CS计算机代考程序代写 flex assembly Network Layer

Network Layer
 Network Layer: the most complex layer
 Requires the coordinated actions of multiple, geographically distributed network elements (switches & routers)
 Must be able to deal with very large scales
 Billions of users (people & communicating devices)
 Biggest Challenges
 Addressing: where should information be directed to?
 Routing: what path should be used to get information there?

Packet Switching
t0
Network
 Transfer of information as payload in data packets
 Packets undergo random delays & possible loss
 Different applications impose differing requirements on the transfer of information
t1

Perspectives of Packet Networks
 External View of the network
 Services that the network provides to the transport layer
 Services are independent of the underlying network
 Whether the network service requires setting up of connections  Whether data transfer requires any quality-of-service guarantees
 Internal Operation of the network
 Considers physical topology of the network and its interconnection  Approaches used to direct information – datagram, virtual circuit
 Addressing and routing procedures
 Deal with congestion inside the network
 Traffic management inside the network

Network Service
Messages
Transport layer
Network service
End system α
Segments
Messages
Network service
End system β
Transport layer
Network layer
Data link layer
Physical layer
Network layer
Data link layer
Physical layer
Network layer
Data link layer
Physical layer
Network layer
Data link layer
Physical layer
 Network layer can offer a variety of services to transport layer  Connection-oriented service or connectionless service
 Best-effort or delay/loss guarantees

Connectionless vs. Connection-oriented
Connectionless :
 Only two basic interactions between transport and network layer  Request to network layer to send a packet
 Indication from network layer that a packet has arrived
 User can request packet transmission at any time  No need to inform network layer ahead of time
 Responsibility for error control, sequencing and flow control on transport-layer Connection-oriented :
 Connection-setup required
 Network layer must be informed about the new flow to be sent to the network  Network layer maintains state information about the flows it is handling
 Allows usage and quality-of-service negotiations  Network resources may be allocated
 Connection-terminationrequired
 Complexthanconnectionlessservice

Network Service vs. Operation
Network Service
 Connectionless (UDP)
 Datagram Transfer
 Connection-Oriented (TCP)
 Reliable and possibly constant bit rate transfer
Internal Network Operation
 Connectionless
 Datagram operation  IP
 Connection-Oriented
 Virtual Circuit operation  Telephone connection  ATM
Various combinations are possible
 Connection-oriented service over Connectionless operation  Connectionless service over Connection-Oriented operation  Context & requirements determine what makes sense

Complexity at the Edge or in the Core?
C 12321
End system αβ
End system 1 12321Medium12321 13
4
3
2
2
4
1 2
Physical layer entity Data link layer entity 3
AB Network
2 1
3 Network layer entity
Network layer entity 4
Transport layer entity
Need for the network to grow to very large scale –
 keep the core of the network simple (connectionless packet network)  provide necessary complexity at the edge

Network Layer Functions
Essential
 Routing: mechanisms for determining the set of best paths for routing packets requires the collaboration of network elements
 Forwarding: transfer of packets from NE inputs to outputs
 Priority & Scheduling: determining order of packet transmission in each NE
Optional: congestion control, segmentation & reassembly, security

Packet-Switching Networks
Datagrams and Virtual Circuits

The Switching Function
 Dynamic interconnection of inputs to outputs
 Enables dynamic sharing of transmission resource
 Two fundamental approaches:
 Connectionless
 Connection-Oriented: Call setup control, Connection control
Backbone Network Switch
Access Network

Packet Switching Network
User
Packet switching network
 Transfers packets between users
 Transmission lines + packet switches (routers)
 Origin in message switching
Two modes of operation:  Connectionless
 Virtual Circuit
Transmission line
Network
Packet
switch

Message Switching
Message
Source
 Message switching invented for telegraphy
 Entire messages multiplexed onto shared lines, stored & forwarded
 Headers for source & destination addresses
Message
Destination
Message
Message
Switches
 Routing at message switches
 Connectionless

Message Switching Delay
Source Switch 1

Switch 2 Destination
T
t t t t
Delay Minimum delay = 3 + 3T
Additional queueing delays possible at each link

Long Messages vs. Packets
source
BER=p=10-6 BER=10-6
1 Mbit message
dest How many bits need to be transmitted to deliver message?
 Approach 1: send 1 Mbit message
 Probability message arrives correctly
 Approach 2: send 10 100-kbit packets
 Probability packet arrives correctly
P (1106)106 e106106 e1 1/3  On average it takes about
3 transmissions/hop
 Total # bits transmitted ≈ 6 Mbits
P(1106)105 e105106 e0.1 0.9 cc
 On average it takes about 1.1 transmissions/hop
 Total # bits transmitted ≈ 2.2 Mbits

Packet Switching – Datagram
 Messages broken into smaller units (packets)
 Source & destination addresses in packet header
 Connectionless, packets routed independently (datagram)
 Packet may arrive out of order
 Pipelining of packets across network can reduce delay, increase throughput
 Lower delay that message switching, suitable for interactive traffic
Packet 1
Packet 1
Packet 2
Packet 2 Packet 2

Packet Switching Delay
Assume three packets corresponding to one message traverse same path
1
12
t t
t t
23
3
12 Delay
3
Minimum Delay = 3τ + 5(T/3) (single path assumed)
Additional queueing delays possible at each link Packet pipelining enables message to arrive sooner

Delay for k-Packet Message over L Hops
Source Switch 1

t t
t t
1
23
Switch 2 Destination
12
3 12
3
3 hops
3 + 2(T/3) first bit received
3 + 3(T/3) first bit released 3 + 5 (T/3) last bit released
L hops
L + (L-1)P first bit received
L + LP first bit released
L + LP + (k-1)P last bit released where T = k P

Routing Tables in Datagram Networks
Destination
address port
Output
 Route determined by table lookup
 Routing decision involves finding next hop in route to given destination
 Routing table has an entry for each destination specifying output port that leads to next hop
 Size of table becomes impractical for very large number of destinations
0785
7
1345
12
1566 6
2458 12

Example: Internet Routing
 Internet protocol uses datagram packet switching across networks
 Networks are treated as data links  Hosts have two-port IP address:
 Network address + Host address
 Routers do table lookup on network address
 This reduces size of routing table
 In addition, network addresses are assigned
so that they can also be aggregated  Discussed as CIDR in Chapter 8

Packet Switching – Virtual Circuit
Packet
Packet
Virtual circuit
Packet
Packet
 Call set-up phase sets up pointers in fixed path along network
 All packets for a connection follow the same path
 Abbreviated header identifies connection on each link
 Packets queue for transmission
 Variable bit rates possible, negotiated during call set-up
 Delays variable, cannot be less than circuit switching

Connection Setup
Connect request
Connect confirm
Connect
request …
Connect confirm
Connect request
Connect confirm
SW 1
SW 2
SW
n
 Signaling messages propagate as route is selected
 Signaling messages identify connection and setup tables
in switches
 Typically a connection is identified by a local tag, Virtual Circuit Identifier (VCI)
 Each switch only needs to know how to relate an incoming tag in one input to an outgoing tag in the corresponding output
 Once tables are setup, packets can flow along path

Connection Setup Delay
Connect request
CR
1
CC
CC
23
12
3
Release
t t
t t
CR Connect 12 confirm
3
 Connection setup delay is incurred before any packet can be transferred
 Delay is acceptable for sustained transfer of large number of packets
 This delay may be unacceptably high if only a few packets are being transferred

Virtual Circuit Forwarding Tables
Input VCI
Output Output port VCI
 Each input port of packet switch has a forwarding table
 Lookup entry for VCI of incoming packet
 Determine output port (next hop) and insert VCI for next link
 Very high speeds are possible
 Table can also include priority or other information about how packet should be treated
12
13
44
15
15
23
27
13
16
58
7
34

Cut-Through switching
Source Switch 1
Switch 2
Destination
1
1
2
2
3
3
t t t t
2 Minimum delay = 3 + T
1
3
 Some networks perform error checking on header only, so packet can be forwarded as soon as header is received & processed
 Delays reduced further with cut-through switching

Message vs. Packet Minimum Delay
 Message:
L  + L T = L  + (L – 1) T + T
 Packet
L+ LP+(k–1)P = L+(L–1)P + T
 Cut-Through Packet (Immediate forwarding after header)
=L+T Above neglect header processing delays

Example: ATM Networks
 All information mapped into short fixed-length packets called cells
 Connections set up across network
 Virtual circuits established across networks  Tables setup at ATM switches
 Several types of network services offered  Constant bit rate connections
 Variable bit rate connections

Asynchronous Tranfer Mode (ATM)
 Packet multiplexing and switching  Fixed-length packets: “cells”
 Connection-oriented
 Rich Quality of Service support
 Conceived as end-to-end
 Supporting wide range of services
 Real time voice and video
 Circuit emulation for digital transport
 Data traffic with bandwidth guarantees
 Detailed discussion in Chapter 9

ATM Networking
Voice Video Packet
Voice Video Packet
ATM Adaptation Layer
ATM Adaptation Layer
ATM Network
 End-to-end information transport using cells
 53-byte cell provide low delay and fine multiplexing
granularity
 Support for many services through ATM Adaptation Layer

TDM vs. Packet Multiplexing
Variable bit
rate Delay Burst traffic Processing
TDM Packet
*In mid-1980s, packet processing mainly in software and hence slow; By late 1990s, very high speed packet processing possible
Multirate only
Easily  handled
Low, fixed Inefficient
Minimal, very high speed
Variable
Efficient
*
Header & packet processing required

ATM: Attributes of TDM & Packet Switching
Voice
Data packets
Images
1 2 3
MUX 4
TDM
ATM
Wasted bandwidth
• Packet structure gives flexibility & efficiency
32143214321
4313221 Packet Header
• Synchronous slot transmission gives high speed & density

ATM Switching
Switch carries out table translation and routing
1
5 6
N
Switch
1
2 3
N
voice
67
video
data
25
video
67
25 32
N
1
75 67
32 61
3 2
39 67
voice
32
32
video
61
video
75
data
39
ATM switches can be implemented using shared memory, shared backplanes, or self-routing multi-stage fabrics


ATM Virtual Connections
 Virtual connections setup across network
 Connections identified by locally-defined tags
 ATM Header contains virtual connection information:  8-bit Virtual Path Identifier (VPI)
 16-bit Virtual Channel Identifier (VCI) – local identifier
 Powerful traffic grooming capabilities
 Multiple VCs can be bundled within a VP (flows that have a common
path through the network are grouped together)
 Similar to tributaries with SONET, except variable bit rates possible Virtual paths
Physical link
Virtual channels

VPI/VCI switching & multiplexing
VPI 3
VPI 5
a
a
b cb
ATM Sw 1
ATM cross- connect
ATM Sw 2
ATM Sw 3
c
d e
VPI 2
VPI 1
d e
Sw = switch
 Connections a,b,c bundled into VP at switch 1
 Crossconnect switches VP without looking at VCIs  VP unbundled at switch 2; VC switching thereafter
 VPI/VCI structure allows creation virtual networks
 Can support large number of connections – provides scalability
ATM Sw 4

MPLS & ATM
 ATM initially touted as more scalable than packet switching
 ATM envisioned speeds of 150-600 Mbps
 Advances in optical transmission proved ATM to be
the less scalable: @ 10 Gbps
 Segmentation & reassembly of messages & streams into 48-byte cell payloads difficult & inefficient
 Header must be processed every 53 bytes vs. 500 bytes on average for packets
 Delay due to 1250 byte packet at 10 Gbps = 1 msec; delay due to 53 byte cell @ 150 Mbps ≈ 3 msec
 MPLS (Chapter 10) uses tags to transfer packets across virtual circuits in Internet