Pulse or digital communications – Synchronizers – Frequency or phase control using synchronizing signal
Reexamination Certificate
1998-12-30
2002-09-17
Chin, Stephen (Department: 2634)
Pulse or digital communications
Synchronizers
Frequency or phase control using synchronizing signal
C375S368000
Reexamination Certificate
active
06452991
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to systems and methods for acquiring channel synchronization and, more particularly, to systems and methods for acquiring such synchronization in time division multiple access communications systems.
BACKGROUND OF THE INVENTION
Many digital cellular telephone networks use time division multiple access (“TDMA”) channels for dividing communications resources among the terminals using the network. By way of example, digital cellular telephone networks operating under the ANSI-136 standard employ a frequency division multiple access (“FDMA”)/TDMA system in which the frequency band is divided into numerous 30 kHz subbands which are each further divided into up to six separate communications time slots. The users involved with a particular phone call in an ANSI-136 network access one of the available time slots on a specific frequency sub-band. Thus, up to six separate calls may occur simultaneously on any given sub-band, as the six sets of users share the sub-band by transmitting only during their assigned time slots.
To establish a phone call or to lock to a digital control channel in an ANSI-136 or other FDMA/TDMA network, a user terminal tunes its transceiver to one of the sub-bands and acquires synchronization to a predetermined time slot in that sub-band. In cellular systems established under ANSI-136, this synchronization is accomplished by determining the location of a synchronization sequence or “syncword.” In traditional TDMA systems, a unique syncword is transmitted during each time slot, typically as the first few modulated symbols in the time slot or “burst.” Thus, for example, under the ANSI-136 protocol, a total of 162 differential quadrature phase shift keyed (“DQPSK”) symbols are transmitted in each time slot or burst, the first 14 of which comprise the above-mentioned syncword. As ANSI-136 specifies up to six time slots per frequency sub-band, a total of six unique syncwords are specified, one for each of the TDMA time slots.
In typical digital cellular telephone systems, a cellular terminal acquires synchronization via the following process. First the local oscillator on the terminal is tuned to the center frequency of the appropriate frequency band or subband, and a portion of the signal being transmitted on that band is received, sampled and digitized so as to provide a “batch” of digitized samples to the processor included within the telephone. Under ANSI-136, a “batch” typically comprises a total of 1408 samples, which corresponds to eight separate samples per symbol period over 176 consecutive symbol periods. Note that a batch of 176 symbols is typically used (instead of the 162 symbols which are transmitted during a time slot) to ensure that the 14-symbol syncword is not divided between the beginning and the end of the batch.
Upon receiving a batch of 1408 samples, the processor performs a series of correlations on the received batch of samples to determine which 14-symbol (112 sample) portion of the batch has the highest correlation with any one of the six, predefined, 14-symbol syncwords. This is typically accomplished by correlating the first 112 samples (samples
1
-
112
out of the 1408 sample batch) with each of the six syncwords, then correlating samples
2
-
113
with each of the six syncwords, and so on, until each consecutive group of 112 samples in the batch has been correlated with each of the six syncwords. This may require a total of 7,776 separate complex correlations. The results of these correlations are then compared to determine the location of the group of 112 samples which had the highest correlation energy with any of the six predefined syncwords.
Unfortunately, the local oscillators provided in many cellular telephones and other wireless communications terminals demonstrate some degree of error in frequency. This error may arise due to component tolerances, temperature changes, or aging or other effects, and may, in many instances, exceed several kHz or more. Additionally, local oscillators typically have varying settling time, which also may introduce a dynamic frequency error. The net effect of these frequency errors is to reduce the correlation energy detected when a group of received samples which generally correspond to a known syncword are correlated with that syncword. Such frequency effects are particularly problematic in FDMA/TDMA networks which have frequent hand-offs, a large number of frequency sub-bands, and/or low cost user terminals (which typically have less accurate local oscillators).
To identify and compensate for anticipated frequency errors, the above described correlation process may be performed multiple times, with the batch of samples “prerotated” in phase a different amount on each iteration to adjust the frequency (and hence offset the frequency error). Depending upon component tolerances and the magnitude of the frequency change which may occur during hand-offs, as many as thirteen iterations (each of which, in typical systems, comprises 7,776 complex correlations) may be required in order to obtain correlations over a sufficient range of frequencies such that there is a high probability of detecting the syncword.
The highest correlation energy obtained during the above described correlation process (which may involve over 100,000 complex correlations) is then compared to a fixed syncword detection threshold. If this correlation energy value does not exceed the fixed threshold, the terminal will typically repeat the process with one or more new batches of digitized samples. Ultimately, if none of these repeated efforts to acquire synchronization are successful, the terminal will conclude that no base station is transmitting on that sub-band (or there is too much noise to acquire it), and synchronization on a different frequency sub-band will then be attempted. However, if the correlation energy exceeds the detection threshold, then the syncword number (e.g., 1, 2, 3, 4, 5 or 6 in ANSI-136 ), the position of the syncword within the batch, and the phase pre-rotation (which corresponds to frequency error) are reported to the microprocessor. Note that the detection threshold typically is set low enough to allow detection of syncword patterns during periods of channel impairment.
If a valid syncword pattern is found, the microprocessor adjusts the local oscillator frequency based on the phase prerotation (frequency error) associated with the identified syncword, and receives and samples a second segment of the signal transmitted by the base station so as to provide a second batch of digitized samples. Typically, the timing for the reception of this second batch of samples is adjusted so that the identified syncword appears as the first group of samples in the batch. The above-described process is repeated in order to determine a new phase pre-rotation, and this process continues until the phase pre-rotation (the frequency error) is within a specified threshold. Once the frequency error is within the specified threshold, the correlation process may be repeated one or more additional times to confirm the presence of a syncword in two or more bursts to reduce the possibility of detecting false synchronization patterns. If initial synchronization is not accomplished within a predetermined amount of time, a failure may be declared on that frequency sub-band.
The above-described synchronization method may require significant time. This adversely affects both the time a user must wait to complete calls and the “muting” time which occurs when a cellular user switches communications between two different base stations. Another traditional method, which requires less processing time, performs complex correlations using only one sample per symbol period. Thus, for example, if the received signal is sampled at a rate of eight samples per symbol, the first correlation would use samples
1
,
9
,
17
, . . .
105
; the second correlation would use samples
9
,
17
,
25
, . . .
113
; and so on. This method reduces the processing time by nearly an order of magnitu
Chin Stephen
Ericsson Inc.
Fan Chieh M.
Myers Bigel & Sibley & Sajovec
LandOfFree
Systems and methods for acquiring channel synchronization in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Systems and methods for acquiring channel synchronization in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Systems and methods for acquiring channel synchronization in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2829886