limits of communicationn3cat.upc.edu/presentations/master_thesis-ingacio_llatser.pdf · microsoft...
TRANSCRIPT
Exploring the ScalabilityExploring the Scalability Limits of CommunicationLimits of Communication Networks at the Nanoscale
Ignacio Llatser MartíIgnacio Llatser Martí
1Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Advisors: Eduard Alarcón Cot and Albert Cabellos‐Aparicio
Nanonetworks
2Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
NanonetworksNanonetworks
Nanotechnology is envisaged to allow theNanotechnology is envisaged to allow the development of nanometer‐scale machines
Nano‐EMNano‐EM
Biological
I F Ak ildi J Mi l J t “El t ti Wi l N N t k ”Ian F. Akyildiz, Josep Miquel Jornet, “Electromagnetic Wireless Nanosensor Networks”, Nano Communication Networks (Elsevier), 2010.
3Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
NanonetworksNanonetworks
Nanotechnology is envisaged to allow theNanotechnology is envisaged to allow the development of nanometer‐scale machines
Nano EMNano‐EM
Biological
I F Ak ildi F d B tti C i ti Blá “N t k A i tiIan F. Akyildiz, Fernando Brunetti, Cristina Blázquez, “Nanonetworks: A new communication paradigm”, Computer Networks (Elsevier), 2008.
4Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
NanonetworksNanonetworks
The capabilities of nanomachines are constrained byThe capabilities of nanomachines are constrained by their limited detection/actuation range.
Nanonetworking is an emerging field studying comm nication among nanomachinescommunication among nanomachines
The resulting nanonetworks will greatly expand the capabilities of a single nanomachine
5Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
NanonetworksNanonetworks
Current network protocols and techniques cannot beCurrent network protocols and techniques cannot be directly applied to communicate nanomachines
Two main paradigms emerge:
Nano‐electromagnetic communication
Molecular communication
6Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
NanoNano‐‐electromagneticelectromagnetic communicationcommunication
Graphene based nano antennas (CNTs and GNRs) areGraphene‐based nano‐antennas (CNTs and GNRs) are envisaged to implement nano‐EM communications
Due to the lower wave propagation speed in grapheneDue to the lower wave propagation speed in graphene, graphene‐based nano‐antennas radiate EM waves in the THz bandTHz band
Josep Miquel Jornet Ian F Akyildiz “Graphene Based Nano Antennas for Electromagnetic NanocommunicationsJosep Miquel Jornet, Ian F. Akyildiz, Graphene‐Based Nano‐Antennas for Electromagnetic Nanocommunications in the Terahertz Band”, Proc. European Conference on Antennas and Propagation, Barcelona, 2010 .
7Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Molecular communicationMolecular communication
Information is encoded insidemoleculesInformation is encoded inside molecules
Ca2+ DNA
8Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Molecular communicationMolecular communication
Molecules are sent among nanomachinesMolecules are sent among nanomachinesBrownian motion
Spontaneous diffusionSpontaneous diffusion
Massimiliano Pierobon Ian F Akyildiz “A physical end to end modelMassimiliano Pierobon, Ian F. Akyildiz, A physical end‐to‐end model for molecular communication in nanonetworks”, IEEE JSAC, 2010.
9Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
ApplicationsApplications of nanonetworksof nanonetworks
Wireless NanoSensor Networks (WNSN)Wireless NanoSensor Networks (WNSN)
Intrabody disease detection and cooperative drug delivery systemsdelivery systems
10Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Ian F. Akyildiz, Josep Miquel Jornet, “The Internet of Nano‐Things”, IEEE Wireless Communications, 2010.
Motivation Motivation of this thesisof this thesis
How different will nanonetworks be from traditionalHow different will nanonetworks be from traditional electromagnetic networks?
We need a scalability theory for nanonetworks
Study the performance metrics of the networkThroughput
Transmission delayTransmission delay
Energy consumption
…
When the network size is reduced to the nanoscale
11Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Main contributionsMain contributions
Scalability analysis of the channel capacity ofScalability analysis of the channel capacity of electromagnetic nanonetworks
Characterization (both analytically and by simulation) of the ph sical channel of diff sion based molec larthe physical channel of diffusion‐based molecular nanonetworks
Scalability analysis of several performance metrics using l b d d l ti i th i ia pulse‐based modulation in the previous scenario
12Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Scalability of the channel capacity of electromagnetic nanonetworksof electromagnetic nanonetworks
13Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Scalability of the channel capacity of EM nanonetworksScalability of the channel capacity of EM nanonetworks
Bandwidth THz very high channel capacityBandwidth THz very high channel capacity
ff h h l h lQuantum effects in the nano‐EM physical channel
v 1
Lower wave propagation speed
kdeA 1
LCvp
Molecular absorption
M l l i
abs eA
kdeTTT 1)1( Molecular noise
Only appears when signal is transmitted
mol eTTT 1)1( 00
14Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
ScalabilityScalability of the channel capacity of EM nof the channel capacity of EM nanonetworksanonetworks
How do these quantum effects affect the channelHow do these quantum effects affect the channel capacity at the nanoscale?
We particularize Shannon’s law for the frequency‐selecti e nano EM channelselective nano‐EM channel
15Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
ScalabilityScalability of the channel capacity of EM nof the channel capacity of EM nanonetworksanonetworks
We obtain analytical expressions of the channel capacityWe obtain analytical expressions of the channel capacity as a function of Δ, d and PT
Δ: nanomachine lengthd: transmission distanced: transmission distancePT: transmitted powerN0: noise power spectral densityc: speed of light
16Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
k1: constant
ScalabilityScalability of the channel capacity of EM nof the channel capacity of EM nanonetworksanonetworks
We find the limits of the previous expressions whenWe find the limits of the previous expressions when 0, d 0 and PT 0
We derive the necessary conditions to keep the net ork feasiblenetwork feasible
The transmission distance needs to scale proportionally to the nanomachine length:nanomachine length:
The scalability of the transmitted power PT depends on whether quantum effects are presentwhether quantum effects are present
17Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
ScalabilityScalability of the channel capacity of EM nof the channel capacity of EM nanonetworksanonetworks
Scalability of the transmitted power P as a function ofScalability of the transmitted power PT as a function of the nanomachine size
With quantum effectsWithout quantum effects
18Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Diffusion‐based channel characterization in molecularcharacterization in molecular nanonetworks
19Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
DiffusionDiffusion‐‐based channel characterizationbased channel characterization
Transmitters encode information into the release pattern of moleculesof molecules
Emitted molecules move according to Brownian motionFick’s laws of diffusion
Receivers measure the local concentration of molecules d d d h i f iand decode the information
Baris Atakan Ozgur B Akan “Deterministic capacity of information flow in
20Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Baris Atakan, Ozgur B. Akan, Deterministic capacity of information flow in molecular nanonetworks”, Nano Communication Networks (Elsevier), 2010.
DiffusionDiffusion‐‐based channel characterizationbased channel characterization
The diffusion based molecular channel is very differentThe diffusion‐based molecular channel is very different from the traditional EM channel
Bandwidth kHz low channel capacityBandwidth kHz low channel capacityLong propagation delay
Very energy efficientVery energy efficient
New sources of noiseBrownian motion
Molecules are discrete
We need to characterize this channel in order to study ythe scalability of diffusion‐based molecular communication
21Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
DiffusionDiffusion‐‐based channel characterizationbased channel characterization
We propose a pulse based modulation schemeWe propose a pulse‐based modulation scheme
Q: number of emitted moleculesD: diffusion coefficientr: transmission distancer: transmission distancet: time
22Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
DiffusionDiffusion‐‐based channel characterizationbased channel characterization
We find analytical expressions for the most relevantWe find analytical expressions for the most relevant metrics from the communication standpoint
Pulse delay
Pulse amplitudePulse amplitude
Q: number of emitted
Pulse width moleculesD: diffusion coefficientr: transmission distance
23Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
DiffusionDiffusion‐‐based channel characterizationbased channel characterization
The results are validated by simulationThe results are validated by simulation
Pulse delay Pulse amplitude
24Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
DiffusionDiffusion‐‐based channel characterizationbased channel characterization
P l idthPulse width
25Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
DiffusionDiffusion‐‐based channel characterizationbased channel characterization
Scalability of the performance metrics of the diffusionScalability of the performance metrics of the diffusion‐based molecular channel compared to the wireless EM channelchannel
26Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Conclusions and outcomes
27Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
ConclusionsConclusions
Nanonetworks will greatly expand the range ofNanonetworks will greatly expand the range of applications of nanotechnology
We lay the foundations of a scalability theory for nanonet orksnanonetworks
The use of graphene‐based antennas gives electromagnetic nanonetworks a scalability advantage over traditional networks
The studied metrics in molecular nanonetworks scale worse than in wireless electromagnetic networks
28Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Research outcomesResearch outcomes
4 papers4 papersI. Llatser, A. Cabellos‐Aparicio, E. Alarcón, J. M. Jornet, I. F. Akyildiz, “Scalability of the Channel Capacity of Electromagnetic Nanonetworks”, to be submitted to IEEE Transactions on Wireless Communications.
I. Llatser, E. Alarcón, M. Pierobon, “Diffusion‐based Channel Characterization in Molecular Nanonetworks”, submitted to IEEE MoNaCom 2011.
I Llatser I Pascual N Garralda A Cabellos‐Aparicio M Pierobon E AlarcónI. Llatser, I. Pascual, N. Garralda, A. Cabellos‐Aparicio, M. Pierobon, E. Alarcón, J. Solé‐Pareta. “NanoSim: A Simulation Framework for Diffusion‐based Molecular Communication”, to be submitted to IEEE GLOBECOM 2011.
N. Garralda, I. Llatser, A. Cabellos‐Aparicio, M. Pierobon “Simulation‐based , , p ,Evaluation of the Diffusion‐based Physical Channel in Molecular Nanonetworks”, submitted to IEEE MoNaCom 2011.
2 co‐supervised master thesisNora Garralda, “Simulation‐based Evaluation of the Diffusion‐based Physical Channel in Molecular Nanonetworks”.
Iñaki Pascual, “NanoSim: Simulation Tool for Diffusion‐based Molecular ,Communication in Nanonetworks”.
29Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Exploring the ScalabilityExploring the Scalability Limits of CommunicationLimits of Communication Networks at the Nanoscale
Ignacio Llatser MartíIgnacio Llatser Martí
30Master in Computer Architecture, Networks and Systems, Universitat Politècnica de Catalunya, 2011.
Advisors: Eduard Alarcón Cot and Albert Cabellos‐Aparicio