a speed adaptive mobile internet protocol over wireless local a rea network jun tian nov. 15 th 2005...
TRANSCRIPT
A SPEED ADAPTIVE MOBILE INTERNET PROTOCOL OVER WIRELESS LOCAL
AREA NETWORK
Jun Tian
Nov. 15th 2005
Final Defense Presentation
1. Instruction2. Related Works 3. Performance of MIP over WLAN at
Different Speeds 4. Quantitative Analysis of the MIP over
Wireless LAN Handoff Latencies 5. A Speed Adaptive MIP and Its Performance
Evaluation 6. Summary and Future Works
Contents
Internet Users by Regions
Source: eTForecasts
1 MotivationTrend 1- more and more mobile users
The worldwide number of Internet users will reach nearly 1 billion in mid 2005. The U.S. continues to lead with over 185M Internet users forecasted for year-end 2004. There is little Internet user growth in the developed countries, but in the next five years many Internet users will be supplementing PC Internet usage with Smartphone and mobile device Internet usage. Internet usage is growing strongly in China, which surpassed Japan for second place in 2003. The growth of Internet users will continue in the developing countries for another decade.
“Much of future Internet users growth is coming from populous countries such as China, India, Brazil, Russia and Indonesia. These countries will also see strong growth of wireless web usage and for many new Internet users the cell phone will be their only Internet access device. ”
-- September 3, 2004 - Computer Industry Almanac
Wireless Internet Usage and Projections
Year-End 2001 2004 2007
Worldwide
Internet Users (millions) 533 945 1,460
Wireless Internet User Share
16.0% 41.5% 56.8%
USA
Internet Users (millions) 149 193 236
Wireless Internet User Share
4.5% 27.9% 46.3%
Asia-Pacific
Internet Users (millions) 115 357 612
Wireless Internet User Share
34.8% 50.9% 60.4%
W. Europe
Internet Users (millions) 126 208 290
Wireless Internet User Share
13.9% 49.6% 67.0%
Source: September 3, 2004 - Computer Industry Almanac
1 MotivationTrend 1- more and more mobile users
IP Backbone/InternetMobile IP
Multi-mode terminal w/MobileIP client& IPSec Client
PublicWLAN
802.11 Access Points
Ethernet
802.11 Access Points
Ethernet
WLANGateway, HA, FA
Home AAA ServerEnterprise
Corporate LAN
VPNFirewall
CDMA2000
Mobile devices can connect to office networks anytime from anywhere….through
CDMA
EDGEWCDMA
GSM
1 MotivationTrend 2- MIP connect the world
1 MotivationTrend 3- Moving speed accelerates wealth accumulation
Japanese Maglev Test speed 581 km/h (361 mph)
Shanghai China's maglev train: 429km/h (267mph )
France's Lignes a Grande Vitesse (LGV): commercial speed of 320km/h(198mph)
German maglev (magnetic levitation): 400 kilometers per hour(248mph)
High Speed Trains
1mbps—550m
11mbps—160m
6mbps—100m
36mbps—20m
1 MotivationTrend 4- small cell to keep a high throughput
802.11b data rate 1mbps 2mbps 5.5mbps 11mbps
Office distance 50m 40m 35m 25m
Outdoors distance 550m 400m 270 160m
Data rate vs distance
Small Wireless CellsAdvantage:•Higher throughput •More number of users
•Diameter of cell decreases by Nnumber of cells in a certain area increases by N²number of available channels increases by N²
Disadvantage:•Mobile host crosses cells more often
More number of handoffs
1 MotivationMobilityMobility
Outdoors
Indoors
Office
Fixed
Walk
Vehicle
0.10.1 11 1010 100 Mbps100 Mbps
Mobility vs data rateMobility vs data rate
Walk
WLAN
GSM
IS-95
IS-136
EDGE
WCDMA
CDMA2000
WPAN
Code Division Multiple Access 2000
Enhanced Data-rate for GSM Evolution
Wide band Code Division Multiple Access
WPAN: Bluetooth(802.15): 0.7-2 Mb/s data WPAN: Bluetooth(802.15): 0.7-2 Mb/s data rates,rates, Cable replacement, Short distances, up Cable replacement, Short distances, up to 10 m.to 10 m.
3G, Wide Area Network(WAN) coverage
WLAN: 802.11WLAN: 802.11
2G
802.11 n
Rapid Mobility?
Could Mobile IP combine all them together
and fulfill its job at high speed
2.1 Network Layer Handoff Management 2.2 Wireless LAN 2.3 Wireless LAN Handoff Management 2.4 Low Latency Handoff Mechanisms for MIP
over 802.11 Network 2.5 Location Tracking
2. Related worksMIP
Hierarchical MIP
Cellular IP
HAWAII
802.11, 11a, 11b, 11g standards
Define the MAC and PHY layer of OSI
Infrastructure, ad hocHandoff management frame
Handoff procedure
Techniques to Reduce IEEE 802.11 Handoff latency:.
Techniques to Reduce IEEE 802.11 Handoff latency
Mishra in [Mish03]: L2 handoff latencies discovery 90% and reauthentication 10%.
In [Mish04], Mishra proposed to reduce reauthentication latency by IAPP.
Kim in [Kim04] and Shin in [Shin04], propose selective scanning algorithms to reduce discovery latency. …..
These are L2 only algorithms
Malki in [Malk02] proposed two mobility protocols, pre- and post-registration, using L2 trigger.
In pre-registration, MN may communicate both oFA and nFA.
In post-registration, data cached in nFA before the registration completed.
These are cross layer solutionsGPS or signal strength to triangulate the location of MN
2.1 Network Layer Handoff ManagementMacro and Micro Mobility
Home Network
INTERNET
Home Agent
GFA
Micro-mobility handoff
Macro-mobility handoff
Macro mobility
Micro mobilitydomain
GatewayFA
Handovers in Micro Mobility transparent outside the domain
Mobile IP
HMIP
CIP
HAWAII
Register (HA)
Home Agent (HA)
Home Network
Correspondent Node (CN)
Foreign Agent (FA)
Foreign Network
Mobile Node(MN)
• When mobile node (MN) moves to a foreign network it obtains acare-of-address (COA) from the foreign agent (FA) that registers it with the home agent (HA)• COA is used by HA to tunnel packets to MN
SolicitationAdvertisement (FA,COA)
Register
Mobile Node
• Each mobile node has a home network, home address and home agent
Packets sent by MN godirectly to CN
- Triangle Routing
2.1 Network Layer Handoff ManagementMobile IP –Macro-mobility management
Specified in RFC3344- C. Perkins, Nokia Research Center, August 2002
- Reverse Tunnel
2.1 Network Layer Handoff ManagementHierarchical Mobile IP –Micro-mobility
• Tree-like structure of FAs • Hierarchical Tunneling • Registration is Regionalized inside a domain To eliminate Handoff Traffic•Regional Registration Request & Reply
FA1
FA2 FA3
FA4 FA5 FA6 FA7
MH
HA
INTERNET
MNFA1
MNFA2
MNFA4
MNCOA
MNFA3
MNFA6
MNCOA
MH MH
CN
MNCOA
MNFA5
Binding
Problems addressed in this dissertation
• Studying the effect of Rapid Mobility on the performance of Mobile Networking Protocol. In particular, we examine:– How speed affects the performance ?
• Performance/speed relationship
– What constructs the handoff latency ? • A global view of the handoff latency
– Breakdown the handoff latency to see: • Where does the latency come from ?• How much ?• What should we do with them ?• A deep examination of the handoff latency
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
Problems addressed in this dissertation
• Studying the effect of Rapid Mobility on the performance of Mobile Networking Protocol. In particular, we examine the effects of:–
• Performance/speed relationship
– What constructs the handoff latency ? • A global view of the handoff latency
– Breakdown the handoff latency to see: • A deep examination of the handoff latency
• Where does the latency come from ?• How much ?• What should we do with them ?
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
How speed affects the performance ?
Experimental Scenario: In this scenario, a rapid moving MN will travel trough 8 APs. Each AP is wired to a FA. The distance between every two consecutive APs is d= 250m, 500m or 1000m. The moving speed of MN is V, varying from 10m/s to 80m/s.
3. Performance at different speeds
FA1
HA
CN
AP1
internet
FA2
AP2
FA3
AP3
FA4
AP4
FA5
AP5
FA6
AP6
FA7
AP7
FA8
AP8
RAMON: a Rapid Mobile Network emulator
3. Performance at different speeds
Originally developed by Edwin A. Hernandez in 2002.
Rebuilt and extended by Jun Tian since 2003.
Architecture of RAMON
3. Performance at different speeds
FA1
192.168.1.1HUB1 AP1
192.168.1.3Attenuator1
Antenna1
ControllerFA2
192.168.2.1HUB2 AP2
192.168.2.3Attenuator2
Antenna2
FA3
192.168.3.1HUB3 AP3
192.168.3.3Attenuator3
Antenna3
Emulator
192.168.1.2
192.168.2.2
192.168.3.2
192.168.4.2
COM
COM
192.168.4.1
HA
10.3.3.14
Internet
MN
192.168.4.5
Attenuators are program controllable
Attenuators are program controllableAttenuators are pro-gram controllable
•manipulates the Attenuators
•A router
•Run emulation
•Control Attenuators
By increasing or decreasing the signal strength of one AP, we can emulate the MN moving towards or away from the AP. By varying the increasing or decreasing speed of the signal strength, we can emulate the speed change of the MN.
Speed 20 m/s distance 1000m
time-sequence graph throughput graph
3. Performance at different speeds
Speed 80 m/s distance 1000m
time-sequence graph throughput graph
3. Performance at different speeds
Speed10 distance 500m
time-sequence graph throughput graph
3. Performance at different speeds
Speed40 distance 500m
time-sequence graph throughput graph
3. Performance at different speeds
3. Performance at different speedsTable 1: Average Throughput at Different Speeds and AP Distances.
Speed (m/s)AP
distance (m)
Bytes transferred (kB)
Travel Time (s)
Average throughput (kB/s)
Total handoff time(s)
Effective time(s)
PMaxavg
(kB/s)
Handoff Rate
(FAs/s)
20 1000 78000 396 196.970 58 338 232.5 0.02
40 1000 33000 197 167.512 57 140 234.31 0.04
60 1000 16700 130.5 127.969 56 74.5 234.07 0.06
80 1000 9200 98.5 94.359 57 41.5 232.673 0.08
10 500 78500 397 197.733 58 339 233.01 0.02
20 500 33100 198 167.172 56 142 234.4 0.04
30 500 16600 129 128.682 56 73 232.86 0.06
40 500 9200 98 93.877 58 40 232.8 0.08
Conclusion 1:
•The total handoff time doesn’t change with speed
Conclusion 2 :
•Effective time/total travel time ratio drops when speed goes up
•This is the reason why high speed has low throughput
3. Performance at different speeds
3. Performance at different speeds
Average Throughputs vs Speeds.
Analysis: Let
Pavg--Average throughput
PMaxavg – Average throughput without handoff
Ttravel– Total travel time
Teffective – Total effective time for ftp transmission
Thandoff- Total handoff time while traveling
Khandoff – The number of handoffs while traveling
thandoff – Average handoff time among 7 times of handoff
Pavg = (Pmaxavg / Ttravel ) x Teffetive
= Pmaxavg (Ttravel – Thandoff ) / Ttravel
= Pmaxavg (1 – Thandoff / Ttravel)
= Pmaxavg( 1 – Khandoff x thandoff / Ttravle)
= Pmaxavg( 1 – (Khandoff / Ttravle ) x thandoff ))
3. Performance at different speeds
Conclusion 3: Above analysis Pavg = Pmaxavg( 1 – (Khandoff / Ttravle ) x thandoff ))
Conclusion 1 thandoff doesn’t change
The change of Pavg is caused by Khandoff / Ttravel ratio.
The performance of MIP over WLAN is related to the ratio of the number of handoffs/ total travel time, which is the MN handoff rate rh .
rh = v/d. The ratio of speed and cell size(AP distance).
This is different from what was previously believed.
3. Performance at different speeds
3. Performance at different handoff rate
At handoff rate 0.02 FAs/s, the average throughput is 197.35 kB/s . When the handoff rate goes up to 0.08 FAs/s, the average throughput drops to 94.118 kB/s
80
100
200
0 0.02 0.04 0.06 0.08
Handoff rate FA/s
Kby
tes/
sec
120
140
160
180
Problems addressed in this dissertation
• Studying the effect of Rapid Mobility on the performance of Mobile Networking Protocol. In particular, we examine the effects of:– How speed affects the performance ?
• Performance/speed relationship
– What constructs the handoff latency ? • A global view of the handoff latency
– Breakdown the handoff latency to see: • A deep examination of the handoff latency
• Where does the latency come from ?• How much ?• What should we do with them ?
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
Answer:
Pavg = Pmaxavg( 1 – rh x thandoff )) equation 2
Problems addressed in this dissertation
• Studying the effect of Rapid Mobility on the performance of Mobile Networking Protocol. In particular, we examine the effects of:– How speed affects the performance ?
• Performance/speed relationship
– • A global view of the handoff latency
– Breakdown the handoff latency to see: • A deep examination of the handoff latency
• Where does the latency come from ?• How much ?• What should we do with them ?
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
What constructs the handoff latency ?
4. MIP over WLAN handoff latency
thandoff = tL2handoff + tL3handoff + tL4handoff (equation 1)
1 MN moves, signal decay or signal interruption causes MN probe request to all channels, available AP’s probe response to MN, MN selects best nAP and sends Authentication request and after gets Authentication ACK sends Reassociation Request
and gets Reassociation Response. L2 handoff finished.
GFA
FA 1 FA 2
HA
CN
AP2AP1
MN
2. MN finds an nFA on its local network by the Agent Discovery process. After received 3 time of Agent Advertisement from nFA, MN sends Registration Request. nFA forward it to HA or GFA. HA or GFA reply with
Registration Reply. L3 handoff finished.
3 Then MN begin to recover interrupted
communication. L4 latency
Global view of MIP/WLAN handoff
Problems addressed in this research• Studying the effect of Rapid Mobility on the performance of Mobile
Networking Protocol. In particular, we examine the effects of:– How speed affects the performance ?
• Performance/speed relationship
– What constructs the handoff latency ?• A global view of the handoff latency
– Breakdown the handoff latency to see: • A deep examination of the handoff latency
• Where does the latency come from ?• How much ?• What should we do with them ?
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
Answer:
thandoff = tL2handoff + tL3handoff + tL4handoff (equation 1)
• Studying the effect of Rapid Mobility on the performance of Mobile Networking Protocol. In particular, we examine the effects of:– How speed affects the performance ?
• Performance/speed relationship
– What constructs the handoff latency ?• A global view of the handoff latency
–
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
4. Quantitative Analysis of the MIP over Wireless
LAN Handoff Latency A deep examination of the handoff latency
Breakdown the handoff latency to see:
•Where does the latency come from ?
•How much ?
•What should we do with them ?
MN
nAP
HA
nFA
L2 movement detection
MIP agent discovery
MIP registration
oAP oFA
MIP Advertisement
Registration Request
Registration Reply
Data packageL3 HO signalL2 HO signal
CN
TCP ACKProbe request
Probe response
Reassociation requestAuthentication
Reassociation responseL2 reassociation
TCP retransmission
L2 AP searchingL2
L3
L4
L2 movement detection delay: The strength of received signal degrades below a certain threshold, this may be caused by collision, radio signal fading, or AP is out of range. The STA(MN) detect the lack of radio connectivity based on weak received signal reported by the physical layer or failed frame transmissions. During this period, the TCP ACKs sent by MN to CN keeping lost, as well as the TCP packets from oFA to MN.
L2 searching delay: If transmission remains unsuccessful, the STA start scanning each channel by bradcasting a probe-request frame and waiting for probe respones.L2 reassociation: MN sends reassociation request to selected AP and waits for reassociation responseL3 agent discovery delay: MN got agent advertisement from nFA through nAP. If MN is using active mode, it will send agent solicitation ask for agent advertisement.L3 registration delay: MN sends registration request to HA or GFA and waits for reply.
L4 TCP retransmission belay: The retransmission timer at the sender is doubled with each unsuccessful retransmission attempt, in order to reduce the retransmission rate. Thus when the MN is reconnected, TCP will take a long time to recover from such a reduction and data will not be transmitted for a period of time. ( TCP exponential backoff retransmission policy)
4. Quantitative Analysis of the MIP over Wireless LAN Handoff Latency
A deep examination of the handoff latency
• Studying the effect of Rapid Mobility on the performance of Mobile Networking Protocol. In particular, we examine the effects of:
–
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
Breakdown the handoff latency to see:
•Where does the latency come from ?
•How much ?
•What should we do with them ?
Answer
tL2handoff = tL2detection + tL2seraching + tL2reassociation (equation 3)
tL3handoff = tmipagentdicovery + tmipregistration (equation 4)
tL4handoff = ttcp-back-off (equation 5)
thandoff = tL2detection + tL2seraching + tL2reassociation + tmipagentdicovery + tmipregistration + ttcp-back-off ( equation 6)
• Studying the effect of Rapid Mobility on the performance of Mobile Networking Protocol. In particular, we examine the effects of:– How speed affects the performance ?
• Performance/speed relationship
– What constructs the handoff latency ?• A global view of the handoff latency
–
• Designing and Implementing high performance and scalability solution or protocols for rapid mobility
Breakdown the handoff latency to see:
•Where does the latency come from ?
•How much are the latencies ?
•What should we do with them ?
4. Quantitative Analysis of the MIP over Wireless LAN Handoff Latency
A deep examination of the handoff latency
Experimental Scenario: In this scenario, a rapid moving MN will travel trough 8 APs. Each AP is wired to a FA. The distance between every two consecutive APs is d= 250m, 500m or 1000m. The moving speed of MN is V, varying from 10m/s to 80m/s. Pick out 20 experiments data.
FA1
HA
CN
AP1
internet
FA2
AP2
FA3
AP3
FA4
AP4
FA5
AP5
FA6
AP6
FA7
AP7
FA8
AP8
4. Quantitative Analysis of the MIP over Wireless LAN Handoff Latency
A deep examination of the handoff latency
exp#
Delay
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 avg avg
L2 movement detection 1.033
1.064
1.133
1.032
1.044
1.131
1.009
1.120
1.023
1.039
1.100
1.013
1.021
1.006
1.104
1.003
1.110
1.100
1.302
1.098
1.074
L2 AP searching 0.061
0.044
0.063
0.100
0.065
0.057
0.056
0.060
0.059
0.076
0.045
0.049
0.051
0.043
0.069
0.064
0.054
0.064
0.056
0.044
0.059
1.143
L2 reassociation 0.005
0.009
0.006
0.008
0.003
0.004
0.010
0.006
0.026
0.005
0.030
0.010
0.009
0.017
0.006
0.013
0.010
0.006
0.009
0.004
0.010
MIP agent discovery 2.996
1.945
3.023
2.563
2.756
2.578
2.436
3.001
2.213
3.008
2.770
2.545
3.001
2.600
2.598
2.674
2.783
3.012
2.349
2.404
2.660 2.746
MIP registration 0.073
0.042
0.052
0.050
0.052
0.043
0.060
0.704
0.054
0.053
0.041
0.042
0.065
0.046
0.047
0.062
0.054
0.057
0.070
0.062
0.086
TCP back_off 5.058
6.01 5.345
5.323
5.125
5.004
5.625
5.002
4.998
5.006
5.728
4.768
5.202
5.312
4.544
4.806
5.705
5.602
5.71 5.172
5.253
5.253
Handoff Latency 9.226
9.511
9.622
9.076
9.045
8.817
9.196
9.893
8.373
9.187
9.714
8.427
8.896
9.024
8.368
8.622
9.716
9.841
9.496
8.784
9.142
9.142
4. Quantitative Analysis of the MIP over Wireless LAN Handoff Latency
A deep examination of the handoff latency
MN
nAP
HA
nFA
1.074 L2 movement detection
2.660 MIP agent discovery
0.086 MIP registration
oAP oFA
MIP Advertisement
Registration Request
Registration Reply
Data packageL3 HO signalL2 HO signal
CN
TCP ACKProbe request
Probe response0.059 L2 AP searching
Reassociation requestAuthentication
Reassociation response0.010 L2 reassociation
5.253 TCP retransmission
L2 delay1.143
L3 delay2.746
L4 delay5.253
Handoff delay : 9.142
Conclusions:
•The largest latency is due to TCP, then Layer 3.
•If we reduce the L2 and L3 delay, L4 delay will be reduced exponentially.
•This dissertation focuses on L3 latency.
4. Quantitative Analysis of the MIP over Wireless LAN Handoff Latency
A deep examination of the handoff latency
5. Speed Adaptive Mobile IP
Handoff Rate rh = v/d MN moves through how many APs per second.
rh = Khandoff / Ttravel (equation 7)
Where Khandoff is the number of handoffs occurred during the MN traveling. Ttotal is MN’s total travel time.
Lower handoff rate has higher throughput
To reduce rh without changing total travel time, have to decrease the number of handoffs. The optimal is Khandoff = 0
Pavg = Pmaxavg( 1 – rh x thandoff )) (equation 2)
Let N be total FA numbers on the way MN traveling. Let’s assume somehow M is the number of FAs MN can communicate with without L3 handoff delay, which means M is the number of FAs MN has registered to at that moment.
The optimal is let M = N. But this costs too much resources, especially when the number of active MNs is large. Also we don’t know how long will the MN travel at the beginning.
5. Speed Adaptive Mobile IP
We call M the FA set size that MN registered.
Questions:
1. How to decide FA set size M
2. How to guarantee MN can communicate with a FA set almost like to do with a single FA.
1 hhandoff rtM
where thandoff is the handoff time, rh is the handoff rate. Here we use the experimental average handoff time 9.142s for thandoff. rh is dynamic. For example, at speed 40m/s, AP distance 500m, M = | 9.142 x 40/500 | + 1 = 2. At speed 80m/s, AP distance 500m, M = 3.
For Q1
5. Speed Adaptive Mobile IP
(equation 8)
For Q2, How to guarantee MN can communicate with a FA set almost like to do with a single FA.
Current FA pre-registers MN with M potential FAs to reduce L3 handoff latency, at the same time let IP packets be multicast to those M FAs in this FA set. So MN won’t feel any handoff delay at the IP level.
By above 2 steps, the set of FAs that MN can talk to without L3 latency was extended from one point in stationary state, to a line at high speed.
5. Speed Adaptive Mobile IP
rh = 0 0<rh < 0.109 0.109<rh< 0.218 0.218<rh< 0.328
M = 1 M = 2 M = 3 M = 4
FA Set Size vs Handoff Rate
Speed extension
MN’s registration message is extended by speed extension. According to Mobile IP Vendor/Organization-Specific Extensions[RFC3115]. Two Vendor/Organization Specific Extensions are allowed for MIP, Critical (CVSE) and Normal (NVSE) Vendor/Organization Specific Extensions. The basic difference is when the CVSE is encountered but not recognized, the message containing the extension must be silently discarded, whereas when a NVSE is encountered but not recognized, the extension should be ignored, but the rest of the Extensions and message data must still be processed. We use the NVSE extension.
5. Speed Adaptive Mobile IP
Normal Vendor/Organization Specific Extension
Type = 134 for NVSE extension.
Length is the size in bytes of the extension, not including the type and length bytes.
Verdor/org-ID is assigned in RFC 1700. set 5205.
Vendor-NVSE-Type Indicates the particular type of Vendor-NVSE-Extension. Set as 12 for speed extension.
Vendor-NVSE-Value: the value of MN moving speed.
5. Speed Adaptive Mobile IP
FA1
HA
CN
AP1
MN
internet
FA2
AP2
FA3
AP3
FA4
AP4
1
2
3
4
5
6 7
89
10 11
12
13
Whenever the MN needs to handoff to a new FA set, after it gets that many times of agent advertisements which is determined by speed(step 1),
it sends a registration request with up-to-date moving speed information to the very first FA in a new FA set (step 2). This FA will calculate the handoff rate and M
The first FA relays the registration request to upper FA or HA(step 3).
Meanwhile, it decapsulates the speed extension, refill the MIP header and authentication extension and then forward it to other FAs(M-1 FAs) in this FA set(step 4).
These other FAs relay the registration request to upper FA or HA as well(step 5).
When the GFA or HA received these registration requests, it builds up tunnels downwards to each FA and responses with registration reply(step 6 and 7). When the FA received the registration reply, it builds up tunnel upwards to the GFA or HA.
Whenever the MN setups the Link-layer contact with the FA, the later forwards the registration reply to the former(step 8, 9 or 10).The MN gets the care-of-address from agent advertisement message(step 10 or 9) or registration reply message(step 9 or 10), and begins data communication.
At the same time, it still sends registration request to the new FA set with up-to-date speed information (step 11). The very first FA in this set decupsulate the message and set up a new FA set. Forward registration request.(12, 13) and repeat the above process.
5. Speed Adaptive Mobile IP
Time-sequence graph at speed 60m/s and AP distance 1000m
Same scenario as above except SA-MIP is installed.5. Speed Adaptive Mobile IP
Time-sequence graph at speed 80m/s and AP distance 1000m
5. Speed Adaptive Mobile IPSame scenario as above except SA-MIP is installed.
Average throughput at different speeds and AP distances.
5. Speed Adaptive Mobile IP
Speed (m/s)
AP distance
(m)
Bytes transferred
(kB)
Travel Time(s)
Arg throughput
(kB/s)
Handoff Rate
(FAs/s)
20 1000 85000 399 213.03 0.02
40 1000 37500 198 189.39 0.04
60 1000 19400 130 149.23 0.06
80 1000 11600 99 117.17 0.08
10 500 84400 398 212.06 0.02
20 500 37400 198 188.89 0.04
30 500 19500 131 148.55 0.06
40 500 11500 98 117.34 0.08
Average Throughput of Speed-Adaptive MIP
5. Speed Adaptive Mobile IP
80
100
200
0 0.02 0.04 0.06 0.08
Handoff rate FA/s
Kby
tes/
sec
120
140
160
180
SA -MIP
MIP
(212.55 - 197.35) /197.35 = 7.69%
(189.14 - 167.34) /167.34 = 13.02%
(148.89 - 128.32) /128.32 = 15.97%
(117.25 - 94.12) /94.12 = 24.58%
220
6. Summery and Future Works
Contributions: Evaluate the rapid mobility of MIP over wireless LAN in a
laboratory environments. Depicted the relationship between the performance and the
moving speed of MN Quantitatively analyzed the handoff latencies of the MIP
over wireless LAN Speed Adaptive MIP is proposed and evaluated
Future WorksSpeed adaptive scheme should be applied to layer 2 and layer 4 handoff latencies.
Publications:
1. J. Tian, A. Helal, "Rapid Mobility of MIP over WLAN," International Conference on Computer Networks and Mobile Computing (ICCNMC'05), Zhangjiajie, China, Aug 2-4, 2005. Lecture Notes in Computer Science, Springer-Verlag GmbH, ISSN: 0302-9743, ISBN: 3-540-28102-9, Volume 3619/2005
2. Jun Tian and A. Helal, "Performance of MIP over WLAN in Rapid Moving Environments," The 4th ACS/IEEE International Conference on Computer Systems and Applications, Dubai/Sharjah, UAE, March 8-11, 2006
3. J. Tian and A. Helal, "Speed Adaptive MIP" submitted to the IEEE Wireless Communication&Network Conference, to be held in Las Vegas, NV, April 2006
4. J. Tian and A. Helal, "Speed Adaptive MIP over Wireless LAN," Submitted to Journal of Wireless Communications and Mobile Computing
• backups
1.1 Network Layer Handoff ManagementCellular IP –Micro-mobility(Campbell )
•Inside the CIP network, Uplink data packets are routed from MN to the gateway on a hop-by-hop basis. •The path taken by these packets is cached in route cache of base stations. •CIP uses mobile originated data packets to maintain reverse path. This path is used to route downlink packets addressed to a mobile host.
CIP routing
CIP BS
IP router
CIP Node
IP routing
IP tunnelling
HAMobile IP enabled Network
MN
GW
Cellular IP network
10Mb2ms
802.11
BS1
BS2
BS3
BS4
1.1 Network Layer Handoff ManagementCellular IP –Micro-mobility
HAMobile IP enabled Network
Active MN
GW
Cellular IP network
Idle MN
Paging update packet
Routing update packet
Routing and Paging Update
1.1 Network Layer Handoff ManagementCellular IP –Micro-mobility
MN listen/send to only one BS(TDMA network )
Hard Handoffpacket loss cannot be eliminated
MN listen/send to two or more BS simultaneously (CDMA network)
Semi-Soft Handoffgood performance eliminates packet loss
Handoff
Rebuilding of RAMON:
• Update hardwire: 3 FA, MN, new CISCO 350 APs, control board(Tarek Kaissi)
•Re-construct the testbed: NAT on HA, routing table on Emulator(help from Edwin)
•Software installation and configuration:
•Linux kernel 2.4.20,
•HUT dynamic MIP implementation version 0.8.1 (Helsinki University of Technology)
•Erase architecture delay
3. Performance at different speeds
L2 movement detection delay: The strength of received signal degrades below a certain threshold, this may be caused by collision, radio signal fading, or AP is out of range. The STA(MN) detect the lack of radio connectivity based on weak received signal reported by the physical layer or failed frame transmissions. During the period, the TCP ACK sent by MN to CN keeps lost.
L2 searching delay: If transmission remains unsuccessful, the STA start scanning each channel by bradcasting a probe-request frame and waiting for probe respones.
L2 reassociation: MN sends reassociation request to selected AP and waits for reassociation response.
tL2handoff = tL2detection + tL2seraching + tL2reassociation (equation 4)
L3 agent discovery delay: MN got agent advertisement from nFA through nAP. If MN is using active mode, it will send agent solicitation ask for agent advertisement.
L3 registration delay: MN sends registration request to HA or GFA and waits for reply.
tL3handoff = tmipagentdicovery + tmipregistration (equation 4)
L4 TCP retransmission belay: The retransmission timer at the sender is doubled with each unsuccessful retransmission attempt, in order to reduce the retransmission rate. Thus when the MN is reconnected, TCP will take a long time to recover from such a reduction and data will not be transmitted for a period of time. ( TCP exponential backoff retransmission policy)
tL4handoff = ttcp-back-off (equation 5)
thandoff = tL2detection + tL2seraching + tL2reassociation + tmipagentdicovery + tmipregistration + ttcp-back-off ( equation 6)
4. Quantitative Analysis of the MIP over Wireless LAN Handoff Latency
A deep examination of the handoff latency
FA1
HA
CN
AP1
MN
internet
FA2
AP2
FA3
AP3
FA4
AP4
1
2
3
4
5
6 7
89
10 11
12
13
Whenever the MN needs to handoff to a new FA set, after it gets that many times of agent advertisements which is determined by speed(step 1),
it sends a registration request with up-to-date handoff rate information to the very first FA in a new FA set (step 2).
The first FA relays the registration request to upper FA or HA(step 3).
Meanwhile, it decapsulates the speed extension, refill the MIP header and authentication extension and then forward it to other FAs(M-1 FAs) in this FA set(step 4).
These other FAs relay the registration request to upper FA or HA as well, just like the request comes from the MN(step 5).
When the GFA or HA received these registration requests, it builds up tunnels downwards to each FA and responses with registration reply(step 6 and 7).
When the FA received the registration reply, it builds up tunnel upwards to the GFA or HA. Whenever the MN setups the Link-layer contact with the FA, the later forwards the registration reply to the former(step8, 9 or 10).
The MN gets the care-of-address from agent advertisement message(step 10 or 9) or registration reply message(step 9 or 10), and begins data communication.
At the same time, it sends registration requests to the new FA with up-to-date speed information (step 11). This new FA decapsulates the message, sets up a new FA set, forwards the request(12,13)
and repeats the above process.
5. Speed Adaptive Mobile IP