cognitive radios and dynamic spectrum access and cognitive radio • software-defined radio (sdr) -...
TRANSCRIPT
![Page 1: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/1.jpg)
Cognitive Radios and Dynamic Spectrum Access
© 2015
Luiz DaSilva
Professor of Telecommunications Trinity College Dublin
![Page 2: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/2.jpg)
Acknowledgement
• These slides are adapted from a set originally prepared by Prof Luiz DaSilva, Trinity College Dublin, and Prof Allen MacKenzie, Virginia Tech, for a tutorial on cognitive radios and networks
![Page 3: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/3.jpg)
Evolution
Traditional (firmware)
Radios
Software- Defined Radios
Cognitive Radios
Cognitive Networks
![Page 4: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/4.jpg)
SDR and Cognitive Radio
• Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system architectures for wireless networks and user terminals (SDR Forum)
• Cognitive radio (CR) – a radio that is aware of
and can sense its environment, and can make decisions about its operating behavior based on that information and pre-defined objectives (IEEE 1900.1)
![Page 5: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/5.jpg)
Cognitive Radio Definition
• A transceiver that is aware, adaptive, and capable of learning from experience
… of the RF environment (communication waveform features, propagation channel characteristics) … of its own capabilities (power consumption, DSP clock rate, device operational status, …) … of policies it needs to follow … of local (link) and global (network) objectives … of network conditions … of other users’ priorities and authorizations
![Page 6: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/6.jpg)
Cognitive Radio Definition
• A transceiver that is aware, adaptive, and capable of learning from experience
… transmit power control … dynamic waveform selection … dynamic spectrum access … routing … negotiation of waveforms and protocols
![Page 7: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/7.jpg)
Cognitive Radio Definition
• A transceiver that is aware, adaptive, and capable of learning from experience
… experience-weighted table lookup … machine learning algorithms
![Page 8: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/8.jpg)
Evolution Revisited
Software Radio
• Conventional Radio +
• Software Architecture
• Reconfigurability • Provisions for
easy upgrades
Conventional Radio
• Traditional RF Design
• Traditional Baseband Design
Cognitive Radio
• SDR + • Intelligence • Awareness • Learning • Observations
![Page 9: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/9.jpg)
Cognitive Radio
Cognitive Engine
Objective Function
Policies
Allowed frequency bands for primary/secondary use
Maximize SNR, data rate
Occupied bands
Signal strength
Meters Knobs
Neighbor list
Frequency of operation
Power
Waveform
![Page 10: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/10.jpg)
Levels of Cognitive Radios (1/2)
Level Capability Comments
0 Pre-programmed
A software-defined radio.
1 Goal Driven Environment awareness. Chooses waveform according to goal.
2 Context Awareness
Knowledge of what the user is trying to do.
3 Radio Aware Knowledge of radio and network components, environment models.
Adapted from: J. Mitola, Cognitive Radio: An Integrated Agent Architecture for Software Defined Radio, PhD Dissertation, RTH, Sweden, May 2000.
![Page 11: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/11.jpg)
Levels of Cognitive Radios (2/2)
Level Capability Comments
4 Capable of Planning
Analyzes situation (levels 2 & 3) to determine goals.
5 Conducts Negotiations
Negotiates operating parameters with another radio.
6 Learns Environment
Autonomously determines structure of new environment.
7 Adapts Plans Generates new goals.
8 Adapts Protocols
Proposes and negotiates new protocols.
Adapted from: J. Mitola, Cognitive Radio: An Integrated Agent Architecture for Software Defined Radio, PhD Dissertation, RTH, Sweden, May 2000.
![Page 12: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/12.jpg)
Cognition Cycle
Ada
pted
from
Mito
la, “
Cog
nitiv
e R
adio
for F
lexi
ble
Mob
ile
Mul
timed
ia C
omm
unic
atio
ns ,”
IEEE
Mob
ile M
ultim
edia
C
onfe
renc
e, 1
999,
pp
3-10
.
![Page 13: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/13.jpg)
Cognition Cycle : Simplified
Observe
Environment
Decide
Orient
Act
![Page 14: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/14.jpg)
Applications (1/2)
Source: Bostian, 2007
![Page 15: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/15.jpg)
Applications (2/2)
• Dynamic spectrum access (the primary driver) • Cooperative medium access and cooperative
communications • Opportunistic switching among available wireless
networks (cellular, WLAN, mesh, etc.) • Adaptive selection of available radio resources
(e.g., cognitive MIMO, selection among different waveforms)
• Public safety interoperability and deployment of emergency infrastructure
![Page 16: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/16.jpg)
Dynamic Spectrum Access
• Temporal and geographical variations in spectrum utilization – Measurements indicate that utilization of
licensed spectrum typically varies between 15 and 85%
• Cognitive approaches are being developed to opportunistically take advantage of gaps in spectrum utilization
• DSA has become a primary motivation for cognitive radios and networks – But DSA does not imply a cognitive approach – And cognition can be applied to other problems
![Page 17: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/17.jpg)
Desirable Cognitive Radio Attributes
• Wideband – “DC to daylight”
• Any waveform • Flexible architecture • High performance and low power consumption • Affordable and accessible • Relatively straightforward to use and innovate
with • Robust • Access to suitable frequency spectrum segments
for innovation and testing
![Page 18: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/18.jpg)
Hardware and Software Any SDR can go
here
Source: Bostian, 2007
![Page 19: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/19.jpg)
Universal Software Radio Peripheral
• Responsible for getting baseband signals to and from RF, and very little else – Up/Down conversion – Filtering – Interpolation and
decimation • ~4 MHz usable bandwidth
via USB 2.0 • USRP and Ettus Research:
http://www.ettus.com
Source: Rondeau and Nolan, 2008
![Page 20: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/20.jpg)
GNU Radio
• http://gnuradio.org/trac/wiki • Communication systems built out of discrete
components – One task per component
• Components cover: – Sources and sinks (USRP, sound, file, vector,
UDP) – Filters (FIR and IIR) – Math (add, subtract, multiply, etc.) – Probes (GUI windows, power) – Signal processing (synchronization, mod/
demod, etc.) • Combined Python/C++
![Page 21: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/21.jpg)
Core Research Areas (1/2)
• Awareness – spectral sensing, signal detection and classification – Evolving towards broader notions of context
awareness • Hardware platforms – development of agile,
adaptive radios • APIs – interfaces between the cognitive engine
and meters, meters, knobs, objective function and policies
• Learning and reasoning mechanisms • Protocols and spectrum etiquette
![Page 22: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/22.jpg)
Core Research Areas (2/2)
• Cognitive networks – convergence, cross-layer effects, dissemination of network state information
• Security and validation • Policies and economic models
![Page 23: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/23.jpg)
Cognitive Process
• At the core of a cognitive network (or cognitive radio) is a cognitive process
• This algorithm is responsible for – Reasoning – Learning – Planning
• Most candidate algorithms from artificial intelligence
![Page 24: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/24.jpg)
Reasoning
• The immediate decision process leading to a choice of an action or actions
• Uses historical knowledge and knowledge about the current environment – Current environment might
include spectrum/signal environment, network conditions, user preferences, and policy considerations
![Page 25: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/25.jpg)
Learning
• The long-term process of accumulating knowledge based on the perceived results of past actions
• Learning enriches the cognitive element’s knowledge base in order to improve the efficacy of future reasoning
• In some cases, learning and reasoning may be closely intertwined
![Page 26: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/26.jpg)
Planning
• The process of developing a strategic sequence of actions in order to meet a goal
• Unfortunately, planning has not yet been applied to the development of a cognitive process
• We will focus on reasoning and learning algorithms
![Page 27: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/27.jpg)
Cognitive Radio: The standard AI system view
27
Cogni)ve engine reads informa)on from the radio and adjusts its knobs through actuators.
![Page 28: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/28.jpg)
Centralized Reasoning • Optimization Algorithms
– Find the provably best solution to a carefully stated optimization problem
– Nonlinear programming, integer programming, combinatorial optimization, etc.
• Bayesian Networks – Not exactly reasoning, per se, but easily expanded to a
reasoning approach. • Metaheuristic Algorithms
– Black-box procedures for solving computationally “hard” problems
– Most use a local or neighborhood search procedure • Neural Networks
– Simple computational elements (neurons) combined and trained to solve a problem
![Page 29: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/29.jpg)
Learning Methods
• Supervised – System trained using a set of known inputs
and desired outputs provided by an expert • Unsupervised
– System trained in an open loop fashion with no input from experts or rewards
• Rewards-based – System uses utility measurement to infer the
correctness of past actions
![Page 30: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/30.jpg)
Credit Assignment Problem
• The most difficult part of rewards-based learning is the credit assignment problem
• This is the problem of how to assign credit or blame for an outcome to a particular prior action or action component
• Two steps – Correlation of past actions with observations – Assignment of credit
• First step is particularly difficult in cognitive radios due to time scale differences
![Page 31: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/31.jpg)
Connection Between Learning and Reasoning
• Learning methods cannot be considered independently of the reasoning method that they support. – Sometimes, it difficult to distinguish between
them • Reasoning algorithm must apply the knowledge
that the learning algorithm has accumulated
![Page 32: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/32.jpg)
Learning in a Bayesian Network
• A key problem with applying the Bayesian Network technique is that the dependency graph must be known a priori
• But, the dependency graph can also be learned, given a set of past observations of inputs and outputs
• Since the search space is large, metaheuristics are often the “learning” algorithm of choice in this case
![Page 33: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/33.jpg)
Learning with a Genetic Algorithm
• Genetic Algorithms are sophisticated heuristic optimization algorithms – But, they are typically slow and incapable of
learning • Idea: Build a knowledge base of past problems
and solutions. Use this knowledge base to seed the GA population with potential solutions
![Page 34: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/34.jpg)
Case-based Decision Theory
• Build a database of cases: – Problem, Action, Result
• Compare new problem and objectives to cases in the knowledge base – Compare problems on similarity (from [0,1]) – Evaluate old results with new utility function – Rank order cases by similarity × utility
• Use actions from best cases in initial GA population
• Allows GA to find good solutions in few generations.
![Page 35: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/35.jpg)
Learning in a Neural Network
• “Learning” in a Neural Network – finding the appropriate weights – is typically done in a supervised fashion
• But, there are many possibilities for using the results of prior outcomes (e.g. the outcomes in a knowledge base) to train a neural net
• Typically, learning here means applying the back propagation algorithm (or another technique) to discover the appropriate weights – But, the structure and/or required size of the
neural network might also be learned
![Page 36: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/36.jpg)
Frequency Usage Policies
• Contiguous range of frequencies subject to some regulatory treatment
• Traditional: Static spectrum licenses guarantee protection from in-band interference from 3rd parties
• DSA: dynamically select operating spectrum currently not in use by incumbent or other secondary users
![Page 37: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/37.jpg)
Static Spectrum Allocation
• Divides spectrum by frequency band, location, and time
• Must consider worst-case scenarios • All frequency bands have been allocated
– But most licensed bands have low utilization – While unlicensed bands are often over-utilized
• Strongly favors the incumbent
![Page 38: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/38.jpg)
Spectrum Allocation Charts U
.S.
Jam
aica
Hon
g K
ong
Afg
hani
stan
![Page 39: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/39.jpg)
Dynamic Spectrum Access
• More flexible coexistence of users within a band • Primary or incumbent
– Holds spectrum access rights protected from interference
• Secondary – Accesses a band without causing interference
to primary users – There are multiple secondary users in a band – Need for spectrum etiquette – Access rights may be traded in a secondary
spectrum market
![Page 40: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/40.jpg)
Expected Market Dynamics
Source: Chapin and Lehr, 2007
![Page 41: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/41.jpg)
Primary/Secondary Cooperation
• Cooperative DSA (LSA, ASA) – Secondary user can only use the band with
permission from the primary – Contract established between parties
determines access rights, payments • Non-cooperative DSA
– Secondary does not require permission – “Easements” specify conditions for non-
cooperative access to a band
![Page 42: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/42.jpg)
Policy Options: Primary User
Source: Peha, 2005
![Page 43: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/43.jpg)
Policy Options: Secondary User
Source: Peha, 2005
![Page 44: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/44.jpg)
Spectrum Broker (1/3)
• A firm that specializes in spectrum trading • In its simplest form, matches buyers to sellers
Source: Chapin and Lehr, 2007
![Page 45: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/45.jpg)
Spectrum Broker (2/3)
• Contracts with users to provide QoS-differentiated spectrum access
• Acquires spectrum rights from license holders or through easements
Source: Chapin and Lehr, 2007
![Page 46: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/46.jpg)
Spectrum Broker (3/3)
• Takes responsibility for spectrum etiquette being followed between primary and secondary users
• Provides access to spectrum on demand
Source: Chapin and Lehr, 2007
![Page 47: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/47.jpg)
Policy Reasoner
• Adaptations to policies changing over time • Adaptations to policies changing over geography • Compliance to spectrum etiquette in secondary
markets • Self-checking policies
– Implications of adaptation decisions worked out in advance
• Regulatory compliance – Deployment may be accelerated due to
simplified accreditation (?)
![Page 48: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/48.jpg)
XG Policy Framework
Source: Krishnan, 2004
![Page 49: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/49.jpg)
Policy Language Requirements
• Must handle complexity of current spectrum policy – A patchwork that evolved over time – Human (engineer/lawyer) readable
• Must be extensible and standards-based – XML and other markup languages
• Must support logical framework for – Validation of completeness and consistency of
policies – Verification of policy-conformance usage
![Page 50: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/50.jpg)
Barriers to DSA
• Conservatism of regulators (in many countries) • Resistance from incumbents
– May rely on spectrum scarcity as a barrier to increased competition
• Hardware platforms capable of reasoning and decision making – Cost, form factor
• Complexity
![Page 51: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/51.jpg)
Static Allocation: Command & Control
• Highly prescriptive management of the radio spectrum
• Licenses are assigned • Licenses have tended to be very specific detailing
what kind of services can be delivered and which kind of technologies should be used to deliver the services
![Page 52: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/52.jpg)
Specified Technologies, Specified Services
• Advantage: Interference is easier to control • Disadvantages: Slow, favors incumbents,
inefficient, regulators pick service and technology winners
![Page 53: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/53.jpg)
overlay underlay?
• Only 5-10% of the spectrum used at any one time.
• Why can’t we share? – Overlay – Opportunistic use of fallow spectrum – Underlay – Non-interfering easement use
Sharing Spectrum
![Page 54: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/54.jpg)
Fitting in White Space is Hard
![Page 55: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/55.jpg)
![Page 56: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/56.jpg)
Property Rights
• Defining spectrum property rights is itself a hard problem
• Goal: Define the parameters associated with a packet/bundle/block of spectrum so that: 1. The user of the spectrum has a clear set of
entitlements 2. The user has a clear set of obligations in
terms of target emission levels outside its own spectrum
3. It is possible to trade spectrum in some kind of market system
![Page 57: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/57.jpg)
A Spectrum Usage Continum
time
now + 1 minute
+ 1 day
+ 1 year
frequency
![Page 58: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/58.jpg)
Traditional Static Allocation
time
space
frequency
![Page 59: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/59.jpg)
time
frequency
space
Pure Market Based Assignment
![Page 60: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/60.jpg)
space
time
frequency
Market Based Assignment with Aggregation
![Page 61: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/61.jpg)
space
time
frequency
Real Life
![Page 62: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/62.jpg)
time
frequency
space
subletting
sub-subletting
Hierarchical Market Mechanism
![Page 63: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/63.jpg)
Spectrum of Options
Exclusive Usage Rights Spectrum Commons
![Page 64: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/64.jpg)
Open Spectrum
• Open access is provided to spectrum with few regulatory restrictions…
• Tragedy of the Commons – Refers to a conflict between individual interests
and the common good. – Examples: Packet Forwarding, Power Control,
etc.
![Page 65: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/65.jpg)
Seven Guide Rules for Spectrum Commons
1. All transmitters must have receiver capability (to enable adaptation)
2. Power is restricted (but on aggregate devices not on individual)
3. Signalling capability is provided (to enable unlicensed devices communicate resource needs)
4. Etiquette is defined (to allocate resources during congestion)
5. Enforcement mechanisms are defined (to enforce compliance with other rules)
6. Assignments are reversible (to allow resources to be de-allocated and deployed for other uses)
7. User security and privacy are protected
![Page 66: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/66.jpg)
Signal Detection and Classification
• Why is spectrum sensing an important aspect of cognitive radio functionality?
• Do all cognitive radios require spectrum sensing?
![Page 67: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/67.jpg)
Hypothesis Testing
• Types of Error – Type I Error (False Alarm): The probability of
detecting a primary user signal when none is present. P(reject H0 | H0 is true)
– Type II Error (Missed Detection): The probability of not detecting a primary user’s signal. P(accept H0 | H0 is false)
Received Signal
Null Hypothesis
Alternative Hypothesis
![Page 68: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/68.jpg)
Fundamental Detection Tradeoffs
• There is a tradeoff between the types of error. – Reducing the false alarm probability will
always increase the missed detection probability
– But, changing the detection method or the length of time for detection may improve matters
![Page 69: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/69.jpg)
Signal Detection Scenario 1: Known Signal
• Goal: Detect the presence/absence of a known signal
• Example: TV Band White Space • Optimal in AWGN: Matched Filter
– Maximizes received SNR – But, requires coherent detection
![Page 70: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/70.jpg)
Signal Detection Scenario 2: No Knowledge
• Goal: Detect the presence/absence of a completely unknown signal in a given band
• If no information about the transmitted signal is available, then the optimal detector in the presence of AWGN is the energy detector – Measure the energy at the output of a
bandpass filter, and compare to threshold – Simplest technique, most popular in literature
but poor performance
![Page 71: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/71.jpg)
Signal Detection Scenario 3: Informed Ignorance
• Goal: Exploit knowledge of the types of signals usually used for communications to improve detector performance without knowledge of the specific modulation scheme
• Modulated signals are usually coupled with periodic elements (sinusoidal carriers, pulse trains, repeating spreading codes, etc.) making them cyclostationary – By analyzing spectral correlation function, can
differentiate noise energy from signal energy – But, computationally complex and long
observation time
![Page 72: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/72.jpg)
Single-Radio Detection is Hard
• Consider the detection of a digital TV signal for a TV-band white space application – Typical DTV receiver can decode a signal level
of -83 dBm without error – Due to the possibility of fading & likely
difference in antenna position, CR should detect presence of DTV signal down to -113 dBm
– But, thermal noise in such bands is typically -106 dBm, so CR must detect DTV signals below the noise floor
![Page 73: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/73.jpg)
Two Radios Are Better Than One
• The amount of signal loss due to multipath and shadowing can change dramatically over relatively short distances – Hence, if several nodes work cooperatively,
there is a high likelihood that at least one of them will have a good instantaneous channel
![Page 74: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/74.jpg)
Some Results from Mishra, et al.
74
![Page 75: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/75.jpg)
Other Results from Mishra
• Whether hard or soft decisions should be exchanged is controversial in the literature. Appears that hard decisions may be sufficient if receivers are not synchronized.
• Correlated fading can have deleterious effects on cooperative performance. A few uncorrelated users are better than many correlated users.
• Cooperation between a small number of radios (~10) is sufficient to achieve good results.
![Page 76: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/76.jpg)
The Classification Problem
• Converts the simple hypothesis test into one with multiple alternative hypotheses:
![Page 77: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/77.jpg)
Likelihood Function
• A likelihood function is a conditional probability function considered as a function of its second argument, with its first argument held fixed:
• In our case, we are looking at the likelihood of a particular modulation given a received signal
• Statistical tests can then be performed by comparing likelihoods
![Page 78: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/78.jpg)
Likelihood-based Modulation Classification
• Attempts to calculate the likelihoods for each of the candidate modulation schemes. – Different approaches to dealing with unknown
parameters (e.g. phase, phase jitter, frequency offset, channel impulse response, etc. lead to different likelihood-based modulation classification approaches.
![Page 79: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/79.jpg)
Feature Based Modulation Classification
• Feature based modulation classification is a two-stage process: – Extract relevant “features” from the received
signal. – Use the chosen features to decide which
modulation scheme was used. • Hence, designing a FBMC requires
– Choosing the relevant features – Selecting a decision-making approach
![Page 80: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/80.jpg)
Feature Based Modulation Classification
• Some features: – Variance of the instantaneous magnitude,
phase, and frequency – The PDF of the phase – Statistics of the signal itself
• Some decision making techniques: – PDF-based – Euclidean distance – Neural networks – Clustering algorithms
![Page 81: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/81.jpg)
Feature Based Modulation Classification
• Often, FBMC is done in a hierarchical fashion, e.g. determine the modulation class (PSK vs. FSK vs. QAM) and then go back to determine the order of the constellation
• Simpler to implement than likelihood approaches, but poorer performance on some modulation classes
![Page 82: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/82.jpg)
Comparing Likelihood and Feature Based Classification
• Qualities of a good modulation classifier: – robust with respect to preprocessing
inaccuracy – able to recognize a large number of
modulations in environments with different propagation characteristics, real-time functionality
– low computational complexity
![Page 83: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/83.jpg)
Cognitive Networks
• In a network of autonomous, adaptive radios, do individual optimizations result in network-wide optimal performance?
• How much information about network conditions must independent radios have to make effective adaptations?
• Learning and reasoning are needed to manage complex cross-layer optimizations
![Page 84: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/84.jpg)
Definition
• A cognitive network has a cognitive process that can perceive current network conditions, and then plan, decide and act on those conditions. The network can learn from these adaptations and use them to make future decisions, all while taking into account end-to-end goals
• Critical components: – Cognition (as opposed to reactive, localized
schemes) – End-to-end goals (as opposed to layer level
goals) • Does not specify application or mechanism
![Page 85: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/85.jpg)
Adaptations at Multiple Levels
Node
Link
Network
Dynamic frequency selection
Beam Forming/ nulling
Power control
MIMO
Topology control
Spectrum negotiation
Cooperative Comms.
Waveform selection
Adaptive routing
![Page 86: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/86.jpg)
Cognitive Radio vs. Networks
Similarities • Cognitive process
• Tunable parameters
– Software Defined Radio – Software Adaptable
Network • Cross-layer aspects
Differences • End-to-end vs point-to-
point scope • Distributed vs centralized
problem solving • Heterogeneity • Cooperation vs Selfishness
– Cognitive network elements may work together to achieve higher level goals
• More complex with many more states, interactions, elements
“Improves its performance through experience gained over a period of time without complete information about the environment in which it operates”
![Page 87: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/87.jpg)
Framework for Cognitive Networks
Cognitive element
element goal
cognitive element
element goal
end-to-end goal end-to-end goal end-to-end goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
element goal
cognitive element
element goal
CSL
Requirements Layer
Cognitive Process
Software Adaptable Network
configurable element
configurable element
configurable element
network status sensor
network status sensor
cognitive element
element goal(s)
element status
transfer
SAN API
![Page 88: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/88.jpg)
Requirements Layer
• End-to-end goals from users, applications, resources
• Cognitive Specification Language (CSL) – Converts end-to-end
goals into cognitive elements goals
– Similar to a QoS specification language
• Element Goals – Specifies behavior of
individual cognitive elements
Cognitive element
element goal
cognitive element
element goal
end-to-end goal end-to-end goal end-to-end goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
element goal
cognitive element
element goal
CSL Application/ User/ Resource Layer
cognitive element
element goal(s)
Cognitive element
element goal
cognitive element
element goal
end-to-end goal end-to-end goal end-to-end goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
element goal
cognitive element
element goal
CSL Application/ User/ Resource Layer
Cognitive Process
Software Adaptable Network
configurable element
configurable element
configurable element
network status sensor
network status sensor
cognitive element
element goal(s)
element status
transfer
SAN API
![Page 89: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/89.jpg)
Cognitive Process
• Cognitive Elements – Can be single
process, centralized or distributed
– Can be multiple processes spanning one more or nodes
– Adapts and “learns” to make decisions that meet end-to-end goals
Observe
Environment
Decide
Orient
Act
Cognitive element
element goal
cognitive element
element goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
cognitive element Cognitive
Process
cognitive element
element status
transfer
Cognitive element
element goal
cognitive element
element goal
end-to-end goal end-to-end goal end-to-end goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
element goal
cognitive element
element goal
CSL Application/ User/ Resource Layer
Cognitive Process
Software Adaptable Network
configurable element
configurable element
configurable element
network status sensor
network status sensor
cognitive element
element goal(s)
element status
transfer
SAN API
![Page 90: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/90.jpg)
Cognitive Process (cont’d)
• Element Status Transfer – Communication of
state between different cognitive elements
• Performs feedback loop – OODA loop as
template – Future decisions
based on observations of network performance
Observe
Environment
Decide
Orient
Act
Cognitive element
element goal
cognitive element
element goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
cognitive element Cognitive
Process
cognitive element
element status
transfer
Cognitive element
element goal
cognitive element
element goal
end-to-end goal end-to-end goal end-to-end goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
element goal
cognitive element
element goal
CSL Application/ User/ Resource Layer
Cognitive Process
Software Adaptable Network
configurable element
configurable element
configurable element
network status sensor
network status sensor
cognitive element
element goal(s)
element status
transfer
SAN API
![Page 91: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/91.jpg)
Software Adaptable Network
• API – Provides control
over configurable network elements
• Configurable Elements – Points of network
control for cognitive process
• Network Status Sensors – Reads status of
the network
Software Adaptable Network
configurable element
configurable element
configurable element
network status sensor
network status sensor
SAN API
Cognitive element
element goal
cognitive element
element goal
end-to-end goal end-to-end goal end-to-end goal
cognitive element
element goal
cognitive element
element goal(s)
cognitive element
element goal
cognitive element
element goal
CSL Application/ User/ Resource Layer
Cognitive Process
Software Adaptable Network
configurable element
configurable element
configurable element
network status sensor
network status sensor
cognitive element
element goal(s)
element status
transfer
SAN API
![Page 92: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/92.jpg)
Cognitive Networks: Issues (1/4)
Cognitive Engine
Objective Function
Policies
How well do local objective functions reflect network-wide goals?
How do selfish objectives impact network performance (the cost of selfishness)?
![Page 93: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/93.jpg)
Cognitive Networks: Issues (2/4)
Cognitive Engine
Objective Function
Policies
How much information regarding network state is necessary for sound local decisions (the cost of ignorance)?
What are the tradeoffs between operating in ignorance and disseminating information throughout the network?
![Page 94: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/94.jpg)
Cognitive Networks: Issues (3/4)
Cognitive Engine
Objective Function
Policies
How do local adaptations affect end-to-end performance?
Do independent, autonomous adaptations converge? If so, do they converge to an efficient point?
![Page 95: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/95.jpg)
Cognitive Networks: Issues (4/4)
Cognitive Engine
Objective Function
Policies
What are the most efficient cognitive techniques?
How fast must cognition operate to be effective?
When is it worth trying to learn about the environment and when are simple, heuristic decisions more effective?
![Page 96: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/96.jpg)
Critical Issues (1/2)
• Abstraction and filtering – How to reduce the amount of information
required by the system • Distributed operation
– Solving multiple goals in a distributed fashion • Learning and cognition
– What differentiates a cognitive network from adaptive, distributed, behavior
• Heterogeneity – Differing network elements increase complexity
![Page 97: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/97.jpg)
Critical Issues (2/2)
• Control Channels – How to distribute control and state information
in the network. • Synchronization
– Need to avoid adaptation loops, keep observations fresh and relevant
• Platforms – Cognitive radio platforms are in their infancy – Most are not ready to support a full network
stack
![Page 98: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/98.jpg)
Read more… (1/3) • B. Le, T. W. Rondeau, C. W. Bostian, “Cognitive radio realities,” Wireless
Comm. Mobile Comp., 6:1-12, 2006. • J. Mitola III, “Cognitive radio architecture,” in Cognitive Networks: Towards
Self-Aware Networks, Q. H. Mahmoud (ed.), John Wiley and Sons, 2007, pp. 147-202.
• S. Haykin, “Cognitive radio: brain-empowered wireless communications,” IEEE JSAC, v. 23, no. 2, Feb. 2005.
• J. Mitola, Cognitive Radio: An Integrated Agent Architecture for Software Defined Radio, PhD Dissertation, RTH, Sweden, May 2000.
• D. H. Friend, R. W. Thomas, A. B. MacKenzie, and L. A. DaSilva, “Distributed Learning and Reasoning in Cognitive Networks,” in Cognitive Networks: Towards Self-Aware Networks, Q. Mahmoud, ed., Wiley, 2007 (and references therein).
• T. G. Dietterich and P. Langley, “Machine learning for cognitive networks: Technology assessment and research challenges,” in Cognitive Networks: Towards Self-Aware Networks (Q. Mahmoud, ed.), pp. 97-120, Wiley, 2007.
• C. Clancy, J. Hecker, E. Stuntebeck, and T. O’Shea, “Applications of machine learning to cognitive radio networks,” IEEE Wireless Communications, vol. 14, pp. 47-52, August 2007.
![Page 99: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/99.jpg)
Read more… (2/3) • L. Gavrilovska, V. Atanasovski, I. Macaluso, and L. A. DaSilva, “Learning
and Reasoning in Cognitive Radio Networks,” IEEE Communication Surveys and Tutorials, vol. 15, no. 4, 4th quarter 2013, pp. 1761-1777.
• J. M. Chapin and W. H. Lehr, “The path to market success for dynamic spectrum access technology,” IEEE Comm. Magazine, pp. 96-103, May 2007.
• J. M. Peha, “Approaches to spectrum sharing,” IEEE Comm. Mag., pp. 10-11, Feb. 2005.
• R. Krishnan, “Towards policy-defined cognitive radio,” NSF Workshop on Programmable Wireless, 2004.
• D. N. Hatfield and P. J. Weiser, “Property rights in spectrum: Taking the next step,” in Proc. IEEE Dynamic Spectrum Access Networks (DySPAN), 2005.
• W. Lehr and J. Crowcroft, “Managing shared access to a spectrum commons,” in Proc. IEEE Dynamic Spectrum Access Networks (DySPAN), 2005.
• J. M. Chapin and W. H. Lehr, “The path to market success for dynamic spectrum access technology,” IEEE Communications Magazine, pp. 96-103, May 2007.
![Page 100: Cognitive Radios and Dynamic Spectrum Access and Cognitive Radio • Software-defined radio (SDR) - a collection of hardware and software technologies that enable reconfigurable system](https://reader031.vdocuments.us/reader031/viewer/2022022502/5aad2cbf7f8b9a8d678dc67f/html5/thumbnails/100.jpg)
Read more… (3/3) • I. F. Akyildiz, W.-Y. Lee, M. C. Vuran, and S. Mohanty, “NeXt generation/
dynamic access/cognitive radio wireless networks: A survey,” Elsevier Computer Networks Journal, vol. 50, pp. 2127-2159, September 2006.
• M. Gandetto and C. Regazzoni, “Spectrum sensing: A distributed approach for cognitive radio terminals,” IEEE Journal on Selected Areas in Communications, vol. 25, pp. 546-557, April 2007.
• O. A. Dobre, A. Abdi, Y. Bar-Ness, and W. Su, “Survey of automatic modulation classification techniques: Classical approaches and new trends,” IET Communications, vol. 1, no. 2, pp. 137-156, April 2007.
• S. M. Mishra, A. Sahai, and R. W. Brodersen, “Cooperative sensing among cognitive radios,” in Proc. IEEE International Conference on Communications, vol. 4, 2006.
• C. R. C. M. da Silva, W. C. Headley, J. D. Reed, and Y. Zhao, “The application of distributed spectrum sensing and available resource maps to cognitive radio systems,” in Proc. of the Information Theory and Applications Workshop, (La Jolla, CA), 2008.
• R. W. Thomas, D. H. Friend, L. A. DaSilva, and A. B. MacKenzie, “Cognitive Networks: Adaptation and Learning to Achieve End-to-end Performance Objectives,” IEEE Communications Magazine, vol. 44, no. 12, Dec. 2006, pp. 51-57.