Trio JR240, Руководство пользователя

Страница: 17 / 76

Page 17
J-Series Data Radio – User Manual
Digital Collision Avoidance
ChannelShare™
In some network applications there exists the potential for over the
air data collisions between REMOTE radios. This can occur when
an exception reporting type of protocol is used (such as DNP3) or
when multiple applications are being transported over the radio
network (such as MODBUS and DNP3).
In both of these scenarios, the possibility exists for two or more
REMOTE radios attempting to access the channel (i.e.: talk to
the Access Point) at exactly the same time. If this occurs, both
REMOTE radios are unaware of the other trying to access the
Access Point and both will attempt to transmit their data. As such,
the Access Point will received a corrupted message from both
radios and a re-try will be required.
ChannelShare™ digital collision avoidance is a type of Access
Point controlled channel access scheme that decreases the
possibility for over the air data collisions to occur. It achieves this
by having any REMOTE (or BRIDGE) with data that needs to
be transmitted request permission from the Access Point before
transmitting the data. The mechanism which is used to achieve this
is called a Token Grant.
Token Grants
When ChannelShare™ digital collision avoidance is enabled, the
Access Point radio (or BRIDGE that is behaving as an Access
Point for its DOWNSTREAM network) issues “Tokens” to each
REMOTE (or BRIDGE) radio that needs to transmit data.
When a REMOTE (or BRIDGE) radio has data packets that need
to be sent to an Access Point, it first requests a TOKEN from the
Access Point (or BRIDGE) radio. When a TOKEN is requested,
the requesting radio includes the amount of data that needs to be
transmitted.
When the Access Point radio receives the TOKEN request, it
determines a suitable time when the requesting radio can access
the channel and transmit its data. The Access Point then grants a
TOKEN to the radio wants to access the channel, and the TOKEN
includes the serial number of the radio being granted, the starting
time and length for which the TOKEN applies. All other radios not
granted the TOKEN remain silent during the TOKEN period.
The function of the Access Point is to consider and manage all
TOKEN requests and only grant TOKENS in such a way that no
two devices accessing the channel will cause a collision.
TOKENs are granted to specific radios, based on their serial
number. As such, their is no opportunity for multiple radios to
become confused about what TOKEN has been granted.
The process of requesting “permission” to send data and receiving
the “OK to Send” response does result in increased data latency
and slightly slower data throughput. However, the practical reality
of systems where large amounts of exception reporting data need
to be handled mean the actual throughput perceived by the user is
much higher because the number of over the air data collisions is
significantly lower.
All radios in the network also employ a small amount of random
channel access backoff timing so that when a TOKEN has expired,
requests for TOKENs to be granted are not themselves the subject
of a large number of over the air data collisions.
Retries and Retransmissions
No Ack Retries
Data transmitted from a REMOTE to an Access Point is
acknowledged by the Access Point. No Ack Retry Limit defines
how many times the REMOTE attempts to transmit data to an
Access Point if NO acknowledgement arrives from the Access
Point for the previous attempt by the REMOTE.
Once the No Ack Retry Limit is reached, the data is discarded by
the REMOTE. The important point to note is that the No Ack Retry
mechanism does not have any bandwidth penalties as it is only
required when interference prevents the successful delivery of data
to the Access Point radio.
Retransmissions
Except in PTP modes, data transmitted from an Access Point (or
BRIDGE) to a REMOTE is not acknowledged by the REMOTE.
Data Retransmissions define how many times the Access Point
(or BRIDGE) duplicates (or re-transmits) data to the REMOTES.
The theory is that even if the REMOTE is subject to interference,
at least one of the messages will be received. Duplicate messages
are discarded by the REMOTE. Each data re-transmission results
in less usable bandwidth in the Access Point to REMOTE direction.
However, in a practical system it is a more efficient method of
preventing lost data than waiting for the SCADA system to timeout
and try again.
Force Retransmissions Across Hops
For increased reliability and protection from other radios operating
in the vicinity, it is possible to force the re-transmissions to occur in
different hops. The data is first transmitted in one hop and then the
second re-transmission takes place on the next hop. This method
ensures that each re-transmission will be sent on a different
frequency within the hop pattern of the radio and so avoids fixed
frequency interference. However, there is a larger decrease in
usable bandwidth in the Access Point to REMOTE direction when
re-transmissions are spread across multiple hops. To minimise
the loss of bandwidth, it is recommended to choose a shorter hop
interval.