The Challenges of Robust Inter-Vehicle Communications IEEE VTC2005-Fall Marc Torrent-Moreno, Moritz Killat and Hannes Hartenstein DSN Research Group Institute of Telematics University of Karlsruhe Marc Torrent-Moreno
Structure of this talk» Vehicular ad-hoc networks» Challenges of inter-vehicle communications» Impact on different types of communications Periodic messages Event driven messages Bidirectional links» Summary and Conclusions Marc Torrent-Moreno - 2
Vehicular Ad-Hoc Networks (VANETs)» Increasing interest: Projects, consortiums, standardization bodies...» Goals Safety Environmental Comfort Infotainment» Main characteristics 802.11p (DSRC) as underlying technology Mainly geo- and broad-cast Lack of centralized entity» Not yet an agreed of applications with specific requirements Reliability Robustness Delay Fairness» Underestimated implications from real radio channel effects Marc Torrent-Moreno - 3
Challenges of Inter-Vehicle Communications Marc Torrent-Moreno
Challenges I» Lack of centralized management entity Difficult to make an efficient use of the resources, 802.11 [1]» High mobility Compromise iterative optimization algorithms (e.g., Li et al. [2], ADHOC- NOW 04) and predefined routes (e.g., Füßler et al. [3], ELSEVIER 04)» Hidden terminal problem Broadcast mediums specially sensible Torrent-Moreno et al. [4], VANET 04 Message reception probability Priority Non-priority Highly populated scenario with 2.56Mbps inside communication range in a 6Mbps medium Distance [m] Marc Torrent-Moreno - 5
Challenges II» Channel congestion Scenarios with high number of nodes when high penetration rates 802.11p [5] considers channel hopping (restricted time for safety) Qu et al. [6], VANET 04, study different repetition strategies and introduce feasibility areas: x Can we achieve? - Prob. Reception Failure = 0.01% - Channel Busy Time = 50% Message generation interval Exemplary Configuration: - Message Rate = 200 ms - Message Size = 250 bytes - 30meter between cars 4 lane highway 150meters message range = 40 Interfers Interferer number Marc Torrent-Moreno - 6
Challenges III» Radio channel characteristics In VANETs: multiple mobile and reflecting objects able to degrade the strength and quality of the receiving signal (Fading) Deterministic radio propagation models are invalid to evaluate the performance of communication protocols (e.g. Takai et al. [7], MOBIHOC 01) No Speed 10 miles 85 miles 125 miles BER Transmission power trade-off when trying to increase safety: Transmission power Prob. Reception of, single transmission Prob. of collision for all transmissions SNR (db) BER vs. SNR at 12Mbps and 400ns RMS delay Yin et al.[8], VANET 04 Prob. Successful reception without interferences Model and adjustments from Taliwal et al.[9], VANET 04 Marc Torrent-Moreno - 7
Challenges IV Link Layer Desynchronization» Why Collisions may occur in a 802.11 medium? Two or more nodes have the same remaining back-off time Two or more nodes can not sense the transmission of each other Nodes far away from each other Multi-path Fading and Shadowing Link Layer Desynchronization Node in the middle X sends a message with: -Communication Range = 200m -Carrier Sense Range = 400m As close as 85m! Marc Torrent-Moreno - 8
Different Types of Inter-Vehicle Communications Marc Torrent-Moreno
Periodic Messages» Motivation: Systems are able to detect non-safe situations with an accurate and updated information of all surrounding vehicles» Requirements: - minimum message inter-arrival time - minimum reliability - minimum coverage area» Trade-offs when trying to increase safety: Packet generation rate: #packets #collisions Transmission power: TxPower Interferences Information aggregation looks like a reasonable solution (safety value of information decreases with distance) Marc Torrent-Moreno - 10
Periodic Messages Fairness» Fairness Ultimate goal is SAFETY Classical ad hoc networks approaches: Capacity Connectivity Not valid to maximize safety All nodes might have different safety requirements or danger level Torrent-Moreno et al [10], VANET 05: Centralized algorithm: Limit channel capacity Fairly share bandwidth Max-min TxPower Per node maximality Marc Torrent-Moreno - 11
Event Driven Messages» Motivation: Needed in emergency situations to disseminate information» A clever strategy required to handle: Broadcast storm problem (Ni et al.[11], MOBICOM 99) Multi-path fading and shadowing effects» Challenge: Make sure all cars receive the information within specific time in the most efficient manner Marc Torrent-Moreno - 12
Bidirectional Links» Motivation: Applications, e.g., for traffic information or access to stationary nodes few hops away» P(2-ways link) = P(1-way link) 2, assuming independence» One hop link: Resulting overhead: Link probability Required prob. of reception» Multi-hop: Avoid unicast flows, e.g., CBF (Füßler et al. [3], Elsevier 04) Marc Torrent-Moreno - 13
Summary and Conclusions» Challenges of robust inter-vehicle communications: Hidden terminal problem High channel load Radio channel characteristics» Impact on: Broadcast-periodic messages Event driven messages (data dissemination) Bidirectional links» PHY/MAC models have crucial impact on protocols performance» Need accurate models, also following technological advances» Consider probabilistic models when designing robust communication protocols Marc Torrent-Moreno - 14
Thank you very much for your attention! http://dsn.tm.uka.de Marc Torrent-Moreno
References 1/2 [1] IEEE Std. 802.11-1999, Part11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications IEEE Std. 802.11, 1999 edition. [2] X. Li, T. Nguyen, and R. Martin, An Analytic Model Predicting the Optimal Range for Maximizing 1-Hop Broadcast Coverage in Dense Wireless Networks in Proc. ADHOC-NOW, July 2004. [3] H. Füßler, J. Widmer, M. Käsemann, M. Mauve, and H. Hartenstein, Contention- Based Forwarding for Mobile Ad-Hoc Networks, Elsevier's Ad Hoc Networks 1 (4), November 2003. [4] M. Torrent-Moreno, D. Jiang, and H. Hartenstein, Broadcast Reception Rates and Effects of Priority Access in 802.11-Based Vehicular Ad-Hoc Networks, in Proc. ACM VANET, October 2004. [5] Dedicated Short Range Communications working group, http://grouper.ieee.org/groups/scc32/dsrc/index.html. [6] Q. Xu, T. Mak, J. Ko, and R. Sengupta, Vehicle-to-Vehicle Safety Messaging in DSRC, in Proc. ACM VANET, October 2004. Marc Torrent-Moreno - 16
References 2/2 [7] M. Takai, J. Martin, and R. Bagrodia, Effects of Wireless Physical Layer Modeling in Mobile Ad Hoc Networks, in Proc. ACM MOBIHOC, October 2001. [8] J. Yin, T. ElBatt, G. Yeung, B. Ryu, S. Habermas, H. Krishnan, and T. Talty, Performance Evaluation of Safety Applications over DSRC Vehicular Ad Hoc Networks, in Proc. ACM VANET, October 2004. [9] V. Taliwal, D. Jiang, H. Mangold, C. Chen, and R. Sengupta, Empirical Determination of Channel Characteristics for DSRC in Proc. ACM VANET, October 2004. [10] M. Torrent-Moreno, P. Santi, H. Hartenstein, Fair Sharing of Bandwidth in VANETs, in Proc. ACM VANET, September 2005. [11] S.-Y. Ni, Y.-C. Tseng, Y.-S. Chen, and J.-P. Sheu, The Broadcast Strom Problem in a Mobile Ad Hoc Network, in Proc. ACM MOBICOM, August 1999. Marc Torrent-Moreno - 17