Network Connectivity of VANETs in Urban Areas

Wantanee Viriyasitavat, Ozan K. Tonguz, Fan Bai

IEEE communications society conference on sensor, mesh and Ad hoc networks

89821006鄭翔升

Outline

Introduction Cellular automata-based traffic mobility

model Network connectivity in urban traffic Conclusion

Introduction

Vehicular Ad Hoc Networks applications Safety relatedapplications non safety-related applications

It is essential to analyze and to have a complete understanding of the network topology and its connectivity pattern

Introduction

Static characteristics network connectivity path redundancy

Dynamic characteristics connection duration Re-healing time

Traffic model

Due to the unavailability of urban vehicular traffic traces

Cellular Automata (CA)-based vehicular mobility model Cellular road structure Vehicle movement Traffic light control

Traffic model

Cellular Road Structure for Manhattan Grid

evenly-spaced horizontal and vertical two-lane ,bi-directional streets each lane is modeled as N cells one vehicle per cell

Traffic model

Vehicle Movement 1. Vehicle’s state: rn : street number where Vehicle n is located Dn : direction of travel of Vehicle n xn and vn : the position and the speed dn : distance to the vehicle in front of it In and sn are the closest intersection and the

distance to that intersection Tn is the turning decision at the intersection In

Traffic model

2. Algorithm for Updating Vehicle’s State Case I: Go straightly

Acceleration step Braking step : front car Randomization step : ??? Vehicle movement step : update Case II: TURN red-light : stop green-light : right or left

Traffic model

Traffic Light Control Cycle duration : green-red-yellow Green light ratio Signal offset between two consecutive

intersections

Network connectivity

Two types of traffic: Non-transit Transit

Four categorized of traffic: Morning Rush Hour traffic Lunch Time traffic : low transit Evening Rush Hour traffic Midnight traffic : high speed

Network connectivity

Two different network characteristics corresponding to two types of application

Static characteristics Network connectivity : reachable of safety

messages Path redundancy

Dynamic characteristics Connection duration Re-healing time

Network connectivity

Static characteristics of network connectivity Network connectivity : Two vehicles can be

connected either directly or indirectly (via a multi-hop route)

Path redundancy between two vehicles – the maximum number of (either node- or edge-) disjoint paths between two connected vehicles.

Network connectivity

network connectivity statistics averaged over 100 simulation runs

Network type Density (veh/km2)

Average network connectivity

Very sparse 40 68.12

Moderately sparse 60 97.97

Sparse 80 99.71

Moderate 160 100

Dense 240 100

Highly dense 320 100

Network connectivity

Average 20 neighboring vehicles => network connectivity 100%

network connectivity is less than 80% in a very sparse network (40 veh/km2)

Disconnected network problem may become a serious problem during the initial deployment of intelligent vehicles

Network connectivity

Path redundancy statistics

Network connectivity

number of redundant paths increases with the traffic density

But does not necessarily decrease with distance

Roughly 20 copies of the same message 8 more on the intersection In most cases, more than one path available

between them

Network connectivity

Dynamic characteristics of network connectivity

Number and duration of connected periods Re-healing time – the duration of time

during which two vehicles are disconnected

Network connectivity

Even in 80 veh/km2 dense network, the connectivity between two vehicles lasts for less than 6 minutes on average.

These statistics become much worse when traffic density decreases 10 sec in 40 veh/km2 network

Network connectivity

Re-healing time

Network connectivity

8 seconds of re-healing time in a very sparse network

less than 3 seconds in a dense network

Network connectivity

The bipolar behavior : connect ? not evenly distributed

Broadcast storm problem becomes much more severe in a moderate or highly dense network

Path redundancy Multi-path routing protocols

Conclusion

Cellular Automata (CA)-based mobility model analyzed the network connectivity pattern of urban traffic

serious disconnected network problem bipolar behavior is observed where both the

broadcast storm and the disconnected network problems coexist