On Cooperative Settlement Between Content, Transit and Eyeball ISPs
Richard T.B. Ma
Columbia University
Dah-ming Chiu, John C.S. Lui
The Chinese University of Hong Kong
Vishal Misra, Dan Rubenstein
Columbia University
Outline
• Current ISP Settlement Problems
• ISP Models
• Profit Sharing Among ISPs
• Implications
A view of Internet Service Providers (ISPs)
• The Internet is composed of Autonomous Systems (ASes).
• An ISP is a business entity.– Might comprise multiple ASes.– Provide Internet access.
– Objective: maximize profits.
ISP
• Eyeball ISPs– Provide Internet access to customers:– Place Large investment in infrastructure.– E.g. AT&T, Verizon …
• Content ISPs– Provide contents via the Internet.– Serve customers like:
• Transit ISPs– Tier 1 ISPs: global connectivity of the Internet.– Provide transit services for other ISPs.– Cover a large geographic area.
Different classes of ISPs
Current ISP settlements
Zero-Dollar Peering Settlement
Customer-Provider Settlement
Transit ISP Transit ISP
Content ISP Eyeball ISP
Transit ISP
ISP positions on current settlement
Not enough revenue to recover investments.Other ISPs are free-riding on our facilities.
Home-users’ monthly fees do not cover costs.We should be able to generate more revenue.
Transit
Eyeball
Content Providers We have paid our fair share for transit and delivery and buy bandwidth from ISPs.
Yes No
Net Neutrality Debate: Provide Content-based Service Differentiation ?
Network Balkanization: De-peering between ISPs
Level 3 Cogent
Issues of the current ISP settlements
How to appropriately share profits amongst ISPs?
Transit Eyeball
Transit Transitzero-dollar peering
Content Providers
Contribution of this work
• Modeling of ISPs– How the revenues are generated– How different kinds of ISPs interact with one another
• Profit Sharing Solution Among ISPs– Efficiency– Fairness– Uniqueness
• Implications on Bilateral Settlements– Why the current settlements failed– What kind of new settlements should emerge
The Network Model: Eyeball Side
• Geographic Regions (r)
• Per Customer Monthly Charge (r)
• Customer Size (Xr)
• Eyeball ISP (Bj)
• Revenue from a region r (rXr)
$
US
UK
₤
X$
X₤
$ X$+₤ X₤
B1
B2
B2
B3
r=$
r= £
Eyeball Side Demand Assumption
• Elastic intra-region demand – Customers can switch among ISPs
within a region.– New eyeballs may take customers from
other eyeballs in the same region.– Customers move to other eyeballs when
the original eyeball leaves the system.
• Inelastic inter-region demand– Customers cannot switch to ISPs in
other regions.– Constant customer size in a region.
B1
US
B2
B3
UK
B3
X$
X₤
The Network Model: Content Side
X$
X₤
{♫, ♣}
$ X$+₤ X₤
• Content Items (q)
• Content ISP (Ci)
• Per Customer Revenue for content q (q)
• Content-side Revenue for uploading content q to region r (qXr)
C2
C1
C3♣
♫
(♫ +♣)(X$+X₤)
How to share profits amongst ISPs?
{♫}
{♣}
{♫,♣}
• One content and one eyeball ISP.
• One region, US, and one content, ♫.
• Egalitarian profit sharing:
X$
{♫}
C1 B1
US$ X$♫ X$
Profit generated: v=($+♫)X$
j(B )=j(C ) = v21
How to share profit? -- the baseline case
• Symmetry: symmetric eyeball ISPs get the same profit.• Efficiency: summation of all ISPs’ profit equals v.
• Fairness: same mutual contribution for any pair of ISPs.
X$
{♫}
C1
B1
US
$ X$♫ X$
B2
j(C ) - v = j(B ) - 021
j(C ) +2 j(B ) = v
j(C )= v32
61j(B )= v
How to share profit? -- multiple eyeballs
Unique solution (Shapley value)
Properties of Shapley Value
The Shapley Value
Efficiency Symmetry Fairness
Myerson 1977
Routing Incentive
Interconnecting Incentive
CoNEXT ‘07
Solution Stability
Shapley 1977
Efficiency Symmetry Dummy Additivity
Shapley 1953
Efficiency Symmetry Strong Monotonicity
Young 1985
• The unique solution (Shapley value) that satisfies Efficiency Symmetry and Fairness:
n eyeball ISPs.
j(B )= v, j(C ) = vn+1n
n(n+1)1
♫ X$
X$
{♫}
C1
B1
US
$ X$
B2
Bn
How to share profit? -- multiple eyeballs
{♫}
C1
B1
US
Bn-1
• The more eyeballs, the more profit the content ISP gets.– Elastic users move between eyeball ISPs.– Multiple eyeball ISPs provide redundancy; – The only content ISP has more leverage.
• When one eyeball leaves the system: • The marginal profit loss of the content ISP:
– If n=1, the content ISP loses everything if the eyeball leaves.
– The content ISP loses only 1/n2 of its original profit.
j(B )= v, j(C ) = vn+1n
n(n+1)1
Results and implications of profit sharing
Bn
nn-1j’ (C )= v
Dj(C )= v - v = - j(C )n+1n
nn-1 1
n2
C2
US
B1
C1
Cm
X$
$ X$
{♫}
{♫}
{♫}
♫ X$
m content ISPs.
j(C )= v, j(B ) = vm+1m
m(m+1)1
• The unique solution (Shapley value) that satisfies Efficiency Symmetry and Fairness:
How to share profit? -- multiple contents
• The more contents, the more profit the eyeball ISP gets.– Content can be obtained by any content ISP.
– Multiple content ISPs provide redundancy;
– The only eyeball ISP has more leverage.
• The marginal profit loss of the eyeball ISP:
– If m=1, the eyeball ISP loses everything if the content leaves.
– The eyeball ISP loses only 1/m2 of its original profit.
j(C )= v, j(B ) = vm+1m
m(m+1)1
Dj(B )= - j(B )1m2
C2
US
B1
C1
Cm
{♫}
{♫}
{♫}
Results and implications of profit sharing
C2
C1
Cm
{♫}
{♫}
{♫}
♫ X$X$
B1
US
$ X$
B2
Bn
• The unique solution (Shapley value) that satisfies Efficiency Symmetry and Fairness:
j(B )= v, j(C ) = vm(n+m)
nn(n+m)
m
Profit share -- multiple eyeballs and contents
• Intuition for elastic demand and supply– The more of the same kind provide redundancy.
– The less of a kind can obtain more leverage.
C2
C1
Cm
B1
B2
Bn
Results and implications of ISP profit sharing
• Each ISP’s profit is– Inversely proportional to the
number of ISPs of its type.
– Proportional to the number of ISPs of the opposite type.
j(B ) = , j(C ) = nm
(n+m)v n
m(n+m)v
Sm=1
m
Sk=1
k
n+m+kv ( )m
m ( )kk ( )n+m+k-1
m+kj(B )=
Sn=1
n
Sk=1
k
n+m+kv ( )n
n ( )kk ( )n+m+k-1
n+kj(C )=
Sm=1
m
Sn=1
n
n+m+kv ( )m
m ( )nn ( )n+m+k-1
m+nj(T )=
♫ X$ C2
C1
Cm
{♫}
{♫}
{♫}
X$
B1
US
$ X$
B2
Bn
T2
T1
Tk
Profit share -- eyeballs, transits and contents
Profit share -- eyeballs, transits and contents
Sm=1
m
Sk=1
k
n+m+kv ( )m
m ( )kk ( )n+m+k-1
m+kj(B )=
Sn=1
n
Sk=1
k
n+m+kv ( )n
n ( )kk ( )n+m+k-1
n+kj(C )=
Sm=1
m
Sn=1
n
n+m+kv ( )m
m ( )nn ( )n+m+k-1
m+nj(T )=
• Intuition– The more of the same kind provide redundancy.
– The less of a kind can obtain more leverage.
• Revenue sources are separable – Eyeball-side components:
– Content-side components:
Profit share -- multiple regions and items
$ X$ ₤ X₤
♫X$ ♣X$ ♫X₤ ♣X₤
B2X$
B1
US
$ X$
C2
C1
{♫}
UK
X₤
₤ X₤
{♣}
C2 B3
T2
T1
T3
{♫,♣}
(♫ +♣)(X$+X₤)
• A specific revenue component is shared by– Content ISPs that provide the item
– Eyeball ISPs that generate the revenue
– Transit ISPs that help the delivery
Profit share -- multiple regions and items
B2X$
B1
US
$ X$
C2
C1
{♫}
UK
X₤
₤ X₤
{♣}
C2 B3
T2
T1
T3
{♫,♣}
(♫ +♣)(X$+X₤)
♣X$
Profit share – general topologies
v = ♫ X$
B2 X$
B1
US
C1
{♫}T1
ji(N, v) = [S ji(N \{j}, v) + v(N )1{i is veto}]j≠i|N |
1
jC1(N, v) = [0 + v + v + v] = v41
31
31
125
Dynamic Programming Procedure!
jC1 = 0
B2
B1
C1 T1
jC1 = 1/3v
B2
B1
C1 T1
jC1 = 1/3v
B2
B1
C1 T1
C1 is Veto.
B2
B1
C1 T1
T1
T2
T3
T4
C2
C1
C3
B2
B3
B1
B2
CB
T
Implications – the value chain
$ $$
$ $
$ $
$$
$ $$
$
$
$$ $$
CRBR
T1
T2
T3
T4
C2
C1
C3
B2
B3
B1
B2
CB
T
Implications – the value chain
• Revenue Flows– Content-side revenue (CR): Content Transit Eyeball
– Eyeball-side revenue (ER): Eyeball Transit Content
$ $$
$ $$
$ $$
$$$
CRBR
T1
T2
T3
T4
C2
C1
C3
B2
B3
B1
B2
CB
T
Implications – equivalent bilateral settlements
$ $$
$ $$
$ $$
$$$
CRBR
• When CR ≈ BR, bilateral implementations:– Customer/Provider: Contents & Eyeballs are customers.
– Zero-dollar Peering: Transit ISPs peer with each other.
– Stable structure for homogenous local ISPs 10 years ago.
Provider CustomerCustomer
Zero-dollarPeering
T1
T2
T3
T4
C2
C1
C3
B2
B3
B1
B2
CB
T
Implications – equivalent bilateral settlements
$ $$
$ $ $
$ $$
$$$
CR
BR$ $$
$ $ $$ $$
$ $ $
• If CR >> BR, bilateral implementations:– Reverse Customer/Provider: Transits compensate Eyeballs. – Paid Peering: content-side compensate eyeball-side.– New settlements are needed to sustain a stable structure.
Customer Provider
Paid Peering
• Content-Transit-Eyeball ISP model– Customer demand, revenue generation.
– Closed-form Shapley value for regular topologies.
– Dynamic Programming for general topologies.
• Implications for current bilateral settlements– Transit ISPs might need to compensate Eyeball ISPs,
which creates a Reverse Customer/Provider settlement.
– Paid Peering settlement might exist among Transit ISPs.
• Guideline for – Government: make regulatory policy for the industry.
– ISPs: negotiate stable and incentive settlements.
Summary