|
RF design guide Radio
control
|
|
|
<< RF design guide Top |
<<BACK
NEXT>> |
|
|
|
Here we explain the general communication methods required for radio
control using radio waves. We will touch on the causes of errors that
create problems with communication using radio waves, and we will explain
briefly how to solve them.
|
 Communication methods
<Menu>
|
Here we will explain briefly about
communication methods. There are two communication methods, two-way
communication and one-way communication, and the thinking behind both is
the same as for radio communication.
Two-way communication
With this communication method, information flows in the direction from
device A → device B, and the direction from device B → device A.
Half duplex
The transmission of information is carried out in the direction from
device A → device B, and from device B → device A. However, with this
communication method, communication is carried out alternately, not
simultaneously.
Full duplex
With this communication method, the transmission of information can be
carried out in the direction from device A → device B, and from device B →
device A simultaneously.
One-way communication
With this communication method, information flows in the direction from
device A to device B only. Information does not flow from B to A. If
control signals and so on besides data are sent, it is by one-way
communication.
Telecontrol and telemetry is carried out using one-way communication or
half-duplex operation. Communication is by half duplex operation with the
user switching between transmitting and receiving with a handy
transceiver. Wireless LANs and the like use apparent full duplex.
Communication method between wireless LANs
Let's consider sending NetMeeting video (standard Windows communication
software) using a wireless LAN. The images from a CMOS camera and the
sound from a microphone both appear to run simultaneously. From the point
of view of the user, it appears as if there is full duplex communication.
With ordinary radio communication, one frequency channel (frequency band)
is used, so the radio equipment can only use half duplex in the radio part
(radio wave). However, between the user and the equipment, the
communications are modified so as to appear to be full duplex
communication.
|
Data transmission
control procedures
<Menu>
|
On the transmission path there are various
possible causes of errors, and in order to transmit the data correctly,
line control between the transmitting equipment and the receiving
equipment is necessary to control synchronization of timing and errors.
For ordinary communication, there are data transmission control procedures
such as non-procedural, basic control, and high level data link control (HDLC),
and these can be applied to two-way communication with radio.
The operation mode of the radio equipment is 1:1, 1:N, or N:N and when the
mode is two-way communication, a system with a special radio protocol is
added to the data transmission control procedure.
|
| Non-procedural |
Transmission is carried out when just the
transmission control code is agreed by the transmitting and receiving
ends. Transmission control such as the control of errors and so on is
performed by the user of the equipment.
|
| Basic control procedure |
This is a character based synchronous system
where data is transmitted while verifying the serial response of the
transmitting and receiving ends using transmission control codes.
Transmission control is performed automatically. Data is transmitted in
block units, and if an error occurs, only that block is resent.
There are two modes, basic mode and extended mode. In basic mode, only
text files are transmitted, and when transmitting transparent data (binary
data), extended mode is used.
|
| High level data link
control protocol (HDLC) |
While the basic procedure uses character
based transmission, HDLC uses bit-oriented transmission, in which data
regarding addresses, control codes, information, frame check codes and so
on are included in frame based transmissions. It is a synchronous flag
system in which the start and end of the transmitted frames are flagged
(7E Hex). Transmission control is automatic with FCS (Frame Check
Sequence) used for error detection. This allows for highly reliable
transmission. Data can be transmitted with full transparency. FCS uses the
CRC
|
|
← Transmission direction
Flag
01111110 |
Address |
Control |
Information |
FCS |
Flag
01111110 |
|
| Data link method |
Data link systems between transmitting and
receiving equipment includes contention, polling, selecting systems and so
on.
|
| Contention system |
A system for establishing a data link with a
point-to-point method of connection. The transmitting end sends an enquiry
code, and data is transmitted after the receiving end sends acknowledgment
to the transmitting end. With point-to-point systems, both devices have an
equal relationship.
|
| Polling / selecting
systems |
A system for establishing a data link with a
multi-point method of connection. This is a linkage between a control
station and subsidiary station.
Polling
The control station periodically sends requests to send data (data sent to
the control station) to the subsidiary stations in the net.
Selecting
If the control station has data to send to a specific subsidiary station,
it enquires whether the subsidiary station is able to receive before
sending the data.
|
Wireless errors
<Menu>
|
With data transmission, for wired (cable)
communication as well as for wireless, error control must be implemented.
There can be no comparison between radio communication using radio waves
and wired communication in terms of the prevalence of causes of errors
such as noise, interference, decay, and so measures against errors must be
put in place. Various methods are possible, but there is always the issue
of balancing the level of processing used with the constraints such as
application, cost, time, and engineering of the system being designed.
Normally, we are not bothered by slight noise or interruptions when we use
a mobile phone, and we can put up with a bit of flicker on a television
screen.
High level error processing is used for the sound in mobile phones, but
errors seem to be inevitable. However, for data transmissions such as
mail, this is a problem.
If we consider radio controlled industrial equipment and devices,
malfunction caused by transmission errors would cause serious and
life-threatening accidents, and might involve the loss of important data.
As a designer of radio equipment, it is necessary to pay very close
attention that this sort of thing does not happen even if mistakes are
made. It is also necessary to incorporate fail safe thinking into designs,
and we will consider this later on.
The following types of error are possible
Random errors
Errors that occur randomly without any temporal relationship with other
errors.
Burst errors
Errors that occur suddenly and relatively consecutively.
|
Processing methods for
errors <Menu>
|
Processing errors in radio communication must
be handled at both the hardware and software levels, but if appropriate
measures are taken, communication that is not inferior to wired
communication can be achieved.
|
| One-way communication
error processing |
For one-way data transmission such as the
commands for telecontrol or sound and images, the following error
processing methods are available.
|
|
No error processing is used. |
The system will be unusable. |
|
Decision using error detecting codes |
The results of some operations can
be checked by human senses, and for relatively unimportant data such
as the collection of temperature data, where the same data is sent
continuously (or it is analog-type data), the data can be put in the
frame structure along with code for detecting errors, and the
receiving equipment can then decide whether or not there are errors
based on this code, and can discard the data if an error is present.
Methods of calculating error detecting codes include the checksum
method, the CRC method and so on. |
|
FEC |
For applications where you want to
keep errors to the minimum, use a method such as FEC (Forward Error
Correction).
This is a system in which the transmitter includes code that allows
the receiver to perform error correction. This code makes the
transmitted data longer than the actual data (redundant data), but
this method allows us to approach error free transmission.
Reed-Solomon code and trellis code are representative types of this
kind of code, and there are other types that incorporate these
codes. Furthermore, interleaving methods are also used to handle
burst errors in radio communication. |
|
| |
| Two-way communication
error processing |
In addition to FEC for two-way communication,
there are methods that use ARQ (Automatic Repeat Request).
With the ARQ system, data is sent as a packet in a frame format, and if an
error occurs, the receiver sends a resend request, thus achieving error
free data transmission.
This is used in most wireless LANs and other wireless systems.
In order to avoid errors, it is necessary to use protocols for data
transmission in the wireless zone. Specifically, with packet
communication, address information, a packet number, packet size,
status/control, frame check and the like is attached to either end of each
packet. The receiving end performs error detection, and if necessary sends
a resend request, thus achieving error free data transmission. A preamble
and start code is attached to the first packet in radio communication that
uses packets. There are cases where a preamble is necessary in order to
synchronize the radio equipment.
| Preamble |
Start code |
Receiver address |
Sender address |
Packet number |
Status/control |
User data |
Frame check |
|
| Frame check
systems (error control systems) |
With the purpose of detecting data errors in
the packet frame, the transmitting end adds a frame check code at the end
of the packet when forming the packet. The receiving end then decides
whether or not there are errors based on this code, and if it detects a
data error, it sends a request to the transmitting end to resend the data,
thus achieving error free data transmission.
This system includes the checksum method, the CRC method (Cyclic
Redundancy Check), the parity check method and so on. The CRC method is
better at detecting errors compared with the checksum method and so on,
and it is used in most data transmission protocols such as wireless LANs,
hard disk controllers and the like.
CRC systems include 5-bit and 12-bit, but currently the most commonly used
systems are 16-bit or 32-bit, and in particular, CRC-CCITT and so on are
frequently used.
CRC-CCITT
With CRC-CCITT the data frame is calculated by dividing by constants, and
the 16-bits (2 bytes) of the result are attached for transmission at the
end of the data. The receiving end performs a similar operation, and if
the result is correct, it determines that there is no error in the data
transmission. If not, it determines that there is an error and sends a
resend request.
To explain in rather more detail, in the CRC operation at the transmitting
end, the bit string of the frame data is treated as a numerical value
(this is called the message polynomial), and the message polynomial is
divided by a generator polynomial (constant) X16 + X12 + X5 + 1, and the
remainder (CRC code: 2 bytes) is attached to the end of the data for
transmission. The rest is the same.
This operation can be performed by the CPU software, but if high speed
processing is necessary, it is performed using hardware. Some CPUs contain
hardware for this CRC operation, while other CPUs form packet data frames
(HDLC function). If the situation permits, it is conceivable to use FPGA
or gate arrays, including peripheral equipment.
|
Transmission line code
<Menu>
|
In the transmission path of wireless
communication there may be noise and interference that causes loss, so
that the code pulse is distorted, and discriminating between data becomes
hard.
In the receiving equipment a clock signal for synchronization is required
for decoding the data, but it is necessary to extract this from the
received data stream. The base band signal normally uses NRZ code, but at
the transmitter end if this is directly input to a modulator as a base
band signal, with data that is a sequence of 0s or 1s, the receiver cannot
duplicate the synchronizing clock from the signal (clock recovery).
In order to overcome this problem, means such as the use of Manchester
code are available.
As shown in the diagram below, with Manchester code the middle of the code
always reverses polarity, so that unbroken strings of 0's or 1's is
avoided, and clock recovery for the receiver is simplified. However,
compared with NRZ code, using Manchester code makes the occupied frequency
band wider. This code converts the signal from 0 to 10, and from 1 to 01.
|
|
|
The fail safe
design philosophy
<Menu>
|
Fails safe means that even if a failure or
interference occurs, the software or hardware exercises control that tends
in the direction of safety. It is a safety philosophy that seeks to limit
any damage to the minimum. This is implemented in all fields of design,
including construction, electricity and the like.
Especially compared with other technologies, the likelihood of errors
arising with products using radio waves is high. However hard we try to
achieve perfect error processing, we cannot guarantee that there will be
no errors at all. Please remember that it is necessary to implement fail
safe thinking with regard to the system overall.
|
|
|
<< RF design guide Top |
<MENU> |
<<BACK
NEXT>> |
|
|
