Broadcast Storm Mitigation Techniques In Vehicular Ad Hoc Network

Author: N. Wisitpongphan and O.K. Tonguz, Carnegie Mellon University J.S. Parikh, P. Mudalige, P. Bai, and V. Sadekar, General Motors Corporation Presented by: Yu Danlei Ajou University
1

Motivation:
Broadcast is being used to discover nearby neighbors or propagate useful traffic information in VANET

High level of contention and collisions at the link layer due to an excessive number of broadcast packets.
2

Introduction:
Differences between VANETs and MANETs mobility dimensional one-dimensional line onea strip broadcast transmission Three broadcast suppression techniques: weighted p-persistence pslotted 1-persistence 1slotted p-persistence p3

Organization:
Background on VANETs and related research Quantify the impact of the broadcast storm problem in VANETs and discussion Three broadcast mitigation algorithms, and explain the model and assumptions The performance of three broadcast techniques
4

Background and Related Work 1.1

VANET:
PriorityPriority-based broadcast scheme [3] RoleRole-based multicast protocol [4] 802.11802.11-based multihop broadcast (UMB) protocol[5]

MANET: Distributed gossip-based routing [8] gossipThresholdThreshold-based techniques [9] SenderSender-based multipoint relay ( MPR) technque[11] Tackle broadcast storm problem by hardware[12.13]
5

Background and Related Work 1.1

VANET:
[3] Broadcast Reception Rates and Effects of Priority Access in 802.11802.11-Based Vehicular Ad Hoc Networks M. Torrent- Moreno, D.Jiang, and H.Hartensten Torrent-

The scheme gives higher priority to nodes that need to transmit timetime-critical messages. It can indirectly mitigate the severity of the storm by allowing nodes with higher priority to access the channel as quickly as possible.
6

Background and Related Work 1.1

VANET:
[4] Role-Based Multicast in highly Mobile but Sparsely Connected Ad Hoc RoleNetworks L. Briesemeister and G. Hommel A role-based multicast protocol that suppressed broadcast redundancy by roleassigning shorter waiting time prior to rebroadcasting to more distant receivers. Disadvantages: a sparsely connected or fragmented network where the broadcast storm is not a main problem

Background and Related Work 1.1

VANET:
[5] Urban Multi-Hop Broadcast Protocol for Inter-Vehicle MultiInterCommunication Systems G. Korkmaz et al., An efficient 802.11-based urban multihop broadcast protocol. 802.11Its designed to suppress broadcast redundancy by only allowing the furthest vehicle from the transmitter to rebroadcast the packet.
8

Background and Related Work 1.1

Sum up:
Our approach aims to reduce the load submitted from the network layer to the data link layer by combining the probabilistic broadcast technique with timer-based timersuppression. This article is on a well connected network, in this work we only consider the broadcast storm problem on major highways.

Ongoing Research and Previous work 1.2

Carnegie Mellon University:
1) GrooveNet disseminate messages to within a prespecified bounded region with minimal handshaking and state sharing information 2) GrooveSim simulator operates in five different modes: Drive mode Simulation mode Playback mode Hybrid simulation mode Test generation mode Disadvantage: Although GrooveNet allows only relevant vehicles to receive and rebroadcast the packets, the current version of the protocol does not have a mechanism for preventing the broadcast storm problem from happening.
10

The Broadcast Storm In MANETs and VANETs 2.1

MANET: 1.RREQ (a route request) 1.RREQ 2.Expanding Ring Search (a node can cache each routing 2.Expanding entry for a longer time and can also reply on behalf of the destination.) VANET: RSU (roadside unit)

11

FourFour-Lane Traffic Case Study

2.2

Different density Total delay can be negligible Dramatic packet loss ratio

12

Broadcast Suppression Techniques 3

DistanceDistance-Based Broadcasting: weighted p-persistence pslotted 1-persistence 1slotted p-persistence pReceived-SignalReceived-Signal-Strength Based Schemes:
13

Weighted p-persistence Broadcasting pRule 3.1.1

Upon receiving a packet from node i, node j checks the packet ID and rebroadcast with probability pij if it receives the packet for the first time; otherwise it discards the packet

farther node gets higher probability

14

Slotted Slotted 1-persistence Broadcasting Rule 3.1.2

a node checks the packet ID and rebroadcasts with probability 1 at the assigned time slot Tsij if receive the packet for the first time and has not received any duplicates before its assigned time slot; otherwise, discards. A shorter waiting time will be assigned to the nodes located in the farthest region.
15

Slotted pSlotted p-Persistence Broadcasting Rule 3.1.3

PrePre-determined probability p The performance of this scheme depends on the value of p

16

Received-Signal-StrengthReceived-Signal-Strength-Based Schemes 3.2

RRS is used where GPS signals may not be received. Nodes obtain the RSS of the broadcast packet received from the DSRC device driver and determine whether or not to rebroadcast the packet To get a closer estimate of the relative distance to the transmitter, each vehicle should periodically probe its neighbors in order to keep track of the time averaged RSS.

17

Performance Analysis 4
SingleSingle-Lane Network:
1) Link Load 2) Packet Penetration Rate

Multilane Network:
18

Link load 4.1.1

The link load measures the amount of broadcast traffic received at each node over a unit time It depends on the number of retransmitting nodes The smaller p, the better performance in terms of link load The link load is reduced dramatically when the slotted scheme is employed 19

Packet penetration rate 4.1.2

Measured from a singlesinglelane network with random traffic distribution Fast packet penetration rate means good transmission in emergency system Slotted 1-persistence 1performs poorly in spars network

20

There is almost no benefit in using the reforwarding probability in light traffic condition At higher traffic density the reforwarding probability should be set to at least 0.5 in the persistence case and 0.8 in the slotted p-persistence case in order to achieve at least 80 21 ppercent of the maximum performance

Multilane Network

4.2

Multilane network is simply a collection of multiple single-lane networks singleN-fold and n-lane with a factor n or 1/n n-

22

Packet Loss Ratio And Delay Analysis And Discussion 4.3

Communication Model and Broadcast Protocol Implementation :
1) OPNET using IEEE 802.11a with 10 MHz bandwidth 2) range is approx. 1km 3) assume that =WAIT_TIME.

Simulation Results:
Packet Loss Ratio Latency
23

Packet Loss Ratio

Four densities Among the three schemes proposed, slotted p-persistence pyields the best performance while the worst scheme is weighted p-persistence p24

Latency

Slotted p-persistence introduces the longest propagation pdelay Traffic density doesn t have much impact on the number of hops chosen by the routing protocol 25

Conclusion:
Routing protocols should be designed to address the broadcast storm problem to avoid unnecessary loss of important safety related packets during a broadcast storm. The proposed slotted 1-persistence and slotted p1ppersistence schemes can reduce broadcast redundancy and packet loss ratio by up to 70 percent while still offering acceptable end-to-end delay for most multihop VANET end-toapplication. No fixed infrastructure in VANET. Further research may located in fixed ones for other application.

26

If you have any question, slowly and loudly !

27