QoS Provisioning in WirelessMobile Networks

1
QoS Provisioning in Wireless/Mobile Networks

• Yaser Khamayseh
• Candidacy Committee
• E. Elmallah
• M. MacGregor
• M. Müller
• M. Buro
• S. Bates
• University of Alberta

2
Agenda

• Introduction
• Literature Review
• Research Directions
• Some Results
• Questions

3
Introduction

• Cellular Networks
• Mobility Models
• Call Admission Control Algorithms
• Scheduling Algorithms

4
Cellular Networks

• 1G (e.g., AMPS uses FDMA)
• Fixed cell Capacity
• 2G (e.g., GSM uses TDMA)
• 2.5G (e.g., GPRS)
• IMT-2000 requirements
• Soft cell capacity
• 3G (e.g., UMTS uses WCDMA)

5
Power Allocation in the Downlink

• Where,
• 1- is the received power from the serving
  BS, is estimated as the following
• 2- is the path loss from the serving BS to
  user i.

6
Power Allocation in the Downlink
7
Large Scale Path Loss Log Normal Shadow Fading
Model
Where, 1- is normally distributed with
mean 0 and standard deviation s 2- is
the average path loss, and it is computed
as 3- is the average path loss
at a reference distance
8
Mobility Models and Prediction

• Random Way Model
• A Model with User Residence and Movement
  Distributions

9
Call Admission Control Algorithms and Handoff

• Accept/Reject new requests
• Literature
• Guard Channel Posner-ITC-85
• Handoff Prediction/Dynamic reservation
  Levine-ITN-97,Hou-WC-01,Naghshineh SAC-96,
  Epstein-VTC-95, Xiao-ITN-01
• CDMA Networks Soh-INFOCOMM-04, Zhao-ITMC-03

10
Scheduling Algorithms

• Decide which user(s) will be served in the next
  time slot.
• Distribute resources among active users
• Example in the literature Cao-IP-01
• FPLS scheduler Huang ITMC-04
• WISPER Akyildiz-ITN-99

11
Literature Review

• QoS Provisioning using an Adaptive Framework
  Kwon-WN-03
• Down Link Scheduling in CDMA Data Network
  Joshi-MobiCom-00
• Dynamic Bandwidth Allocation with Fair Scheduling
  for WCDMA Systems Xu-WC-02

12
Literature Review QoS Provisioning using an
Adaptive Framework

• Adaptive framework
• Overload probability
• Target bandwidth
• Mobility Model
• Number of users in a region by the end of the
  prediction period is assumed to be binomial
  distribution.
• BAA
• Reduction
• Expansion

13
Literature Review Down Link Scheduling in CDMA
Data Network
14
Literature Review Down Link Scheduling in CDMA
Data Network

• Cost Functions

15
Literature Review Dynamic Bandwidth Allocation
with Fair Scheduling for WCDMA Systems

• GPS scheduler
• BAA algorithm
• Estimate the backlogged rate
• Calculate Si(k) (for not backlogged users) to the
  minimum rate guarantee.
• Assign rates for the users
• Distribute the remaining resource between users
  fairly
• Leaky Bucket regulator

16
Framework
Multimedia Server
17
Total Frame Delay Ratio

• Denote the number of time slots
  during which frame f of connection i is delayed.
• is bounded by .
• The total frame delay ratio for
  connection i.
• is bounded by .

18
Total Frame Delay Ratio (TFDR)
System Capacity C 3 Channels
t 1
time
19
Frame Delay (FD)
t 1
time
20
Research Directions

• Optimal resource management algorithms for fixed
  capacity cell
• Optimal resource management algorithms for soft
  capacity cell
• Scheduling Algorithm
• Prediction-based CAC for soft capacity cell

21
Research Directions Optimal Algorithms Fixed
Cell Capacity

• System capacity C
• Time is slotted
• Each user is assigned 1 channel per
• , , and are know
  for each connection.

22
Research Directions Optimal Algorithms Fixed
Cell Capacity
Users
t 1
time
1- Delay Frames 2- Reject Connection
23
Research Directions Optimal Algorithm Fixed Cell
Capacity

• Maximize Effective Throughput
• subject to
• Capacity constraint System capacity is C
  channels.
• For each connection i the
• For each frame in each connection
• Only time units in the future is
  considered by the algorithm.

24
Research Directions Optimal Algorithm Fixed
Cell Capacity

• I propose to investigate
• The computational complexity of the problem.
• Useful algorithms for solving the problem.
• The design and performance of an on-line resource
  algorithm to deal with the MAX-ET-TFDR problem
  for fixed capacity system
• Compare the performance achieved by the on-line
  algorithm proposed in step 3 versus the more
  idealized algorithm obtained in step 2.

25
Research Directions Optimal Algorithms Soft
Cell Capacity

• Soft System capacity (Power and Interference
  Limited)
• Time is slotted
• , , and are know
  for each connection.
• Path Loss values for each connection

26
Research Directions Optimal Algorithm Soft Cell
Capacity

• Similar to the previous formulation
• The system is power and interference limited
• Uses the power equation to estimate the required
  amount of power for transmitting to any user

27
Research Directions Optimal Algorithm Soft Cell
Capacity
Users
time
t 1
1- Delay Frames 2- Reject Connection
28
Research Directions Scheduling Algorithm

• Soft Capacity System
• Path Loss for each user at the beginning of the
  time slot
• TFDR and FD

29
Research Directions Scheduling Algorithm

• Actions
• Estimate the required transmit power for each
  user
• If no enough resources
• Decide on which connection to delay

30
Research DirectionsScheduling Algorithm

• Minimize Forced Termination
• subject to
• Capacity constraint (power constraints).
• For each connection
• For each frame in each connection

31
Research Directions Scheduling Algorithm

• I propose to investigate
• The design of an ideal scheduler for the above
  mentioned problem.
• The computational complexity of the problem.
• The design of an effective heuristics algorithm.
• Integrate the heuristics developed in the
  previous step with a CAC algorithm.

32
Research Directions Prediction-based CAC

• Soft Capacity System (CDMA)
• Mobility Model
• Estimate the overload probability by the end of
  the prediction period.
• User stays in the same region with probability Pc
• User moves to the right region with probability
  Pr
• User moves to the left region with probability Pl

33
Research Directions Prediction-based CAC
Mobility Model
Region M
Region 1
34
Research Directions Prediction-based CAC
Transition Diagram
35
Research Directions Prediction-based CAC

• All users in the same region are assumed to
  experience the worst case path loss in their
  respective region.
• Cell Configuration
• Feasible Configuration (i.e., The cell has enough
  power to serve all users).
• Infeasible Configuration (i.e., The desired power
  is more than the available power at the BS).

36
Research Directions Prediction-based CAC

• Accept new connection if
• Use limited sampling to estimate overload
  probability.

37
Some Results

• Prediction-based CAC results
• Percentage of Completed calls
• Percentage of Blocked calls
• Percentage of Forced Terminated calls
• Effective Throughput

38
Results (CAC)
39
Results (CAC)
40
Results (CAC)
41
Results (CAC)
42
References

• Joshi-MobiCom-00 N. Joshi, S. Kadbas, S. Petel,
  and G. Sundaram. Down link scheduling in cdma
  data network. In MobiCom 2000, August 2000.
• Kwon-WN-03 T. Kwon, Y. Choi, C. Bisdikian and
  M. Naghshineh. Qos provisioning in
  wireless/mobile multimedia networks using an
  adaptive framework. Wireless Networks, 951-59,
  2003.
• Xu-WC-02 L. Xu, X. Shen, and J. Mark. Dynamic
  bandwidth allocation with fair scheduling for
  wcdma systems. IEEE Wireless Communication, pages
  26-32, April 2002.
• Posner-ITC-85 E. C. Posner and R. Guerin.
  Traffic policies in cellular radio that minimize
  blocking of handoff calls. In Proc. 11th ITC,
  Kyoto, Japan, 1985

43
References

• Soh-INFOCOMM-04 W. Soh and H. Kim. Dynamic
  bandwidth reservation in cellular networks using
  road topology based mobility prediction. In
  INFOCOMM 04, 2004.
• Huang ITMC-04 V. Huang and W. Zhuang.
  Qos-oriented packet scheduling for wireless
  multimedia cdma communications. IEEE Transaction
  on Mobile Computing, 373-85, January-March 2004.
• Levine-ITN-97 David A. Levine, Ian F. Akyildiz,
  and Mahmoud Naghshineh. A resource estimation and
  call admission algorithm for wireless multimedia
  networks using the shadow cluster concept. IEEE
  Transactions on Networking, February 1997.

44
References

• Hou-WC-01 J. Hou and Y. Fang. Mobility-based
  call admission control schemes for wireless
  mobile networks. Wireless Communications and
  Mobile Computing, 1269-282, 2001.
• Akyildiz-ITN-99 I. Akyildiz, D. Levine, and I.
  Joe. A slotted cdma protocol with ber scheduling
  for wireless multimedia networks. IEEE
  Transactions on Networking, 7146-158, April
  1999.
• Naghshineh SAC-96 M. Naghshineh and M.
  Schwartz. Distributed call admission control in
  mobile/wireless networks. J. Selected Areas
  Communications, May 1996.
• Epstein-VTC-95 B. Epstein and M. Schwartz.
  Reservation strategies for multimedia trac in a
  wireless environment. In 45th IEEE Vehicular
  Technology Conf. (VTC'95), July 1995.

45
References

• Xiao-ITN-01 M. Xiao, N. Shro, and E. Chong.
  Distributed admission control for
  power-controlled cellular wireless systems.
  IEEE/ACM Transactions on Networking, 9790800,
  December 2001.
• Zhao-ITMC-03 D. Zhao, X. Shen, and J. Mark.
  Radio resource management for cellular cdma
  system supporting heterogeneous services. IEEE
  Transaction on Mobile Computing,
  2147-160,April-June 2003.
• Cao-IP-01 Y. Cao and V. Li. Scheduling
  algorithms in broad-band wireless networks. IEEE
  Proceedings of the IEEE, 8976-86, January 2001.

46
(No Transcript)