wilamowski-b-m-irwin-j-d-industrial-communication-systems-2011

53-4 Industrial Communication Systems
TsTXOffset
Time slot starts here
TsRxTx
TsACKWait
TX
Data
ACK
TsCCAOffset
TsCCA
≤TsMaxPacket
TsRxACKDelay
TsACK
RX
Data
ACK
TsRxOffset TsRxWait TsTxACKDelay
FIGURE 53.3 Packet timing and time slot format. Brighter box is for reception state and darker box is for
transmitting state of devices.
For successful and efficient TDMA communications, all nodes within the same network must share a
common sense of time; in other words, synchronization of clocks between devices is critical. Common
sources of clock drift include temperature and aging. Consequently, tolerances on timekeeping and time synchronization
mechanisms are specified to ensure that devices “exactly” know when the start of a slot occurs.
The format of a timeslot is shown in Figure 53.3.
In this figure, it is shown how a unicast message exchange occurs between a transmitting source
device (call it “A”) that is transmitting to a destination device (call it “B”), which is listening. Assuming
that both nodes are already synchronized, then “B” knows when to expect the first bit of the preamble
of the message coming from “A” (TsMaxPacket is the duration of the longest packet allowed by the IEEE
802.15.4 standard, i.e., 4.256.ms).
There is an initial CCA (which starts with a TsCCAOffset = 1.8 ± 0.1.ms delay and lasts TsCCA =
0.128.ms) that is not strictly needed but has been taken into account in order to improve coexistence
with other wireless networks. Within the slot, the actual transmission of the source message has a
delay with respect to the beginning of a slot (TsTxOffset = 2.12 ± 0.1.ms in the Figure 53.3). This short
time delay allows the source and destination to set their frequency channel and allows the receiver to
begin listening on the correct channel. Also, the receiver has a delay with respect to time slot beginning
(TsRxOffset = 1.12 ± 0.1.ms in Figure 53.3), needed to set up the radio. In addition, since there is a tolerance
on clocks, the receiver has a guard interval (TsRxWait = 2.2 ± 0.1.ms in Figure 53.3), i.e., it must
start to listen before the ideal transmission start time and continue listening after that ideal time.
Once the transmission is complete, the communication direction is reversed and the destination
device sends an acknowledgment packet (ACK) whether it received the source device message successfully
or with a specific class of detected errors. The switching among the transmitting and the receiving
state of the radio must be no longer than TsRxTx = 0.192.ms, while the source node waits TsRxAckDelay =
0.8 ± 0.1.ms from the end of the message to the listening of the ACK (the guard time for the ACK arrival
can be small, since the two nodes are tightly synchronized by the arrival of the message from the source
to the destination); on the contrary, the destination node must send the ACK after TsTxAckDelay =
1.0 ± 0.1.ms from the end of the message.
Communicating devices are assigned not only to a superframe and time slot but have also a specific
channel offset, by means of which channel hopping is implemented. This 3-tuple forms the so-called
communications link, i.e., the opportunity to establish a connection between communicating devices.
All devices must support multiple links. The number of possible links is, typically, equal to the number
of channels utilized by a network times the number of slots in the superframe. For example, the use of
15 channels and 9,000 slots per superframe results in 135,000 possible links.
© 2011 by Taylor and Francis Group, LLC