 This means that a station, on receipt of the token, starts to transmit a waiting frame and
concurrently receives and discards any frame fragments that may be circulating around
the ring.

Timed token rotation protocol

 Unlike the transmission control method used with the basic token ring — which is based
on the use of the priority and reservation bits in the access control field of token and
information frames — an FDDI ring uses a scheme that is based on a timed token
rotation protocol that is controlled by a preset parameter known as the target token
rotation time (TTRT).
 For each rotation of the token, each station computes the time that has expired since it
last received the token. This is known as the token rotation time (TRT) and includes the
time taken by a station to transmit any waiting frames plus the time taken by all other
stations in the ring to transmit any waiting frames for this rotation of the token.
 Clearly, if the ring is lightly loaded then the TRT is short but, as the loading of the ring
increases, so the TRT measured by each station increases also. The TRT is thus a
measure of the total ring loading.
 The timed token rotation protocol ensures access to the ring is shared fairly between all
stations by allowing a station to send any waiting frames only if its measured TRT is
less than the preset TRT for the ring.
 Hence, when a station has frames waiting to send, on receipt of the token it computes
the difference between the TTRT and its current TRT. This is known as the token hold
time (THT) since it determines for how long the station may transmit waiting frames
before releasing the token.
 If the THT is positive, then the station can transmit for up to this interval. If it is
negative, then the station must forego transmitting any waiting frames for this rotation
of the token. A positive THT is known as an early token while a negative THT is known
as a late token.

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok

 Multimedia Information Networks

Dept. of ECE, ATMECE, Mysuru Dr. Shalini Hanok