Wednesday, January 19, 2011

Communication model of CDMA

Information Stream ----------------->
Speech coding --> Channel coding --> Scrambling --> Spread Spectrum --> Modulation --> RF Transmit

<------------------------------ Information Stream
Speech decoding <-- Channel decoding <-- De-Scrambling <--De-Spread Spectrum <-- De-Modulation <-- RF Receive


 Speech coding

Speech coding is critical to digital transmission. CDMA system uses an efficient method of speech coding and extensive error recovery techniques to overcome the harsh nature of the radio channel.

The objective of speech coding is not only to maintain speech quality but also to reduce the quantity of transmitting data.

Speech coding algorithms (digital compression) are necessary to increase cellular system capacity.

Coding must also ensure reasonable fidelity, that is, a maximum level of quality as perceived by the user.

Coding can be performed in a variety of ways (for example, waveform, time or frequency domain).

Vocoders transmit parameters which control reproduction of voice instead of the explicit, point-by-point waveform description.

Variable Rate Vocoding
  • CDMA uses a superior Variable Rate Vocoder which includes Full rate during speech, Low rates in speech pauses, Increased capacity and More natural sound.
  •  Voice, signaling, and user secondary data may be mixed in CDMA frames. 
  • The output is 20 ms frames at fixed rates:  Full Rate, 1/2 Rate , 1/4 Rate , 1/8 Rate, & Blank. 
  • CRC is added to all the frames for the 13 kb vocoder, but only to the Full and 1/2 rate frames for the 8 kb vocoder.
  • CRC is not added to the lower rate frames in the 8 kb vocoder, but that is ok because they consist mostly of background noise and have a higher processing gain.
  • Current vocoder rates are 8kbps, 13kbps, and 8kbps EVRC (Enhanced Variable Rate Coder)



 Channel coding

Channel coding usually falls into two classes: Block interleave codes and Convolution codes. The objective of channel coding is adding additional supervising bits in the information stream to ensure get correct signal at receive side.

Convolution Encoder & Interleave Encoder

Convolution Encoder: It increases the reliability but reduce the transmitting efficiency, because each code stream adds supervising bit for rectified
Block Interleave Encoder: It does not change the efficiency but have some delays, because the transmitter and receiver must process to writing first and then reading


Scrambling

The paging channel includes many import information such as user’s IMSI, In order to keep the user’s information secret, we use the data scrambling.

Data scrambling function:

Data scrambling is accomplished by modulo-2 addition (XOR),one input is a modulation symbol(19.2ksps) coming out of the block interleaver, another input is a random sequence, which created by decimator on long code generation. That means, Use the 64 times decimator to pick up the first chip of each 64 chips to form a random sequence. So the random sequence rate is 19.2kcps. (1.2288/64)


Spread Spectrum

In CDMA we use Spread code rate: 1.2288Mcps
Following codes are used to make Spread code
  • Forward Link: Walsh code
  • Reverse Link: Long PN code


Direct Sequence Spreading

Output of the randomizer is direct sequence spread by the long code
Each mobile station spreads its reverse traffic channel using the same long PN code but with a different offset, which is determined by a unique 42-bit mask.

The mobile station can use one of two unique long code masks:
  • A public long code mask based on the ESN
  • A private long code mask


Orthogonal Spreading
  • Each symbol output from the Mux is exclusive XORed by the assigned Walsh function
  • Walsh function has fixed chip rate of 1.2288 Mcps
  • Channels are distinguished from each other by Walsh function
  • Bandwidth used greatly exceeds source rate



Modulation

QPSK & OQPSK are the types of modulation
The forward traffic channel is combined with two different PN sequences: “I” and “Q”
Baseband filtering ensures the waveforms are contained within the 1.25 MHz frequency range
The final step is to convert the two baseband signals to radio frequency (RF) in the 800 MHz or 1900 MHz range

Quadri-Phase Shift Key (QPSK) Modulation
  • BASEBAND: The total frequency band occupied by the aggregate of all the information signals used to modulate a carrier
  • FILTER: Electronic circuit devised to modify the frequency distribution of a signal spectrum
  • BASEBAND FILTER: filter (used in quadrature modulation) that limits the input signal to the SyQuest band +-T/2, where T is the transmitted pulse rate.
  • GAIN CONTROL: the gain of the overhead channels (pilot, sync and paging) in the composite I and Q is set. The gain of each forward traffic channel is constantly adjusted by the reverse link power control process.

OQPSK
The reverse traffic channel data after direct sequence spreading is spread in quadrature by adding modulo-2.This stream with the zero-offset I and Q PN short code sequences is used on the forward CDMA channel.

Why a half chip delay in the Q Component?
The data spread by the Q PN short code sequence is delayed by half a PN chip time, 406.901 ns, with respect to the data spread by the I PN short code sequence. This prevents the I and Q to change value simultaneously, thus eliminating diagonal transitions

No comments:

Post a Comment