Communications in Distributed Autonomous Vehicles

Download Report

Transcript Communications in Distributed Autonomous Vehicles

Cooperative Control of Distributed
Autonomous Vehicles in
Adversarial Environments
2002 MURI Minisymposium
Ameesh Pandya
Prof. Greg Pottie
2002 MURI Minisymposium
Overview
• Fault tolerant communication network supporting
hierarchical distributed communication network.
• Robust network algorithm for highly dynamic
mobile nodes (say, UAVs).
• Providing minimum communications between
mobile nodes to minimize the probability of
jamming.
• Working closely with Prof. Speyer’s group to
develop the communication model according to
control traffic.
2002 MURI Minisymposium
Wireless Communication Model
Application Layer
Transport layer
IP
Network
Link Layer
MAC Layer
Radio
Channel
2002 MURI Minisymposium
Our Concentration
Application Layer
Transport layer
IP
Network
Link Layer
MAC Layer
Radio
Channel
2002 MURI Minisymposium
Area of Concentration
QoS Constraints for Control Traffic
• Data Rate for the control traffic : 2 Mbps
– This could be considered as the upper bound.
– Achieved by using 2 Mbps modem.
• Latency for control traffic: 0 – 100 ms
– Worst latency is 100ms for control traffic.
2002 MURI Minisymposium
Channel Capacity
• Capacity constraint for the control traffic.
• Channel capacity in terms of received and
transmitted power, jamming power, spread
factor, bit rate.
• Goal is to know the reliable transmitting
distance between the nodes at 2Mbps for the
given parameters.
2002 MURI Minisymposium
Channel Capacity
• Assumptions:
– Isotropic antenna
– Spread spectrum modulation.
– For Low probability of intercept (LPI),
Pr/WsN0 = 0.1, where Pr is the received power
and Ws is the band width of spread spectrum
signal.
2002 MURI Minisymposium
Channel Capacity
• Shannon’s Equation:
Pr
C  W log 2 (1 
)
WN 0
where, Pr is the received power, W is the channel bandwidth.
• For isotropic antenna,
Pt
Pr 
4d 2
where Pt is the transmitted power
• Spread factor, f = Ws/R, where Ws is the band
width of the spread spectrum signal and R is the
information rate in bps.
2002 MURI Minisymposium
Channel Capacity
• If we do not consider broadband jammer, then
C  W log 2 (1 
1
Pt
)
WN 0 4d 2
• In presence of broadband jammer capacity becomes:
C  W log 2 [1 
Pt
4d 2
1
]
5 Jav' Eb
W ( N0  2 2
)
r f R N 0
'
where, J av is the average jamming power at distance r from the
receiver
• If we use CDMA, then in presence of jammer for Nu simultaneous
users, channel capacity is given by (assuming identical signal power):
C  W log 2 [1 
Pt
1
]
4d 2 W ( N  2 N u  1 E )
0
b
f
2002 MURI Minisymposium
Simulation Result
•
•
•
•
2002 MURI Minisymposium
Achievable transmitting
distance at 2 Mbps for
different values of
transmitting power.
Here, the transmitting
power is assumed to be
1 Watt and 2 Watts.
Assuming available
channel bandwidth to
be 100Mbps.
Simulation carried out
with the assumption of
ideality i.e. no jammer
and propagation loss.
MAC Layer Clustering
• Considering n nodes
(UAVs).
• Selecting clusters (cluster
heads).
• Each cluster having back
bone node.
• Using optimal cluster
algorithm.
2002 MURI Minisymposium
Future Objectives
• Developing clustering algorithms for mobile
nodes in dynamic environment.
• Clustering algorithms:
– UAV - UAV
– UAV - UGV
• Obtaining simulation results on the performance
and robustness of the algorithm.
2002 MURI Minisymposium
Insight
• The solution to the communication network model
for this particular problem “may” be very close to
IPv6.
• Looking into this possibility.
2002 MURI Minisymposium