CSI defines the largest allowable range of a high resolution number to be 99999.

Interpretation of the decimal locator for a 4 byte data value is given below. The decimal equivalent of bits GH is the negative exponent to the base 10.

BITSDECIMAL FORMAT
G H A5 digits

0 0 0

XXXXX.

0 0 1

XXXX.X

0 1 0

XXX.XX

0 1 1

XX.XXX

1 0 0

X.XXXX

1 0 1

.XXXXX

C.3 GENERATION OF SIGNATURE

At the end of a binary transmission, a signature is sent. The signature is a 2 byte integer value which is a function of the data and the sequence of data in the Output Array. It is derived with an algorithm that assures a 99.998% probability of detecting a change in the data or its sequence. The CR10 calculates the signature using each transmitted byte beginning with the Final Storage format data (for K command, echo and carriage return line feed are not included) until the 2 byte signature itself. By calculating the signature of the received data and comparing it to the transmitted signature, it can be determined whether the data was received correctly.

SIGNATURE ALGORITHM

S1,S0 - represent the high and low bytes of the signature, respectively

M - represents a transmitted data byte

n - represents the existing byte

n+1 - represents the new byte

T - represents a temporary location

C - represents the carry bit from a shift operation

1.The signature is initialized with both bytes set to hexadecimal AA.

S1(n) = S0(n) = AA

APPENDIX C. BINARY TELECOMMUNICATIONS

2.When a transmitted byte, M(n+1), is received, form a new highsignature byte by setting it equal to the existing low byte. Save the old high byte for later use.

T1 = S1(n)

S1(n+1) = S0(n)

3.Form a temporary byte by shifting the old low signature byte one bit to the left and adding any carry bit which results from the shift operation. A "shift left" is identical to a multiply by 2. Ignore any carry bit resulting from the add.

T2 = shift left (S0(n)) + carry

4.Form the new low signature byte by adding the results of operation 3 to the old high signature byte and the transmitted byte. Ignore any carry bits resulting from these add operations.

S0(n+1) = T2 + S1(n) + M(n+1)

As each new transmitted byte is received, the procedure is repeated.

C-5