Lab 2 Blocking and Interference Statistics in Cellular Networks

The purpose of this lab demonstrate the interplay of noise and interference in wireless cellular systems using statistical models. You will use the the MATLAB based RUNE simulator that allow you to analyze larger cellular systems to produce statiscally more reliable results.

The lab consists of two parts. The rst part of the lab is a walk-through tutorial for the basic RUNE functionality. At the end of this part you are expected to understand which are the basic models available in RUNE and how  to use the RUNE functions and parameters in order to control the models.  In the second part of the lab you will perform analysis of the blocking and
interference statistics in a F/TDMA-based cellular network. The purpose of this part is to give an in-depth understanding of the fundamental trade-o between capacity and the quality provided to the served users.

  • Students should work in groups of 2
  • Book a timeslot for one of the lab session in the Lab booking system in Daisy 
  • Download the lab manual from the student download area
  • Write a small report in which you answer the questions, give solutions for the exercises, explain the procedure, etc. to be submitted on 27th of september.

Lärare Jens Zander skapade sidan 17 juli 2012

kommenterade 29 september 2012

In exercise 12, we have a reuse factor of 9. This means that km or lm can be equal to 0?

Anmäl missbruk
Lärare kommenterade 29 september 2012

Yes.

Also note that in question 12, "reuse 3" means that there are 3 cells in one cluster, "reuse 3/9" means there are 3 sites, 9 cells in one cluster. So, remember to use three sectors per site in the latter case.

Anmäl missbruk
kommenterade 29 september 2012

I have one problem with exercise 12 because when I set kn=4 ,ln=1 i get  NaN  as result in sta.iul  and  sta.sirul; morover this appears to be also the median (sta.sirul) and therefore plotting median of Sir to answer the question about the number of cluster we need to use then i have gap (no value)  for the combination (4,1) so 21cluster .....how is possible??? thanks for the reply  

Anmäl missbruk
Lärare kommenterade 30 september 2012

I don't know which parameters you used for cluster size, sps, offered traffic, etc, but my guess is that you are using a high reuse factor and high load value. Therefore some users in a cell cannot be assigned a channel because all channels in that cell were already in use. In that case this user will not be admitted to the system, so its interference and SINR values will be NaN. When you calculate the median of an array which includes NaN values, Matlab returns NaN. These NaN values cannot be plotted, therefore you see gaps in your plot. So, when you calculate the interference and SINR values, do not include the users which are not admitted to the system. See the help for "isfinite" function.

Anmäl missbruk
kommenterade 30 september 2012

to be brief this what my par contains at the begining: (the only diffrence is for that case Kn= 4 and ln =1)

par =

frametime: 1
nframes: 1
slottime: 0.12500000000000
nslots: 8
seed: 1
cellradius: 400
km: 1
lm: 1
kn: 1
ln: 1
sps: 1
gainconst: -28
noise: -128
alpha: 4
sigma: 0
raa: 0.50000000000000
corrdist: 110
offtraf: 10
mht: Inf
amean: 0.20000000000000
vmean: 1
homargin: 3
pinit: 33
pmax: 33
rbermax: 1
usefastf: 0
kpc: 20
usefh: 0

Anmäl missbruk
En användare har tagit bort sin kommentar
kommenterade 30 september 2012

In exercise 14, we have to make a suitable choice of number of clusters. This number should be the one we found in exercise 12? If not, what is the criteria to choose that number?

Anmäl missbruk
Lärare kommenterade 30 september 2012

Yes, you should use the result you obtained in exercise 12. If you think that your answer to ex. 12 may be problematic for some reason, use a slightly larger number than what you obtained in ex.12 so that you do not underestimate the effect of interference.

Anmäl missbruk
kommenterade 30 september 2012

Also in exercise 15 we have to use the same cluster number?

Anmäl missbruk
Lärare kommenterade 30 september 2012

Yes. When you run simulations in ex. 13, 14, 15 you need to use your result from ex.12.

Anmäl missbruk
kommenterade 1 oktober 2012

In question 15, in order to plot SIR, should we use cdfplot or cdfplotmed?

Anmäl missbruk
Lärare kommenterade 1 oktober 2012

Cdfplot and cdfplotmed do almost the same job. The difference is that cdfplotmed can handle matrix input and plots all colums as separate data series. Depending on how your input data is structured, you could use both functions.

Anmäl missbruk
Feedback Nyheter