RFID data signals from RFID tags may be recovered by determining the probabilities of transitions between data states between a series of a pairs of signal samples using a set of predetermined probabilities related to data, timing, baud rate and/or phase variables affecting the received signal and processing those determined probabilities to determine the sequence of such transitions that has the highest probability of occurrence. A second set of predetermined probabilities related to transitions in the opposite direction may be used to sequence in a reverse direction. The determination of the sequence representing the RFID tag data may be iterated in both directions until further iterations do not change the determined probabilities.
1. A method for detecting a data sequence from a sequence of symbols backscattered by an RFID tag, comprising:
receiving the sequence of backscattered symbols at a receiving antenna, where each backscattered symbol has a random symbol duration, and a random phase resulting from random variations introduced by instability in the RFID backscatter process;
using a receiver to:
over sample the received symbols to produce a set of samples;
estimate data values for short sequences of consecutive symbols that are most likely to have occurred based upon the set of samples, where the duration of each symbol and the phase of each symbol are assumed fixed during each short sequence and are assumed randomly varying between the short sequences due to the instability in the RFID backscatter process, by:
determining a chosen fixed number of samples per symbol;
selecting a first sample from the set of samples corresponding to the start of a first symbol in the short sequence of consecutive symbols;
selecting a sample from the set of samples that is half the chosen fixed number of samples from the first sample in the short sequence of consecutive symbols, where the selected sample is chosen to be the first sample within the second half of the first symbol in the short sequence of consecutive symbols;
selecting the chosen fixed number of samples starting from the first sample within the second half of the first symbol in the short sequence of consecutive symbols;
selecting a first sample and a first sample within the second half of each symbol within the short sequence of consecutive symbols based on the selected first sample of the first symbol and the chosen fixed number of samples;
applying a matched filter with a length determined based upon the chosen fixed number of samples per symbol to the selected chosen fixed number of samples starting from the first sample within the second half of the first symbol in the short sequence of consecutive symbols; and
determining the following:
(
d
^
k
,
d
^
k
-
1
)
=
arg
max
d
k
,
d
k
-
1
(
r
2
,
k
-
2
-
r
1
,
k
-
1
)
d
k
-
1
+
(
r
2
,
k
-
1
-
r
1
,
k
)
+
(
r
2
,
k
-
r
1
,
k
+
1
)
d
k
where {circumflex over (d)} k denotes the kth decoded bit, d k denotes an hypothesis for the kth decoded bit and r 1h , r 2h are matched filter outputs for the first sample and the first sample within the second half of the hth symbol and where the predetermined duration of consecutive symbols is three, starting and ending at half symbol boundaries; and
determine a sequence of bits by combining the estimates of short sequences of data values for consecutive symbols; and
outputting the sequence of bits from the receiver as the detected data sequence.
2. The method of claim 1 , wherein using the receiver to estimate data values for short sequences of consecutive symbols that are most likely to have occurred based upon the set of samples, where the duration of each symbol and the phase of each symbol are assumed fixed during each short sequence and are assumed randomly varying between the short sequences due to the instability in the RFID backscatter process further comprises estimating each short sequence of consecutive symbols by:
determining data values for a first short sequence of symbols that are most likely to have occurred based upon samples selected from the set of samples, where the determination includes choosing a fixed number of samples per symbol;
using the data value determined for the last symbol in the first short sequence of symbols as the first symbol of a next short sequence of symbols; and
using the chosen fixed number of samples per symbol to determine the starting point for the estimation of the data values of the next short sequence of symbols.
3. The method of claim 1 , wherein the matched filter is implemented using an integrate and dump circuit.
4. The method of claim 1 , wherein the real parts of the matched filter outputs are used as the values of the matched filter outputs in the equation:
(
d
^
k
,
d
^
k
-
1
)
=
arg
max
d
k
,
d
k
-
1
(
r
2
,
k
-
2
-
r
1
,
k
-
1
)
d
k
-
1
+
(
r
2
,
k
-
1
-
r
1
,
k
)
+
(
r
2
,
k
-
r
1
,
k
+
1
)
d
k
.