1. CMPT 771/471: Internet Architecture & Protocols
School of Computing Science
Simon Fraser University
CMPT 771/471: Internet Architecture & Protocols
Network Layer
Instructor: Dr. Mohamed Hefeeda

2. Review of Basic Networking Concepts
Internet structure
Protocol layering and encapsulation
Internet services and socket programming
Network Layer
Network types: Circuit switching, Packet switching
Addressing, Forwarding, Routing
Transport layer
Reliability and congestion control
TCP, UDP
Link Layer
Multiple Access Protocols
Ethernet

3. The Network Core Mesh of interconnected routers
The fundamental question: how is data transferred through net?
circuit switching: dedicated circuit per call: telephone net
packet-switching: data sent thru net in discrete “chunks”

4. Network Core: Circuit Switching
Network resources (e.g., bandwidth) divided into “pieces” using
Frequency division multiplexing (FDM)
Time division multiplexing (TDM)
Pieces allocated to “calls” (connections)
 guaranteed performance
Resource piece idle if not used by owning call
no sharing
Connection setup is required
Examples
(Traditional) Telephone network

5. Circuit Switching: Dedicated Circuits

6. Network Core: Packet Switching
each end-end data stream divided into packets
packets from different users share network resources
each packet uses full link bandwidth
resources used as needed
store and forward: packets move one hop at a time
Node receives complete packet before forwarding
resource contention:
aggregate resource demand can exceed amount available
congestion: packets queue, wait for link use
Bandwidth division into “pieces”
Dedicated allocation
Resource reservation

7. Packet Switching: Statistical Multiplexing
10 Mb/s
Ethernet C A
statistical multiplexing
1.5 Mb/s B
queue of packets
waiting for output
link D E
Sequence of A & B packets does not have fixed pattern, shared on demand  statistical multiplexing
In contrast, in TDM each host gets same slot in revolving TDM frame

8. Packet Switching: Efficiency
Packet switching allows more users to use network!
1 Mb/s link
each user:
100 kb/s when “active”
active 10% of time
circuit-switching:
10 users
packet switching:
with 35 users, probability > 10 active less than
N users
1 Mbps link
P = 0.1
Prob of more than 10 active = 1 – prob of 10 or less are active = 1 – [ sum_{i=0}^{10} C(35, i) * p^i * (1-p)^(35-i) ]
Q: how did we get value ?

9. Packet Switching Advantages Disadvantages no call setup  simpler
resource sharing (statistical multiplexing) 
better resource utilization
more users or faster transfer (a single user can use entire bw)
Well suited for bursty traffic (typical in data networks)
Disadvantages
Congestion may occur 
packet delay and loss
need protocols to control congestion and ensure reliable data transfer
Bursty traffic: if we have few active users, circuit switching cannot achieve full link utilization (some slots will remain idle). In packet switching, a single active user can use the entire bandwidth. This leads to better resource utilization and shorter transfer time of data. In some sense, this statistical multiplexing could also be a bad thing: one user sends a huge amount of traffic at once making everybody else suffers long delay

10. Packet Switching: Two Classes
Datagram network
Example: The Internet
Virtual-circuit network
Examples: ATM (Asynchronous Transfer Mode), frame relay, X.25

11. Packet-switched Datagram Networks
no call setup at network layer
routers: no state about end-to-end connections
no network-level concept of “connection”
packets forwarded using destination host address
packets between same source-dest pair may take different paths
application
transport
network
data link
physical
1. Send data
2. Receive data

12. Packet-switched VC Networks
Source-to-dest path behaves much like telephone circuit
 performance-wise
connection setup, teardown for each call before data can flow
each packet carries VC identifier (not destination address)
every router on source-dest path maintains state for each passing connection
link, router resources (bandwidth, buffers) may be allocated to VC
Examples:
ATM (Asynchronous Transfer Mode), frame relay, X.25

13. VC Networks: Connection Setup
Signaling protocols are used to
setup, maintain, and teardown VCs
Note: not widely used in the current Internet
application
transport
network
data link
physical
application
transport
network
data link
physical
5. Data flow begins
6. Receive data
4. Call connected
3. Accept call
1. Initiate call
2. incoming call

14. Network Taxonomy Telecommunication networks Circuit-switched
FDM
TDM
Packet-switched
Networks
with VCs
Datagram

15. Review of Basic Networking Concepts
Internet structure
Protocol layering and encapsulation
Internet services and socket programming
Network Layer
Network types: Circuit switching, Packet switching
Addressing, Forwarding, Routing
Transport layer
Reliability and congestion control
TCP, UDP
Link Layer
Multiple Access Protocols
Ethernet

16. Network Layer We focus on datagram networks (Internet)
Network layer protocols in every host and router
Network layer’s goal
transport data from sending host to receiving host
We focus on datagram networks (Internet)
application
transport
network
data link
physical
network
data link
physical

17. Network Layer in the Internet
Host, router network layer functions:
Transport layer: TCP, UDP
IP protocol
addressing conventions
datagram format
packet handling conventions
Routing protocols
path selection
RIP, OSPF, BGP
Network
layer
forwarding
table
ICMP protocol
error reporting
router “signaling”
Link layer
physical layer

18. Routing vs. Forwarding Routing Forwarding1 2 3
0111
value in arriving
packet’s header
routing algorithm
local forwarding table
header value
output link
0100
0101
1001
Routing
determine route taken by packets from source to destination
Routing algorithms, e.g., RIP, OSPF, BGP
Forwarding
move packets from router’s input to appropriate output
use forwarding table populated by routing algorithm
E.g., IP forwarding function

19. 32 bit destination IP address
IP Datagram Format
IP protocol version
number
32 bits
total datagram
length (bytes)
header length
(bytes)
head.
len
type of
service
ver
length
for
fragmentation/
reassembly
Provides some QoS
fragment
offset
16-bit identifier
flgs
max number
remaining hops
(decremented at
each router)
time to
live
upper
layer
Internet
checksum
32 bit source IP address
32 bit destination IP address
upper layer protocol
to deliver payload to
Options (if any)
E.g. timestamp,
record route
taken, specify
list of routers
to visit.
data
(variable length,
typically a TCP
or UDP segment)
IP ver 4.0

20. IP Addressing: Introduction
32-bit identifier for each host, router network interface
Represented in Dotted-decimal notation
223 1

21. IP Addressing How do we assign IPs? Divide network into subnets,
Network interface:
connection between host/router and physical link
routers typically have multiple interfaces
host typically has one interface
Unique IP addresses associated with each interface
How do we assign IPs?
Divide network into subnets,
each has a common ID

22. Subnets
/24
/24
/24
Subnet is:
a group of devices that can reach each other without intervening router
identified by high order bits of IP addresses
Subnet ID
Host ID
/24
/24: # bits in subnet portion of address, subnet mask

23. Subnets How many subnets? 6 subnets Recipe:
How many subnets?
6 subnets
Recipe:
detach each interface from its host or router, creating isolated networks
Each isolated network is a subnet

24. IP Addressing: CIDR CIDR: Classless InterDomain Routing
subnet portion of address of arbitrary length
address format: a.b.c.d/x, where x is # bits in subnet portion of address
Old Classful Addressing:
Subnet length had to be /8 (class A), /16 (class B), /24 (class C)
Why CIDR?
Finer control over address allocation  reduce waste of addresses
Ex: company with 2000 machines would have to get class B, wasting 63,000+ addresses
subnet
part
host
/23

25. IP Addresses: How to Get One?
Q: How does host get IP address?
hard-coded by system admin in a file
WIN: control-panel->network->configuration->tcp/ip->properties
UNIX: /etc/rc.config
DHCP: Dynamic Host Configuration Protocol: dynamically get address from as server
“plug-and-play”

26. IP Addresses: How to Get One?
Q: How does network get subnet part of IP addr?
A: gets allocated portion of its provider ISP’s address space
ISP's block /20
Organization /23
Organization /23
Organization /23
… … ….
Organization /23
ICANN: Internet Corporation for Assigned Names and Numbers
ISPs get their address space from ICANN
ICANN: Internet Corporation for Assigned Names and Numbers
allocates addresses, manages DNS and assigns domain names

27. Hierarchical Addressing: Route Aggregation
Hierarchical addressing allows efficient advertisement of routing
information:
“Send me anything
with addresses
beginning
/20”
/23
/23
/23
Fly-By-Night-ISP
Organization 0
Organization 7
Internet
Organization 1
ISPs-R-Us
/16”
/23
Organization 2 .

28. Review of Basic Networking Concepts
Internet structure
Protocol layering and encapsulation
Internet services and socket programming
Network Layer
Network types: Circuit switching, Packet switching
Addressing, Forwarding, Routing
Transport layer
Reliability and congestion control
TCP, UDP
Link Layer
Multiple Access Protocols
Ethernet

29. Routing algorithm: find the least-cost path
Graph Abstraction u y x w v z 2 1 3 5
Graph: G = (N,E)
N = set of routers = {u, v, w, x, y, z }
E = set of links ={(u,v), (u,x), (v,x), (v,w), (x,w), (x,y), (w,y), (w,z), (y,z)}
cost of link (x1, x2):
Metric value, e.g., c(w,z) = 5
could be 1 (typical), or
inversely related to bandwidth, or
inversely related to congestion
Routing algorithm: find the least-cost path

30. Classification of Routing Algorithms
Global or local information?
Global:
all routers have complete topology, link cost info
“link state” algorithms
Local:
each router knows physically-connected neighbors, link costs to neighbors
“distance vector” algorithms

31. A Link-State Routing Algorithm
Dijkstra’s algorithm
net topology, link costs known to all nodes
accomplished via “link state broadcast”
all nodes have same info
computes least cost paths from one node (source) to all other nodes
gives forwarding table for that node

32. A Link-State Routing Algorithm
Notation:
c(x,y): link cost from node x to y;
c(x,y) = ∞ if not direct neighbors
D(v): current value of cost of path from source to dest. v
p(v): predecessor node along path from source to v
N': set of nodes whose least cost path definitively known

33. Dijsktra’s Algorithm 1 Initialization: 2 N' = {u} 3 for all nodes v
if v adjacent to u
then D(v) = c(u,v)
else D(v) = ∞ 7
8 Loop
9 find w not in N' such that D(w) is a minimum
10 add w to N'
11 update D(v) for all v adjacent to w and not in N' :
D(v) = min { D(v), D(w) + c(w,v) }
13 /* new cost to v is either old cost to v or known
shortest path cost to w plus cost from w to v */
15 until all nodes in N'
iterative: after k iterations, know least cost path to k destinations

34. Dijkstra’s algorithm: example
Step 1 2 3 4 5 N' u ux
uxy
uxyv
uxyvw
uxyvwz
D(v),p(v)
2,u
D(w),p(w)
5,u
4,x
3,y
D(x),p(x)
1,u
D(y),p(y) ∞
2,x
D(z),p(z) ∞
4,y u y x w v z 2 1 3 5

35. Dijkstra’s algorithm: example (2)
Resulting shortest-path tree from u: u y x w v z
Resulting forwarding table in u: v x y w z
(u,v)
(u,x)
destination
link

36. Distance Vector Algorithm
Bellman-Ford Equation (dynamic programming)
Define
dx(y) := cost of least-cost path from x to y
Then
dx(y) = min {c(x,v) + dv(y) }
where min is taken over all neighbors v of x v

37. Bellman-Ford example Determine du(z) How would you use BF equationv z 2 1 3 5
u has 3 neighbors: v, x, w and
dv(z) = 5, dx(z) = 3, dw(z) = 3
B-F equation says:
du(z) = min { c(u,v) + dv(z),
c(u,x) + dx(z),
c(u,w) + dw(z) }
= min {2 + 5,
1 + 3,
5 + 3} = 4
How would you use BF equation
to construct shortest paths?

38. Distance Vector Algorithm: Idea
Basic idea:
Each node periodically sends its own distance vector estimate to neighbors
When a node x receives new DV estimate from neighbor, it updates its own DV using B-F equation:
Dx(y) ← minv{c(x,v) + Dv(y)} for each node y ∊ N
Dx(y) = estimate of least cost from x to y
c(x,y) = cost from node x to its neighbor y
Under minor, natural conditions, the estimate Dx(y) converge to the actual least cost dx(y)

39. Distance Vector Algorithm: Notes
Dx(y) = estimate of least cost from x to y
Distance vector: Dx = [Dx(y): y є N ]
Node x knows cost to each neighbor v: c(x,v)
Node x maintains Dx = [Dx(y): y є N ]
Node x also maintains its neighbors’ distance vectors, that is:
x maintains Dv = [Dv(y): y є N ] for every neighbor v

40. Distance Vector Algorithm
Each node:
Iterative
Continues until no more info is exchanged
Each iteration caused by:
local link cost change
DV update message from neighbor
Asynchronous
Nodes do not operate in lockstep
Distributed
Each node receives info only from its directly attached neighbors
NO Global info
wait for (change in local link cost or msg from neighbor)
recompute estimates
if DV to any dest has changed, notify neighbors

41. z y x Dx(y) = min{c(x,y) + Dy(y), c(x,z) + Dz(y)} = min{2+0 , 7+1} = 2
Dx(z) = min{c(x,y) + Dy(z), c(x,z) + Dz(z)} = min{2+1 , 7+0} = 3
node x table
x y z x y z ∞
from
cost to
cost to
cost to
x y z
x y z x x
from y
from y z z
node y table
cost to
cost to
cost to x z 1 2 7 y
x y z
x y z
x y z x ∞ ∞ x ∞ x y
from y
from
from y z z ∞ ∞ ∞ z
node z table
cost to
cost to
Example
cost to
x y z
x y z
x y z x x x
∞ ∞ ∞
from y
from y
from y ∞ ∞ ∞ z z z 7 1
time

42. Comparison of LS and DV algorithms
Message complexity
LS: with n nodes, E links, O(nE) msgs sent
DV: exchange between neighbors only
But send entire table
Speed of Convergence
LS: O(n2) algorithm requires O(nE) msgs
may have oscillations
DV: convergence time varies
may be routing loops
count-to-infinity problem
Robustness: what happens if router malfunctions?
LS: node can advertise incorrect link cost
each node computes only its own table  some degree of robustness
DV: node can advertise incorrect path cost
each node’s table used by others error propagates thru network
In The Internet:
LS: OSPF (recent, more features)
DV: RIP (old, small nets)

43. Hierarchical Routing Our routing study thus far - idealization
all routers identical
network “flat” … not true in practice
scale: with 200 million destinations:
can’t store all dest’s in routing tables!
routing table exchange would swamp links!
administrative autonomy
internet = network of networks
each network admin may want to control routing in its own network

44. Hierarchical Routing
aggregate routers into regions, “autonomous systems” (AS)
routers in same AS run same routing protocol
“intra-AS” routing protocol
routers in different ASes can run different intra-AS routing protocols
Gateway router
Direct link to router in another AS, must use same inter-AS routing protocol

45. Interconnected ASes 3c 3a 2c 3b 2a AS3 2b 1c 1a 1b AS1 1d
Intra-AS
Routing
protocol
Inter-AS
Forwarding
table 3c
Forwarding table is configured by both intra- and inter-AS routing protocols
Intra-AS sets entries for internal destinations
Inter-AS & Intra-As sets entries for external destinations

46. Inter-AS tasks AS1 needs:
Suppose router in AS1 receives datagram for which dest is outside of AS1
Router should forward packet towards one of the gateway routers, but which one?
AS1 needs:
to learn which dests are reachable through AS2 and which through AS3
to propagate this reachability info to all routers in AS1
Job of inter-AS routing! 3b 1d 3a 1c 2a
AS3
AS1
AS2 1a 2c 2b 1b 3c

47. Example: Choosing among multiple ASes
Now suppose AS1 learns from the inter-AS protocol that subnet x is reachable from AS3 and from AS2
To configure forwarding table, router 1d must determine towards which gateway it should forward packets for dest x
Hot potato routing: send packet towards closest of two routers
Learn from inter-AS
protocol that subnet
x is reachable via
multiple gateways
Use routing info
from intra-AS
protocol to determine
costs of least-cost
paths to each
of the gateways
Hot potato routing:
Choose the gateway
that has the
smallest least cost
Determine from
forwarding table the
interface I that leads
to least-cost gateway.

48. Internet inter-AS routing: BGP
BGP (Border Gateway Protocol): the de facto standard
BGP provides each AS a means to:
Obtain subnet reachability information from neighboring Ases (reachability = AS path)
Propagate the reachability information to all routers internal to the AS
Determine “good” routes to subnets based on reachability information and policy
BGP allows a subnet to advertise its existence to rest of the Internet: “I am here”
Why reachability? Why not exact route? Because exact route will impose too much overhead. Routers would
Have to handle a huge amount of information. With reachability (i.e., AS path), core routers have in their tables
Around 140,000 entries!
This is a tradeoff between scalability and route optimality.
BGP is concerned with finding any route to destination, not necessarily the optimal route.

49. BGP basics
Pairs of routers (BGP peers) exchange routing info over semi-permanent TCP connections: BGP sessions
Note: BGP sessions do not correspond to physical links
When AS2 advertises a prefix to AS1, AS2 is promising it will forward any datagrams destined to that prefix towards the prefix
AS2 can aggregate prefixes in its advertisement 3b 1d 3a 1c 2a
AS3
AS1
AS2 1a 2c 2b 1b 3c
eBGP session
iBGP session

50. Distributing reachability info
With eBGP session between 3a and 1c, AS3 sends prefix reachability info to AS1
1c can then use iBGP to distribute this new prefix reachability info to all routers in AS1
1b can then re-advertise the new reachability info to AS2 over the 1b-to-2a eBGP session
When router learns about a new prefix, it creates an entry for the prefix in its forwarding table. 3b 1d 3a 1c 2a
AS3
AS1
AS2 1a 2c 2b 1b 3c
eBGP session
iBGP session

51. Path attributes & BGP routes
When advertising a prefix, advert. includes BGP attributes
prefix + attributes = “route”
Two important attributes:
AS-PATH: contains ASes on the path to the prefix
NEXT-HOP: Indicates the specific internal-AS router to next-hop-AS. (There may be multiple links from current AS to next-hop-AS.)
When gateway router receives route advert., it uses import policy to accept/decline

52. BGP messages BGP messages exchanged using TCP BGP messages:
OPEN: opens TCP connection to peer and authenticates sender
UPDATE: advertises new path (or withdraws old)
KEEPALIVE keeps connection alive in absence of UPDATES; also ACKs OPEN request
NOTIFICATION: reports errors in previous msg; also used to close connection

53. BGP Route Selection
Router may learn about more than 1 route to some prefix. Router must select a route
Elimination rules:
Local preference value: policy decision
(Routes are assigned values by AS administrator based on import policy)
Shortest AS-PATH
Closest NEXT-HOP router: hot potato routing
Additional criteria

54. BGP Routing: Route Advertising
A,B,C are provider networks
X,W,Y are customer (of provider networks)
X is dual-homed: attached to two provider networks
X does not want to route traffic from B to C
… so X will not advertise to B its route to C
BGP export policy

55. BGP Routing: Route Advertising (cont’d)
A advertises to B the path AW
B advertises to X (its client) the path BAW
Should B advertise to C the path BAW?
No way! B gets no “revenue” for routing CBAW since neither W nor C are B’s customers
Rule of thumb: a provider wants to route only to/from its customers! (unless there is a mutual peering deal)

56. Why different Intra- and Inter-AS routing ?
Policy:
Inter-AS: admin wants control over how its traffic routed, who routes through its net.
Intra-AS: single admin, so no policy decisions needed
Scale:
hierarchical routing saves table size, reduced update traffic
Performance:
Intra-AS: can focus on performance
Inter-AS: policy may dominate over performance
Is routing in the Internet optimal?
NO, because:
- BGP policies may mandate longer paths
- BGP advertises only AS path. ASes have different sizes. Even if they have the same size, we do not know the cost of
paths within ASes.
- A route may (typically) crosses multiple ASes, each with its own intra-domain routing algorithm with different metric.
This makes it hard to even define an optimality criterion.

57. Unicast, multicast, broadcast
Unicast: one source, one destination
E.g., web session
Multicast: one source, multiple destinations
Subset of all possible destinations
E.g., streaming a hockey game to interested fans
Broadcast: one source, all destinations
E.g., broadcasting link state info to ALL routers in a domain in OSPF protocol
Anycast: multiple possible sources, one destination
Sources have same (anycast) address
Request is forwarded to appropriate source
(Still in research phases)