Transcript Title of Presentation Second Line of Title Subheadings (if

MOTA: Engineering an Operator Agnostic
Mobile Service
Supratim Deb, Kanthi Nagaraj, Vikram Srinivasan
Bell Labs
Mobile Data Explosion and Need for New Technological Innovations
FCC National Broadband Plan:
 500 MHz of additional spectrum
 Technical and Business innovations that increase efficiency of spectrum utilization
Wireless Service Today and the End User Perspective
Src: opensignalmaps.com
Takeaways:
 Spectrum shortage exacerbated by deployment practices
 Users demand choice
Wireless service provider should depend on location, pricing and user
preferences
Challenges in User’s Making Appropriate Choices?
 Option 1: Centralized entity makes choices.
 Operators unlikely to share network planning
information.
 Option 2: User’s use signal strength from
different base stations
 This is insufficient and can result in poor user
experience.
 Additional signaling information needed.
VF may be
better choice
 Choice of network should depend on user
mobility pattern
 Switching at fine time scales incurs huge overhead
in core network
Everyone
joins O2
 Goal:
 Distributed decisions by each user
 Concise network signaling that accounts for
mobility
 Evolutionary over current standards.
Src: opensignalmaps.com
MOTA Service Model
BTS
 Service Aggregator: New intermediary
between users and operators
 Responsible for maintaining customer
relationships
PGW
CoreNet1
Service Aggregator Cloud
BTS
AAA
Server
 Handles all control plane operations that
cannot be handled by a single operator
BTS
 Tracking and paging
PGW
 Billing and authentication
 Seamless switching across operators at Layer 3
CoreNet2
BTS
Network layer
and above
Module
for
Switching
Decisions
MIH Layer
MAC and lower
layers
Tracking
& Paging
Mobile
IPv6
Anchor
MOTA Framework
 What information should each operator
maintain?
 What aggregate information should be
broadcast by each base station?
User mobility?Network experience
Session duration?
Network load?
Operator 1
Operator 2
Price?
2G Interface
3G Interface
Application 1
Application 2
 What information should each user maintain?
Operator m
 How should a client decide the following:
 What operator to associate with each interface
(2G, 3G, 4G)?
 What applications to associate with each interface
(Voice, video, data etc.)?
4G Interface
Application n
User experience?
Battery status
User behavior
Utilities and Proportional Fairness – The Framework for All Seasons!
 User utility
Price sensitivity of application a
U a ( Ra , pa )  wa ln Ra  a pa
Weight of application a
Price of application a
Rate of application a
 User’s Objective:
Maximize E[U a ( Ra , pa )]
aA
 Subject to:
– Each application associates with only one interface
– Each interface associates with only one operator.
Comment: Price for each operator is constant. Operators sets a single price per unit
weight per technology across all cells
Signaling and Algorithm for Static Clients
Fact:
 Proportional fair scheduling is typically used by most cellular technologies.
 If total weight of applications associated with a base station j is Wj, and PHY rate of user u is ru
and weight of his application is wa, then aggregate rate user receives under proportional fair
scheduling is:
wa
Tj 
rr
Wj
 Network Signaling for Static Users: Each base station only needs to transmit its
aggregate load Wj and its price pj.
Recall
Operator 1
Operator 2
2G Interface
3G Interface
Application 1
Application 2
4G Interface
Operator m
Base station conveys
1. Price per unit weight
2. Total load
Application n
User computes
1. Which operator to select for each technology
2. Which application goes to each technology
Based on
1. Signaling information
2. Energy considerations
3. Application characteristics
Greedy User Algorithm
 Utility of associating application of weight w to base station j = w f(pj, Wj, w)
 Utility of operator that offers maximum utility is Gl
Gl ( w)  max f ( p j ,W j , w)
operators
 Order application weights in increasing order w1 <= w2,… <= wn
 Assign applications in this order.
 Greedy Algorithm:
 Iterate over all applications
 In the rth step
 Assign application r to interface that maximizes
(curr _ weight  wr )G(curr _ weight  wr ) 
(curr _ weight)G(curr _ weight)
Price of Anarchy – Global Efficiency versus Selfish Strategy
Theorem:
 Let r be vector of PHY data rates of all users.
 There exists a constant K, 1  K   such that
SELFISH( K .r )  GLOBAL(r )

Comment: Proportional fair scheduling at base stations ensures that local
decisions are not very bad.
Signaling for Mobile Users
 Question: What signaling information should the base station send that is useful from a
user’s perspective?
 Answer: Something that will allow the user to compute her net utility when she
associates with this operator and moves around.
 Question: Isn’t this dependent on each user’s individual mobility pattern?
 Answer: Clearly yes. Hence convey only aggregate information based on average usage
patterns. This could depend on time of day etc.
Signaling for Mobile Clients
Base station tracks:
R j , k ( a)
= aggregate log(PHY rate)
over the time spent in cell-k by user u’s
application a, when it is initiated in cell j.
Cell k
T j , k ( a)
= aggregate time spent in cell k,
by users u’s application a initiated in cell-j
Cell j
For each application class, Base station k
conveys:
 E[R
k
j ,k
(a)]  ln(Wk ) E[T j ,k (a)]
User Utility and Algorithm
 Recall user utility in static case = w f(pj, Wj, w)
Static algo cannot be
applied
 In mobile case = E[ R j ,k (a)]  ln(Wk ) E[T j ,k (a)] /(application duration)
k
 Assumption: Total load at base station much larger than individual weight of user
applications
 Can now apply standard Maximized Generalized Assignment algorithm
 E.g.: Local Greedy Search with ½ - e factor approximation.
Difficult to quantify price of anarchy. In mobile case, scenario is more dynamic.
Similar to multiple agent learning. Difficult to prove strong guarantees.
Putting it together in practice
Implementation over Existing IEEE, IRTF and IETF proposals:
 Use IEEE 802.21 for signaling
 IRTF MPA framework for authentication and acquiring IP address and network
resources.
 Fast Handover in MIPv6 to simultaneously establish tunnel to gateway of new network
and forward packets.
Gathering network state information:
 Needs to be managed carefully depending on FDD versus TDD systems to minimize
overhead.
Evaluation
Network Topology:
 Cell tower location of a major operator in
Indian city (5Km X 5Km area)
 Clutter information along with RF tool used to
generate RF map
 We assume two operators share the same cell
tower locations.
 Each offers HSDPA and LTE
Application Models:
 3 classes, voice, video and data
 Generated according to guidelines for next
generation mobile networks
User Mobility:
 Manhattan and random waypoint
Performance Improvement as Fraction of Mobile Users is Varied
Area Spectral Efficiency improves by 2.5X-4X
Performance Gain over Optimized Single Operator
At least 60% gain over single operator with load balancing across technologies
User Utility
What’s in it for the Operators?
MOTA Model
Traditional Model
Operator incentive
Price
Simulations imply 20% incentive. Far more research required.
Reflections
 Are there alternative simpler
architectures possible that just
exploit roaming agreements between
operators?
 How can this be combined with ideas
of dynamic spectrum access? Do
operators really need to swap
spectrum at fine time scales?
 Is operator signaling really required?
Can end users learn appropriate
association over time?
 A phone app that makes these decisions
for you.
Questions?