Download Mobile Communications Networks

Contents
Mobile Communications Networks
ƒ Mobile Communications Networks and NGMN
ƒ Why Bio-Inspired Networking?
− Evolving through BiologicallyBiologically-Inspired Technologies −
ƒ Bio-Inspired Defined
ƒ Bio-Inspired Techniques in Communications
ƒ New Elements Incorporating NGMN
Abbas Jamalipour, PhD; Fellow IEEE, Fellow IEAust
ƒ Closing Remarks
IEEE Distinguished Lecturer
Email: [email protected]
October 2009
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Mobile Communication Networks
Communications Technologies
ƒ Mobile communications networks are getting more and
more complex with
ƒ Emergence of several access technologies has resulted
in a multitude of heterogeneous systems targeting
different service types, data rates, and users
2
ƒ 1G to 2G migration: A transition from analog to digital
ƒ Variety of services they offer
ƒ 2G to 3G evolution: Popularity of Internet and need for higher mobile data
rates
ƒ Variety of devices connected to the network
ƒ Complementing service from several access technologies:
ƒ Cellular: 2G (GSM and IS-95), 3G (UMTS and cdma2000)
ƒ Variety of environment and channel conditions they work in
ƒ High speed data networks: IEEE 802.11, HiperLAN
ƒ Variety of possible interconnections they have to make
ƒ WiMAX, Mobile WiMAX, MobileFi, …
ƒ and more recently, variety of network topologies they can use
ƒ Digital broadcasting systems: DAB, DVB, DMB
ƒ The missing bit
ƒ A single architecture to integrate all these and future systems, enabling
users to have global reliable connectivity
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
3
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
4
Main Motivations for NGMN
Universal Ubiquitous Coverage
ƒ Demand for better availability of services and applications
ƒ Global connectivity for any-type services at anytime, anywhere and
anyhow
ƒ Rapid increase in the number of wireless subscribers who want to
make use of the same handheld terminal while roaming
ƒ Support for bandwidth intensive applications such as real-time
multimedia, online games and videoconferencing as well as
traditional voice service (e.g., through VoIP)
ƒ Universal ubiquitous coverage across different radio
technologies is the ultimate objective of the future mobile
networks
ƒ Answering the increasing demand for higher transmission rates
and flexible access to diverse services
ƒ Offering a rich range of services with variable bandwidth and
service quality
ƒ Satisfying users’ mobility and traffic service requirements
ƒ Covering different geographic areas and accessing to different
types of service
ƒ The scalable and distributed next generation mobile network
architecture is expected to offer any-type services over a diverse set
of indoor, outdoor, pedestrian, and vehicular
ƒ The universal ubiquitous coverage need to be realized
through
ƒ These services will be offered over a large range of overlapping
access networks that offer different data rates, coverage, bandwidth,
delay and loss, and other QoS requirements
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
ƒ Connectivity across multiple networks
ƒ Interoperability across different radio technologies
5
Next Generation Mobile Network
ƒ
ƒ
ƒ
ƒ
ƒ
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Next Generation Mobile Network
To offer an integrated system
To promote interoperability among networks
To offer global coverage and seamless mobility
To enable the use of a universal handheld terminal
To enhance service quality compared to current wired networks
NGMN will be an
integrated platform
interconnecting multiple
networks for seamless
user connectivity for
multimedia applications
anytime and anywhere
server farm,
gateways, proxies
broadcast
PSTN, CS core
gateways
MSC
IP-based
core
SGSN
BSC
firewall, GGSN,
gateway
router
Internet
NGMN to:
ƒ be an open system
ƒ be an access-independent to underlying transport technologies
ƒ be an access-independent with service oriented functionalities
ƒ include seamless mobility across networks
ƒ aim toward providing a guaranteed end-to-end service quality
Mobile Communications Networks: Evolving through biologically-inspired technologies
SS7 signalling
GSM
ITU Recommendations
A. Jamalipour, 2009
6
access
points
It is the ultimate
solution to the problem
of ubiquitous mobile
communications!
7
A. Jamalipour, 2009
private
WPAN
RNC
private
WLAN
UMTS
public
WLAN
Mobile Communications Networks: Evolving through biologically-inspired technologies
8
Heterogeneous Mobile Technologies
Networking Issues in NGMN
ƒ Internetworked routing
and address allocation
ƒ Resource management
ƒ Traffic and congestion
control
ƒ Mobility among many
networks
ƒ Network address
translation
ƒ Network protocol
translation
ƒ Communication networks differ based on their
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
Air interface and spectrum requirements
Offered services
Data rates and QoS requirements
Modulation and coding scheme
Core network functionalities
Signaling requirements between terminal and network
ƒ Service across other networks is not guaranteed
ƒ Lack of interoperability
ƒ Lack of service agreement
ƒ Users require
I-CSCF
9
Complex NGMN Design
Third party applications
and value added services
Security
Network
Access network interfaces,
connecting reconfigurable
SDR-based end terminals
A. Jamalipour, 2009
APIs
Resource Mgmt.
Cross-layer
coordination
Mobility Mgmt.
&
OAM&P
SGSN
Visitor UMTS CN
Visitor UMTS CN
WiMAX
WiMAX
ƒ
ƒ
ƒ
ƒ
ƒ
Convergence sublayer
WLAN
Mobile Communications Networks: Evolving through biologically-inspired technologies
10
Emerging
networks
Physical
Mobile Communications Networks: Evolving through biologically-inspired technologies
Architecture supporting heterogeneous networks
Adaptive protocol stack
Enhanced mobility management addressing different handoffs
Enhanced RRM techniques to admit horizontal and vertical handoffs
Open Design Issues
QoS Mgmt.
3G
A. Jamalipour, 2009
ƒ
ƒ
ƒ
ƒ
Application
2G
WiMAX
Core
WiMAX
ASN G/W
MIP-FA
NGMN System Requirements
Services
Access independent network
functionalities in a transparent
manner hiding access specific
signaling requirements
P-CSCF
P-CSCF
P-CSCF
GGSN
MIP-FA
Visitor UMTS
N/W
Data Flow
Signaling Flow via UMTS Interface
Mobile Communications Networks: Evolving through biologically-inspired technologies
Service control and mechanism
essential for the smooth
operation of the network
architecture
MIP HA
PDSN
MIP-FA
Visitor
CDMA
2000
N/W
Destination
N/W
S-CSCF
IMS
Home
N/W
ƒ Different handheld terminals
ƒ Separate subscriptions
A. Jamalipour, 2009
CN
HSS
11
Adaptive NGMN Architecture
Cross layer coordination
Vertical handoff
Call admission control
Mobile terminal
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
12
Broadband Convergence Network
NGMN’s Inter-Domain Service
ƒ Broadband Convergence Network
(BcN) is the way for the NGMN
ƒ Service transition from fixed domain into mobile domain and vice
versa seamlessly
ƒ No sensible change of service quality received by a user while
moving from a fixed domain into a mobile domain
ƒ Integration of heterogeneous fixed
and mobile networks with varying
transmission characteristics
regional
ƒ Fixed-to-mobile or mobile-to-fix domain
ƒ From one mobile network to another mobile network
ƒ Service independency to the radio access technology
metropolitan area
Vertical
Handover
campus-based
ƒ No dramatic change in QoS particularly in data rate (i.e., the most humanly
sensible quality measure)
ƒ Logically followed by the delay requirement
ƒ Such service availability will need modifications at all layers of the
network protocol stack
ƒ A system of authentication and authorization that supports access
across different network is also needed
Horizontal
Handover
in-car,
in-house,
personal area
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
13
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
14
IP Architecture
IP Architecture – Conventional
ƒ If we agree on the assumption that IP will be the core part
of the next generation mobile networks, then the
traditional protocol architecture seems to be inadequate
ƒ Link layer (e.g. Ethernet): Providing connectivity to other network
segments; i.e. not to hosts in different networks
ƒ Network layer (e.g. IP): Delivering datagram packets across multiple
networks
ƒ Transport layer
ƒ A modular architecture designed based on stack of protocols
ƒ Using services provided by the lower module
ƒ Providing new services to the upper layers
ƒ TCP: Providing connection-oriented communication services, making
communication reliable, avoiding network congestion
ƒ UDP: Providing simple and unreliable transport for quicker
communications (required for real-time applications)
ƒ Communications mainly between adjacent layers
Application
ƒ Where to put the main elements necessary for NGMN?
RTP
Transport (TCP)
ƒ QoS: So that IP network could be used for voice, video, and other
multimedia real-time services
ƒ Mobility: Among APs of the same technology (micro-mobility) or across
networks of different technologies (macro-mobility)
Transport (UDP)
Network (IPv4/IPv6)
LLC and MAC
Physical Access
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
15
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
16
NGMN Protocols Re-Design Need
IP Architecture – Modified
ƒ Physical layer
ƒ Adding mobility features at network layer
ƒ Increasing address space and easing routing at IP layer
ƒ Modifying TCP for better performance at high error rate
environments (e.g. over wireless channel)
ƒ Additional sub-layers for adaptive selection of a network and
associated MAC protocol
ƒ Multiple physical network interfaces (cellular, W-LAN, WiMax, …)
ƒ Link layer
ƒ Establishment of concurrent connections via different access networks
ƒ Packet scheduling and optimum network selection mechanisms
ƒ Network layer
ƒ Accommodating mobility in IP protocol
ƒ Faster and easier routing techniques with less signaling
ƒ IP global (and heterogeneous) address translations
Application
RTP
Transport (UDP)
ƒ Transport layer
Network (MIPv6)
ƒ More wireless friendly transport protocols (than TCP and UDP)
Adaptive MAC and Scheduler/Selector
ƒ Application layer
ƒ Management of optimum compression and data rate control
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Modified TCP
17
A. Jamalipour, 2009
W-LAN
W-CDMA
TDMA
N-CDMA
IEEE
802.11
UMTS
GSM/
GPRS
cdmaOne
WiMax
IEEE
802.16
Mobile Communications Networks: Evolving through biologically-inspired technologies
Protocol Stack Enhancement
AdaptNet
ƒ Modifications of individual layer protocols so that the
overall architecture can handle the heterogeneity
ƒ Making link, transport, and application layers adaptive
(not changing the network layer) at mobile host
ƒ Also inclusion of some cross-layer interactions
ƒ Application layer
ƒ Example: The AdaptNet
18
ƒ Handling data and bit error rate fluctuations of the wireless
channel by means of adaptive source and channel coding
ƒ Transport layer
ƒ Modification of overall protocol stack, removing the
modularity character from it and allowing interaction of
protocol layers with layers other than the adjacent one
ƒ Use of an adaptive mobile-host-centric transport protocol called
Radial Reception Control Protocol
ƒ Link layer
ƒ Use of an adaptive MAC for seamless medium access control
over heterogeneous networks
ƒ Use of an adaptive error correction scheme which changes the
coding rate in accordance with the channel condition
ƒ Example: The Cross-Layer architecture design
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
19
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
20
Cross-Layer Architecture Design
Coordination Planes
ƒ Concept: By leaving protocol stack strictly modular, it
will be inefficient with respect to performance, QoS,
and energy consumption, etc.
ƒ Solution: Proving information from non-adjacent
layers in a cross-layer structure
ƒ Four separate vertical planes that coordinate the
information exchange and actions to be done by
individual layer protocols
ƒ QoS: For distribution of QoS requirements and constraints
and coordination of efforts by layers to achieve QoS
ƒ Security: For elimination of encryption duplication at several
layers
ƒ Mobility: Enhancing interactions among TCP, IP, and link
layers in handling mobility in different environments
ƒ Wireless Link Adaptation: Providing adaptive bit error rate
and data rate depending on different wireless channel
conditions and different mobile environment
Wir
ele
Ada ss Link
ptat
ion
y
rity
Mob
ilit
Network
Sec
u
Transport
QoS
Application
Link
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
21
Cross-Layer Coordination
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Cross-Layer Coordination Methods
ƒ Cross-layer coordination between different entities within
the architecture would be necessary in NGMN
ƒ For wireless system discovery to provide a list of access networks and
their associated QoS parameters
Interlayer Signaling Pipe
(ISP) Approach
Cross-layer information (TCP/RLP
related) are stored in the wireless
extension header (WEH)
ƒ To support QoS enabled application, direct communication between
application layer and QoS sub-layer are essential
Application
Interlayer Pipe
ƒ To provide services in a visited network based on service policy and
subscriber profiles signaling between mobility management sub-layer
and services sub-layer as well as between services sub-layer with
resource management and QoS management sub-layer are essential
ƒ When service is no more possible after a vertical handover
ƒ Also for accounting purpose using information related to the resources
used, QoS provided, time and duration of provision of network
resources, etc
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
22
TCP
IPv6 (WEH)
RLC
WMAC
WPHY
• Interlayer Pipe passes through all layers
• Suffers from longer processing delay
23
A. Jamalipour, 2009
Direct Connectivity
Approach
- Direct connectivity among non-adjacent layers
- Separate definition of APIs
Application
Transport
Network
Link
PHY
• How controllable QoS parameters form individual
layers translate into parametric quantities?
• How to optimize the decision process?
Mobile Communications Networks: Evolving through biologically-inspired technologies
24
Cross-Layer Coordination Methods
ICMP Approach
Cross-Layer Coordination Views
Ad Hoc Network Approach
ƒ Cross-layer coordination not yet standardized
Use of
of watermark
watermark based
based mechanisms
mechanisms
Use
to initiate
initiate ICMP
ICMP messages
messages whenever
whenever
to
network conditions
conditions change
change
network
Information exchange
exchange through
through external
external servers
servers
Information
Application
report
ƒ Will be key to offering enhanced services
Application
Socket layer
Table lookup
Adaptive application
server
ƒ Mainly confined to the network level functionalities
TCP
Protocol specific
action
Transport
Network
ICMP messages
ƒ Room for contributions
Network
Link
WCI server
PHY
Wireless
Access
network
ƒ Therefore concentric to the network layer in the
proposed system model
Link
• May not be suitable for time-critical applications
PHY
• Suffers from longer processing delay
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
25
Broadband Wireless Internet
26
ƒ Cellular
ƒ In line with current infrastructure and coverage
ƒ Ad hoc networks
ƒ Distribution of responsibilities of network elements
ƒ To add coverage, capacity, and new services for example
through vehicular communications (VANET)
ƒ The path toward Broadband Wireless IP therefore
crosses multiple networks with heterogeneous
characteristics consisting of several technologies with
multiple configurations consisting of
ƒ Wireless Mesh networks
ƒ In appose to the existing star topology of cellular networks
ƒ The main limitations of a wireless network is high level of
transmission power and multipath transmission
ƒ Wireless mesh network can remove those limits through
ƒ Cellular based networks (centralized)
ƒ Ad hoc networks (decentralized)
ƒ Mesh networks (mixed centralized-decentralized)
Mobile Communications Networks: Evolving through biologically-inspired technologies
Mobile Communications Networks: Evolving through biologically-inspired technologies
Different Topologies
ƒ Implementation of a true Broadband Wireless IP requires
an efficient integration of heterogeneous BcN elements
A. Jamalipour, 2009
A. Jamalipour, 2009
ƒ Covering short range, so low power transmission
ƒ No ugly towers
ƒ Mostly LoS, so no multipath transmission problem
27
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
28
Challenges
Why Seek Inspiration from Biology?
ƒ Too much of complexity in the heterogeneous network
ƒ Resource management of multiple interconnected
networks and their topology creation
ƒ Traffic management and load balancing/sharing among
heterogeneous mobile and fixed networks
ƒ Network optimization, organization of efficient
interconnection, and incorporation among multiple
networks
ƒ Living organisms are complex adaptive systems
ƒ Artificial systems are going in that direction too
ƒ Look for new solutions to difficult problems
ƒ Life has many self-* features which are also desirable in
artificial systems:
ƒ
ƒ
ƒ
ƒ
ƒ
Biological systems may give some hints toward
dealing with these challenges …
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
29
Self-organization
Self-adaptation
Self-healing ability
Self-optimization
Self-robustness
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
NGMN and Biological Systems
The Term “Bio-Inspired”
ƒ Heterogeneous networking technologies, devices, and
services keep changing in NGMN over the time
ƒ Bio-inspired demonstrates the strong relation between
30
ƒ A particular system or algorithm
ƒ which has been proposed to solve a specific problem
ƒ A centralized control/management tends to be an infeasible
approach for NGMN
ƒ And a biological system
ƒ which follows a similar procedure or has similar capabilities
ƒ The dynamic NGMN architecture and protocols need to
have
ƒ Bio-inspired computing represents a class of algorithms focusing on
efficient computing
ƒ scalability, self-organization, self-adaptation, and sustainability
ƒ e.g. in optimization processes and pattern recognition
ƒ By appropriately mapping key biological principles to
NGMN architectures it is possible to create a scalable,
self-organizable, self-adaptable, and sustainable network
ƒ Bio-inspired systems rely on system architectures for massively
distributed and collaborative systems
ƒ Such network is motivated by the inspirations from various
biological systems’ abilities to naturally adapt to the changing
environment
ƒ Bio-inspired networking is a class of strategies for efficient and
scalable networking under uncertain conditions
ƒ e.g. in distributed sensing and exploration
ƒ e.g. in delay tolerant networking
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
31
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
32
Biologically Inspired Problem Solving
Why Bio-Inspired Networking?
ƒ Structural view: communication
is an intrinsic part of an
organization
ƒ Typical problems that can be tackled with bio-inspired
solutions are characterized by the:
ƒ Absence of a complete mathematical model
ƒ Example organizations:
ƒ Brain (organization of neurons)
ƒ Animal “super organisms”
(ant/bee colonies)
ƒ Human society
ƒ Large number of (inter-dependent) variables
ƒ Non-linearity
ƒ Combinatorial or extremely vast solution space
ƒ Those natural and living
organizations seem better
organized than the current
Internet!
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
33
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Design of Bio-Inspired Solutions
Example of Bio-Inspired Research Fields
ƒ Identification of analogies
ƒ Evolutionary Algorithms (EAs)
ƒ In swarm or molecular biology and ICT systems
ƒ Artificial Neural Networks (ANNs)
ƒ Understanding
ƒ Swarm Intelligence (SI)
ƒ Computer modeling of realistic biological behavior
ƒ Engineering
ƒ Artificial Immune System (AIS)
ƒ Model simplification and tuning for ICT applications
Identification of
analogies
between biology
and ICT
A. Jamalipour, 2009
34
Understanding
Modeling of
realistic biological
behavior
Engineering
ƒ Cellular Signaling Pathways
Model simplification
and tuning for ICT
applications
Mobile Communications Networks: Evolving through biologically-inspired technologies
35
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
36
Evolutionary Algorithms (EAs)
Evolutionary Algorithms (EAs)
ƒ Mainly rooted on the Darwinian theory of evolution
ƒ An EA uses some mechanisms inspired by biological
evolution
ƒ EAs can be categorized into the following Classes
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ Reproduction, Mutation, Recombination, Selection
ƒ EAs represent a set of search techniques used in
computing to find the solutions to optimization problems
Genetic Algorithms (GAs)
Evolution strategies
Evolutionary programming
Genetic programming
Classifier systems
ƒ Examples
ƒ Simulated annealing
ƒ Generic probabilistic meta-algorithm for the global optimization
problem
ƒ Simulated hill-climbing
ƒ A mathematical optimization technique which belongs to the family
of local search
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
37
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Artificial Neural Networks (ANNs)
Artificial Neural Networks (ANNs)
ƒ A Neural Network is a network of biological neurons
ƒ A neuron k that connects n inputs can be described as:
38
ƒ ANNs are non-linear statistical data modelling tools
ƒ Used to acquire knowledge from the environment
(known as self-learning property)
Neuron's bias (b)
Input: x1
Input: x2
ƒ The weights of the neurons are determined in a learning
process
w2
…
ƒ They can be used to model complex relationships
between inputs and outputs or to find patterns in data
Input: xn
Activation/Squashing
function
Wt(w1)
Σ
u
f(u)
Output: yk
wn Summing
junction
⎛ n
⎞
yk = f (uk ) = f ⎜⎜ ∑ wkj x j + bk ⎟⎟
⎝ j =1
⎠
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
39
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
40
Swarm Intelligence (SI)
SI – Example of Ant Colonies
ƒ An Artificial Intelligence (AI) technique based on the
observations of the collective behavior in decentralized
and self-organized systems
ƒ Ants solve complex tasks by simple local means
ƒ Ant productivity is better than the sum of their single
activities
ƒ Ants are “grand masters” in search and exploration
ƒ Typically made up of a population of simple agents
interacting locally with one another and with their
environment (no centralized control)
ƒ Local interactions between autonomously acting agents
often lead to the emergence of global behavior
ƒ Examples: Ant/bee/termite colonies, bird flocking, animal
herding, bacteria growth, and fish schooling
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
41
SI – Example of Ant Colonies
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
42
SI Properties
ƒ Properties of Swarm Intelligence
ƒ Agents are assumed to be simple
ƒ Indirect agent communication
ƒ Global behavior may be emergent
ƒ Specific local programming not necessary
ƒ Behaviors are robust
ƒ Required in unpredictable environments
ƒ Individuals are not important
“The emergent collective
intelligence of groups of simple
agents.” (Bonabeau)
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
43
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
44
What is Stigmergy?
SI Example: Collective Foraging by Ants
ƒ Stigmergy is a mechanism of spontaneous, indirect
coordination between agents or actions, where the trace
left in the environment by an action stimulates the
performance of a subsequent action, by the same or a
different agent
ƒ Starting from the nest, a random search for the food is
performed by foraging ants
ƒ Pheromone trails are used to identify the path for
returning to the nest
ƒ The significant pheromone concentration produced by
returning ants marks the shorted path
ƒ Produces complex, apparently intelligent structures, without
need for any planning, control, or even communication between
the agents
ƒ supports efficient collaboration between extremely simple
agents, who lack any memory, intelligence or even awareness of
each other
Nest
Food
ƒ Stigmergy is a form of self-organization first observed in
social insects
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
45
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Principles of Swarm Intelligence
Principles of Swarm Intelligence
ƒ What makes a Swarm Intelligent system work?
ƒ Randomness allows new solutions to arise and directs
current ones
• Positive Feedback
• Negative Feedback
• Randomness
• Multiple Interactions
ƒ Ant decisions are random
ƒ Exploration probability
ƒ Positive Feedback reinforces good solutions
ƒ Food sources are found randomly
ƒ Ants are able to attract more help when a food source is found
ƒ More ants on a trail increases pheromone and attracts even
more ants
ƒ Multiple Interactions: No individual can solve a given
problem. Only through the interaction of many can a
solution be found
ƒ Negative Feedback removes bad or old solutions from
the collective memory
ƒ One ant cannot forage for food; pheromone would decay too fast
ƒ Many ants are needed to sustain the pheromone trail
ƒ More food can be found faster
ƒ Pheromone decay
ƒ Distant food sources are exploited last
ƒ Pheromone has less time to decay on closer solutions
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
46
47
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
48
SI Routing Overview
SI Application in MANET Routing
ƒ Ant-Based Control
ƒ Routing in MANETs is an extension of Ant Foraging!
ƒ Ant Based Control (ABC) is introduced to route calls on a circuitswitched telephone network. ABC is the first SI routing algorithm
for telecommunications networks
ƒ Ants looking for food…
ƒ Packets looking for destinations…
ƒ AntNet
ƒ AntNet is introduced to route information in a packet switched
network
ƒ AntNet is related to the Ant Colony Optimization (ACO) algorithm
for solving Traveling Salesman type problems
ƒ AntHocNet
ƒ Can routing be solved with SI?
ƒ Can routing be an emergent behavior from the
interaction of packets?
ƒ A MANET routing algorithm based on AntNet which follows a
reactive routing approach
ƒ Termite
ƒ Also a MANET routing algorithm
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
49
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Ant-based Task Allocation
Bee Colony Optimization (BCO)
ƒ Combined task allocation according to ACO paradigm
has been investigated for MANETs
ƒ The BCO algorithm is inspired by the behavior of a
honey bee colony in nectar collection
ƒ This biologically inspired approach is currently being employed
to solve continuous optimization problems
ƒ The proposed architecture for MANETs is completely
dependant on probabilistic decisions
ƒ training neural networks, job shop scheduling, server optimization
BCO provides a population-based search
procedure in which individuals called
foods positions are modified by the
artificial bees with time and the bee’s aim
is to discover the places of food sources
with high nectar amount and finally the
one with the highest nectar
ƒ During the lifetime of the MANETs, nodes adapt the
probability to execute one task out of a given set
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
50
51
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
52
Bee Colony Optimization (BCO)
Artificial Immune System (AIS)
ƒ Artificial bees fly around in a multidimensional search space and
some (employed and onlooker bees) choose food sources
depending on their experience of and their nest mates, and adjust
their positions
ƒ Some (scouts) fly and choose the food sources randomly without
using experience
ƒ If the nectar amount of a new source is higher than that of the
previous one in their memory, they memorize the new position and
forget the previous one
ƒ Thus, ABC system combines local search methods, carried out by
employed and onlooker bees, with global search methods, managed
by onlookers and scouts, attempting to balance exploration and
exploitation process
ƒ Artificial immune systems are computational systems
inspired by theoretical immunology and observed
immune functions, principles and models, which are
applied to complex problem domains
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
53
ƒ The primary goal of an AIS is to efficiently detect
changes in the environment from the normal system
behavior in complex problem domains
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
Why the Immune System?
AIS – Application Examples
ƒ Recognition
ƒ Fault and anomaly detection
ƒ Ability to recognize pattern that are different from previously
known or trained samples, i.e. capability of anomaly detection
54
ƒ Data mining (machine learning, pattern recognition)
ƒ Robustness
ƒ Agent based systems
ƒ Tolerance against interference and noise
ƒ Diversity
ƒ Autonomous control and robotics
ƒ Applicability in various domains
ƒ Scheduling and other optimization problems
ƒ Reinforcement learning
ƒ Inherent self-learning capability that is accelerated if needed
through reinforcement techniques
ƒ Security of information systems
ƒ Memory
ƒ Misbehavior detection for MANETs based on the DSR
protocol
ƒ System-inherent memorization of trained pattern
ƒ Distributed
ƒ Autonomous and distributed processing
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
55
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
56
Human Immune System
Danger Theory and NGMN
Mapping HIS to NGMN
Danger Zone (DZ)
The immune system is
ƒ distributed in several parts of the
body
ƒ Governed
Mapping
Dangerby Lymphotic laws
ƒ Distressed
Theory
to NGMN cell sends initiation (alarm)
signal (IS) to its neighborhood
ƒ The IS is captured by antigen presenting
cell (APC) – starts the Danger Theory
ƒ Danger Zone (DZ) creation – handle the
invaders locally
ƒ Only cells within the DZ get stimulated
ƒ fault-tolerance (autonomous)
ƒ enabling survivability
ƒ using selective response
ƒ having memory
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
57
Survivability Framework in NGMN
Co-stimulation
Process
A. Jamalipour, 2009
ƒ Basis of all biological systems
ƒ Specificity of information transfer
ƒ Similar structures in biology and
in technology
Æ Especially in computer networking
Security Control
Framework
Attack
Localization
Intra-domain
Isolation
58
ƒ Properties
Attack Detection
Framework
Recognition
Process
Mobile Communications Networks: Evolving through biologically-inspired technologies
Molecular and Cell Biology
Survivability
Framework
Initiation
Process
A. Jamalipour, 2009
Attack
Recovery
ƒ Lessons to learn from biology
ƒ Efficient response to a request
ƒ Shortening of information pathways
ƒ Directing of messages to an applicable destination
Inter-domain
Isolation
Mobile Communications Networks: Evolving through biologically-inspired technologies
59
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
60
Info Exchange in Cellular Environments
Intracellular Signaling Pathways
Reception of signaling molecules (ligands such as hormones, ions, small molecules)
ƒ Signaling in biological systems occurs at multiple levels
and in many shapes
(1-a)
ƒ Signaling describes interactions between individual molecules
Different
cellular answer
ƒ The information processing capabilities of a single cell
ƒ Received information particles initiate complex signaling cascades
that finally lead to the cellular response
Nucleus
DNA
mRNA translation
into proteins
Mobile Communications Networks: Evolving through biologically-inspired technologies
ƒ
After processing
the information, a
specific cellular
answer is initiated
ƒ
The effect could be
the submission of a
molecule
DNA
ƒ Intercellular signaling
A. Jamalipour, 2009
Reorganization of
intracellular
structure
Nucleus
Gene
transcription
Secretion of
hormones
etc.
(3-a)
ƒ Communication among multiple cells is performed by intercellular
signaling pathways
ƒ Objective is to reach appropriate destinations and to induce a
specific effect at this place
ƒ
Communication with other
cells via cell junctions
(2)
ƒ Intracellular signaling
Transfer via
receptors on cell
surface
(3-b)
Intracellular
signaling
molecules
ƒ Main cellular signaling techniques
ƒ
Neighboring cell
(1-b)
Submission of signaling molecules
61
Intracellular Information Exchange
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
62
Intercellular Information Exchange
ƒ Local: A signal reaches only cells in the neighborhood.
The signal induces a signaling cascade in each target
cell resulting in a very specific answer which vice versa
affects neighboring cells
Tissue 1
DNA
DNA
Tissue 2
DNA
Signal
(information)
DNA
Receptor
DNA
A. Jamalipour, 2009
Remote: A signal is released in the
blood stream, a medium which carries it
to distant cells and induces an answer
in these cells which then passes on the
information or can activate helper cells
(e.g. the immune system)
Gene transcription
results in the
formation of a
specific cellular
response to the
signal
Mobile Communications Networks: Evolving through biologically-inspired technologies
63
A. Jamalipour, 2009
od
Blo
DNA
Tissue 3
DNA
Mobile Communications Networks: Evolving through biologically-inspired technologies
64
Lessons to be Learnt
Adaptation to Networking
ƒ The adaptation of mechanisms known from molecular
and cell biology promises to enable more efficient
information exchange
ƒ New concepts for behavior patterns of network nodes
ƒ Local mechanisms
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ Improved efficiency and reliability of the entire communication
system
ƒ Flexible self-organizing infrastructures
ƒ Remote mechanisms
ƒ Localization of significant relays,
helpers, or cooperation partners
ƒ Semantics of transmitted
messages
ƒ Cooperation across domains
ƒ Internetworking of different
technologies
ƒ Authentication and authorization
Adaptive group formation
Optimized task allocation
Efficient group communication
Data aggregation and filtering
Reliability and redundancy
ƒ Main concepts to be exploited in the context of
communication networks
ƒ Signaling pathways based on specific signal cascades with
stimulating and inhibitory functionality used for intracellular
communication
ƒ Diffuse (probabilistic) communication with specific encoding of
the destination receptors for intercellular communication
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
65
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
New elements incorporating the NGMN
Hierarchical Wireless Sensor Networks
Some examples include:
ƒ Sensor clustering for efficient routing
66
ƒ Layered topology design for better
data aggregation
ƒ Secure sensor networking
ƒ Wireless sensor networks
ƒ Monitoring elements for NGMN efficient operation
CH3
ƒ Mobile and vehicular ad hoc networks
Sink
CH 1
CH4
CH5
CH2
ƒ Increasing the coverage and capacity of the NGMN
ds1
FL2
Destination
n1
d1
ƒ Wireless mesh networks
n5
n1
n4
Node
n1
source
Cluster
n3
n3
ƒ Increasing reliability and providing an alternative backbone
n2
FL1
n2
Source
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
67
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
68
Vehicular Ad Hoc Networks
Multi-layered Wireless Mesh Networks
North
ITS and
infotainment
applications of
V2V and V2I
ƒ Developing a
new backbone
network for the
Internet
Group 2
Group 3
Vehicle A
S2
S3
S4
S1
Group 1
Internet
IEEE 802.20 Tower
IEEE 802.20 Tower
Router
N
Group 4
A
Source
A. Jamalipour, 2009
ƒ Applications:
D
B
F
ion
Destinat
C
Mobile Communications Networks: Evolving through biologically-inspired technologies
Router
IEEE 802.20 Tower
69
Closing Remarks
ƒ Emergency
ƒ Fault tolerance
ƒ Increased
throughput
ƒ Reliability
A. Jamalipour, 2009
Router
Router
Router
Router
IEEE 802.20 Tower
Router
Router
Fixed Links
802.20 Cover
area
Wi-Fi Cell
Mobile Communications Networks: Evolving through biologically-inspired technologies
70
Mobile Networks & Topologies
Networks
ƒ Advanced cellular networks
ƒ GSM, GPRS, UMTS, cdmaOne, cdma2000, HSDPA, LTE, …
ƒ Wireless LAN IEEE 802.11 family
ƒ Wireless MAN IEEE 802.16 family
ƒ Other emerging technologies
Topologies
ƒ Cellular based systems (centralized)
ƒ Ad hoc networks fixed vehicular nodes (decentralized)
ƒ Mesh networks (mixing centralized and decentralized)
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
71
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
72
NGMN Developments
Bio-Inspiration Role
ƒ Development trend of NGMN and Wireless IP has been
separated into two distinct ways:
ƒ Very little number of studies on biologically inspired
network models exist in the literature
ƒ Cellular based – moving from CS to PS and all IP-based
ƒ IP-oriented standards oriented around IEEE 802.1x and 802.2x
ƒ Available models mainly imitate some biological coordination
aspects
ƒ No matter how these rather exclusive directions develop, the
future of mobile data will hang around a heterogeneous
solution that will include both approaches
ƒ Providing QoS and security in NGMN and Wireless IP will be
the task of all layers of the network protocol stack, with
particular attention at the higher layers in order to be aligned
with the heterogeneous nature of the future networks
ƒ Bandwidth and resource management of large number of
network users will eventually push W-LAN and W-MAN
standards into licensed spectrum
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
ƒ As for the nature, however, they could have great
potential to assist with better and more efficient network
management in mobile communications networks,
particularly for the future dynamic non-centralized
heterogeneous NGMN environment
ƒ To provide scalability, self-organization, self adaptation,
sustainability, and added network security
73
Mobile Communications Networks
Evolving through biologicallybiologically-inspired technologies
Thank You
AJ
October 2009
A. Jamalipour, 2009
Abbas Jamalipour
[email protected]
Mobile Communications Networks: Evolving through biologically-inspired technologies
75
A. Jamalipour, 2009
Mobile Communications Networks: Evolving through biologically-inspired technologies
74