Seamless Handoff in Personal Area Networks Ben Cheung Duke Nguyen December 12, 2003.

Seamless Handoff in Personal Area Networks

Ben Cheung

Duke Nguyen

December 12, 2003

Presentation Outline

General Overview Design Test Cases Performance and Experimental Results Future work to be done

What is a Personal Area Network?

A personal area network (PAN) is a computer network used for communication among computer devices close to one person.

The device may or may not belong to the person in question. A key capability for a PAN is to enable devices to autonomously detect and acquire one another

Wireless PANs can be made possible with network technologies such as IrDA, Bluetooth, and 802.11.

What is seamless handoff?

End-to-end IP continuity without failures in the midst of a download activity due to link outages or handovers

A vertical roaming scheme for switching across alternative wireless technologies (e.g., WiFi, WPAN and cellular)

Ensures a successful termination of a data transfer even when:

– The underlying link connections are damaged or disrupted due to the device mobility

– A mobile user passes through different alternative wireless technologies

Seamless Handoff Goals

Continuous connectivity Low latency Minimum packet loss Minimum infrastructural modifications

Horizontal and Vertical Handoffs

Horizontal– When the two network access points are based on the

same wireless link technology– The link is damaged or disrupted solely due to device

mobility

Vertical– When the two network access points are based on different

wireless technologies– Moving to an area covered by a different wireless

technology

Example Horizontal vs. Vertical Handoff

Horizontal– User has a laptop with a standard 802.11 network card.– He walks from SEAS to CS UCLA department and

continues to have internet access.

Vertical– User has a laptop with 802.11 network card and a 1xRTT

card.– He connects to the Internet using 802.11 while indoors, but

when he steps outside he is seamlessly handed off to a 1xRTT connection automatically.

Example scenario

Joe Shmoe is downloading a large 30 GB video in Boelter Hall.

During the transfer, he decides to grab lunch in North Campus.

He doesn’t want to stop the download, and at the same time, he doesn’t want to leave his laptop behind.

What does he do?

Example scenario (cont.)

Joe Shmoe connects his laptop to a handoff server

Now, he can bring his laptop with him to North Campus, and the download will be continuous.

How is this done?

Example scenario (cont.)

The handoff server freezes the download activity at the session layer when horizontal and/or vertical handoffs are detected.

This permits the download to continue as if it were constantly connected to the Internet, despite roaming across different wireless networks.

Handoff Diagram

Virtual Private Network (VPN)

Handoff Server

Internet

Handoff Diagram

VPN

Bluetooth

Handoff Server

Internet

Test Scenarios

Horizontal Handoff– Switch from static IP address to dynamic IP address DHCP in a different

network/subnet and vice versa– 1 Laptop using 802.11– UCLA CS Department– wget

http://download.microsoft.com/download/speechSDK/SDK/5.1/WXP/EN-US/speechsdk51.exe

– 68 MB file Vertical Handoff

– Switch from wired ethernet to Bluetooth, switch at 180s, and vice versa– 1 Laptop using 100 Mbps ethernet and Xircom Bluetooth adapter– UCLA CS Department– wget

http://download.microsoft.com/download/speechSDK/SDK/5.1/WXP/EN-US/speechsdk51.exe

– 68 MB File

Test Scenarios (cont.)

Vertical Handoff– Switch from 1xRTT to 802.11, switch at 60s, and

vice versa– 1 Laptop using standard 802.11 and Aircard

1xRTT card– UCLA CS Department– wget

http://download.microsoft.com/download/speechSDK/SDK/5.1/WXP/EN-US/speechsdk51.exe

– 68 MB File

Performance Data

Trial Run

Duration (s)

Delay (s)

Throughput KB/s

1 123.6 2.6 550.16

2 122.4 2.5 555.56

3 124.5 2.6 546.18

4 121.6 2.7 559.21

5 126.9 2.4 535.86

Still User, 1 interruption, 802.11 static -> dynamic IP different network/subnet (Avg throughput = 549.39 KB/s), switch at 60s

Performance Data

Trial Run

Duration (s)

Delay (s)

Throughput KB/s

1 121.9 0.9 557.83

2 120.9 1.1 562.45

3 124.5 1.0 546.18

4 124.6 1.1 545.75

5 122.2 1.1 556.46

Still User, 1 interruption, 802.11 dynamic -> static IP different network/subnet (Avg throughput = 553.74 KB/s), switch at 60s

Performance Data

Trial Run

Duration (s)

1xRTT Throughput KB/s

Delay (s) 802.11 Throughput KB/s

1 181.32 37.21 0.8 542.12

2 180.29 38.21 1.3 546.23

3 178.41 37.65 1.3 555.21

4 179.00 36.11 1.2 553.21

5 182.18 33.89 1.1 539.91

Still User, 1 interruption, 1xRTT-> 802.11 different network/subnet (Avg throughput = 291.98 KB/s), switch at 60s

Performance Data

Trial Run

Duration (s)

802.11 Throughput KB/s

Delay (s) 1xRTT Throughput KB/s

1 1051.99 544.42 0.7 35.62

2 1182.08 512.26 0.8 33.21

3 1179.32 532.26 0.8 32.22

4 1229.24 522.21 0.8 31.36

5 1092.21 510.22 1.0 36.22

Still User, 1 interruption, 802.11 -> 1xRTT different network/subnet (Avg throughput = 66.08 KB/s), switch at 60s

Performance Data

Trial Run

Duration (s)

Wired Throughput KB/s

Delay (s) Bluetooth Throughput KB/s

1 321.54 342.22 1.1 45.22

2 315.19 343.26 0.9 45.96

3 242.21 362.53 0.8 44.12

4 271.04 355.55 0.9 43.95

5 291.21 348.95 1.2 46.66

Still User, 1 interruption, Wired Ethernet -> Bluetooth (Avg Throughput = 197 KB/s) switch at 180s.

Performance Data

Trial Run

Duration (s)

Bluetooth Throughput KB/s

Delay (s) Wired Throughput KB/s

1 358.54 44.16 0.9 336.35

2 359.05 46.96 0.9 332.58

3 356.01 45.32 0.7 339.99

4 353.14 44.48 0.6 346.51

5 369.01 49.56 0.7 312.56

Still User, 1 interruption, Bluetooth -> Wired Ethernet (Avg Throughput = 190 KB/s) switch at 180s.

Performance Analysis

Handoffs were transparent to the user Congestion window behavior as expected Throughput remained consistent File integrity maintained VPN connection to handoff server maintained

with tunnel security

Future Work to be Done

Optimizations Scalability Testing Analysis

Summary

Seamless Handoff: Continuous connectivity, low latency, minimal infrastructural modifications

Horizontal and vertical handoff scenarios tested

Measured/evaluated performance of data transmission in different PAN scenarios

References

Mark Stemm, Randy H. Katz, " Vertical Handoffs in Wireless Overlay Networks," Mobile Networks and Applications, 1996.

R. Hsieh, Zhe Guang Zhou, A. Seneviratne, " S-MIP: a seamless handoff architecture for mobile IP," In Proceedings of IEEE INFOCOM 2003.

V. Ghini, G. Pau, P. Salomoni, M. Roccetti, M. Gerla, " Smart Download on the Go: A Wireless Internet Application for Music Distribution over Heterogeneous Networks, " Submitted to ICC2004, Paris.

Pangalos, P.A.; Boukis, K.; Burness, L.; Brookland, A.; Beauchamps, C.; Aghvami, A.H., " End-to-end SIP based real time application adaptation during unplanned vertical handovers, " Global Telecommunications Conference, 2001. GLOBECOM '01. IEEE , Volume: 6 , 25-29 Nov. 2001 Page(s): 3488 -3493.