A Congestion-Sensitive Model to Manage a NGN Environment with - - PowerPoint PPT Presentation

▶

Feb 07, 2023 274 likes •430 views

A Congestion-Sensitive Model to Manage a NGN Environment with Heterogeneous Mobile Customers Jose Moura Lancaster University Supervised by Dr. Christopher Edwards Agenda: Motivation How to deal with Congestion Work Plan NGN

SLIDE 1

A Congestion-Sensitive Model to Manage a NGN Environment with Heterogeneous Mobile Customers

Jose Moura Lancaster University

Supervised by Dr. Christopher Edwards

SLIDE 2

Jose Moura 2 Agenda:

Motivation How to deal with Congestion Work Plan NGN Model Business Model Network Selection Conclusions References

SLIDE 3

Jose Moura 3 Next Generation Network (NGN) is a highly

dynamic & competitive environment

Increase on the data traffic transported by mobile

broadband networks creates a huge stress on the deployed network infrastructure

I aim to address the NGN congestion on the edge

f the network

The scenario is a public area (e.g. train station)

covered by a heterogeneous wireless network that suddenly becomes congested due to a flash crowd

f users that are commuting

Motivation (1)

SLIDE 4

Jose Moura 4 If the deployed network infrastructure could not

conveniently support traffic congestion, then

Unsatisfied users choose an alternative operator Operator's reputation is degraded Operator's profit is reduced

Motivation (2)

SLIDE 5

Jose Moura 5 Some possible solutions

New access cellular technology (e.g. LTE [1]) Dynamic spectrum [2] Problem: Both very expensive because force

perator to buy new equipment

So, …

How to deal with congestion

SLIDE 6

Jose Moura 6 For modelling and solving the congestion problem

in the NGN environment:

Develop a model to understand how operators and

users should interact [3]

Study a new business model [4] Study a new network selection scheme [5]

Work Plan

SLIDE 7

Jose Moura 7

Develop a NGN model using Game Theory [3]

Operators and users have opposite goals

Identify the strategies for each player

operator accepts or blocks user
user stays attached, handovers or leaves

Handover management

signaling overhead
latency
trigger (network vs. terminal)

After the network becomes congested

new user user’s perceived quality

user churn

perator’s reputation operator’s profit

NGN Model (1)

SLIDE 8

Jose Moura 8

NGN Model (2)

0,00 5,00 10,00 15,00 20,00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 N Payoffs

a13 a23 b12 b13

SLIDE 9

Jose Moura 9 Model results [3] pointed out some stable vector

f players’ strategies, from which no player has

any incentive to choose a different strategy, but its stability depends on several aspects, namely network congestion

Results suggest that network stability depends

n the way the congestion is controlled

How to control the congestion?

NGN Model (3)

SLIDE 10

Jose Moura 10

Business Model (1)

Business model characteristics [4]:

uses a dynamic tariff sent to each user through a

DHCP/DNS/IPv6 header option mechanism

there is an incentive to enhance user cooperation

essentially when network becomes congested

supports heterogeneous users (leisure and

business users). Leisure users are price sensitive and business users are quality sensitive

finds the user distribution, constrained by channel

capacity, which optimizes operator’s profit

models how channel access contention at MAC

layer influences the user’s quality and operator’s profit

SLIDE 11

Jose Moura 11

Business Model (2)

Results [4] shown that

dynamic tariff offers a maximum operator's profit for

a specific channel capacity

5 0,0 1 00,0 1 5 0,0 200,0 25 0,0 1 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 Channel Capacity (Mbps)

S tat_ no p t S tat_ F a ir_n

t D yn_

SLIDE 12

Jose Moura 12

Network Selection

Alternative way to attenuate the congestion problem [5]

cooperation between operators + stored historic data

predicts congestion occurrence

handover of multimode handsets freeing some strategic

Network Attachment Points (NAPs)

some connections downgrade their mode of access technology

(e.g. from 3G to GSM)

new customer’s admission is managed between operators

network selection (alternative to signal strength) based on a

quality parameter used on each host

distribute the traffic load as best as possible between all local

available network attachment points not necessarily from the same technology

For example, beacon delay inspired in SIGCOMM paper [6]

SLIDE 13

Jose Moura 13

Conclusions

Current conclusions

GT identifies the best strategies for operators and users

when they interact in a NGN environment

Wireless congestion could be controlled by

business model
network selection scheme

Future work

We are evaluating how the network selection based on

quality deals with the congestion problem

We aim to model and study a business model with two

perators

SLIDE 14

Jose Moura 14

References

[1] Ekstrom, H.., et al., Technical Solutions for the 3G Long-Term

Evolution, IEEE Comm. Magazine, vol. 44, no. 3, pp. 38-45, 2006

[2] Zhao, Q., Sadler, B., A survey of dynamic spectrum access,

IEEE Signal Processing Mag., 2007

[3] Moura, J., Dunmore, M., Edwards, C., Next Generation Network

Management of Heterogeneous Mobile Users, In Proc. MobiWac,

pp. 111-118, 2008
[4] Moura, J., Dunmore, M., Edwards, C., A Novel Pricing

Approach to Support Heterogeneous Users in Wireless Networks, Under Submission to WMuNeP, 2009

[5] Moura, J., Dunmore, M., Edwards, C., NGN model with

heterogeneous mobile customers, Trilogy Summer School, 2009

[6] Vasudevan, S., et al., Facilitating access point selection in IEEE

802.11 wireless networks, In Proc. of ACM SIGCOMM Conference

n internet Measurement, 2005

SLIDE 15

Jose Moura 15