Molecular Communication UsingTiming & Payload
Christopher Rose1
I. Saira Mian2
1Rutgers University, WINLAB2 University College London
WINLAB IABDecember 12, 2014
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)– Identical tokens (ACh, glutamate, epinephrine, Ca++ ... )
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)– Identical tokens (ACh, glutamate, epinephrine, Ca++ ... )– “Inscribed” tokens (DNA, mRNA, glycans, proteins ... )
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)– Identical tokens (ACh, glutamate, epinephrine, Ca++ ... )– “Inscribed” tokens (DNA, mRNA, glycans, proteins ... )
• Nanosystems will intercommunicate
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)– Identical tokens (ACh, glutamate, epinephrine, Ca++ ... )– “Inscribed” tokens (DNA, mRNA, glycans, proteins ... )
• Nanosystems will intercommunicate
– RF/Optical/Acoustic – often antenna mismatch/power issues
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)– Identical tokens (ACh, glutamate, epinephrine, Ca++ ... )– “Inscribed” tokens (DNA, mRNA, glycans, proteins ... )
• Nanosystems will intercommunicate
– RF/Optical/Acoustic – often antenna mismatch/power issues– Token exchange might be more efficient
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)– Identical tokens (ACh, glutamate, epinephrine, Ca++ ... )– “Inscribed” tokens (DNA, mRNA, glycans, proteins ... )
• Nanosystems will intercommunicate
– RF/Optical/Acoustic – often antenna mismatch/power issues– Token exchange might be more efficient
• Maybe macrosystems too
Rutgers WINLAB IAB Fall 2014 C. Rose
1 Bio-Inspired Wireless Overview
Biology-Inspired Molecular Communication
• Eu/Prokaryotic systems intercommunicate
– Emission/reception of special molecules (tokens)– Identical tokens (ACh, glutamate, epinephrine, Ca++ ... )– “Inscribed” tokens (DNA, mRNA, glycans, proteins ... )
• Nanosystems will intercommunicate
– RF/Optical/Acoustic – often antenna mismatch/power issues– Token exchange might be more efficient
• Maybe macrosystems too
Intriguing Science & Engineering
Rutgers WINLAB IAB Fall 2014 C. Rose
2 Bio-Inspired Wireless Overview
LOTS of recent work in the area
Rutgers WINLAB IAB Fall 2014 C. Rose
2 Bio-Inspired Wireless Overview
LOTS of recent work in the area
It’s a Snake!
Rutgers WINLAB IAB Fall 2014 C. Rose
2 Bio-Inspired Wireless Overview
LOTS of recent work in the area
It’s a Snake!It’s a Tree!
Rutgers WINLAB IAB Fall 2014 C. Rose
2 Bio-Inspired Wireless Overview
LOTS of recent work in the area
It’s a Snake!It’s a Tree!It’s a Wall!
Rutgers WINLAB IAB Fall 2014 C. Rose
2 Bio-Inspired Wireless Overview
LOTS of recent work in the area
It’s a Snake!It’s a Tree!It’s a Wall!
It’s a Spear!
Rutgers WINLAB IAB Fall 2014 C. Rose
2 Bio-Inspired Wireless Overview
LOTS of recent work in the area
It’s a Snake!It’s a Tree!It’s a Wall!
It’s a Spear!It’s a Rope!
Rutgers WINLAB IAB Fall 2014 C. Rose
3 Bio-Inspired Wireless Overview
Chris Is Getting Old (and cranky?)
Rutgers WINLAB IAB Fall 2014 C. Rose
3 Bio-Inspired Wireless Overview
Chris Is Getting Old (and cranky?)
Is there a
Rutgers WINLAB IAB Fall 2014 C. Rose
3 Bio-Inspired Wireless Overview
Chris Is Getting Old (and cranky?)
Is there a
Unifying Elephant?
Rutgers WINLAB IAB Fall 2014 C. Rose
3 Bio-Inspired Wireless Overview
Chris Is Getting Old (and cranky?)
Is there a
Unifying Elephant?
Framework+
Fundamental Limits+
Applications
Rutgers WINLAB IAB Fall 2014 C. Rose
4 Bio-Inspired Wireless Overview
A (stab at a) Unified Framework
Rutgers WINLAB IAB Fall 2014 C. Rose
4 Bio-Inspired Wireless Overview
A (stab at a) Unified Framework
Energy Use Is Fundamental
Rutgers WINLAB IAB Fall 2014 C. Rose
4 Bio-Inspired Wireless Overview
A (stab at a) Unified Framework
Energy Use Is FundamentalToken construction + Transport
Rutgers WINLAB IAB Fall 2014 C. Rose
4 Bio-Inspired Wireless Overview
A (stab at a) Unified Framework
Energy Use Is FundamentalToken construction + Transport
Inscribed Matter Is Fundamental
Rutgers WINLAB IAB Fall 2014 C. Rose
4 Bio-Inspired Wireless Overview
A (stab at a) Unified Framework
Energy Use Is FundamentalToken construction + Transport
Inscribed Matter Is Fundamentalm-RNA→ 3.6× 1024bits
kg
Rutgers WINLAB IAB Fall 2014 C. Rose
4 Bio-Inspired Wireless Overview
A (stab at a) Unified Framework
Energy Use Is FundamentalToken construction + Transport
Inscribed Matter Is Fundamentalm-RNA→ 3.6× 1024bits
kg
Timing Is Fundamental
Rutgers WINLAB IAB Fall 2014 C. Rose
4 Bio-Inspired Wireless Overview
A (stab at a) Unified Framework
Energy Use Is FundamentalToken construction + Transport
Inscribed Matter Is Fundamentalm-RNA→ 3.6× 1024bits
kg
Timing Is FundamentalMean first passage time is key
Rutgers WINLAB IAB Fall 2014 C. Rose
5 Bio-Inspired Wireless Overview
Today’s Talk
Rutgers WINLAB IAB Fall 2014 C. Rose
5 Bio-Inspired Wireless Overview
Today’s Talk
Timing Channel Abstraction
Rutgers WINLAB IAB Fall 2014 C. Rose
5 Bio-Inspired Wireless Overview
Today’s Talk
Timing Channel Abstraction
Information-Theoretic Modeling︸ ︷︷ ︸(for Roy, Narayan, Predrag, Waheed and Anand )
Rutgers WINLAB IAB Fall 2014 C. Rose
5 Bio-Inspired Wireless Overview
Today’s Talk
Timing Channel Abstraction
Information-Theoretic Modeling︸ ︷︷ ︸(for Roy, Narayan, Predrag, Waheed and Anand )
Energy
Rutgers WINLAB IAB Fall 2014 C. Rose
5 Bio-Inspired Wireless Overview
Today’s Talk
Timing Channel Abstraction
Information-Theoretic Modeling︸ ︷︷ ︸(for Roy, Narayan, Predrag, Waheed and Anand )
EnergyBounds
Rutgers WINLAB IAB Fall 2014 C. Rose
5 Bio-Inspired Wireless Overview
Today’s Talk
Timing Channel Abstraction
Information-Theoretic Modeling︸ ︷︷ ︸(for Roy, Narayan, Predrag, Waheed and Anand )
EnergyBounds
Ball Park Calculations
Rutgers WINLAB IAB Fall 2014 C. Rose
6 Bio-Inspired Wireless Overview
Diffusion Cartoon
(T ,B )k k
j j(S ,B )~
captureemission
Coding→ Emission→ Transport→ Capture→ Decoding
Rutgers WINLAB IAB Fall 2014 C. Rose
7 Bio-Inspired Wireless Overview
Diffusion with Drift Cartoon
(T ,B )k k
j j(S ,B )~
captureemission
Coding→ Emission→ Transport→ Capture→ Decoding
Rutgers WINLAB IAB Fall 2014 C. Rose
8 Bio-Inspired Wireless Timing Channel Details
Mathematical Abstraction
SM
1S
1D
1T
MT
MD
SSort
+
+
Rutgers WINLAB IAB Fall 2014 C. Rose
8 Bio-Inspired Wireless Timing Channel Details
Mathematical Abstraction
SM
1S
1D
1T
MT
MD
SSort
+
+
S = T + D
Rutgers WINLAB IAB Fall 2014 C. Rose
8 Bio-Inspired Wireless Timing Channel Details
Mathematical Abstraction
SM
1S
1D
1T
MT
MD
SSort
+
+
S = T + D~S = Sort[S]
Rutgers WINLAB IAB Fall 2014 C. Rose
8 Bio-Inspired Wireless Timing Channel Details
Mathematical Abstraction
SM
1S
1D
1T
MT
MD
SSort
+
+
S = T + D~S = Sort[S]
First passage time: E[D] = 1/µ
Rutgers WINLAB IAB Fall 2014 C. Rose
9 Bio-Inspired Wireless Timing Channel Details
Mutual InformationM tokens on an interval τ(M)
Rutgers WINLAB IAB Fall 2014 C. Rose
9 Bio-Inspired Wireless Timing Channel Details
Mutual InformationM tokens on an interval τ(M)
I(S; T) = h(S)− h(S|T)
= h(S)− h(D)
≤ M (h(S)− h(D)) , (i.i.d. D)
Rutgers WINLAB IAB Fall 2014 C. Rose
9 Bio-Inspired Wireless Timing Channel Details
Mutual InformationM tokens on an interval τ(M)
I(S; T) = h(S)− h(S|T)
= h(S)− h(D)
≤ M (h(S)− h(D)) , (i.i.d. D)
Easy, Right?
Rutgers WINLAB IAB Fall 2014 C. Rose
9 Bio-Inspired Wireless Timing Channel Details
Mutual InformationM tokens on an interval τ(M)
I(S; T) = h(S)− h(S|T)
= h(S)− h(D)
≤ M (h(S)− h(D)) , (i.i.d. D)
Easy, Right?
I(~S; T) = h(~S)− h(~S|T) = ?
Rutgers WINLAB IAB Fall 2014 C. Rose
10 Bio-Inspired Wireless Timing Channel Details
Hypersymmetries
Rutgers WINLAB IAB Fall 2014 C. Rose
10 Bio-Inspired Wireless Timing Channel Details
Hypersymmetries
∃M ! TΩ→ ~T
(permutation operator PΩ(), index Ω)
Rutgers WINLAB IAB Fall 2014 C. Rose
10 Bio-Inspired Wireless Timing Channel Details
Hypersymmetries
∃M ! TΩ→ ~T
(permutation operator PΩ(), index Ω)
T and PΩ(T) are indistinguishable at output
Rutgers WINLAB IAB Fall 2014 C. Rose
10 Bio-Inspired Wireless Timing Channel Details
Hypersymmetries
∃M ! TΩ→ ~T
(permutation operator PΩ(), index Ω)
T and PΩ(T) are indistinguishable at output
We can balance any given I()-maximizing fT() so that:
fT(T) = fT(PΩ(T)) ∀Ω
Rutgers WINLAB IAB Fall 2014 C. Rose
10 Bio-Inspired Wireless Timing Channel Details
Hypersymmetries
∃M ! TΩ→ ~T
(permutation operator PΩ(), index Ω)
T and PΩ(T) are indistinguishable at output
We can balance any given I()-maximizing fT() so that:
fT(T) = fT(PΩ(T)) ∀ΩConsider Only Hypersymmetric T
maxfT
I(~S,T)
Rutgers WINLAB IAB Fall 2014 C. Rose
11 Bio-Inspired Wireless Timing Channel Details
More Symmetry
fT() hypersymmetry→ fS() hypersymmetry
Rutgers WINLAB IAB Fall 2014 C. Rose
11 Bio-Inspired Wireless Timing Channel Details
More Symmetry
fT() hypersymmetry→ fS() hypersymmetry
fD() non-singular→ fS() continuous
Rutgers WINLAB IAB Fall 2014 C. Rose
11 Bio-Inspired Wireless Timing Channel Details
More Symmetry
fT() hypersymmetry→ fS() hypersymmetry
fD() non-singular→ fS() continuous
∴ “Edges and Corners” of fS() have zero measure
Rutgers WINLAB IAB Fall 2014 C. Rose
11 Bio-Inspired Wireless Timing Channel Details
More Symmetry
fT() hypersymmetry→ fS() hypersymmetry
fD() non-singular→ fS() continuous
∴ “Edges and Corners” of fS() have zero measure
M ! identical (permuted) patches of fS()
Rutgers WINLAB IAB Fall 2014 C. Rose
11 Bio-Inspired Wireless Timing Channel Details
More Symmetry
fT() hypersymmetry→ fS() hypersymmetry
fD() non-singular→ fS() continuous
∴ “Edges and Corners” of fS() have zero measure
M ! identical (permuted) patches of fS()
h(~S) = h(S)− logM !
Rutgers WINLAB IAB Fall 2014 C. Rose
12 Bio-Inspired Wireless Timing Channel Details
Channel Redux
SM
1S
1D
1T
MT
MD
S
+
+
Ω
SΩ⇒ ~S
Rutgers WINLAB IAB Fall 2014 C. Rose
13 Bio-Inspired Wireless Timing Channel Details
A Useful Equivalence
Rutgers WINLAB IAB Fall 2014 C. Rose
13 Bio-Inspired Wireless Timing Channel Details
A Useful Equivalence
~S,Ω ↔ S
Rutgers WINLAB IAB Fall 2014 C. Rose
13 Bio-Inspired Wireless Timing Channel Details
A Useful Equivalence
~S,Ω ↔ S
h(S|T) = h(~S,Ω|T))
= h(~S|T) +H(Ω|~S,T)
Rutgers WINLAB IAB Fall 2014 C. Rose
13 Bio-Inspired Wireless Timing Channel Details
A Useful Equivalence
~S,Ω ↔ S
h(S|T) = h(~S,Ω|T))
= h(~S|T) +H(Ω|~S,T)
I(~S; T) = I(S; T)−(
logM !−H(Ω|~S,T))
Rutgers WINLAB IAB Fall 2014 C. Rose
13 Bio-Inspired Wireless Timing Channel Details
A Useful Equivalence
~S,Ω ↔ S
h(S|T) = h(~S,Ω|T))
= h(~S|T) +H(Ω|~S,T)
I(~S; T) = I(S; T)−(
logM !−H(Ω|~S,T))
I(~S; T) = h(S) +H(Ω|~S,T)︸ ︷︷ ︸The Money!
− (logM ! + h(D))︸ ︷︷ ︸constant
Rutgers WINLAB IAB Fall 2014 C. Rose
14 Bio-Inspired Wireless Timing Channel Details
TENSION!
Rutgers WINLAB IAB Fall 2014 C. Rose
14 Bio-Inspired Wireless Timing Channel Details
TENSION!
Entropy maximized by independent T
h(S) ≤∑m
h(Sm)
Rutgers WINLAB IAB Fall 2014 C. Rose
14 Bio-Inspired Wireless Timing Channel Details
TENSION!
Entropy maximized by independent T
h(S) ≤∑m
h(Sm)
H(Ω|~S,T) maximized by correlated T
H(Ω|~S,T) = logM !
identical launch times T1 = T2 = · · · = TM
Rutgers WINLAB IAB Fall 2014 C. Rose
15 Bio-Inspired Wireless Timing Channel Details
My Past Personal Struggles
∃ closed form results/bounds for H(Ω|~S,T)
Rutgers WINLAB IAB Fall 2014 C. Rose
15 Bio-Inspired Wireless Timing Channel Details
My Past Personal Struggles
∃ closed form results/bounds for H(Ω|~S,T)
maxfT()
h(S) +H(Ω|~S,T) ≥ ? (ISIT’13)
Rutgers WINLAB IAB Fall 2014 C. Rose
15 Bio-Inspired Wireless Timing Channel Details
My Past Personal Struggles
∃ closed form results/bounds for H(Ω|~S,T)
maxfT()
h(S) +H(Ω|~S,T) ≥ ? (ISIT’13)
maxfT()
h(S) +H(Ω|~S,T) ≤ ? (ISIT’14)
Rutgers WINLAB IAB Fall 2014 C. Rose
16 Bio-Inspired Wireless Timing Channel Details
Channel Use Formalities Handwaving
Rutgers WINLAB IAB Fall 2014 C. Rose
16 Bio-Inspired Wireless Timing Channel Details
Channel Use Formalities Handwaving
PUNCHLINEρ ≡ M
launch epochall ok if mean first passage time E[D] <∞
Rutgers WINLAB IAB Fall 2014 C. Rose
17 Bio-Inspired Wireless Timing Channel Details
Capacity Per Token
Define:Cm(M) =
1
MmaxfT()
I(~S; T)
Rutgers WINLAB IAB Fall 2014 C. Rose
17 Bio-Inspired Wireless Timing Channel Details
Capacity Per Token
Define:Cm(M) =
1
MmaxfT()
I(~S; T)
Then:Cm = lim
M→∞Cm(M)
Rutgers WINLAB IAB Fall 2014 C. Rose
17 Bio-Inspired Wireless Timing Channel Details
Capacity Per Token
Define:Cm(M) =
1
MmaxfT()
I(~S; T)
Then:Cm = lim
M→∞Cm(M)
And:Ct = ρCm
Rutgers WINLAB IAB Fall 2014 C. Rose
18 Bio-Inspired Wireless Timing + Payload
Construction Energy
Rutgers WINLAB IAB Fall 2014 C. Rose
18 Bio-Inspired Wireless Timing + Payload
Construction Energy
Identical Tokens: c0 joules per token
Rutgers WINLAB IAB Fall 2014 C. Rose
18 Bio-Inspired Wireless Timing + Payload
Construction Energy
Identical Tokens: c0 joules per token
Inscribed Tokens:
Rutgers WINLAB IAB Fall 2014 C. Rose
18 Bio-Inspired Wireless Timing + Payload
Construction Energy
Identical Tokens: c0 joules per token
Inscribed Tokens:substrate: c1 joules per token
payload B bits: B∆c1 joules per tokensequence# K bits: K∆c1 joules per token, where
Rutgers WINLAB IAB Fall 2014 C. Rose
18 Bio-Inspired Wireless Timing + Payload
Construction Energy
Identical Tokens: c0 joules per token
Inscribed Tokens:substrate: c1 joules per token
payload B bits: B∆c1 joules per tokensequence# K bits: K∆c1 joules per token, where
1MH(Ω|~S,T) ≤ K ≤ 1
M logM !
Rutgers WINLAB IAB Fall 2014 C. Rose
19 Bio-Inspired Wireless Timing + Payload
And Now ...
Rutgers WINLAB IAB Fall 2014 C. Rose
19 Bio-Inspired Wireless Timing + Payload
And Now ...
Define:χ ≡ µ (first passage rate)
ρ (token launch rate)
Rutgers WINLAB IAB Fall 2014 C. Rose
19 Bio-Inspired Wireless Timing + Payload
And Now ...
Define:χ ≡ µ (first passage rate)
ρ (token launch rate)
Min Max Lower BoundParade
Rutgers WINLAB IAB Fall 2014 C. Rose
19 Bio-Inspired Wireless Timing + Payload
And Now ...
Define:χ ≡ µ (first passage rate)
ρ (token launch rate)
Min Max Lower BoundParade
exponential first passage
Rutgers WINLAB IAB Fall 2014 C. Rose
19 Bio-Inspired Wireless Timing + Payload
And Now ...
Define:χ ≡ µ (first passage rate)
ρ (token launch rate)
Min Max Lower BoundParade
exponential first passage(it’s kinda the timing channel’s “Gaussian”)
Rutgers WINLAB IAB Fall 2014 C. Rose
20 Bio-Inspired Wireless Bounds
Timing-Only Bits/Joule
Rutgers WINLAB IAB Fall 2014 C. Rose
20 Bio-Inspired Wireless Bounds
Timing-Only Bits/Joule
Theorem 1.
CT ≥1
c0
logχ+ e−1χ
∞∑k=2
(1
χ
)k(kχ− 1)
log k!
k!︸ ︷︷ ︸per token order uncertainty
Rutgers WINLAB IAB Fall 2014 C. Rose
21 Bio-Inspired Wireless Bounds
Payload-Only Bits/JouleTheorem 2.
CP =B
c1 + ∆c1
(B + mint
1MH(Ω|~S, t)
)Lemma 3.
CP ≥B
c1 + ∆c1
B + e−1χ
∞∑k=2
(1
χ
)k(kχ− 1)
log k!
k!︸ ︷︷ ︸per token order uncertainty
Rutgers WINLAB IAB Fall 2014 C. Rose
22 Bio-Inspired Wireless Bounds
Payload + Timing Bits/Joule Lower Bound
Theorem 4.
RP+T ≈log
(1 + χM
e
)+B
c1 + ∆c1
B + e−1χ
∞∑k=2
(1
χ
)k(kχ− 1)
log k!
k!︸ ︷︷ ︸per token order uncertainty
where RP+T ≤ CP+T .
Rutgers WINLAB IAB Fall 2014 C. Rose
22 Bio-Inspired Wireless Bounds
Payload + Timing Bits/Joule Lower Bound
Theorem 4.
RP+T ≈log
(1 + χM
e
)+B
c1 + ∆c1
B + e−1χ
∞∑k=2
(1
χ
)k(kχ− 1)
log k!
k!︸ ︷︷ ︸per token order uncertainty
where RP+T ≤ CP+T .
ASIDE: dumb header ( 1M logM !): CP+T → 0 in M
Rutgers WINLAB IAB Fall 2014 C. Rose
23 Bio-Inspired Wireless Bounds
Info per Unit Energy
χ↔ passage rate per launch ratec0 = 1, c1 = 0, ∆c1 = 1
Rutgers WINLAB IAB Fall 2014 C. Rose
24 Bio-Inspired Wireless Bounds
Info per Passage per Unit Energy
1χ ↔ launch rate per passage rate
c0 = 1, c1 = 0, ∆c1 = 1
Rutgers WINLAB IAB Fall 2014 C. Rose
25 Bio-Inspired Wireless Play Time
Play Time Setup
Rutgers WINLAB IAB Fall 2014 C. Rose
25 Bio-Inspired Wireless Play Time
Play Time Setup
Rsource sink
Rutgers WINLAB IAB Fall 2014 C. Rose
25 Bio-Inspired Wireless Play Time
Play Time Setup
Rsource sink
Protein Token Construction 4BATP = 3.2B × 10−19J
Rutgers WINLAB IAB Fall 2014 C. Rose
25 Bio-Inspired Wireless Play Time
Play Time Setup
Rsource sink
Protein Token Construction 4BATP = 3.2B × 10−19J
Diffusion Coefficient, D:≈ 10−5m2/s in air≈ 10−5cm2/s in water
Rutgers WINLAB IAB Fall 2014 C. Rose
25 Bio-Inspired Wireless Play Time
Play Time Setup
Rsource sink
Protein Token Construction 4BATP = 3.2B × 10−19J
Diffusion Coefficient, D:≈ 10−5m2/s in air≈ 10−5cm2/s in water
Mean First Passage Time, E[D] =≈ R2
2D
Rutgers WINLAB IAB Fall 2014 C. Rose
25 Bio-Inspired Wireless Play Time
Play Time Setup
Rsource sink
Protein Token Construction 4BATP = 3.2B × 10−19J
Diffusion Coefficient, D:≈ 10−5m2/s in air≈ 10−5cm2/s in water
Mean First Passage Time, E[D] =≈ R2
2D
Across a table (1m): E[D] ≈ 14hrs (need fan )
Rutgers WINLAB IAB Fall 2014 C. Rose
25 Bio-Inspired Wireless Play Time
Play Time Setup
Rsource sink
Protein Token Construction 4BATP = 3.2B × 10−19J
Diffusion Coefficient, D:≈ 10−5m2/s in air≈ 10−5cm2/s in water
Mean First Passage Time, E[D] =≈ R2
2D
Across a table (1m): E[D] ≈ 14hrs (need fan )
Across a synapse (20nm): E[D] = 0.2µs
Rutgers WINLAB IAB Fall 2014 C. Rose
26 Bio-Inspired Wireless Play Time
Play Time Numbers
Rutgers WINLAB IAB Fall 2014 C. Rose
26 Bio-Inspired Wireless Play Time
Play Time Numbers
1χ
= ρµ
= 1 = B
Rutgers WINLAB IAB Fall 2014 C. Rose
26 Bio-Inspired Wireless Play Time
Play Time Numbers
1χ
= ρµ
= 1 = B
Across a table: ≈ bits/day/attojoule
Rutgers WINLAB IAB Fall 2014 C. Rose
26 Bio-Inspired Wireless Play Time
Play Time Numbers
1χ
= ρµ
= 1 = B
Across a table: ≈ bits/day/attojouleAcross a synapse: ≈ Mb/s/attojoule
Rutgers WINLAB IAB Fall 2014 C. Rose
26 Bio-Inspired Wireless Play Time
Play Time Numbers
1χ
= ρµ
= 1 = B
Across a table: ≈ bits/day/attojouleAcross a synapse: ≈ Mb/s/attojoule
1χ
= ρµ
= 1000 = B:
Rutgers WINLAB IAB Fall 2014 C. Rose
26 Bio-Inspired Wireless Play Time
Play Time Numbers
1χ
= ρµ
= 1 = B
Across a table: ≈ bits/day/attojouleAcross a synapse: ≈ Mb/s/attojoule
1χ
= ρµ
= 1000 = B:Across a table: ≈ Kb/day/femtojoule
Rutgers WINLAB IAB Fall 2014 C. Rose
26 Bio-Inspired Wireless Play Time
Play Time Numbers
1χ
= ρµ
= 1 = B
Across a table: ≈ bits/day/attojouleAcross a synapse: ≈ Mb/s/attojoule
1χ
= ρµ
= 1000 = B:Across a table: ≈ Kb/day/femtojoule
Across a synapse: ≈ Gb/s/femtojoule
Rutgers WINLAB IAB Fall 2014 C. Rose
27 Bio-Inspired Wireless Fantasy
Tantalizing
Rutgers WINLAB IAB Fall 2014 C. Rose
27 Bio-Inspired Wireless Fantasy
Tantalizing
Suppose token construction energy cost fan energy cost
Rutgers WINLAB IAB Fall 2014 C. Rose
27 Bio-Inspired Wireless Fantasy
Tantalizing
Suppose token construction energy cost fan energy cost
1mg RNA per second⇒ 3.6× 1018 bits/sec
Rutgers WINLAB IAB Fall 2014 C. Rose
28 Bio-Inspired Wireless Fantasy
Appropriately Awed Response
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Lower Bounds
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Lower Bounds
Need Bit Efficiency?
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Lower Bounds
Need Bit Efficiency?Slow release with timing &/or small payload
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Lower Bounds
Need Bit Efficiency?Slow release with timing &/or small payload
Need Rate Efficiency?
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Lower Bounds
Need Bit Efficiency?Slow release with timing &/or small payload
Need Rate Efficiency?Fast release with payload + timing or large payload
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Lower Bounds
Need Bit Efficiency?Slow release with timing &/or small payload
Need Rate Efficiency?Fast release with payload + timing or large payload
Scary Efficiencies and Rates
Rutgers WINLAB IAB Fall 2014 C. Rose
29 Bio-Inspired Wireless Summary
Molecular Communication
Timing + Payload Framework
Lower Bounds
Need Bit Efficiency?Slow release with timing &/or small payload
Need Rate Efficiency?Fast release with payload + timing or large payload
Scary Efficiencies and Rates(beware transport latency)
Rutgers WINLAB IAB Fall 2014 C. Rose