cognitiveradio_14052008_dipti.ppt
TRANSCRIPT
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
1/22
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
2/22
Cognitive Capabilities
Awareness:should be aware of its
own abilities, the regulatingpolicies that govern it, its
neighbours and their abilities etc.
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
3/22
Cognitive Capabilities
Perception:should be able to
sense its environment
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
4/22
Cognitive Capabilities
Learning:consequent to
perception it should be able tolearn about the general
characteristics of its environment
and trends
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
5/22
Cognitive Capabilities
Reasoning:relationships between
the various entities should beunderstood and sound decisionsinferred, obviating theoverwhelming, and perhapsimpossible, task of enumeratingevery single alternative
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
6/22
Cognitive Capabilities
Memory:should demonstrate
improved performance afteroperating in the same environment
over an extended period of time
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
7/22
Architecture
Software Radio(perception)
Cognition Engine Learning Engine
(learning)
Reasoning Engine
(reasoning)
Knowledge Base
(awareness, memory)
SoftwareRadio
Cognition Engine
Knowledge
Base
Reasoning
Engine
Learning
Engine
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
8/22
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
9/22
Layer-wise distribution of Meters
and Knobs Contd.METERS
Interference,
BER, receivedsignal power,
noise power,
SNR, fading
statistics,
doppler spread,
delay spread,
angle of arrival,
dynamic range
KNOBSPower, frequency bandof operation, carriermodulation type,
baseband modulationtype, pulse shaping,data rate, number ofchannels, bandwidth,
equalization, antennatuning, antenna steering,antenna heightadjustment, type ofantenna (if more thanone type of antennaavailable)
Physical Layer
Data L ink Layer
(MAC + LLC)
Network
Transport
Security
Application
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
10/22
Layer-wise distribution of Meters
and Knobs Contd.METER
Frame error rate
KNOBS
Frame format,
Frame size,multiple access,
duplexing, FEC,
ARQ
(enable/disable)
Physical Layer
Data L ink Layer
(MAC + LLC)
Network
Transport
Security
Application
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
11/22
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
12/22
MTU Size vs ThroughputRef: Anna Calveras Auge, Josep Paradells Aspas. Performance Optimisation Evaluation of
TCP/IP over Wireless Networks., IEEE Performance, Computing and Communications, 1998
For each BER value,
an optimal MTU
exists!
BER: meterat the
physical layer. MTU: knobat the
network layer.Neural
Network
Classifier
BER
ThroughputMTU
Figure is a graph for Point to Point Protocol (PPP) anddeterministic errors.
Graphs would be different for other protocols like Frame
Relay, Asynchronous Transfer Mode (ATM), Ethernet etc.
They would also vary depending on the error type, namely,deterministic and burst.
It is proposed that the graphs can be learnt by a Multi-
Layer Perceptron (MLP) or some other suitable mechanism.
Reinforcement Learning mechanisms are also potentially
relevant.
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
13/22
OFDM Transmitter
Modulator(QAMQPSKBPSK)
Parallelto
serial
incomingbits
C1
C2
C3
Cn
IFFT
f1
f2
f3
fn
OFDM transmission offers opportunity
for adaptation!
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
14/22
OFDM Receiver
Serial toparallel
De-modulator
incomingbits
C2
C3
Cn
IFFT
f1
f2
f3
fn
C1
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
15/22
Capacity maximization in a non-AWGN channel
can be modeled as a Reinforcement Learning
problem.
Mmodulation types:M1,M2, ,MM.
N coding types: C1, C2, , CN.
Mican transmit didata bits per symbol and has a
probability of bit error ei(S) for signal to noise
ratio S.
Coding type Cjhas rate rjand can correct cjbit
errors per block of size bjbits. TS: symbol time, a constant.
B: corrected bit error rate coming out of the
decoder; measured by the radio.
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
16/22
For modulation typeMiand coding type Cj,
the resulting capacity is given by:
Ci,j= ((dirj)/TS ) (1-B)
Ci,j can act as a performance measure.
Objective function:
max f(Mi,Cj,B)= dirj (1-B)
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
17/22
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
18/22
Knowledge Representation
Each node or relation is associated to an
ontology that defines the concept. An
ontology consists of slots representingvarious attributes of that concept. For e.g.
here is the Device ontology created in
OWL (Web Ontology Language). (We usethe SWOOP editor to create and edit ontologies
and RDF Gravity to visualize them.)
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
19/22
Device Ontology
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
20/22
Reasoning
Reasoning involves traversing the semanticgraphs to obtain relevant conclusions.
Some ontologies for the cognitive radio: radio,channel, spectrum, power, coding, modulation,etc.
Some inferences:frequency fcis sparsely used
from time t1to time t2; for channel c, capacity ismaximized with modulation type miand codingmethod cjand so on.
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
21/22
A Simple Example using
Predicate Logic (papers include authors from US DoD) Software Radio (SR) exports predicates to the
knowledge-base regarding detected signalss1,s2,,sN, of the form
signalFreq(si,fi) signalBW(si,Wi)
Goal: To find somefcand Wthat does not overlapany detected signal, while maximizing W andhence the radios capacity.
Define, notOverlap(fc,W,si)
= (fi+Wi/2fc+W/2)
-
8/10/2019 CognitiveRadio_14052008_dipti.ppt
22/22
A Simple Example using
Predicate Logic Define predicate:
action: moveBand(fold,Wold, fnew, Wnew)
precond: i