Transcript Base Station

Cross-Layer Design
-Kalpana Uppalapati
Agenda:








Introduction
Problem definition (3 issues)
Background
Description of the 3 issues
Solutions for the issues
Results of the methods used
Future Research
Conclusion
Introduction:


Wireless Networking:
o Applications: wireless Internet access, ad hoc
networks, sensor networks
o Diverse requirements: high-bandwidth video and
data, low-bandwidth voice and data
o Goal: reliable communication-on-the-move in highly
dynamic environments, QoS provisioning
A central problem: How to efficiently transmit
heterogeneous traffic over wireless links?
Why CLD?

Design Approach:
•
•

Layered design
Each layer can be optimized independently
Changes in one layer do not require changes in the
other layers
Cross-layer design
Layered approach is not adequate in wireless!!
•
•
Inflexible
No Adaptation
Would the advantages of cross-layer design
lead to new network architecture?
Looking at the Status of Lower Layer
May Improve Performance
send
receive
get
status
Expose lower layer status may improve performance
Configuring the Lower Layer to Improve
Performance
send
receive
get
set
status
Control lower layer behavior may improve performance
Examples of Cross-layer Design
Application+Link+Physical: instead of
using “target bER” power adaptation,
consider modifying the target bER based
on application requirements
APPLICATION
TRANSPORT
ROUTING/CONTROL PLANE
Application+Link:perf. gain in e2e
SNR by joint channel coding and
compression for video over wireless
MAC
LINK
Routing+Link: perf. improvement in
percentage of packets not meeting a
delay deadline through joint rate
allocation and routing (10x)
PHYSICAL
MAC+Link: multiuser diversity gain
of throughput if scheduling is based
on link availability instead of FIFO (2x)
(MAC+PHY) perf. gain in channel
utilization through channel
reservation based on physical layer
parameters
MAC+Routing: perf. gain in channel
capacity via joint MAC/routing
protocol design. 3x difference for
802.11
Workdone:



At application-layer, a media server can track packet
losses and adjust media source rate accordingly.
At transport-layer, several cross-layer approaches, such
as EBSN, Snoop TCP, and freeze TCP, have been
proposed as TCP alternatives to distinguish congestion
loss from non-congestion loss and invoke different flow
control mechanisms.
At link and network-layer, the persistence level of the
MAC layer ARQ mechanism should adapt to each
application’s latency and reliability requirements.
3 ISSUES:
Issue 1:
Cross-layer Design to achieve Quality of Service and
resource efficiency in a multi-user context(PHY+MAC+App)
Issue 2:
TCP–over Wireless CDMA Links;SnoopTCP(Link+Transport)
Issue 3:
Cross-Layer Approach for Video over Time-Varying CDMA
Channels
ISSUE 1
ISSUE:
A joint cross-layer design for achieving QoS
(Quality of Service) and Resource Efficiency
Proposed Scheme:
QoS-awareness scheduler and power adaptation scheme
is used at both uplink and downlink MAC layer to
coordinate the behavior of the lower layers for resource
efficiency and QoS
• Central to the proposed cross-layer design is the concept
of adaptation
•
•
•
•
•
The conventional protocol stack is inflexible and the
layers doesn’t adapt to the changing conditions.
Adaptation represents the ability of the network
protocols to observe and respond to channel variation
Central to adaptation is the cross-layer design
The proposed QoS-awareness scheduler and power
adaptation scheme deals with Application, MAC and
physical layers
Purpose of a Scheduler:

MAC scheduler selects
appropriate transmission
power/format and priorities of
the packets for each user
depending on its present
channel condition and the
associated QoS requirements


At MAC layer, different QoS-aware MAC states for both
uplink and downlink transmission are used. The IP-base
station dynamically schedules users in different MAC
states based on resource availability and overall QoS
QoS (combined QoS criteria) = QoS class* QoS stream
QoS class is determined by service pricing
QoS stream is determined by characteristics of data traffic
QoS-aware & Power-adaptive MAC States
a. High-QoS state: Users are actively sending and receiving
traffic. In the high-QoS state, users have a dedicated
control channel with both control and traffic channel
power and timing controlled. Traffic segments are
instantaneously assigned to any high-QoS user when
there is data to send or receive
b. Media-QoS state: While in media-QoS state, users have
contention-free uplink request slots to indicate to an IPbase station that they have data to send. Users also
have shared downlink message slots that are timing
c. Low-QoS state:
Users only maintain connectivity to an IP-base station.
Low QoS mode users have shared downlink paging slots,
where they know to wake up periodically and listen for
incoming pages from the IP-base station
Transition among different QoS states
QoS-aware & Power-adaptive
Transmission:

Different applications have
different QoS network
requirements. QoS vector
constitutes Cost network of a
stream .The network parameters
combined with QoSclass
andQoSstream compose a multi
criteria decision for a given user
and overall cost measurement is
defined as:
Cost =
QoSclass ∗ QoSstream ∗ Costnetwork
QoS-aware and Power-adaptive
Transmission:


The MAC scheduler selects appropriate transmission
power/format and priorities of the packets for each user
depending on its present channel condition and the
associated QoS requirements.
MAC scheduler dynamically computes Cost real based on
the current channel condition.
Simulation Results:
Four different SNR environments,
Location A(Dis=0.5Miles & SNR=20dB)
B (Dis=1.4 Miles & SNR=20dB)
C (Dis=2.0 Miles & SNR=15dB)
D(Dis=1.6 Miles & SNR=10dB), were located within one
cell.
Three applications are chosen to run the test:
Mobile1: A FTP download of 50 distinct 2MB files (B-C)
Mobile 2: A webpage of 205kb was periodically refreshed
(C-D)
Mobile 3: A 128 kb/s media stream served from within the
core network (D-A)

Test #1:
Equal QoS assignment, they
simultaneously request resources.
QoS “Gold” user (mobile#1)
with FTP download
Mobile #1 has throughput 2.5Mb/s
initially
Mobile#2 has 160Kb/s initially
When Mobile #3 comes,
MAC scheduler arbitrates
resource assignments so that
all users are supported.
So, Mobile#1-dropped to 1.4Mb/s
When Mobile #3 completes ,Mac
Scheduler reallocates
bandwidth back to Mobile#1

Test #2:
When different QoS are
assigned to mobiles,
Mobile #1 was assigned low
QoS
Low QoS=> never receive more
than 150Kb/s
Figure shows throughput drop
The available throughput to the
user was reduced from 2.6Mb/s
to 150Kb/s
Issue-2:TCP Over CDMA Wireless
Links
TCP
has been optimized for wired networks. Any packet
loss is considered as a congestion and hence the
window size is reduced dramatically as a precaution,
however wireless links are known to experience sporadic
and usually temporary losses due to fading, shadowing,
handoff etc. which cannot be considered as congestion.
Problem: wireless corruption mistaken for congestion
Solution: Snoop Protocol
Channel Errors
Internet
Router
Loss  Congestion
Burst losses lead to coarse-grained timeouts
23
2121
Loss ==> Congestion
Result: Low throughput
0
Performance Degradation
Sequence number (bytes)
2.0E+06
Best possible
TCP with no errors
(1.30 Mbps)
1.5E+06
TCP Reno
(280 Kbps)
1.0E+06
5.0E+05
0.0E+00
0
10
20
30
40
50
60
Time (s)
2 MB wide-area TCP transfer over 2 Mbps Lucent WaveLAN
Cross-layer protocol design &
optimizations


At the transport-layer, several cross-layer approaches
such as freeze TCP, snoop TCP have been proposed as
TCP alternatives to distinguish congestion-loss from noncongestion loss.
In snoop TCP, TCP layer knowledge is used by link layer
schemes
Transport
Network
Transport-aware link
(Snoop agent at BS)
Link
Physical
Link-aware transport
(Explicit Loss Notification)
Solution: Snoop Protocol

Shield TCP sender from wireless vagaries


Eliminate adverse interactions between protocol
layers
Congestion control only when congestion occurs
Fixed to mobile: transport-aware link protocol
Mobile to fixed: link-aware transport protocol
Snoop Protocol: FH to MH
6
4 3 2
1
Snoop agent
5
Base Station
FH Sender
1
Snoop agent: active interposition agent



Snoops on TCP segments and ACKs
Detects losses by duplicate ACKs and
timers
Suppresses duplicate ACKs from FH sender
Cross-layer protocol design: snoop agent
Mobile Host
Snoop Protocol: FH to MH
1
Snoop Agent
Base Station
FH Sender
Mobile Host
5
Snoop Protocol: FH to MH
4
3
2
1
Base Station
FH Sender
Mobile Host
Snoop Protocol: FH to MH
6
4 3 2
1
5
Base Station
FH Sender
1
Mobile Host
6
Snoop Protocol: FH to MH
4 3 2
5
Sender
1
Base Station
3
2
21
Mobile Host
Snoop Protocol: FH to MH
5 4 3 2
1
6
Base Station
4
3
Sender
2
Duplicate ACK
ack 0
Mobile Host
1
Snoop Protocol: FH to MH
6 5 4 3 2
1
6
Base
5 Station
1
Sender
Retransmit from cache
at higher priority
ack 0
4 3 2
ack 0
ack 0
Mobile Host
1
Snoop Protocol: FH to MH
6 5 4 3 2
1
Base Station
5
Sender
ack 0
Suppress
Duplicate Acks
1 4 3 2
ack 4
Mobile Host
1
Snoop Protocol: FH to MH
6 5
Clean cache on new ACK
Base Station
6
Sender
ack 4
5 1 4 3 2
ack 5
Snoop Protocol: FH to MH
6
Base Station
Sender
ack 4
ack 5
6
1 5 4 3 2
ack 6
Mobile Host
Snoop Protocol: FH to MH
7
9
8
Base Station
Sender
ack 5
ack 6
6
Active soft state agent at base station
Transport-aware reliable link protocol
Preserves end-to-end semantics
1 5 4 3 2
Mobile Host
Snoop Performance
Improvement
Sequence number (bytes)
2.0E+06
Best
possible
TCP
(1.30
Mbps)
1.5E+06
Snoop (1.11 Mbps)
TCP Reno
(280 Kbps)
1.0E+06
5.0E+05
0.0E+00
0
10
20
30
40
50
60
Time (s)
2 MB wide-area TCP transfer over 2 Mbps Lucent WaveLAN
ISSUE 3:Cross-Layer Approach for
Video over Time-Varying CDMA
Channels
Problem:
Applying multi-user diversity to real-time traffic is very
challenging due to the delay requirement of such traffic
Solution:
A Cross-Layer approach,DWGPS(Dynamic Weight
Generalized Processor Sharing) is proposed.
The proposed cross-layer approach can benefit from
information in both the application and physical layers
Existing Technology:GPS
•
•
•
GPS is an ideal fair scheduling discipline originally
proposed for wire line networks
Principle of GPS:
A fixed-weight is assigned to each session and bandwidth
is allocated to all sessions according to the weights and
traffic loads.
GPS can provide each session with a minimum service
rate.
•
•
•
•
The minimum service rate and tight delay bound
guaranteed in GPS may seem attractive to real-time
video transmission
A large weight should be assigned to a video session in
order to guarantee the peak rate
This means a video session will get a large portion of the
total capacity whenever it has traffic to transmit, thus
leading to service degradation of other sessions
In order to apply GPS discipline to video transmission
and extend it to wireless networks, dynamic weights in
GPS are used(DWGPS)


The link layer resource allocation benefits from the video
compression information such as the batch class and
batch arrival size from the application layer; and in
return, the link layer provides the video compression
with the desired service differentiation.
This DWGPS scheduling procedure in the link layer can
achieve low computation complexity and small signaling
overhead.
Application Layer
Video compression
information
Provides video
compression for the
desired service
LINK Layer
Conclusion:
Though Cross-layer design has many advantages the
Achilles heel of cross-layer design is its potential to destroy

modularity, and make the overall system fragile.
Tradeoff between efficiency and modularity
 Future work on CLD implementation can be aiming at
replacing the traditional layered structures completely.
However this might not even be possible because of the
demand for compatibility with every other network using
the IP.
References:
ISSUE 1:
http://ieeexplore.ieee.org/iel5/35/27698/01235598.pdf?tp=
&arnumber=1235598&isnumber=27698
http://ieeexplore.ieee.org/iel5/79/29371/01328087.pdf?tp=
&arnumber=1328087&isnumber=29371
 ISSUE 2:
http://ieeexplore.ieee.org/iel5/9255/29377/01329247.pdf?t
p=&arnumber=1329247&isnumber=29377
http://ieeexplore.ieee.org/iel5/9179/29132/01313348.pdf?t
p=&arnumber=1313348&isnumber=29132
 ISSUE 3:
http://ieeexplore.ieee.org/iel5/35/33162/01561929.pdf?tp=
&arnumber=1561929&isnumber=33162
