mirror of
https://github.com/opsxcq/mirror-textfiles.com.git
synced 2025-08-18 23:01:34 +02:00
234 lines
11 KiB
Plaintext
234 lines
11 KiB
Plaintext
|
|
|
|
|
|
Straight from the Bellyard: Coin Services Update
|
|
|
|
|
|
By: Brian Oblivion
|
|
|
|
|
|
-=Restricted -=Data -=Transmissions
|
|
|
|
"Truth is cheap... but information costs."
|
|
|
|
|
|
|
|
Public Telephones, we all know them, and how vulnerable they are
|
|
to fradulent abuse. Well, so does Bellcore and New England
|
|
Telephone and they have come up with a few methods of how to combat
|
|
fraudulent users.
|
|
|
|
Coin deposited by a typical patron of the public telephone are
|
|
communicated to the Stored Program Control System (SPCS) or Operator
|
|
Services System (OSS) by bursts of certian dual-tone signals. These
|
|
signals consist of the two frequencies 1.7 KHz and 2.2 KHz simul-
|
|
taneously, which are detected in end-office and tandem-office environ-
|
|
ments to provide accurate coin deposit detection.
|
|
|
|
There are two types of interference that are encountered by outside
|
|
origin. These are talkdown and talkoff. Talkdown is the failure to
|
|
recognize a valid dual-tone signal because of interference from
|
|
superimposed speech or other noise that may be from either the
|
|
originating coin line or the terminating end. Takeoff is the false
|
|
acceptance of speech, music, or background noises as valid coin
|
|
deposits. This obviously includes redbox tones, whistles, tape
|
|
reproductions, etc.
|
|
|
|
In order to combat this erroneous acceptance of false credit the
|
|
following methods have been defined and are to be employed.
|
|
|
|
Frequency Limits:
|
|
|
|
The detector must accept dual-tone signals if the frequencies
|
|
of both tones are within +/- 5% of their respective nominal
|
|
values of 1.7 KHz and 2.2 KHz. The detector must also reject
|
|
dual-tone signals if the frequency of either or both tones is
|
|
offset more than +/- 3.5% from its nominal value.
|
|
|
|
Frequency offsets of +/- 1.5 percent are the extremes expected from
|
|
properly designed and maintained coin telephones. The +/- 3.5 percent
|
|
constraint is imposed to minimize the width of the stop-band notch
|
|
filters in anti-fraud notch filters.
|
|
|
|
|
|
Signal Power and Twist:
|
|
|
|
The detector must accept dual-tone signals if the individual
|
|
power levels of both tones are within 0 to -25 dBm0 and within
|
|
5 dB of each other. (Twist is the dB difference between
|
|
power levels of the two tones of a dual-tone signal)
|
|
|
|
The detector must reject dual-tone signals if the individual
|
|
poser level of either or both tones is below -30 dBm0.
|
|
|
|
The range 0 to -25 dBm0 is the maximum expected from properly designed
|
|
and maintained coin telephones and loops. The -30 dBm0 limit is
|
|
imposed to minimize the depth of the stop-band needed in anti-fraud
|
|
notch filters.
|
|
|
|
|
|
Pulse Timing
|
|
|
|
Nickels, dimes, and quaters are represented by one, two, and five
|
|
dual-tone TONE-ON pulses. The dollar signal is represented by a
|
|
single long pulse. Each pulse is followed by a TONE-OFF interval.
|
|
The following chart shows the timing specifics for all coin generated
|
|
tones.
|
|
|
|
+----------------------------------------------------------------------------+
|
|
| |Within-Coin |End-of-coin Tone OFF (ms)|
|
|
|---------------------+-------------+-------------+-------+------+-----------|
|
|
| | TONE-ON (ms)| TONE-OFF(ms)| Must | Must |Allowed to |
|
|
| | Must Accept | Must Accept | Accept|Reject|Acc. or Rej|
|
|
|---------------------+-------------+-------------+-------+------+-----------|
|
|
| Nickel | 1 Pulse | 35 - 160 | ---- | >175 | <160 | 160 - 170 |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Dime | 1st Pulse | 35 - 160 | 25 - 110 | | | |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Dime | 2nd Pulse | 35 - 160 | ---- | > 75 | < 60 | 60 - 75 |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Quater | 1st Pulse | 20 - 100 | 20 - 110 | | | |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Quater | 2nd Pulse | 20 - 60 | 20 - 60 | | | |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Quater | 3rd Pulse | 20 - 60 | 20 - 60 | | | |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Quater | 4th Pulse | 20 - 60 | 20 - 60 | | | |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Quater | 5th Pulse | 20 - 100 | ---- | > 75 | < 60 | 60 - 75 |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Dollar | 1 Pulse | 600 - 700 | ---- | > 75 | < 60 | 60 - 75 |
|
|
|---------+-----------+-------------+-------------+-------+------+-----------|
|
|
| Note: The detector is allowed to accept or reject tone-on and within-coin |
|
|
| tone-off durations outsie their repective must-accept ranges. |
|
|
+----------------------------------------------------------------------------+
|
|
|
|
Methods of controlling talkdown and talkoff.
|
|
|
|
Talkdown, is more of a transmission error problem that is generated
|
|
by interference from superimposed speech or ambient noise. This
|
|
noise comes from the originating coin line or from the terminating
|
|
end. Talkdown results in the failure of the detector to recognize
|
|
a valid dual-tone signal.
|
|
|
|
One of the main sources of talkdown is the improper muting of the voice
|
|
path during tone generation. This antenuated speech or ambient noise
|
|
power is superimposed on the coin signals which corrupts the signal.
|
|
|
|
Muting is accomplished by shorting out the speech path with a cap-
|
|
acitor. Minimum muting attenuation is 15 dB at 300 Hz, 25 dB at
|
|
1000 Hz, and 35 dB at 3000 Hz. The attenuated speech or ambient
|
|
noise power is superimposed on the coin signals which could cause
|
|
talkdown.
|
|
|
|
Some talkdowm is also generated from the called party, operator
|
|
speech, or automated announcements. These signals can only reach
|
|
the detector by reflection through the transhybrid path at a 3-to-4
|
|
wire junction, which is usually located in the originating end
|
|
office.
|
|
|
|
|
|
Talkoff
|
|
|
|
Talkoff is the false acceptance of speech, music or background
|
|
noises as valid coin deposit signals by a detector. This is
|
|
what registers a valid deposit when no actual money is deposited.
|
|
|
|
In order to combat, a few countermeasures are to be employed.
|
|
|
|
o The call must be terminated or routed to an operator if the
|
|
total amount due for an initial deposit or a subsequent deposit
|
|
has been received within an overall session timeout period.
|
|
|
|
o The overall session timeout must be an adjustable parameter with
|
|
a default value of 45 to 50 seconds.
|
|
|
|
Intentional talkoff can be induced by playing loud music into the
|
|
coin telephone by doing this one hopes to get the detectors to produce
|
|
too many talkoff clusters. A talkoff cluster is a group of talkoffs
|
|
occurring within any 5-minute interval. Single quater or dollar
|
|
talkoffs are also included in this definition of a cluster. In order
|
|
to combat this the following measures are being taken.
|
|
|
|
o Total talkoffs must not exceed 12 dollars in 70 hours of continous
|
|
speech or music at a volume leve of 0 vu referred to 0 TLP or,
|
|
instead, at an active speech level of +1 dBm0.
|
|
|
|
o The detector must not produce more than 12 talkoff clusters that
|
|
exceed 20, or 40, or 95 cents in 70 hours of continuous speech or
|
|
music at a volume level of 0vu referred to 0 TLP or, instead, at
|
|
an active speech level of +1 dBm0.
|
|
|
|
o Total talkoffs must not exceed 2 dollars in 200 hours of continuous
|
|
speech or music at -17 vu referred to 0 TLP or, instead, at an active
|
|
speech level of -16 dBm0.
|
|
|
|
|
|
Coin-Operated Telephone Fraud
|
|
|
|
Fraud is possible by acoustically coupling accurate coin deposit
|
|
signals into the mouthpiece of the originating coin telephone, or
|
|
from the called party end. Fraudulent signals comming from the
|
|
coin telephone end can only be stopped within the coin telephone
|
|
with a notch filter to filter out any external fraudulent signals.
|
|
|
|
In order to combat Fraudulent usage the following will be enacted.
|
|
|
|
o End- or tandem-office switching systems will be equipped with
|
|
detectors that monitor signals from the originating coin line.
|
|
|
|
The following methods combat far-end fraud:
|
|
|
|
o Sensing the direction of received coin signals and only reg-
|
|
istering those comming from the coin line.
|
|
|
|
o Increasing the transhybrid path loss (using echo cancelers to
|
|
knock down the fraudulent signals signal strength to that under
|
|
the detector's recognition threshold.)
|
|
|
|
o Inserting a 2200-Hz notch filter in the voice transmission path
|
|
from the far end only when the detector is expecting coin deposits.
|
|
The filters 3 dB bandwidth must be less than 600 Hz centered on
|
|
2200 Hz. and the out-of-band attenuation below 1500 Hz and above
|
|
3000 Hz must be less than 1 dB.
|
|
|
|
|
|
The BOC's are now taking action against 'Redboxing' as well as
|
|
agains some of their own equipment defects. It was only a matter
|
|
of time, of course. I have no idea how long this conversion will
|
|
take nor if it has yet started. There has been some indication
|
|
as some start of a anti-fraud campaign in my local area. I have
|
|
gotten word from some people that coin operated telephones have
|
|
not been accepting tones. I am unaware if these machines were
|
|
BOC operated or if they were COCOTS. I believe them to be BOC
|
|
coin operated telephones. Use it while it lasts. ;)
|
|
|
|
|
|
Brian Oblivion
|
|
@ Black Crawling Systems
|
|
@ atdt.org
|
|
|
|
|
|
The following has been a presentation of...
|
|
|
|
Restricted Data Transmissions
|
|
|
|
-=R -=D -=T
|
|
|
|
"Truth is cheap... but information costs"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|