LTE Radio Access Networks 2013

192620010 Mobile & Wireless Networking Lecture 10: Mobile Transport Layer & Ad Hoc Networks [Schiller, Section 8.3 & Section 9] [Reader, Part 8] Geert Heijenk

Mobile and Wireless Networking 2013 / 2014

Outline of Lecture 10 Mobile transport layer    

Motivation Approaches for improvement        

 

Indirect TCP Snooping TCP Mobile TCP Selective retransmission

Recommended TCP improvements

Ad hoc networks      

Concept Addressing and forwarding in ad-hoc networks Routing in ad-hoc networks          

Problem description DSDV (Destination Sequenced Distance Vector) Ad-hoc On-demand Distance Vector (AODV) DSR (Dynamic Source Routing) Further alternatives

2 Mobile and Wireless Networking 2013 / 2014

Motivation I Transport protocols typically designed for    

Fixed end-systems Fixed, wired networks

TCP congestion control      

 

packet loss in fixed networks typically due to (temporary) overload situations router have to discard packets as soon as the buffers are full TCP recognizes congestion only indirect via missing acknowledgements, retransmissions unwise, they would only contribute to the congestion and make it even worse slow-start algorithm as reaction


Motivation II TCP slow-start algorithm          

sender calculates a congestion window for a receiver start with a congestion window size equal to one segment exponential increase of the congestion window up to the congestion threshold, then linear increase missing acknowledgement causes the reduction of the congestion threshold to one half of the current congestion window congestion window starts again with one segment

TCP fast retransmit/fast recovery        

TCP sends an acknowledgement only after receiving a packet if a sender receives several acknowledgements for the same packet, this is due to a gap in received packets at the receiver however, the receiver got all packets up to the gap and is actually receiving packets therefore, packet loss is not due to severe congestion, continue with (half of) current congestion window (do not use slow-start)


Influences of mobility on TCP-mechanisms TCP assumes congestion if packets are dropped    

typically wrong in wireless networks, here we often have packet loss due to transmission errors furthermore, mobility itself can cause packet loss, if e.g. a mobile node moves from one access point (e.g. foreign agent in Mobile IP) to another while there are still packets in transit and forwarding is not possible

Additional problem:  

it takes a long time to increase the congestion window if the latency of the wireless link is high

The performance of an unchanged TCP degrades severely    

but TCP cannot be changed due to the large base of installation in the fixed network (end-to-end protocol) therefore TCP for mobility has to remain compatible


Solutions: Indirect TCP / Performance Enhancing Proxy (RFC 3135) Indirect TCP or I-TCP segments the connection        

no changes to the TCP protocol for hosts connected to the wired Internet, millions of computers use (variants of) this protocol optimized TCP protocol for mobile hosts splitting of the TCP connection at, e.g., the foreign agent into 2 TCP connections, no real end-to-end connection any longer hosts in the fixed part of the net do not notice the characteristics of the wireless part

mobile host

performance enhancing proxy: access point / foreign agent

„wireless“ TCP

„wired“ Internet

standard TCP


I-TCP socket and state migration        

Socket contains current TCP connection information (seq.num, ports) During handover, current AP buffers packets Since these packets have been acknowledged, they must be forwarded to new AP after handover is executed Fixed-side connection must be maintained (socket)

access point1

socket migration and state transfer

Internet

access point2 mobile host 7 Mobile and Wireless Networking 2013 / 2014

Indirect TCP III Advantages        

no changes in the fixed network necessary, no changes for the hosts (TCP protocol) necessary, all current optimizations to TCP still work transmission errors on the wireless link do not propagate into the fixed network simple to control, mobile TCP is used only for one hop between, e.g., a foreign agent and mobile host therefore, a very fast retransmission of packets is possible, the delay on the mobile hop is known

Disadvantages  

 

loss of end-to-end semantics, an acknowledgement to a sender does now not any longer mean that a receiver really got a packet, foreign agents might crash higher latency possible due to buffering of data within the foreign agent and forwarding to a new foreign agent


Solutions: Snooping TCP I „Transparent“ extension of TCP within the foreign agent    

   

buffering of packets sent to the mobile host lost packets on the wireless link (both directions!) will be retransmitted immediately by the mobile host or foreign agent, respectively (so called “local” retransmission) the foreign agent therefore “snoops” the packet flow and recognizes acknowledgements in both directions, it also filters ACKs changes of TCP only within the foreign agent local retransmission

correspondent host

foreign agent „wired“ Internet

mobile host

snooping of ACKs

buffering of data

end-to-end TCP connection 9 Mobile and Wireless Networking 2013 / 2014

Snooping TCP II Data transfer to the mobile host    

FA buffers data until it receives ACK of the MH, FA detects packet loss via duplicated ACKs or time-out fast retransmission possible, transparent for the fixed network

Data transfer from the mobile host    

FA detects packet loss on the wireless link via sequence numbers, FA answers directly with a NACK to the MH MH can now retransmit data with only a very short delay

Integration of the MAC layer    

MAC layer often has similar mechanisms to those of TCP thus, the MAC layer can already detect duplicated packets due to retransmissions and discard them

Problems    

snooping TCP does not isolate the wireless link as good as I-TCP snooping might be useless depending on encryption schemes 10 Mobile and Wireless Networking 2013 / 2014

Solutions: Mobile TCP Special handling of lengthy and/or frequent disconnections M-TCP splits as I-TCP does    

unmodified TCP fixed network to supervisory host (SH) optimized TCP SH to MH

Supervisory host    

no caching, no retransmission monitors all packets, if disconnection detected    

 

set sender window size to 0 sender automatically goes into persistent mode

old or new SH reopen the window

Advantages  

maintains semantics, supports disconnection, no buffer forwarding

Disadvantages    

loss on wireless link propagated into fixed network adapted TCP on wireless link


Selective retransmission TCP acknowledgements are often cumulative    

ACK n acknowledges correct and in-sequence receipt of packets up to n if single packets are missing quite often a whole packet sequence beginning at the gap has to be retransmitted (go-back-n), thus wasting bandwidth

Selective retransmission as one solution    

RFC2018 allows for acknowledgements of single packets, not only acknowledgements of in-sequence packet streams without gaps sender can now retransmit only the missing packets

Advantage  

much higher efficiency

Disadvantage  

more complex software in a receiver, more buffer needed at the receiver 12 Mobile and Wireless Networking 2013 / 2014

Recommended TCP Improvements (RFC 3481) •  •  •  •  •  •  •  •  • 

Appropriate Window Size (Sender & Receiver) Increased Initial Window (Sender) Limited Transmit (Sender) IP MTU Larger than Default Path MTU Discovery (Sender & Intermediate Routers) Selective Acknowledgments (Sender & Receiver) Explicit Congestion Notification (Sender, Receiver & Intermediate Routers) TCP Timestamps Option (Sender & Receiver) Disabling RFC 1144 TCP/IP Header Compression (Wireless Host) (RObust Header Compression, ROHC, RFC3095, is OK)




 







Ad hoc network concept  

Networks of wireless terminals that do not necessarily rely on existing infrastructure  

     

Although interworking with infrastructure is possible

Direct communication between terminals when needed Multi-hop communication Extended concept of mobility: network mobility (moving routers)


Mobile ad hoc networks Standard Mobile IP needs an infrastructure    

Home Agent/Foreign Agent in the fixed network DNS, routing etc. are not designed for mobility

Sometimes there is no infrastructure!    

remote areas, ad-hoc meetings, disaster areas cost can also be an argument against an infrastructure!

Main topic: routing    

no default router available every node should be able to forward

A

B

C 16


Addressing & forwarding in ad-hoc networks •  •  •  •  • 

Broadcast (single hop) Beaconing Broadcast (multihop) Geocast Unicast


Beaconing Beaconing: •  periodic single-hop broadcast Problems: •  • 

risk of overload no feedback from medium because of broadcast

Solutions:  

adapt load to number of users, e.g., inter-beacon time / power control


Broadcast (multihop) Broadcast (multihop) Forwarding approach:  

flooding: all nodes forward all newly received messages

Problems:  

broadcast storm:      

redundant transmissions synchronization of transmissions no acks, no feedback from medium

Solutions:      

(random) delay before rebroadcasting, not all nodes rebroadcast e.g., distance-based forwarding 19 Mobile and Wireless Networking 2013 / 2014

Geocast Geocast:  

like broadcast, 1. Illustration of geocast flooding algorithm. but packet is only forwarded toFigure certain geographic region

Forwarding approach    

flooding to and in region routing to region + flooding in region

Problems:  

trade-off between accuracy and efficiency

Figure 2. Geocast region and geocast group. 20


geocast region at that time, forming the geocast group at time

ah me geo inc ins

Unicast Unicast: single destination with known address Forwarding approach  

based on routing information




 







Routing examples Routing is a major topic   in principle, every node should be able to forward   dynamic topology   asymmetric links   redundant links: too many links when terminals are close

N1

N1

N2 N3 N4 time = t1

N3

N2 N4

N5 good link weak link

N5 time = t2


Traditional routing algorithms  

Distance Vector  

 

 

periodic exchange of messages with all physical neighbors that contain information about who can be reached at what distance (monodirectional) selection of the shortest path if several paths available

Link State    

periodic notification of all routers about the current state of all physical links (bidirectional) router get a complete picture of the network


Problems of traditional routing algorithms  

Dynamic of the topology  

 

Limited performance of mobile systems      

 

frequent changes of connections, connection quality, participants

periodic updates of routing tables need energy without contributing to the transmission of user data, sleep modes difficult to realize limited bandwidth of the system is reduced even more due to the exchange of routing information links can be asymmetric, i.e., they can have a direction dependent transmission quality

Problem  

protocols have been designed for fixed networks with infrequent changes and typically assume symmetric links


Ad hoc routing algorithms  

Pro-active Example:   Destination Sequenced Distance Vector (DSDV)

 

Re-active Example:   Ad-hoc On-demand Distance Vector (AODV)   Dynamic Source Routing (DSR)


DSDV (Destination Sequenced Distance Vector)    

Extension of distance vector routing Sequence numbers for all routing updates    

 

assures in-order execution of all updates avoids loops and inconsistencies

Decrease of update frequency    

store time between first and best announcement of a path inhibit update if it seems to be unstable (based on the stored time values)

See [Schiller] for details


Ad-hoc On-demand Distance Vector (AODV) • 

Specified in IETF: RFC 3561

• 

Forms the basis for DYMO (Dynamic On-demand MANET routing protocol) which is a planned IETF reactive routing protocol.

• 

Uses destination sequence numbers to avoid loops, and ensure proper updating of routes Storage of routes in Route Table Uses only symmetric links

•  • 


urce broadcasts

(RREQ)

Route Discovery EQ contains: • 

Source broadcasts Route Request (RREQ):

>

node can reply tocan thereply RREQ if if •  A node to the RREQ •  It is the destination It is the destination •  It has a “fresh enough” route

to the destination It has a “fresh enough” route to •  Otherwise it rebroadcasts the RREQ the destination •  Nodes keep track of

herwise it

Source

rebroadcasts the RREQ and discard redundant broadcasts

des keep track of

Destination

> d discard redundant broadcasts

Route Request Propagation Source: Perkins & Royer


Reverse Path Setup

Reverse Path Setup

Nodes update their Route Table with source node information before forwarding RREQ Nodes update their Route Table with Reverse path entry used to forward source•  node information before Route Reply (RREP) back to source forwarding if oneRREQ is received •  Expiration time is long enough for Reverseapath used toand forward RREPentry to be received forwarded • 

Source

(RREP) back to source if one is received Expiration time is long enough for a RREP to be received and forwarded

Destination

Reverse Path Formation 30


destination, unicasts

(RREP) back to source

Forward Path Setup RREP contains: < •  Destination, or intermediate node with current route to

>

destination, unicasts Route Reply (RREP) back to source:

Nodes along path record in Route Table, use to forward RREP Nodes along path record forward route Source can begin sending data in Route Table, use reverse route to forward RREP when it receives first RREP •  Source can begin sending data when it receives first RREP If it later receives a RREP •  If it later receives a RREP withmetric, better metric, with better it updates its route entry • 

it updates its route entry

Perkins, Royer

Source

Destination

Forward Path Formation 31 Mobile and Wireless Networking 2013 / 2014

Sun, UCSB

Route Table

Fields: estination•  IP Address •  Destination IP Address estination Sequence •  Destination Number Sequence Number Next Hop •  HopCount A op Count •  Next Hop IP Address Destination •  Active Neighbors ext Hop IP Address •  Expiration time Source ctive Neighbors Source Active Upstream Neighbors xpiration time

Each time a route entry is used to transmit data, the expiration time is updated to current time + active_route_timeout me a route entry is used to transmit data, the expiration •  Route entries may be updated if a route with greater sequence updated tonumber or smaller hopcount is discovered • 

ntries may be updated if a route with greater sequence 32 r or smaller hopcount is discovered Mobile and Wireless Networking 2013 / 2014

Path Maintenance •  •  • 

Movement of nodes not along active path does not trigger protocol action If source node moves, it can reinitiate route discovery When destination or intermediate node moves, node upstream of break sends unsolicited RREP to all active upstream neighbors •  • 

•  •  • 

∞ metric, incremented Seq# Used to flush stale routes

RREP is propagated to their active neighbors, and so on back to source Source can reinitiate route discovery if route is still needed RREQs for reinitiated route discovery contain destination sequence number of one greater than last known number


Maintenance (cont.) PathPath Maintenance (cont.) Maintenance (example) 3’ 3’

3’ 3’

33 11

11

22

22

Destination Destination Source Source

44

The The Initial Initial Route Route •  •  •  • 

Destination Destination Source Source

44

After After Route Route Reconstruction Reconstruction

Node 3 moves to new location 3’ Node 33 moves Node moves to to new new location location 3’ 3’

Node 2 notices loss of link, sends link failure to Node 1 Node1 forwards link failure to Source Node 22 notices Node notices loss loss of of link, link, sends sends link link failure failure to to Node Node 11 Source reinitiates route discovery, finds new route through Node 4

Node Node 11 forwards forwards link link failure failure to to Source Source

34 Source throu Source reinitiates reinitiates route route discovery, discovery, finds finds new new route route thro Node Node 44 Mobile and Wireless Networking 2013 / 2014

Local Connectivity Management •  •  •  • 

Node must periodically hear from active neighbors to know they are still within range Eavesdrop on neighbor transmissions If no other transmissions within hello_interval, broadcast Hello packet Failure to hear from a neighbor for (1+ allowed_hello_loss) * hello_interval

indicates loss of link


Optimizations • 

Expanding Ring Search • 

• 

RREP generated by intermediate node • 

•  • 

by limiting TTL at first attempt, and increasing it at successive attempts. only if Seq# for route to destination ≥ Dest_Seq# of RREQ

Maintaining Local Connectivity by means of layer 2 info. Local Repair • 

node upstream of link failure tries to find a new route to destination by sending a RREQ (with reduced TTL, and incremented Dest_Seq#)


DSR (Dynamic Source Routing) • 

Similar to AODV

• 

Big difference: DSR uses Source Routing •  AODV relies on storing routing table entries in intermediate nodes  RREQ and RREP carry addresses of all intermediate nodes • 

See [Schiller] for details


A plethora of ad hoc routing protocols Flat  

proactive   DSDV – Destination Sequenced Distance Vector   FSLS – Fuzzy Sighted Link State   FSR – Fisheye State Routing   OLSR – Optimised Link State Routing Protocol   TBRPF – Topology Broadcast Based on Reverse Path Forwarding

 

reactive    

DSR – Dynamic Source Routing AODV – Ad hoc On demand Distance Vector

Hierarchical        

CGSR – Clusterhead-Gateway Switch Routing HSR – Hierarchical State Routing LANMAR – Landmark Ad Hoc Routing ZRP – Zone Routing Protocol

Geographic position assisted        

DREAM – Distance Routing Effect Algorithm for Mobility GeoCast – Geographic Addressing and Routing GPSR – Greedy Perimeter Stateless Routing LAR – Location-Aided Routing 38 Mobile and Wireless Networking 2013 / 2014