(1) How can we justify the bandwidth of an arbitrary signal in BPSK
and ASK?

(2) Since GSM has error correction before encryption, how does error
correction work?

This could be considered poor design, but since GSM uses stream
ciphers (which are essentially XORing the data with a specific
string), errors in the encrypted data carry over as errors in exactly
the same bit positions in the decrypted data.

(3) What happens in GSM/CDMA phones when a mobile unit roams?

(4) In Viterbi's algorithm, why do we check whether the shortest paths
to all states at a particular level have the same first edge?

(5) How do you quantify the number of bits that a convolutional code
can correct?

(6) Why does 802.11 not use CDMA?

Data traffic tends to be bursty, so it is better to allocate bandwidth
in a statistically multiplexed manner (as opposed to CDMA, TDMA, or
FDMA).

The spreading code in WLANs needs to be short to make the
synchronization easy at the receiver.  CDMA codes are typically long
to avoid correlation.

CDMA also requires power control to be effective, which makes it
difficult to use in ad hoc environments.  This is not to say that CDMA
cannot be used in WLANs.

(7) In Bluetooth, when does a slave get the 3-bit ID?