Quality of Service Support

1
Quality of Service Support
2
QOS in IP Networks

• IETF groups are working on proposals to provide
  QOS control in IP networks, i.e., going beyond
  best effort to provide some assurance for QOS
• Work in Progress includes RSVP, Differentiated
  Services, and Integrated Services
• Simple model for sharing and congestion
  studies

3
Principles for QOS Guarantees

• Consider a phone application at 1Mbps and an FTP
  application sharing a 1.5 Mbps link.
• bursts of FTP can congest the router and cause
  audio packets to be dropped.
• want to give priority to audio over FTP
• PRINCIPLE 1 Marking of packets is needed for
  router to distinguish between different classes
  and new router policy to treat packets accordingly

4
Principles for QOS Guarantees (more)

• Applications misbehave (audio sends packets at a
  rate higher than 1Mbps assumed above)
• PRINCIPLE 2 provide protection (isolation) for
  one class from other classes
• Require Policing Mechanisms to ensure sources
  adhere to bandwidth requirements Marking and
  Policing need to be done at the edges

5
Principles for QOS Guarantees (more)

• Alternative to Marking and Policing allocate a
  set portion of bandwidth to each application
  flow can lead to inefficient use of bandwidth if
  one of the flows does not use its allocation
• PRINCIPLE 3 While providing isolation, it is
  desirable to use resources as efficiently as
  possible

6
Principles for QOS Guarantees (more)

• Cannot support traffic beyond link capacity
• Two phone calls each requests 1 Mbps
• PRINCIPLE 4 Need a Call Admission Process
  application flow declares its needs, network may
  block call if it cannot satisfy the needs

7
Summary
8
Scheduling And Policing Mechanisms

• Scheduling choosing the next packet for
  transmission
• FIFO
• Priority Queue
• Round Robin
• Weighted Fair Queuing
• We had a lecture on that!

9
(No Transcript)
10
Discussion of RED

• Advantages
• Early drop
• TCP congestion
• Fairness in drops
• Bursty versus non-Bursy
• Disadvantages
• Many additional parameters
• Increasing the loss

11
Policing Mechanisms

• (Long term) Average Rate
• 100 packets per sec or 6000 packets per min??
• crucial aspect is the interval length
• Peak Rate
• e.g., 6000 p p minute Avg and 1500 p p sec Peak
• (Max.) Burst Size
• Max. number of packets sent consecutively, ie
  over a short period of time
• Units of measurement
• Packets versus bits

12
Policing Mechanisms

• Token Bucket mechanism, provides a means for
  limiting input to specified Burst Size and
  Average Rate.
• Bucket can hold b tokens
• tokens are generated at a rate of r token/sec
• unless bucket is full of tokens.
• Over an interval of length t, the number of
  packets that are admitted is less than or equal
  to (r t b).

13
Token bucket example
parameters b5 r3
14
Integrated Services

• An architecture for providing QOS guarantees in
  IP networks for individual application sessions
• relies on resource reservation, and routers need
  to maintain state info (Virtual Circuit??),
  maintaining records of allocated resources and
  responding to new Call setup requests on that
  basis

15
Call Admission

• Session must first declare its QOS requirement
  and characterize the traffic it will send through
  the network
• R-spec defines the QOS being requested
• T-spec defines the traffic characteristics
• A signaling protocol is needed to carry the
  R-spec and T-spec to the routers where
  reservation is required
• RSVP is a leading candidate for such signaling
  protocol

16
RSVP request (T-Spec)

• A token bucket specification
• bucket size, b
• token rate, r
• the packet is transmitted onward only if the
  number of tokens in the bucket is at least as
  large as the packet
• peak rate, p
• p gt r
• maximum packet size, M
• minimum policed unit, m
• All packets less than m bytes are considered to
  be m bytes
• Reduces the overhead to process each packet
• Bound the bandwidth overhead of link-level
  headers

17
Call Admission

• Call Admission routers will admit calls based on
  their R-spec and T-spec and base on the current
  resource allocated at the routers to other calls.

18
Integrated Services Classes

• Guaranteed QOS this class is provided with firm
  bounds on queuing delay at a router envisioned
  for hard real-time applications that are highly
  sensitive to end-to-end delay expectation and
  variance
• Controlled Load this class is provided a QOS
  closely approximating that provided by an
  unloaded router envisioned for todays IP
  network real-time applications which perform well
  in an unloaded network

19
R-spec

• An indication of the QoS control service
  requested
• Controlled-load service and Guaranteed service
• For Controlled-load service
• Simply a Tspec
• For Guaranteed service
• A Rate (R) term, the bandwidth required
• R ? r, extra bandwidth will reduce queuing delays
• A Slack (S) term
• The difference between the desired delay and the
  delay that would be achieved if rate R were used
• With a zero slack term, each router along the
  path must reserve R bandwidth
• A nonzero slack term offers the individual
  routers greater flexibility in making their local
  reservation
• Number decreased by routers on the path.

20
QoS Routing Multiple constraints

• A request specifies the desired QoS requirements
• e.g., BW, Delay, Jitter, packet loss, path
  reliability etc
• Two type of constraints
• Additive e.g., delay
• Maximum (or Minimum) e.g., Bandwidth
• Task
• Find a (min cost) path which satisfies the
  constraints
• if no feasible path found, reject the connection

21
Example of QoS Routing
D 24, BW 55
D 30, BW 20
A
B
D 5, BW 90
D 14, BW 90
D 5, BW 90
D 5, BW 90
D 7, BW 90
D 10, BW 90
D 5, BW 90
D 3, BW 105
Constraints Delay (D) lt 25, Available Bandwidth
(BW) gt 30
22
Differentiated Services

• Intended to address the following difficulties
  with Intserv and RSVP
• Scalability maintaining states by routers in
  high speed networks is difficult sue to the very
  large number of flows
• Flexible Service Models Intserv has only two
  classes, want to provide more qualitative service
  classes want to provide relative service
  distinction (Platinum, Gold, Silver, )
• Simpler signaling (than RSVP) many applications
  and users may only want to specify a more
  qualitative notion of service

23
Differentiated Services

• Approach
• Only simple functions in the core, and relatively
  complex functions at edge routers (or hosts)
• Do not define service classes, instead provides
  functional components with which service classes
  can be built

24
Edge Functions at DiffServ (DS)

• At DS-capable host or first DS-capable router
• Classification edge node marks packets according
  to classification rules to be specified (manually
  by admin, or by some TBD protocol)
• Traffic Conditioning edge node may delay and
  then forward or may discard

25
Core Functions

• Forwarding according to Per-Hop-Behavior or
  PHB specified for the particular packet class
  such PHB is strictly based on class marking (no
  other header fields can be used to influence PHB)
• BIG ADVANTAGE
• No state info to be maintained by routers!

26
Classification and Conditioning

• Packet is marked in the Type of Service (TOS) in
  IPv4, and Traffic Class in IPv6
• 6 bits used for Differentiated Service Code Point
  (DSCP) and determine PHB that the packet will
  receive
• 2 bits are currently unused

27
Classification and Conditioning

• It may be desirable to limit traffic injection
  rate of some class user declares traffic profile
  (eg, rate and burst size) traffic is metered and
  shaped if non-conforming

28
Forwarding (PHB)

• PHB result in a different observable (measurable)
  forwarding performance behavior
• PHB does not specify what mechanisms to use to
  ensure required PHB performance behavior
• Examples
• Class A gets x of outgoing link bandwidth over
  time intervals of a specified length
• Class A packets leave first before packets from
  class B

29
Forwarding (PHB)

• PHBs under consideration
• Expedited Forwarding departure rate of packets
  from a class equals or exceeds a specified rate
  (logical link with a minimum guaranteed rate)
• Assured Forwarding 4 classes, each guaranteed a
  minimum amount of bandwidth and buffering each
  with three drop preference partitions

30
Differentiated Services Issues

• AF and EF are not even in a standard track yet
  research ongoing
• Virtual Leased lines and Olympic services are
  being discussed
• Impact of crossing multiple ASs and routers that
  are not DS-capable

31
DiffServ Routers

DiffServ Edge Router
Classifier
Meter
Policer
Marker

DiffServ Core Router
PHB
PHB
Select PHB
Local conditions
PHB
PHB
Extract DSCP
Packet treatment
32
IntServ vs. DiffServ
IP
IntServ network
DiffServ network
"Call blocking" approach
"Prioritization" approach
33
Comparison of Intserv Diffserv Architectures
34
Comparison of Intserv Diffserv Architectures
35
Diffserv Theoretical Model
36
Basic Theoretical Model

• Single FIFO queue.
• Bounded capacity holds up to B packets
• All packets have same size
• Packet Arrival arbitrary
• Packet Send 1 packet/time unit
• Actions
• Non-Preemptive model accept or reject
• Preemptive model also preempt

37
Packet Values

• Goal
• Each packet has an intrinsic value
• maximize the total value of packet sent!
• Cheap and expensive packets (two values)
• low value of 1 and high value of ?
• Continuous packet values
• any value in 1,?

38
Competitive Analysis

• Analysis for online algorithms
• For a given sequence S VA(S) / Vopt(S)
• Competitive Ratio MINS VA(S) / Vopt(S)
• Worse case guarantee

39
Non-Preemptive Policies

• Fixed Partition(x)
• At most xB low value and (1-x)B high value.
• Flexible Partition (x)
• At most xB low value and any high value.
• Round Robin(x)
• Like fixed partition.
• send x low and (1-x) high fractional!
• Simulate it using FIFO queue.

40
Implementing Round Robin

• Implementation
• Maintain two variables
• high
• low
• If low packet arrives tests low 1 lt xB
• IF YES ACCEPT
• IF NO REJECT
• High packets the same
• Sending
• low low x
• high high (1-x)
• Main observation
• once a packet is accepted it will be sent
  eventually.
• Sending order not important!

41
Analysis of Round Robin

• Consider the case that all packet values are 1.
• Claim
• For any input sequence
• The number of packet a buffer of size B/2 accepts
• is at least half of a buffer of size B
• Let x ½
• Consider Low and High packets separately
• RR(½)
• Accepts at least half High and half Low
• Benefit at least half

42
Preemptive Policies

• Greedy
• Always accept if the buffer is not full
• Preempt a low value packet to accept a high one
• COMPETITIVE RATIO 2
• ?-Preemptive
• Drop from the head packets with total value ?/?
• Active queue management (AQM)

43
Preemptive Model ?1/2 -Preemptive

• We consider ?1/2-Preemptive Policy
• There are two packet values 1 and ?
• For ?9 each high value packet preempts 3 low
  value packets (pro-active preemptions)

44
?1/2-Preemptive Theorem

• Claim 1 VA(Slow) ?VOPT(Slow) 1/?1/2
  VOPT(Shigh)
• Claim 2 VA(Shigh) ? VOPT(Shigh) 1/?1/2
  VOPT(Shigh)
• Theorem VA(S) ? VOPT(S) 2/?1/2
  VOPT(S)

45
Optimal Offline

• Process the packet in decreasing order of value.
• Accept a packet if possible.
• otherwise reject
• Two values
• Maximizes the number of high value packets
• Given a buffer of size B
• Maximizes the total number of packets
• Using the remaining buffer space.

46
Proof Outline Claim 2

• We partition the schedule to intervals
• Intervals ends when the buffer is empty.
• Overloaded intervals some high value packet is
  lost and only high value packets are scheduled.
• Underloaded intervals no high value packet is
  lost

47
Proof (Claim 1)

• We show VA(Slow) ?VOPT(Slow) 1/?1/2 VA(Shigh)
• Low packet loss overflow Preemption
• Low packet lost in overflow
• Opt also lost a packet.
• Low packet preempted by a high packet
• Value of high ?
• Preempted ?1/2
• Value is 1/?1/2 V(high)
• Recall VA(Shigh) ? VOPT(Shigh)

48
Proof Outline (Claim2)

• We divide the HIGH packet loss into two subsets
• The packets lost by OPT (easy case)
• The packets scheduled by OPT

49
Proof Outline (Claim 2)

• Observation 1
• When some high value packet is lost the buffer is
  full of high value packets

50
Proof Outline (Claim 2)
Observation 2 If there are at least B/?1/2 high
value packets in the buffer then the next packet
to be scheduled is a high value packet.
51
Proof Outline (Claim 2)

• Observation 1 ? The length of an overloaded
  interval is at least B
• Observation 2 ? An optimal offline policy could
  have scheduled at most B/?1/2 additional high
  value packets
• The ratio between the additional loss and the
  benefit of the overloaded interval is bounded by
  1/?1/2
• VA(Shigh) ? VOPT(Shigh) 1/?1/2 VOPT(Shigh)

52
Lower bound (Non-Preemptive)

• Scenario
• B low value packets
• maybe B high value packets
• Online accepts xB low value
• Case I only low values
• Online xB Offline B
• Case II Both low and high value
• online xB (1-x) aB offline aB
• Competitive ratio ? a/(2a-1)
• For large values of a we have a/(2a-1) ? ½

53
Lower bound Preemptive model

• Scenario
• B low value packets
• For zB time units
• one high value packet arrives each time unit
• Maybe B high value packets
• Let zB be the time the Online sends the last low
• (1) No more packets arrive
• (2) B high value packets arrive
• Online Benefit (1) zB z?B (2) zB ?B
• Offline Benefit (1) B z?B (2) z?B ?B
• Solving for best z gives a lower bound (about 0.8)

54
Fixed vs. Flexible Partition
55
Summary of Results Non-preemptive
Two values
Multiple Values
Competitive ratio 1/(2 ln a) 1/(1 ln a)
Policies cont RR Impossibility
56
Summary of Results Preemptive
Multiple Values
Policies Greedy Better Than G Impossibility
Competitive ratio ½ 1/(1.98..) 0.8
2 Values
Policies ?1/2-Preemptive Impossibility
Competitive ratio 1-2/?1/2 1-1/(2?1/2)