Scarica Network Infrastructures Theory e più Appunti in PDF di Sicurezza delle reti solo su Docsity!
1. Single cable - > loss of connection
2. Amplifier to refresh the signal - > cable connection between the amplifiers
Solution: provider or carrier
○ It a network that can be shared and that it is not direct between the hosts
○ Provides connectivity on the demand of the two routers
○ Ask to a carrier to provide/ create a connection between different clients
- Transport Networks (TNs) are public infrastructures operated by service providers named carriers
Carriers provide a variety of services:
○ telephone and leased line services
○ interconnect Internet Service Providers
○ provide bulk bandwidth to other carriers
Introduction
lunedì 25 settembre 2023 10:
The network can be broken up into:
metro network : is the part of the network that lies within a large city or a region
▪ metro access network: extends from a central office out to individual businesses or homes
▪ interoffice network: connects groups of central offices within a city or region
○ long-haul: network interconnects cities or different regions
- Different parts of the network may be owned and operated by different carriers
- The nodes in the network are central offices, sometimes also called points of presence (POPs)
Links between the nodes consist of fiber pairs and, in many cases, multiple fiber pairs - > links in the long-haul network
tend to be very expensive to construct
Two topologies are used
Ring:
Sharing resources
▪ Failure resistent for providing advanced services
2. mesh
Service can be:
Connection oriented
○ sender and receiver connect each other before communication happen
An example is a phone call which
1. Request a connection
2. Create a connection
3. Send information
○ Waste time to communicate
○ Avoid overflow of connections
Guarantee the connection with some requestes in terms of bandwidth, packetloss, exc…, so it says if it can
provide this type of connection or not
Connectionless
○ source sends messages to the receiver whenever it has something to send
○ An example is the IP Protocol
○ Has no guarantees
Core devices work according to one of the following switching paradigm:
circuit switching
○ static multiplexing to divide the shared resource
It is composed by several operations:
1. Divide the channel in sub-channels by TDM or FDM
2. Assign a sub-channel to a specific communication in every link of the communication
3. Instruct each device on how to do the routing part/ switch the channels and the interfaces
4. If the network has not the resources, it refuses the connection - > circuit setup
5. Take out the communication
○ It is good when the communication process is long enough to waste time in creating the circuit
- Optical Networks (ONs) can deliver bandwidth in a flexible manner where and when needed
Optical fiber
○ offers much higher bandwidth than copper cables
○ is less susceptible to various kinds of electromagnetic interference and other undesirable effects
Two generations of optical networks
First generation
▪ optics essentially used for transmission and simply to provide capacity
▪ switching and other intelligent network functions handled by electronics
- Second generation: routing, switching, and intelligence in the optical layer
Two ways of increasing the transmission capacity on a fiber using multiplexing techniques:
Time Division Multiplexing (TDM)
▪ increase the bit rate (requires higher-speed electronics)
▪ many lower-speed data streams are multiplexed into a higher-speed stream
Wavelength Division Multiplexing (WDM)
▪ transmit data simultaneously at multiple carrier wavelengths over a fiber
▪ virtual fibers
It's necessary to assign the use of a timeslot to every client to share
the communication
- Always requires the change from optical to electric domain
- We can always distinguish the signal divided between the two clients
Used for switching directly in the optical domain without changing to
electronic domain
Optical Networks
venerdì 29 settembre 2023 08:
- Also known as wavelength routed networks
- Main idea: incorporate some of the switching and routing functions into the optical part of the network
- The network provides lightpaths to its users
Lightpaths are optical connections
○ carried end to end from a source node to a destination node
○ over a wavelength on each intermediate link
- At intermediate nodes the lightpaths are switched from one link to another link
- Lightpaths may be converted from one wavelength to another wavelength
The sub-channel is a wavelenght: if the wavelenght is available than the network can create the circuit between the two
clients
The wavelenght doesn't have to be necessarily the same , since currently to assign another wavelenght, so to switch
between two sub-channels, it is necessary to recover the binary string (Wavelenght Conversion) in the electronic signal and
than transform it in the optical signal to assign another wavelenght
- The is unique on the channel , so two same can't exists on the same channel
Transparency property: the same infrastructure can be used to realize complete different networks also using complete
different technologies, for instance on the same infrastructure we can realize a voice service and an internet service
- OLTs are used at either end of a point-to-point link to multiplex and demultiplex wavelengths
- Used to pass from an optical domain to an electronic domain
It is composed by three functional elements:
○ Transponders
○ Wavelength multiplexers
○ Optical amplifiers
Transponder (optical-to-electrical-to-optical, O/E/O) adapts the signal coming in from a client of the optical network, and
vice versa
○ converts the signal into a wavelength that is suited for use inside the optical network (from 1.3 μm to 1.55 μm)
○ adds OTN overhead (OPU, ODU, OTU, FEC, etc.) - > add an header to the incoming frames
monitors the bit error rate of the signal at the ingress and egress points in the network, since it transforms the optical
signal to digital signal to pass from optical domain to electronic domain
- OLT also terminates an optical supervisory channel (OSC)
- Transponders typically constitute the bulk of the cost, footprint, and power consumption in an OLT
- Therefore, reducing the number of transponders helps minimize both the cost and the size of the equipment deployed
○ Protection: detect failures and rapidly reroute lightpaths around the failure
○ bit rate transparency: switch signals with arbitrary bit rates and frame formats
○ wavelength conversion: change the wavelength of an incoming signal before transmitting it
multiplexing and grooming
▪ multiplexing and grooming capabilities to switch traffic internally at much finer granularities
▪ this time division multiplexing has to be done in the electrical domain
- Network architectures can be organized by means of the ISO/OSI model
- A more realistic layered model for today’s networks would employ multiple protocol stacks residing one on top of the other
- Lightpaths are service transparent
once the lightpath is set up, it can accommodate different types of services - > for instance, the telephone network had this
property (a channel can be used to transfer voice, data, fax, etc.)
Advantages:
○ data is carried from its source to its destination in optical form
○ no optical-to-electrical conversions along the way
- Hard to realize: analog signals require higher SNR with respect to digital ones
- Optical networks almost always include a fair amount of electronics
- Electronics plays a crucial role in performing the intelligent control and management functions
Electronic is required
○ at the edge of the network: to adapt the signals entering the optical domain
i n the core of the network
▪ for regeneration and wavelength conversion
Since it deals with different data rates signals and different technologies, we need to regenerate the signal to the
original form - > so generally it is necessary to pass to the electronic domain
- Electronic regenerators reduce the transparency of the network
Three types of electronic regeneration techniques for digital data
1R: regeneration (can be seen as an Optical Amplifier)
▪ PRO: supports analog signals
▪ CONS: poor performance
2R: regeneration with reshaping
▪ PRO: offers transparency to bit rates
▪ CONS: limits the number of regeneration steps allowed due to the accumulated jitter
3 R: regeneration with retiming and reshaping (can be seen as the transponder)
▪ PRO: produces a “fresh” copy of the signal
CONS
□ it eliminates transparency to bit rates and the framing protocols
□ Refresh the signal only for one device and not for the other ones
a.
The goal of performance management is to enable service providers to provide guaranteed quality of service to the users of
their network
This usually requires:
○ monitoring of the performance parameters for all the connections
○ taking any actions necessary to ensure that the desired performance goals are met
Fault management involves:
detecting problems in the network
○ alerting the management systems appropriately through alarms
- Fault management also includes restoring service in the event of failures
The performance parameters that the fault management monitors are:
BER
▪ the bit error rate (BER) is the key performance attribute associated with a lightpath
The BER can be detected only when the signal is available in the electrical domain, typically at regenerator or
transponder locations
▪ The system has access to the BER in the OTU or ODU layer of the transponder, so where the domain is electrical
▪ Overhead inserted in OTN frames, which consists of parity check bytes, allows for BER computation
Optical Trace:
▪ Lightpaths pass through multiple nodes and through multiple cards within the equipment deployed at each node.
▪ It is desirable to have a unique identifier associated with each lightpath
▪ This identifier is called an optical path trace
▪ The trace enables the management system to identify, verify, and manage the connectivity of a lightpath
▪ This identifier is a binary string, so the problem now is how to decode the binary string in the optical layer
In a network, a single failure event may cause multiple alarms to be generated - > Es. in a network with 32 lightpaths on
a given link, each traversing through two intermediate nodes, the failure of a single link could trigger a total of 129 alarms
- Alarm management it is required to identify the root-cause alarm of the failure and suppress the redundant alarms
Alarm suppression is accomplished by using a set of special signals, called the forward defect indicator (FDI) and the
backward defect indicator (BDI)
The overhead is used to
○ detect the failures thanks to the BER in it
○ to send FDI/ BDI messages
- When a link fails, the node downstream of the failed link inserts an FDI signal downstream to the next node
- The FDI signal propagates rapidly, and nodes further downstream receive the FDI and suppress their alarms
- The node also sends a BDI signal upstream to the previous node, to notify that node of the failure
- FDI and BDI are sent at different sublayers of the optical layer
Performance and Fault Management
lunedì 2 ottobre 2023 10:
Supporting the optical path trace, defect indicators, and BER measurement requires the use of some sort of overhead
in the optical layer
- How can we add overhead in the optical layer?
Corresponds to different - > so different clients
Through modulations techniques it adds a pilot tone to add overhead
information
The pilot tone is dedicated to every specific user and it is added by the
transponder
- It is a overhead of the Och
- It is the optical supervisor of the channel
It's a special to monitor a specific secction
of the network added by a dedicated
transponder
Signal to monitor the specific link which is
different from the other link
A pilot tone assigns a trace to a single optical
channel (OCh)
Why the pilot tone can't be assign in the OTS layer?
Because different OTS have a different trace, so if we
assign it in the OTS the trace could not change onto
the different links
To notify another subnetwork we can rely on the
electronic layer
Monitoring the optical power means monitoring the pilot tone to measure attenuation
Connection management deals with setting up connections, keeping track of them,
and taking them down when they are not needed anymore
Two different approaches
client-server model or overlay model
the client layer (IP, SDH, ATM, etc.) asks to the server layer the
establishment of a connection (lightpath) without having knowledge of
the internal structure of the optical network
▪ centralized control
▪ suitable as long as lightpaths are set up fairly infrequently
peer model
tight coupling between the client and optical layers: the optical layer
primarily serves a single client (IP)
▪ distributed control
▪ useful if there is a need to set up and take down connections rapidly
The routers participate together to create the lightpath - > so they are
the owners of the optical infrastructure
Cooperation between control plane of the clients and control plane of
the optical network
The needs of a TLC provider of data services
Offer services to its customers
▪ good “Quality of Service” ( QoS )
▪ Service Level Agreements ( SLAs )
Run its backbone in a cost-effective way: convergence of services (example voice and data services) ○
For its Data backbone the provider needs good support for:
Virtual Private networks
▪ Traffic Engineering
▪ Protection/restoration mechanism
Connection-oriented packet switching technologies represent a good answer to these needs ○
- MPLS over Ethernet
- The role of the carrier is to provide connectivity to the clients through the ISP with an IP Address and though an IP Network
- Sometimes Dijsktra for SPT is not the best option
- Assume that there are very different clients on the net
This two packets are not different since they
have the same destination address
We want a technology on the Net that supports
the traffic engineering that manage the traffic in
the best way possible
There is a lot of congestion on the paths since
this is the shortest path for all users
In VPN each host has a private address so the packet can't reach the
destination using only a private address
What the VPN does is encapsulate the packet with the public address to
reach the destination
- Suppose that this link fails
- We have to understand how the packets can change the path
Multi Protocol Label Switching (MPLS)
lunedì 9 ottobre 2023 10:
- The two tunnels are built up running two times the connection-oriented procedure
If the network is pure IP the packets can't be distinguished since the destination machine are connected to the same IP network - > so
they are indistinguishable
But we want to distinguish them and the IN router has to classify them to distinguish them to permit the two packets to follow different
paths
The classification aims to distinguish two different application flows formed by 5 fields in header, called 5 - Tuple Classification:
○ SRC IP
○ DST IP
Protocol ○
○ SRC PORT
DST PORT
- This fields are inspected for every incoming packet to make them follow different paths
The Ingress LER of a MPLS domain analyzes the IP header of the packet, classifies the packet, adds the MPLS label and forwards
to the next hop LSR
- In the MPLS domain the packet is forwarded along the LSP according to the Labels
- The Egress LER removes the Label and the packet is forwarded based on the IP destination address
Three basic actions:
- PUSH operation: add the MPLS label on the label stack
- POP operation: remove the MPLS label to deliver to destination a regular IP packet
- SWAP operation: change the label (in an intermediate node)
- path message to notify that a router wants to create a LSP and travel to the network over the decided path
Resv message to notify which label it expects to receive
a. F does 33, POP since F is the engress router
E receives the message and chooses a label 45 | F, 33 - > the label by which the router wants to receive the packet has to not
be assigned already in the net
b.
c. D receives the message and chooses a label 70 | E, 45
d. C receives the message and chooses a label 2 | D, 70
e. B receives the message and chooses a label 50 | C, 2
f. A receives the message and does FEC | B, 50
We know how many nested tunnels there are
watching the number of the labels in the
packets
- Why ISPs have adopted MPLS?
The main advantage of MPLS for an ISP is that it provides tools to better control the networking layer, useful to:
○ build new services
simplify some procedures ○
○ optimize network utilization
In particular, MPLS has been used by ISPs in their backbone for:
▪ Virtual Private Networks (VPN)
Fault protection
▪ Traffic Engineering
▪ Quality of Services
Move 2: Define a signalling mechanism to distribute customer prefixes among Pes
Use BGP to distribute customer prefixes
▪ Multi-Protocol BGP (MP-BGP) used as signaling protocol to distribute reachability information about customer prefixes
▪ MP-BGP treats VPNs as a separate address family - > the concept of the “customer” (Es. the “L3VPN identifier”) is defined
PE routers establish a full-mesh of iBGP peerings: a PE announces to all the other PEs the customer prefixes that it can reach via
the CE router it is connected to
How can the PE make everyone know that they can access to n private networks through it? Thanks to BGP (Border Gateway
Protocol)
Have the same IP, but they are in different subnets
- The "unreachable" nodes are sent to R2 - > [P1, P2, P3]
- A BGP Session between provider equipments is created
Problem:
○ If more routers say that there is a way to reach a private network, this is a problem
○ Since they are private network, even though they are different subnet, they can have the same IP Address
○ We must identify them
Solution:
○ MP-BGP (Multi Protocol BGP)
○ Insted of specifing thorugh subnet, the PE Routers announce that they can reach the entity
○ The ID of the VPN is specified called L3VPN Identifier
Move 3: Define an encapsulation mechanism to transport packets from one PE to another across the network
○ Use MPLS encapsulation among Pes
○ The PE-PE LSPs can be “Hop-by-Hop” or “Explicitly Routed”
- How can the router understan the next-hop?
- An ad-hoc routing table for every customer has to be built
Two types of routing tables:
- Global routing table
- VRF (VPN Routing & Forwarding)
Shortest path is the same for more nodes => congestion This time shortest path with the evaluation of the efficiency of the
links => congestion
Label Switched Path (LSP) permit to divide the path in an efficient way
