Table of Contents
Advertisement
Chapter 10 - Appendixes
Network protocols
Clock operation mode
Delay measurement mechanism
RT430/434
IEEE1588 standard defines the network layers where the PTP protocol will be applied.
It is possible to use PTP protocol in a network layer with IEEE 802.3 Ethernet (layer 2)
or UDP/IPv4 (layer 3) connection.
The layer 3 (UDP/IPv4) is used in more environments facilitating the compatibility of
sending and receiving messages between the devices connected to the network.
Once the PTP protocol has low traffic when compared to other protocols, the network
traffic is not limiting factor of the use of layer 3.
To use the layer 2 it is necessary that the network has E
all master and slave clocks, which are not common when the network is divided into
subnets and there is not an interconnection between them. The advantage of using
layer 2 is that the traffic through the network is smaller because the sent packets do
not require including IP and UDP address.
PTP protocol requires the master clock sending Sync messages periodically to all
slave clocks connected to the network. Furthermore, master clocks must register and
communicate to the slave clocks the exact timestamp in which the data packets
were sent. This information can be sent in a single packet or two packets separately.
In One-step operation mode, the Sync information is sent in the same data packet as
the timestamp of the message. In Two-step operation mode, the Sync information is
sent in a data packet and the timestamp information of the message is sent in
another one. The accuracy of both modes is the same.
RT430/434 allows sending messages in both One-step and Two-step modes.
According to IEEE1588 a slave clock is capable of measuring the delay of message
propagation that represents the time that a message takes to cross the master-slave
path. The measurement of this delay is necessary to perform a time correction of the
time of receipt of the message in relation to the time it was sent. The delay
measurement is performed by sending messages containing the timestamp of the
time of receipt to the master clock which sends a reply with information of the delay.
The second review of IEEE1588 standard, in 2008, specifies two ways of
compensating delay: End-to-end and Peer-to-peer:
End-to-end: measurement of delay across the network between master and
slave clocks.
Peer-to-peer: measurement of delay only between master and slave clocks
as neighbors.
The advantage of P2P is that the time accuracy is immune to change in the network
topology, since the delay between each master-slave connection is calculated for
each packet sent. However, the P2P solution is possible only when all devices in the
network are transparent, i.e., they can perform delay measurement between one
point and another. In network applications where the network comprises devices
without measurement of delay, it is necessary to use the E2E mode, which calculates
the delay in a general way between the two ends of the network.
RT430/434 is capable of measuring the time a Sync message takes to cross a
master-slave path, using E2E and P2P mechanisms.
RT430/434
connections between
THERNET
91
Advertisement
Table of Contents
