Call Admission Control in Mobile Cellular CDMA Systems using Fuzzy Associative Memory Rupenaguntla Naga Satish Dilip Sarkar Computer Science University of Miami

1
Call Admission Control in Mobile Cellular CDMA
Systems using Fuzzy Associative Memory
Rupenaguntla Naga SatishDilip SarkarComputer
ScienceUniversity of Miami
2
Outline of the talk

• What is QoS?
• Review of CAC Algorithms
• Proposed CAC System
• Assumptions
• Design of FAM for Mobile Cellular Systems
• Operation of the FAM-based CAC System
• Evaluation of the FAM-based CAC System
• Conclusion

3
What is QoS?

• QoS parameters
• Blocking probability
• Handoff failure probability
• Forced termination or dropping probability
• All of these parameters cannot be improved
  simultaneously
• Our objectives
• Keep forced termination rate below a desired
  level
• Maximize channel utilization

4
Review of CAC Algorithms

• Handoff prioritization Schemes
• Guard channels
• Early channel reservation

5
Review Call Pre-blocking Scheme
New alls may be pre-blocking even when resource
is available
6
Review Call Pre-blocking scheme
Forced termination prob. at or below desired
level Very high channel utilization
7
Review Call pre-blocking scheme variable call
holding time
8
Review Call pre-blocking scheme Variable cell
dwell time
9
Call Preblocking Criteria for CDMA systems

• Pft (?, ?, µ, C)
• C is the capacity of a cell using CDMA and is
  given by
• C (W/Ra)/(Eb/N0) (NTW)/S 1, where
• W is the available bandwidth
• R is the data rate
• S is the signal strength
• NT is thermal noise spectral density
• Eb/N0 is the ratio of bit energy to noise power
• spectral density

10
Call Preblocking Criteria for CDMA systems

• TIM total interference margin
• TIM (C 1) NTW
• CIM current interference margin
• CIM IC((C 1) N)/(C - N)
• CIM lt TIM for avoiding outage of ongoing calls

11
Proposed CAC Variable Parameters

• Call arrival rate
• Call holding time
• Cell dwell time
• Number of channels/codes remains constant

12
Proposed CAC System using FAM
13
Design assumptionParameter values change
gradually, not abruptly
Pft
Load
14
Fuzzification of call holding time
Membership functions
15
FAM Fuzzy Associative Memory
16
Inference rules from the FAM

• IF ltfuzzy propositiongt THEN ltfuzzy
  propositiongt
• Example
• IF cell dwell time is VerySmall and call holding
  time is High THEN slope shall be VeryVeryHigh

17
Operation of the System

• Estimation of the parameters
• Fuzzification of call holding time and cell dwell
  time
• Firing of fuzzy rules using FAM inference
  mechanism
• Estimation of pre-blocking loads slope
• Computation of pre-blocking load from the
  estimated slope

18
Firing of Rules, and Inference

• Cell dwell time 18 sec, call holding time 120
  sec
• IF cell dwell time is Small and call holding time
  is Medium THEN slope shall be High

1.0
0.7
0.7
19
Defuzzification
20
Evaluation of the FAM-based CAC System variable
call holding time
21
Evaluation of the FAM-based CAC scheme Variable
cell dwell time
22
Conclusion and Future directions

• FAM-based CAC System can
• Keep forced termination at or below desired level
  FOR
• variable call arrival rate
• variable call holding time
• variable cell dwell time
• Future directions
• Extend for multiple service classes
• voice, data, video, and multimedia
• Extend for CDMA systems

23

• THANK YOU

24

• ?
• Comments!!!