1. CS 5565 Network Architecture and Protocols
Lecture 22
Godmar Back

2. Announcements Project 2B due in 2 parts: Extra Credit Opportunities:
Apr 29 and May 6
Extra Credit Opportunities:
Expand simulator (and your implementation) to introduce multiple link failures and link resurrection
Additional, requiring reading posted
Andersen et al [SIGCOMM’08]: Accountable Internet Protocol (AIP)
Casado et al [HotNets’08]: Rethinking Packet Forwarding Hardware
CS 5565 Spring 2009
4/18/2019

3. Other Routing Protocols
Ad-hoc Routing
Broadcast Routing
Multicast Routing
CS 5565 Spring 2009
4/18/2019

4. creation/transmission
Broadcast Routing
Motivation:
Use in-network duplication (b) rather than source-duplication (a)
(a)
(b) R1 R2 R3 R4
duplicate
creation/transmission
CS 5565 Spring 2009
4/18/2019

5. Broadcast Routing (2) Simplest approach: simple flooding
Forward every packet from every link to all other links every time
Inefficient, loops, “broadcast storms”
Sequence-number controlled flooding
Only forward new packets to all other links
CS 5565 Spring 2009
4/18/2019

6. Reverse Path Forwarding
Only forward packets from link that lies on shortest path to the source
Assume unicast routing has run & every node knows shortest path to source A B G D E C F
CS 5565 Spring 2009
4/18/2019

7. Broadcast using spanning treeF A B G D E C F
Same spanning tree can be used for all sources!
CS 5565 Spring 2009
4/18/2019

8. Spanning Tree ConstructionB G D E C F 1 2 3 4 5 A B G D E C F
Center-based: all nodes send “tree-join” message to known or elected center node
CS 5565 Spring 2009
4/18/2019

9. Multicast Routing
Goal: find a tree (or trees) connecting routers having local mcast group members
tree: not all paths between routers used
source-based: different tree from each sender to rcvrs
shared-tree: same tree used by all group members
Source-based trees
Shared tree
Notes:
3.3 Network Layer: Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

10. Shortest Path Tree
mcast forwarding tree: tree of shortest path routes from source to all receivers
Dijkstra’s algorithm
S: source
LEGEND R1 2 R4
router with attached
group member 1 R2 5
router with no attached
group member
Notes:
3.3 Network Layer: Multicast Routing Algorithms 3 4 R5 6
link used for forwarding,
i indicates order link
added by algorithm R3 i R6 R7
CS 5565 Spring 2009
4/18/2019

11. Reverse Path Forwarding
S: source
LEGEND
router with attached
group member
router with no attached
group member
datagram will be forwarded
datagram will not be
forwarded
Notes:
3.3 Network Layer: Multicast Routing Algorithms
result is a source-specific reverse shortest path tree (SPT) – unless costs are asymmetric
CS 5565 Spring 2009
4/18/2019

12. Reverse Path Forwarding: Pruning
S: source P
LEGEND
router with attached
group member
router with no attached
group member P
prune message
links with multicast
forwarding
no need to forward datagrams down subtrees with no group members
“prune” msgs sent upstream by router with no downstream group members
Notes:
3.3 Network Layer: Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

13. Shared-Tree: Steiner Tree
Steiner Tree: minimum cost tree connecting all routers with attached group members
problem is NP-complete (if intermediate nodes must be found)
excellent heuristics exists
not used in practice:
computational complexity
information about entire network needed
monolithic: rerun whenever a router needs to join/leave
Notes:
1. See L. Wei and D. Estrin, “A Comparison of multicast trees and algorithms,” TR USC-CD , Dept. Computer Science, University of California, Sept 1993 for a comparison of heuristic approaches.
3.3 Network Layer: Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

14. Center-Based trees single delivery tree shared by all
one router identified as “center” of tree
to join:
edge router sends unicast join-msg addressed to center router
join-msg “processed” by intermediate routers and forwarded towards center
join-msg either hits existing tree branch for this center, or arrives at center
path taken by join-msg becomes new branch of tree for this router
Notes:
1. It’s always nice to see a PhD dissertation with impact. The earliest discussion of center-based trees for multicast appears to be D. Wall, “Mechanisms for Broadcast and Selective Broadcast,” PhD dissertation, Stanford U., June 1980.
3.3 Network Layer: Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

15. Center-based Trees Suppose R6 chosen as center: R1 LEGEND R4 32 3 1
LEGEND
router with attached
group member
router with no attached
group member 1
path order in which join messages generated
Notes:
3.3 Network Layer: Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

16. Internet Multicasting Routing: DVMRP
DVMRP: distance vector multicast routing protocol, RFC1075
flood and prune: reverse path forwarding, source-based tree
RPF tree based on DVMRP’s own routing tables constructed by communicating DVMRP routers
no assumptions about underlying unicast
initial datagram to mcast group flooded everywhere via RPF
routers not wanting group: send upstream prune msgs
Notes:
D. Waitzman, S. Deering, C. Partridge, “Distance Vector Multicast Routing Protocol,” RFC 1075, Nov The version of DVMRP in use today is considerably enhanced over the RFC1075 spec.
A more up-to-date “work-in-progress” defines a version 3 of DVMRP: T. Pusateri, “Distance Vector Multicast Routing Protocol,” work-in-progress, draft-ietf-idmr-v3-05.ps
3.4 Network Layer: Internet Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

17. DVMRP: continued…
soft state: DVMRP router periodically (1 min.) “forgets” branches are pruned:
mcast data again flows down unpruned branch
downstream router: reprune or else continue to receive data
routers can quickly regraft to tree
following IGMP join at leaf
odds and ends
commonly implemented in commercial routers
Mbone routing done using DVMRP
Notes:
1. See for a (slightly outdatet) list of multicast capable routers (supporting DVMPR as well as other protocols) from various vendors.
2. ftp://parcftp.xerox.com/pub/net-research/ipmulti for circa 1996 public copy “mrouted” v3.8 of DVMRP routing software for various workstation routing platforms.
3.4 Network Layer: Internet Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

18. IGMP Internet Group Management Protocol
Local protocol used by hosts to inform their routers that they’d like to join a mcast group
MCast addresses are 224.x.x.x
28bits for groups, address indirection
Simple protocol
Join
Leave (optional)
Membership Query (still interested?)
CS 5565 Spring 2009
4/18/2019

19. Tunneling
Q: How to connect “islands” of multicast routers in a “sea” of unicast routers?
physical topology
logical topology
mcast datagram encapsulated inside “normal” (non-multicast-addressed) datagram
normal IP datagram sent thru “tunnel” via regular IP unicast to receiving mcast router which undoes encapsulation
Notes:
For a general discussion of IP encapsulation, see C. Perkins, “IP Encapsulation within IP,” RFC 2003, Oct
The book S. Bradner, A Mankin, “Ipng: Internet protocol next generation,” Addison Wesley, 1995 has a very nice discussion of tunneling
Tunneling can also be used to connect islands of IPv6 capable routers in a sea IPv4 capable routers. The long term hope is that the sea evaporates leaving only lands of IPv6!
3.4 Network Layer: Internet Multicast Routing Algorithms
CS 5565 Spring 2009
4/18/2019

20. Status of IP Multicast MBone exists
PIM: ‘Protocol Independent Multicast’ protocol
alternative to DVMRP
Sporadically deployed
Has not taken off
Despite need (?)
CS 5565 Spring 2009
4/18/2019

21. Data Link Layer

22. The Data Link Layer Our goals:
understand principles behind data link layer services:
error detection, correction
sharing a broadcast channel: multiple access
link layer addressing
reliable data transfer, flow control: done!
instantiation and implementation of various link layer technologies
CS 5565 Spring 2009
4/18/2019

23. Link Layer: Introduction
hosts and routers are nodes
communication channels that connect adjacent nodes along communication path are links
wired vs. wireless links, LANs
provide different services
layer-2 packet is a frame, encapsulates datagram
data-link layer has responsibility of
transferring datagram from one node to adjacent node over a link
CS 5565 Spring 2009
4/18/2019

24. Adaptors Communicating
datagram
link layer protocol
rcving
node
sending
node
frame
frame
adapter
adapter
link layer implemented in “adaptor” (aka NIC)
Ethernet card, PCMCIA card, card
sending side:
encapsulates datagram in a frame (header + data + trailer)
adds error checking bits, rdt, flow control, etc.
receiving side
looks for errors, rdt, flow control, etc
extracts datagram, passes to rcving node
adapter is semi-autonomous
link & physical layers
CS 5565 Spring 2009
4/18/2019

25. Example: Gigabit Ethernet
We’ve covered this so far
Could teach an entire (EE) course on what’s below
LLC Logical Link Control
MAC Media Access Control
PCS Physical Coding Sublayer
PMA Physical Media Access
GMII Gigabit Media Independent Interface
MDI Media Dependent Interface
(Twisted Pair, Fiber)
Source: [Noseworthy 1998]
CS 5565 Spring 2009
4/18/2019

26. Link Layer Services Framing, link access:
encapsulate datagram into frame, adding header, trailer
channel access if shared medium
“MAC” addresses used in frame headers to identify source, dest
different from IP address!
Reliable delivery between adjacent nodes
same ideas as at Layer 4 apply
seldom used on low bit error link (fiber, some twisted pair)
wireless links: high error rates
Q: why both link-level and end-end reliability?
CS 5565 Spring 2009
4/18/2019

27. Link Layer Services (more)
Flow Control:
pacing between adjacent sending and receiving nodes
Half-duplex and full-duplex
with half duplex, nodes at both ends of link can transmit, but not at same time
Error Detection:
errors caused by signal attenuation, noise.
receiver detects presence of errors:
signals sender for retransmission or drops frame
Error Correction:
receiver identifies and corrects bit error(s) without resorting to retransmission
CS 5565 Spring 2009
4/18/2019