Performance of Real-Time Wireless Communication for Railway - - PowerPoint PPT Presentation

▶
performance of real time wireless communication for
SMART_READER_LITE
LIVE PREVIEW

Performance of Real-Time Wireless Communication for Railway - - PowerPoint PPT Presentation

Apr 09, 2023 350 likes •714 views

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p Daniel Richter, Jossekin Beilharz, Lukas Pirl, Christian Werling, and Andreas Polze Operating Systems & Middleware Group Hasso Plattner Institute at

slide-1
SLIDE 1

Daniel Richter, Jossekin Beilharz, Lukas Pirl, Christian Werling, and Andreas Polze

Operating Systems & Middleware Group Hasso Plattner Institute at University of Potsdam, Germany

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

slide-2
SLIDE 2

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 2

IEEE 802.11

Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications

▪ set of standards for wireless communication ▪ local networks with little device mobility ▪ modes of operation

▪BSS/AP (Basic Service Set/Access Point) ▪IBSS (Independent BSS, ad-hoc) ▪…

Motivation

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

slide-3
SLIDE 3

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 3

IEEE 802.11p

802.11 Amendment 6: Wireless Access in Vehicular Environments

▪ high mobility of devices in rapidly changing environments ▪ connection and communication in very short periods of time ▪ OCB mode (Outside the Context of a BSS)

▪think: broadcast; no authentication, no association

Motivation

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

slide-4
SLIDE 4

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 4

▪ vehicular environments

Motivation

slide-5
SLIDE 5

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 5

Car-to-X → Rail-to-X ▪ transfer of maintenance data of critical infrastructure elements

▪e.g. largely self-sufficient units such as level crossings with monitoring signals, diagnostic systems of switches, or weather stations close to the tracks

Motivation

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

slide-6
SLIDE 6

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 6

▪ data transfer via passing trains ▪ via dedicated short-range communications ▪ potentially higher vehicular speeds (vs. Car2X) ▪ potentially higher disturbances

Motivation

Photo: Sven Teschke / License: Creative Commons CC-by-sa-3.0 de

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

slide-7
SLIDE 7

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 7

▪ focus on an extremely short timespan for connection and communication (approx. 100 ms) ▪ compare latencies of networks

▪OCB mode 802.11p 5.9 GHz ▪IBSS (ad-hoc) 802.11n 5 GHz ▪BSS/AP (access point) 802.11n 5 GHz

▪ response to disturbances of OCB mode ▪ real-time applications → latencies

Motivation

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

slide-8
SLIDE 8

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

Background

slide-9
SLIDE 9

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 9

Background

V2X standard in Europe and US for vehicular ad hoc networks ▪ PHY layer (OSI layer 1)

▪frequency: ~5.9 GHz ▪7 channels of 10 MHz bandwidth (Europe)

▪ MAC layer (OSI layer 2)

▪OCB mode (Outside the Context of a BSS, think: broadcast) ▪no authentication ▪no association

IEEE 802.11p

slide-10
SLIDE 10

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 10

Background

802.11 working on a 2.4 GHz for many years, switched to less crowded 5 GHz band ▪ Basic Service Set (BSS)

▪Access Point (AP) infrastructure mode ▪Ad-Hoc / Independent BSS (IBSS) mode ▪Mesh, Monitor, …

802.11p facilitates the 5.9 GHz band ▪ OCB mode (Outside the context of a BSS)

Modes of Operation

slide-11
SLIDE 11

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 11

Background

▪ performance of 802.11p under various conditions

▪analytically,

  • A. A. A. Almohammedi, N. K. Noordin, A. Sali, F. Hashim, and S. Saeed, “A comprehensive performance analysis of IEEE 802.11p based MAC

for vehicular communications under non-saturated conditions,” Journal of ICT Research and Applications, vol. 11, no. 1, pp. 91–112, 2017.

▪in simulations,

  • S. Gräfling, P. Mähönen, and J. Riihijärvi, “Performance evaluation of IEEE 1609 WAVE and IEEE 802.11p for vehicular communications,” in

Ubiquitous and Future Networks (ICUFN), 2010 Second International Conference on. IEEE, 2010, pp. 344–348.

  • Z. Hameed Mir and F. Filali, “LTE and IEEE 802.11p for vehicular networking: a performance evaluation,” EURASIP Journal on Wireless

Communications and Networking, vol. 2014, p. 89, May 2014.

  • K. Bilstrup, E. Uhlemann, E. G. Strom, and U. Bilstrup, “Evaluation of the IEEE 802.11p MAC method for vehicle-to-vehicle communication,” in

Vehicular Technology Conference, 2008. VTC 2008-Fall. IEEE 68th. IEEE, 2008, pp. 1–5.

▪experimentally

  • S. Demmel, A. Lambert, D. Gruyer, A. Rakotonirainy, and E. Monacelli, “Empirical IEEE 802.11p performance evaluation on test

tracks,” in Intelligent vehicles symposium (IV), 2012 IEEE. IEEE, 2012, pp. 837–842.

  • V. Shivaldova, G. Maier, D. Smely, N. Czink, A. Alonso, A. Winkelbauer, A. Paier, and C. F. Mecklenbräuker, “Performance

evaluation of IEEE 802.11p infrastructure-to-vehicle tunnel measurements,” in Wireless Communications and Mobile Computing Conference (IWCMC), 2011 7th International. IEEE, 2011, pp. 848–852.

  • M. Kloc, R. Weigel, and A. Koelpin, “SDR implementation of an adaptive low-latency IEEE 802.11p transmitter system for real-

time wireless applications,” in 2017 IEEE Radio and Wireless Symposium (RWS), Jan. 2017, pp. 207–210.

  • B. Bloessl, M. Segata, C. Sommer, and F. Dressler, “Performance Assessment of IEEE 802.11 p with an Open Source SDR-based

Prototype,” IEEE Transactions on Mobile Computing, vol. 17, no. 5, pp. 1162–1175, 2018.

▪ most studies only look at throughput, packet loss, and signal range ▪ latencies (especially incl. establishing a connection) rarely investigated

Related Work

slide-12
SLIDE 12

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 12

Background

▪ Lisový et al.: implementing 802.11p’s amendments into different Linux subsystems

  • R. Lisový, M. Sojka, and Z. Hanzálek, “IEEE 802.11p Linux Kernel Implementation,” Czech Technical University in

Prague, Tech. Rep., 2014.

▪first-hand practical info on the Linux 802.11p support ▪concise summary of 802.11p’s changes and their motivation

▪ Fernández: compatible hardware

  • J. Fernández Pastrana, “802.11p standard and v2x applications on commercial wi-fi cards,” Master’s thesis,

Universidad de Valladolid. Escuela Técnica Superior de Ingenieros de Telecomunicacin, 2017.

▪states positively tested cards with 802.11p support ▪gives insights into the physical requirements to such cards in general

Related Work – T estbed Setup

slide-13
SLIDE 13

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

T estbed Setup

slide-14
SLIDE 14

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 14

▪ since no off-the-shelf hardware is sold with 802.11p support, we had to patch drivers and adjust the

  • perating system configuration

T estbed Setup

Hardware ▪ wireless cards officially supporting 802.11p scarcely found online ▪ several 5 GHz wireless physically able to support 5.9 GHz band [Fernández] Software ▪ some minor patches to the drivers are needed to make 802.11p usable for applications [Lisový et al.]

slide-15
SLIDE 15

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 15

T estbed Setup

Testbed setup A ▪ Wireless card

▪Qualcomm Atheros AR5B22 Mini PCI-e ▪Chipset: AR9462

▪ Mini PCI-e to PCI-e adapter: adaptare 49006 ▪ Dell Workstation

▪Intel Core i5-3470 CPU ▪8 GB RAM

Hardware

Testbed setup B ▪ Wireless card

▪ Qualcomm Atheros AR9462 ▪ 2.4/5 Ghz WLAN + Bluetooth

▪ HPE GL20 IoT Gateway

▪ Intel I5-4300U CPU ▪ 8 GB RAM In both setups the distance between the communicating nodes is 0.5 meters due to laboratory conditions.

slide-16
SLIDE 16

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 16

T estbed Setup

▪ Ubuntu 16.04 LTS with Linux kernel 4.13.0-31- generic x86 64

▪Linux because existing implementation of 802.11p’s changes into its 802.11 subsystems

Software

▪ 802.11p / OCB mode awareness:

▪ kernel (BSSID in mac80211, commands in cfg80211 and nl80211 to virtually join & leave OCB network and sending and receiving) ▪ tools (e.g. iw 4.0 and later) ▪ regulatory database (wireless-regdb)

[Lisový et al.]

slide-17
SLIDE 17

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

Benchmarks

slide-18
SLIDE 18

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 18

▪ initial connection speed and the latency of ICMP packets’ round-trip times

▪802.11p 5.9 GHz OCB network (OCB) ▪802.11n 5 GHz open ad-hoc network (IBSS) ▪802.11n 5 GHz open access point network (BSS/AP)

▪ for both establishing a connection and exchanging data we require a time frame of around 100 ms

Benchmarks

slide-19
SLIDE 19

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 19

Benchmarks

▪ IBSS: joining via iw ibss join an existing station in IBSS (ad-hoc) mode on a given channel. ▪ BSS/AP: connecting in managed mode to an existing station in open BSS (AP) mode. ▪ OCB: joining via iw ocb join an OCB band on a given frequency.

Connection Speed

Testbed setup A, actions performed 100 times, average duration

slide-20
SLIDE 20

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 20

Benchmarks

ICMP packet round trips (with testbed setup A) ▪ IBSS: between two stations in an open ad-hoc network on 5 GHz band ▪ BSS/AP: between two stations in managed network (one AP mode, one managed mode) on 5 GHz band ▪ OCB: between two stations in an OCB network on 5.9 GHz band

Latency

slide-21
SLIDE 21

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 21

Benchmarks

Latency

20 40 60 80 100 120 140 160

IBSS BSS/AP OCB

slide-22
SLIDE 22

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 22

Benchmarks

▪ IBSS (ad-hoc) 802.11n 5 GHz

Latency

20 40 60 80 100 120 140 160

IBSS BSS/AP OCB

slide-23
SLIDE 23

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 23

Benchmarks

▪ BSS/AP (access point) 802.11n 5 GHz

Latency

20 40 60 80 100 120 140 160

IBSS BSS/AP OCB

slide-24
SLIDE 24

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 24

Benchmarks

▪ OCB 802.11p 5.9 GHz

Latency

20 40 60 80 100 120 140 160

IBSS BSS/AP OCB

slide-25
SLIDE 25

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 25

Benchmarks

▪ real-time constraints → worst-case is important!

Latency

slide-26
SLIDE 26

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 27

Benchmarks

plain latencies in different modes → sensitivity of OCB to disturbances ▪ duration of a full round trip between two nodes, while a third node tries to disturb this connection by sending arbitrary data packets with different intensities on the same channel as the first two nodes (with testbed setup B)

Interference Sensitivity

slide-27
SLIDE 27

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 28

Benchmarks

▪ disturbance intensity: size in bytes of packets sent by third node ▪ OCB is best at average RTT and worst case RTT

Interference Sensitivity

slide-28
SLIDE 28

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p

Current & Future Work

slide-29
SLIDE 29

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 30

Current & Future Work

IEEE 802.11p in ns-3 ▪ implausible behavior

▪e.g. sudden, strong degradation of signal quality (90 meters: 100%; 103 meters: 0%)

▪ needs comparison with real-live measurements ▪ configure ns-3 to match reality in simple scenarios ▪ then, use ns-3 for complex scenarios

▪integrated with a traffic simulator (SUMO) ▪SUMO controls movements of nodes in ns-

Improve Simulations

slide-30
SLIDE 30

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 31

Current & Future Work

▪ Routing with train operating schedules

Data Transfer via Passing Trains

slide-31
SLIDE 31

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 32

Current & Future Work

▪ heavy metal playground of Deutsche Bahn in Annaberg Buchholz (Germany)

Rail2X Field T est

image by Steffen Schmidt via https://www.eisenbahnforumvogtland.de

slide-32
SLIDE 32

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 33

Current & Future Work

▪ 802.11p in OCB in "realistic environment“ ▪ first results:

▪amount of data transferred: some MiB ▪partly strong influence by obstacles ▪high variance of latency, but within valid limits

802.11p on the Road

slide-33
SLIDE 33

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 34

▪ 802.11p’s OCB is suitable within the time constrains

  • f the railway use cases we envision

▪ measurements of the latency have shown OCB to be superior to both BSS/AP and IBSS modes in

▪average latency, ▪maximum latency, ▪and standard deviation.

▪ findings still hold true in saturated wireless environments ▪ future work: use it for more accurate simulations

Conclusion

slide-34
SLIDE 34

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 35

▪ http://gitlab.hpi.de/osm/802.11p-benchmarks

Conclusion

▪ mailto:daniel.richter@hpi.de

slide-35
SLIDE 35

Performance of Real-Time Wireless Communication for Railway Environments with IEEE 802.11p | HICSS-52 | Daniel Richter | January 11, 2019 36

End of Slides for T

  • day…

Q?A!