ELEC ENG 3108 Telecommunications Principles 3: Wideband CDMAMatlab

Java Python ELEC ENG 3108 Telecommunications Principles

Assignment 3: Wideband CDMA

Due 23:59 Monday 6 November – via MyUni only. This assignment is worth 50% of your final grade. What to submit

□   Your working and results

□   Some of your marks will be allocated to discussion, insight and original experimentation or analysis. Your report is expected to be written to a professional standard and demonstrate your personal critical insight.

Question 1 (15/50): WCDMA Planning

Part 1: Implement the Walfisch-Ikegami path-loss model as a function in Matlab. Two functions are required:

(a)  Given environmental parameters and distance, calculate the mean path loss

(b)  Given environmental parameters and the maximum mean path loss, calculate the maximum coverage distance

Part 2:

Using the link budget data given in the notes and these link budgets as a guide, use your path loss model to explore link budgets under different scenarios. You should extend your software to consider data rate, indoor versus outdoor versus vehicular use, etc.

Part 3:

Consider the peak up-link capacity of a third generation W-CDMA cell carrying three types of traffic with the following characteristics:

Service	Bearer data Rate Rj	Activity Factor vj	Bit Energy to Noise Ratio (Eb/N0)j	Number of Channels in use
Video phone	64 kbit/s	100%	1dB	3
File upload	384 kbit/s	25%	3dB	4
Voice	12.2 kbit/s	67%	1dB	Unknown

•     Interference ratio from other cells i=0.65

•     Maximum noise rise = 3 dB

•     Assume that signalling occupies the equivalent of one voice circuit

•     Chip Rate W=3.840 Mchip/s

a) Calculate the Load Factor for each of the 3 services.

b) Hence determine the maximum number of channels available for voice on this cell and calculate the aggregate user data rate on the uplink.

c) Assuming that the following channel data rates are used for each application and that all three services are in use as above, determine the number of Orthogonal Variable Spreading Factor codes available for voice circuits.

– Video phone: 240 kbit/s

– File upload: 480 kbit/s

– Voice: 30 kbit/s (6 marks)

d) How many voice circuits would be available if the cell were only used for voice?  Also calculate the aggregrate up-link user data rate in this case.

Question 2: WCDMA Implementation (35/50)

The aim of this question is to decode the WCDMA signal in the attached Matlab file. The signal consists of

□   An unknown number n of blocks, each of 38400 chips (10ms), a total of 10n ms of signal

□   A normalised noise-free single path downlink signal (perfect path assumption)

□  ELEC ENG 3108 Telecommunications Principles Assignment 3: Wideband CDMAMatlab   A 38400 chip scrambling code which is the real part only of one of the 512 primary downlink scrambling codes Sdl,n where n is in [0..511] as defined in the standard.  The actual scrambling code used is unknown.

□   A strong common pilot channel under the scrambling code. The power level of the pilot channel is unknown.

□   An unknown number of traffic channels, each with a spreading factor between 16 and 512 containing ASCII text. The actual spreading codes used are unknown.

Your challenge is to:

□   identify the scrambling code and signal strength of the Pilot CPICH

□   identify the valid channel codes and their respective signal strength

□   decode the traffic channels and read the enclosed message

□   from what you learn, demonstrate some open-ended investigation. It is quite acceptable to collaborate and for different members of the class to explore different aspects of WCDMA including multipath, additive noise, encoding, etc.

Some things you need to know:

□   The scrambling codes, OVSF codes and common pilot channel (CPICH) have been generated consistent with 3GPP Technical Specification 25.213 Release 4. This standards document is  available on the MyUni website. Pay particular attention to sections 4.3.1, 5.2.1 and 5.2.2.

□   The only common channel is the CPICH.  No other common channels (such as synch channel etc) are in the signal.  So you can identify the scrambling code by correlating the 512 primary codes against the signal to see which one has the highest correlation. Only the real component of the downlink scrambling code is used.

□   The signal is synchronised - that is the first sample corresponds to the beginning of a 38400 chip frame.

□   The traffic channel is modulated with BPSK.  A positive correlation (+1) corresponds to "1", a negative correlation (-1) to "0".  No correlation (0) indicates that the channel is not present.

□   Valid traffic channels contain only 8-bit ASCII characters, beginning with the most significant  bit.  The first two characters are an ASCII representation of numerical digits, corresponding to a number between "01" and "99".  The third character is a space.  This will be sufficient to identify valid channels and to put the traffic channels in the correct order to read the message.

□   The message in the signal is up to 99 lines of ASCII text, each of up to 1000 characters

(including padding in the form. of spaces (character 32)) and beginning with the 2-digit and space header. Control characters such as Line Feed (10) and Carriage Return (13) may be included.

□   The traffic channels all begin at the start of the first frame but do not continue for the entire signal.  In particular, signals with fast codes will finish early.

Marking:

□   20 marks will be based on your technical answers, including your code, signal analysis and the encoded message.

□   5 marks will be based on how well you explain your approach.

□    10 marks will be based on any additional detail, analysis or experimentation you describe in your report. For example, you might want to explore how the exercise compares with the real world (added noise and interference, multipath etc) and see how your decoder works when you add noise to the signal.  You might also like to discuss computational complexity.

Remember, I'm not just looking for some code and some numbers. I want a clear description of how you solved the problem and some exploration of WCDMA using the tools provided by this assignment.

Don't panic! The standard is very explicit in its description of how to generate the scrambling and channelisation codes. One important thing to note is that Matlab starts its vector indices from 1; the standard is consistent with C and starts its vector indices from 0, so you'll need to allow for this.

Some helpful code to get started:

function string=bits2str(bitsequence);

bitsequence=(bitsequence>0.001);

% note: +1 = binary 1, -1 = binary 0; 8-bit ASCII, convert to

% 0 or 1

block=char(reshape(bitsequence,8,length(bitsequence)/8)+48)';

string=char(bin2dec(block));

end

%--------------------

function data=recover(spread,codenumber,tx,scramblecode,nbytes);

data=zeros(1,8*nbytes);

channel=ovsf(spread,codenumber); %you'll need to write the ovsf fn

for j=1:nbytes*8;

k=(j-1)*spread+1;

data(j)=sum(tx(k:k+spread-1).*scramblecode(1+mod(k-1,38400):...

1+mod(k+spread-2,38400)).*channel);

end

The signal file includes the signal (transmitsignal) and also scrambling code 0

(samplescramblingcode0) so that you can verify that your scrambling code function is working properly