2012 Exam Flashcards
How does OFDM simplify the process of high data rate transmission + reception in a highly time dispersive channel? (4)
-
High bit rate single carrier transmission is difficult to achieve
- Time delay spread introduces ISI = BER
- OFDM data is multiplexed over a large # sub-carriers
- Data on each sub-carrier is low enough to avoid ISI
- A FDE uses a complex multiplier to correct the ampl. + phase distortion of each sub-carrier
Method for signal processing in an OFDM system to make a cyclical process (2)
Extend the data vector
Resolve the extended lin. conv, it can appear cyclical
Processing Gain Definition
Ratio of the Spread Spectrum b/w to message signal
DS-SS Definition
take narrowband message + convert to wideband via directly modulating w/ wideband spreading sequence.
Synchronised de-spreading is used to recover the narrowband message
Spreading Sequences used in DS-SS? Explain why an arbitrary selection of code can not be used for multi-user networks (3)
- Pseudo Random Binary Sequences (m-sequences)
- Sharp auto-correlation + pseudo-random (Gaussian like) run length distr. of 1’s + 0’s
- Arbitrary selection of maximal length shift register sequences can give rise to very poor cross-correlation + inhibit code separation
Describe rank deficient, in terms of cap. + why? (2)
Suggest soln
- Due to correlation in composite antenna + propgating medium, not alltheorthogonalspatial modes are available for thearray dimension exciting the channel
- Reduce the potential spatial multiple gain + channel cap.
Soln - use a wider spaced antenna / polarisation
Structure of LTE (2)
- i/p data is passed through the filter memory (feedforward filter) + multiplied w/ appropriate weight
- An error term is generated on a symbol by symbol basis based on the difference between software filter o/p +
- training symbol
- hard decision o/p
Markov Process
next state is defined by the previous state + current i/p
Viterb Eq.
state is defined by data stored in channel memory. Next state is then based on the previous state + i/p symbol (Markov Process)