A brief introduction to my work
Zhengyang Qu (5080309950)
My interest
• Cognitive Radio (CR) MAC protocol design
• CR testbed
Provide a physical environment
Reliable evaluation of algorithms
Enable modification and simplification
The value of testbed
• A large number of advanced CR algorithms and
models have been proposed.
e.g. Auction, Learning.
• Most of these algorithms have not been evaluated
in a practical system.
The real performance of these algorithms remain unknown.
• Testbed is essential to evaluate the proposed
algorithms and to advance the development of
cognitive radio network
The implementation （CC1110）
Advantages
• CC1110 is a true low power SoC processor designed for wireless
applications
• It can be programmed in high level programming language, C++,
which can support much quicker prototyping than the traditional
testbeds based on FPGA
• There are three 8-bit timer and one 16-bit timer in the processor
that can be used to precisely control the PU and SU.
• The RF frequency is fully programmable in the three ISM bands
and several modulation methods are supported by the RF
transceiver
e.g. 300.0-348.0 MHz, 391.0-464.0 MHz and 782.0-928.0 MHz.
2-FSK, GFSK, MSK, ASK
A power control MAC protocol for
moving target based on testbed
•Transmission power control (TPC) technology
using feedback RSSI value
•RSSI: received signal strength indicator,
provides a reliable estimation on the channel
condition
TPC scheme for stationary nodes
C = W log2 (1+ SNRrec ),
(1)
denotes the lower bound of the available capacity
SNRrec means the lower bound of the received SNR value for
reliable transmission
C
When transmission starts, sender transmits a preamble
signal with the upper bound of output power Pout
1.Announce the pending data frame
2. Request the channel condition
After receives the preamble signal, the receivers
feedback the RSSI value on the preamble signal,
which contains the received power Precpre.
Then the sender could decides the optimal
data
transmission power Pout
for the coming data
packets
SNR ´ N
data
out
P
=
rec
pre
rec
P
Pout ,
where N means the AWGN noise.
(3)
The extension for mobility
• We introduce the power offset DP , because
the feedback RSSI only reflects the channel
condition when RSSI is sent.
• The channel variance during the time gap
between RSSI signal and DATA transmission
could cause bad impact on the link quality.
• Thus the output power for moving target is
given as:
t
data
data
Pmob-out
= Pout
+ DP.
(7)
How to achieve the power offset?
• We assume the maximal speed of the nodes
to be v .
• And we deduce the maximal power offset DPmax
under the worst condition.
Under the worst case, the maximal distance
variance between the sender and the receiver is
Ddmax = 2vt .
(4)
Then the maximal power offset can be expressed
as :
a
Ddmax a
data
DPmax = [(1+
) -1]Pout .
ddet
(6)
Here,
denotes the pass loss factor and ddet
means the distance between sender and receiver
when RSSI is transmitted
The maximal power
offset performs as a
metrics and
normalizes the
power offset.
We monitor the
packet loss rate, and
increase/reduce the
power offset when
the packet loss rate
is above/below the
threshold.
TPC is not enough
TPC allows the nodes to control their output
power dynamically, which is energy efficient and
provide a reliable transmission.
We plan to add some scheduling algorithm
based on game.
PW-MAC
(1) Receiver-initiated protocol
(2) Sender waits in awake and ask for the prediction word from the receiver at first
(3) Sender wakes up shortly before receiver’s wake up in the following transmission
(4) Prediction leads to the reduction on duty-cycle.
Apply the game
• Basically, we define a game parameter
(1)The remaining energy
(2)The data buffer saturation degree
• Cooperative game
• Non-cooperative game: cheating
(1)Introduce the punishment
(2)The punishment should not have a large effect on PU’s
performance
• Thank you