Improving Reliability of Platooning Control Messages Using Radio - PowerPoint PPT Presentation

Improving Reliability of Platooning Control Messages Using Radio and Visible Light Hybrid Communication Susumu Ishihara (Shizuoka University) Vince Rabsatt, Mario Gerla (UCLA)

To the direct follower Leader to followers Platooning Camera/ RADAR/ LIDAR, etc. • Autonomous Platooning 
 or Cooperative Adaptive Cruise Control (CACC) • Improves the tra ffi c flow • Reduces the fuel consumption • Improves the drivers’ safety 2

http://www.nedo.go.jp/activities/FK_00023.html http://www.sartre-project.eu/en/about/ news/Sidor/roadtrains_video.aspx Recent Platooning Projects Sartre (EU)-2012 1Truck and 3 cars Demo 6m gap @90km/h 5.9GHz DSRC 40Hz Message Energy ITS (Japan) -2013 4 Truck-Demo 4.7m gap @80km/h 5.8GHz DSRC + IR 50Hz Message 3

Simulation of Platooning on Scenargie 1 , a discrete event simulator (Same scenario provided with plexe 2 , a platooning extension for Veins ) Adaptive Cruise Control (ACC) Uses information of the preceding car Cooperative Adaptive Cruise Control (CACC) Uses information of the preceding car and the leader car [1] Product of Space-Time Engineering, https://www.spacetime-eng.com/ [2] Michele Segata, Stefan Joerer, Bastian Bloessl, Christoph Sommer, Falko Dressler and Renato Lo Cigno, 4 "PLEXE: A Platooning Extension for Veins," IEEE VNC 2014

Many issues for realizing platooning • Control Theory • Vehicle Dynamics • Communications • Sensing • Image Processing • Localization • Traffic Engineering • etc. 5

Communication Issues • Radio Communication Capacity • If many vehicles are on the road, radio communication capacity may be exhausted. • Communication Rate Control • Transmission Power Control • Security • What if attackers make a radio signal jam? • What if a malicious vehicle pretends other cars? • What if a malicious vehicle generates wrong update information? 6

Slow down Fragile Radio Communication • If there is a malicious machine that sends a jam signal, messages transmission from the leader vehicle can be easily damaged. • We can easily make jamming machine using 
 software radio platforms, 
 such as GNU Radio/USRP , WARP , etc. • This is a potential problem of CACC based on radio communications. What if we use Visible Light Communication? 7

Contributions of this paper • We proposed an RF and VLC hybrid communication protocol for platooning control messages for reliable message delivery under RF jamming attacks . • We developed a simulation model supporting both radio communication and VLC on Scenargie • The proposed protocol works e ff ectively to decrease the end-to-end delay of the leader messages and improve the message delivery ratio under jamming attack 8

J Jam S R Tx Power Jam Jam Jam Jam Jam Jam Jam 20dBm Tx Power Jam 20dBm Simulation: E ff ect of jamming attack on Scenargie 64us 10us [Punal2015] Periodic Jammer Our 10~400us 10~400us simulation Jam part is 802.11p Compliant PLCP Preamble + PLCP Header + Part of MAC Header Jammer d r d s CBR Broadcast on IEEE 802.11p 6Mbps 
 (Payload 300bytes / 10Hz) d sr : 10m and 50m 9

Simulation Results: Dist (Receiver, Packet Delivery Ratio Color = Jammer) Packet Delivery Ratio under Jamming Attack Dist (Source, Jammer) 75m Dist(S, R) 15m = 10m 75m 75m 15m 65m 65m Dist(S, R) = 50m Idle duration 10us 120us 122us 150us 200us of Jam signal • If the idle duration is long, the sender can find the channel idle state longer 
 than DIFS (=58us) + Backo ff . Thus, the sender can send a frame. • But if the distance between the sender and the receiver is long, 
 due to the bad SINR, the receiver cannot decode the frame. 10

Using Visible Light Communication (VLC) • Uses LED lights for communication between adjacent vehicles • Cheap – Low Additional Equipment Cost • Every vehicle has tail lights and head lights. • Di ffi cult to attack - Highly directional • VLC jammer has to be in the field of view of the receiver device to attack • Jamming light has to be concentrated to the receiver device • But, long distance communication over multiple vehicles is di ffi cult • Multi-hop communication is needed – Long message delivery delay 11

Shortening message delivery delay • Hybrid of RF Communication and VLC • RF • Wide range – Propagates beyond multiple vehicles • Vulnerable to Jamming • VLC • Short range – Hop by hop communication • Strong against Jamming 12

Radio and VLC Hybrid Message Delivery for Platooning RF Broadcast RF Broadcast VLC Leader • Leader sends a message via both VLC and RF interfaces • If a vehicle receives a new message from either of VLC and RF interfaces, it forwards the message via VLC (and RF). • Only If a condition is satisfied , the vehicle forwards the message via both VLC and RF to avoid RF channel congestion. 13

Condition for forwarding a message via RF • Candidates of conditions • Distance from the previous RF hop vehicle • Elapsed time from the transmission by the leader • Number of hops from the leader • SINR of the signal from previous radio-hop vehicle • etc. In this simulation: “If a new leader message arrives from VLC, forward it via RF and VLC.” Intuition : If an RF message does not arrive, the leader may be su ff ering jamming attack and cannot send the message. Thus other vehicles have not received the message. …. RF broadcast is needed. 14

Simulation Model We implemented multiple interfaces with different antenna models (pos. and direction) and an application model using these interfaces on Scenargie RF IEEE802.11p Interface (6Mbps, AC_VI) Omni antenna: 5.9GHz 5m 30° VLC interfaces with a directional antenna model Comm. range: 5m (the same as the inter vehicular distance) MAC: ALOHA Bitrate: 50kbps (assuming cheap off-the-shelf devices) Note: This model is just an abstract model of VLC. 
 It does not simulate realistic visible light propagation. 15

Periodic Jammer Simulation Scenario 12.5m 17.0m/s … 10m 5m 500m … … 10m t =0 10 vehicles 10m Jammer: Periodic Jammer [64us JAM and 10us idle period] • Platoon of 10 vehicles: 1 platoon scenario and 4 x 4 platoon scenario • Leader sends 200 byte message to the members every 0.1 sec. • Tx power of All vehicles and Jammer: 20 dBm • Propagation: Free Space , Fading: Nakagami • 16

Compared Strategies Forward only when Use receiving a new Use VLC RF Forwarding message via VLC F F F Only Direct RF (FFF) T F F RF Forward (FTF) VLC + No RF Fwd. T F F (TFF) VLC + unconditional T T F RF Fwd. (TTF) VLC + conditional RF T T T Fwd. (TTT) 17

End-to-End delay at 5 th car / 1 platoon Long delay 2.5 2.5 2.5 Long 
 End-to-end delay [sec] End-to-end delay [sec] End-to-end delay [sec] 2 2 2 blind period Blind period VLC’s (4sec.) 1.5 1.5 1.5 (2sec.) multi-hop delay 1 1 1 0.5 0.5 0.5 0 0 0 20 25 30 35 40 45 50 20 25 30 35 40 45 50 20 25 30 35 40 45 50 Time [sec] Time [sec] Time [sec] Only Direct RF (FFF) VLC + No RF Fwd. (TFF) RF Fwd. (FTF) 2.5 2.5 End-to-end delay [sec] End-to-end delay [sec] 2 2 1.5 1.5 Long delay period No positive e ff ect of 1 1 is shortened conditional RF Fwd. 0.5 0.5 0 0 20 25 30 35 40 45 50 20 25 30 35 40 45 50 Time [sec] Time [sec] VLC+unconditional RF Fwd. (TTF) VLC + conditional RF Fwd. (TTT) 18

Other cars / Many platoons cases 2.5 2.5 2.5 2.5 End-to-end delay [sec] End-to-end delay [sec] End-to-end delay [sec] End-to-end delay [sec] VLC 2 2 2 2 1.5 1.5 1.5 1.5 1 1 1 1 0.5 0.5 0.5 0.5 0 0 0 0 20 25 30 35 40 45 50 20 25 30 35 40 45 50 20 25 30 35 40 45 50 20 25 30 35 40 45 50 Time [sec] Time [sec] Time [sec] Time [sec] (a3) 5th car True/False/False (b3) 10th car True/False/False (a3) 5th car True/False/False (b3) 10th car True/False/False 2.5 2.5 2.5 2.5 End-to-end delay [sec] VLC+RF End-to-end delay [sec] End-to-end delay [sec] End-to-end delay [sec] 2 2 2 2 1.5 1.5 1.5 1.5 1 1 1 1 0.5 0.5 0.5 0.5 0 0 0 0 20 25 30 35 40 45 50 20 25 30 35 40 45 50 20 25 30 35 40 45 50 20 25 30 35 40 45 50 Time [sec] Time [sec] Time [sec] Time [sec] (a4) 5th car True/True/False (b4) 10th car True/True/False (a4) 5th car True/True/False (b4) 10th car True/True/False 2.5 2.5 2.5 2.5 5th car/ 10th car/ 5th car/1platoon 10th car/1platoon 4 x 4 platoons 4 x 4 platoons

Gap TTF TTT 10 Packet deliveried in 1 sec # of new messages TFF 8 delivered in 1sec. 6 with VLC 4 FFF False/False/False False/True/False 2 True/False/False 5 th car FTF (RF Fwd.) True/True/False Why packet delivery ratio < 1 True/True/True 0 20 25 30 35 40 45 50 with VLC? Time [sec] Leader Member (a) 5th car / 1 platoon 10 Packet deliveried in 1 sec RF 8 6 4 False/False/False False/True/False 2 VLC 10 th car True/False/False True/True/False True/True/True 0 20 25 30 35 40 45 50 Time [sec] 1 platoon (b) 10th car / 1 platoon 20

Conclusion • We proposed an RF and VLC hybrid communication protocol for platooning control messages for reliable message delivery under RF jamming attacks . • Findings • Long blind under a jamming attack only with RF . • RF + VLC Hybrid communication can better message delivery ratio and short delay • Future work • E ff ect of di ff erent communication speed of VLC • E ff ect of RF jamming on the dynamics of platoons • VLC jamming attack 21

Thank you