Transcript Dynamic Spectrum Allocation System

An Efficient Method For Hop Selection
And Capacity Enhancement In MultiHop Wireless AD-Hoc Networks
Xiaohua (Edward) Li and Juite Hwu
Department of Electrical and Computer Engineering
State University of New York at Binghamton
[email protected]
1.Introduction
• Multi-hop transmission
– Benefit:
• Make network flexible,
• Create larger network
• Save the energy of source node.
– Deficit:
• Increase the complex of system
• Interference from other nodes
1.Introduction (cont' )
• Our objective for this paper
– Deal with mutual interference
– Derive the SINR of each node
– Locate the best path which provides
the maximum capacity under mild
conditions.
2.System Model
• Consider a H-hop wireless
•
•
network with J+1 nodes in a
LхL meters square (H≤J)
Let each node have a
transmission power p
Node i receives signal power
pdij-α node j
L
hop3
Node 2
hop2
Node 1
hop1
0 Source
Node 0
dij: distance between node i and j
α： path-loss exponent
Destination
Node J
L
2.System Model: slotted
transmission protocol
Slot #
0
1
2
…..
k
…..
Node 0
Tx Pk 0
Tx Pk 1
Tx Pk 2
…..
Tx Pk k
…..
Rx Pk 0
Tx Pk 0
Rx Pk 1
Tx Pk 1
Rx Pk 2
Node 1
Tx Pk k-1
Rx Pk k
:
Node j
:
…..
Tx Pk k-j
Rx Pk k-j+1
…..
2.System Model: received signal of
node i
H 1
x j k    pqi , j e
j i , j
i 0
uk  i   N v j k 
dij: distance between node i and j
α： path-loss exponent
u: signal of packet k
v: noise
N: power of noise
3. SINR Analysis and Optimization
Slot #
…..
k-1
k
…..
:
Node j-1
Tx Pk k-j
Rx Pk k-j+1
Node j
…..
Rx Pk k-j
…..
Tx Pk k-j
Rx Pk k-j+1
Node j+1
:
Simplify the received signal
j 1
xˆ j  k    pqi , j e
i 0
ji , j
Signal known by
Node j
u  k  i   Nv j  k 
3. SINR Analysis and
Optimization (cont’)
• SINR of node j at time slot k
pg j 1, j
s j k  
j 2
i0 pgi, j  N
• SINR of node j at time slot k-1
pg j  2, j
s j  k  1 
j 3
 i 0 pgi, j  N
Notice: here we track the same packet u(k-j+1)
3. SINR Analysis and
Optimization (cont’)
• For the detection of u(k-j+1)
sj k  l  
pg j l 1, j

j l  2
i 0
pgi , j  N
• Apply maximal ratio combing (MRC) to find
the maximum SINR of yj(k)
j 1
y j  k    al x j  k  l 
l 0
xˆ 
x
normalized
al  s j  k  l e
 j j l 1, j
3. SINR Analysis and
Optimization (cont’)
• Calculate the overall SINR with MRC
j 1
sj  
l 0
pg j l 1, j

j l  2
i 0
pgi , j  N
Notice: This SINR is for a node j in an H-hop when
detecting packets
3. SINR Analysis and
Optimization (cont’)
• Calculate the Capacity
C1, , H  H   min log 2 1  s j 
1 j  H
Like a water pipe, the capacity is
limited by the minimum tunnel
• Find the max one, we can get the best
transmission hopping path
C H  
max
nodes1, , H 11, J 1
C1,
,H
H 
4. Hop optimization and
node selection
• First term of sj is the dominating one
(when l=0)
j 1
pg j 1, j

j 2
i 0

pgi , j  N
l 0
pg j l 1, j

j l  2
i 0
pgi , j  N
• Formulate the problem to a max-min scheme
p m j l 1 dm 
i
j 1
max min 
dk 
1 j  H
l 0

j l  2
i 0
p m j l 1 dm  N
i
4. Hop optimization and
node selection (cont’)
• Apply previous method
•
•
and locate the points
M nodes are closed to
these points
We have MH-1 possible
paths
Destination
Node J
L
Node i+1
Node i
0
Source
Node 0
M Nodes
L
5. Simulations
Numerical optimization
and Monte-Carlo
simulations
 L=100 meters
 Exhaustive: locate the
best one from MH-1
possible paths
5. Simulations (cont’)
Numerical optimization
and Monte-Carlo
simulations
 Same parameters
 Capacity raises with
the increasing of H
 H increase analysis
performance is better
than proposed method
5. Simulations (cont’)
Numerical optimization
and Monte-Carlo
simulations
 Maximum capacity
we may gain
6. Conclusion
• Advantage of our method
– More efficient in large networks
– Feasible to approximate the optimal path by
numerical evaluation
• Disadvantage
– Results becomes more suboptimal in larger
networks
– We need a good initial condition
Thank you
and
Any question?