Synchronisation Flashcards
Explain the difference between data-aided and non-data-aided synchronisation methods
Data-aided (Schmidt & ML Approx):
- Transmitter inserts symbols already agreed upon the receiver within the message so the receiver can find unknown through knowledge of this ‘data’.
- The disadvantage is that it’s a waste of data and could used to transmit more data. Therefore, spectral effincey is reduced but is non-negligible.
Non-Data-Aided (Phase/Freq. Locked Loop, Square Timing Recovery)
Neither a pre-amble nor decision can be used. Some particular characteristics of the incoming signal can be employed to estimate the unknown parameters.
- 2 unknowns -> v. unreliable
- spectral effincey is the same BUT
- need LARGE amount of data processed to average the effects of noise and find reliable estimate.
What is synchronisation in a communication scheme and name the different types of synchronisation.
It’s the process of aligning offsets of transmission and switching equipment.
The parameters to be considered are the following: frequency offset, phase offset and time offset.
Describe the difference between incoherent and coherent detection
Incoherent detection:
Incoherent detection is when the transmitter and receiver aren’t synchronised. in this case, the receiver knows the general shape of the signal.
Therefore, signals of two different frequencies are used for encoding and decoding.
Coherent detection:
Coherent detection is when the signals are available at the receiver. Therefore, the receiver knows the channels phase.
In coherent detection, we use the signal of the same frequency as that of the signal used in the encoding/modulation technique.
Why can’t QPSK use incoherent detection? Is there a modification to QPSK that can enable incoherent detection? How so?
QPSK cannot use incoherent detection because it’s modulation scheme is a DSBC, which is a coherent detection scheme.
What can be done to enable incoherent QPSK is using a DPSK. A differential detector uses the signal received in the previous symbol interval as a phase reference for the received signal in the current interval.
Describe the important of Low Pass Filter’s bandwidth in the presence of noise.
noise power is proportional to bandwidth.
Therefore, using a LPF, which has the effect of attenuating any frequencies higher than its cut-off frequency. The LPF keeps the bandwidth restricted as so no more noise power creeps in.
In what order should time, frequency and phase be synchronised?
frequency (w/o this, you wouldn’t have the signal), time and phase (need this for constellation diagram)
You are designing the synchronization algorithm for a new single-carrier communication system with an overall bandwidth of 200 kHz. Your company is providing modules for phase recovery with a preamble, for the Frequency Locked Loop (FLL) and for the Square Timing Recovery.
Put the three modules in the correct order and elaborate your answer based on their effects on the received signal.
Order: 1st FLL, 2nd timing recovery, 3rd phase estimation.
Reason: In contrast to the other methods, the FLL does not require any prior information or synchronization.
After the frequency synchronization has taken place, the clock can be recovered. Once this has taken place, (potentially still phase-shifted) Symbols are scanned cleanly.
The remaining phase shift can then be corrected using the preamble-based phase estimator, which however relies on it is that the symbols are as ISI-free as possible and without a frequency offset, i.e. linearly time-dependent phase offset.
By which term is an estimator that achieves the Cramér-Rao Bound referred to? What does this mean in practical terms?
Such an estimator is called efficient.
This means that every other estimator for this parameter based on a certain observation length the same or a higher one has variance.
