Transcript Click

QoS issues and solutions provided
by DiffServ, IntServ, MPLS, and
802.11e
By Khaleel Ali
COMP 529
Presentation outline
• Issue
– Quality of Service (QoS) Problems of internet
• Solutions for network
– Integrated services or InterServ
– Differentiated services or DiffServ
– Multi-Protocol Label Switching Networks or MPLS
• Last mile
– 802.11e
What is QoS?
The quality of voice service is specified by:
• Grade of service (GoS)
– is the probability of a call being blocked or delayed
for more than a specified interval.
• Traffic on a network
• Quality of service (QoS)
– probability of the network meeting a given traffic
contract.
– On telephone it refers to lack of noise and tones on
the circuit, appropriate loudness levels.
How is QoS measured
• Three measurements are used to determine the
quality of service
– Dropped packets
• Percentage of packets lost as they move from end to end.
– Jitter
• Unpredictable variable in delay caused by congestion.
– Latency
• Time it takes a signal to move through a unit in test.
• Low latency must be designed into a network from the
start and it can not be changed later.
Latency
Time it takes from a packet to get from point A to B
Jitter
Packets are queued when network is busy. The packet coming out of
port B will vary in the amount of delay. This variation is inconsistent
and unpredictable.
Dropped Packets
Dropped Packets (Cont..)
QoS problem: The internet
•
•
•
•
When the Internet was created, there was no need for a QoS.
Entire internet ran on a "best effort" system.
Internet was never intended to be for real time applications.
Type of service (ToS) exists in IPv4 and IPv6 but has not been
utilized by networks.
• IP transport is unreliable because as packets travel from origin to
destination they can be
– Dropped
• Routers buffer was full when packet arrived.
• Depends on state of a network.
– Delayed
• Routers in middle had long queues.
• Packet took a longer router to avoid congestion.
– Out of order
• Packets took different routes with different delays.
– Error
• packets are misdirected, or combined together, or corrupted, while in route.
ToS for IPv4 and IPv6
Internet problem: Applications
requiring QoS
A traffic contract guarantees for a network ability of
• Performance
• Throughput
• Latency
Examples of applications that require such guarantee
• Streaming multimedia may require guaranteed throughput
• IP telephony may require strict limits on jitter and delay
• Dedicated link emulation requires both guaranteed
throughput and imposes limits on maximum delay
• Safety-critical application, such as remote surgery may
require a guaranteed level of availability (this is also called
hard QoS).
Internet myth about Bandwidth
• It is more cost effective to "buy" 200% more
bandwidth than a network requires than it is to
worry about QoS.
– Standards are being developed that will change this.
– Internet is still growing and bandwidth alone can’t
provide solutions needed due to its growth.
• Over designing a network and throwing
bandwidth at the QoS problem is only a
temporary fix -- not a solution.
• The “if you build it they will come” phenomenon.
The faster the network, the more user traffic it
will have.
Obtaining QoS
The following provide QoS guarantees:
1. Integrated Services(InterServ)
•
Application that requires some kind of guarantees has to make an
individual reservation.
2. Differentiated Services(DiffServ)
•
Categorizes traffic into different classes, also called class of service
(CoS), and applies QoS parameters to those classes.
3. Multiprotocol Label Switching(MPLS)
•
Tagging each packet to determine priority.
4. 802.11e
•
•
Packets carry their priority code.
For wireless last mile.
Integrated Services(InterServ)
• Is a system that attempts to guarantee QoS on
networks by reserving resources.
• Is a layer 3 implementation and works over IP.
• Is ideal for end-to-end delay and jitter sensitive
applications, such as Voice and Video.
• Every router in the system implements IntServ.
• Every application that requires some kind of
guarantees has to make an individual reservation.
• "Flow Specs" describe what the reservation is for.
• InterServ uses ReSerVation protocol (RSVP) to make
reservations.
Flow Specs
• There are two parts to a flow spec.
• Traffic SPECification (TSPEC)
– What does the traffic look like
• TSPECs specifies mean data rate, the token rate, bucket
depth, nominal MAC frame size and maximum service
interval.
– EX, A video with a refresh rate of 75 frames per second,
with each frame taking 10 packets, might specify a token
rate of 750Hz, and a bucket depth of only 10.
– A conversation would need a lower token rate, but a much
higher bucket depth.
– There are often pauses in conversations, so they can make do
with less tokens by not sending the gaps between words and
sentences.
RSPEC
• Request SPECification (RSPEC)
– What guarantees does the traffic need.
• Best effort - no reservation is needed.
• Controlled Load - mirrors the performance of a lightly
loaded network. Both delay and drop rate are fairly
constant at the desired rate.
• Guaranteed – absolutely bounded service, the delay is
promised never to go above a desired amount, and packets
never dropped, provided the traffic stays within spec.
ReSerVation protocol (RSVP)
• RSVP is used by routers to deliver QoS requests to all nodes along the
path of the flow and to establish and maintain state to provide the
requested service.
• RSVP requests resources in only one direction.
• To make reservations, the RSVP daemon communicates with two local
decision modules:
– Admission control - determines whether the node has sufficient available
resources to supply the requested QoS.
– Policy control - determines whether the user has administrative
permission to make the reservation.
• If either check fails, the RSVP program returns an error notification to
the application.
• If both checks succeed, the RSVP daemon sets parameters in a packet
classifier and packet scheduler to obtain the desired QoS.
– Packet classifier - determines the QoS class for each packet.
– Packet scheduler - orders packet transmission to achieve the promised
QoS for each stream.
RSVP Daemon
InterServ
How does InterServ work?
• An RSVP sender sends PATH messages downstream
towards receivers.
• Each node along the unicast or multicast path stores
information about the flow, based on the sender's TSpec
(bit rate, token bucket size).
• Upon receiving the PATH message, the receiver sends a
reservation (Resv) request message back upstream to the
sender, reversing the path taken by the PATH messages.
• The Resv message indicates the desired QoS with an
RSpec.
• The routers between the sender and listener have to
decide if they can support the reservation being
requested.
How does InterServ work? (cont..)
• If they cannot they send a reject message to let the
listener know about it.
• If they accept the reservation they have to carry the
traffic and store the nature of the flow.
• The Resv message creates and maintains soft state in
the nodes along the path.
– If not refreshed every 30 seconds, the soft state ages out and
is deleted.
– This allows RSVP to adjust and alter the path between RSVP
end systems to recover from route changes
– If also prevents router resources from being tied up due to
receivers that quietly vanish.
– This solves the problem if either the sender or the receiver
crash or are shut down incorrectly without first cancelling the
reservation.
How does InterServ work? (cont..)
• InterServ Uses a token bucket filter method.
• Tokens are sent on the InterServ network using the TSPECs
of reservations.
• Stations have a token bucket which slowly fills up with
tokens, arriving at some constant rate.
• Every packet has to have a token or it cannot be sent.
• The rate at which tokens arrive dictates the average rate of
traffic flow.
• The depth of the bucket dictates how 'bursty' the traffic is
allowed to be.
• The bucket depth should be sufficient to accommodate the
'burst' associated with sending an entire frame all at once.
InterServ Network
InterServ Positives
•
•
•
•
•
Strict bandwidth guarantee - for Voice and Video
traffic—Includes end-to-end maintenance, so a loss
of bandwidth in the core is handled appropriately.
Admission Control —Admitting calls based on
available resources. So rather than having 12 calls
with bad voice quality, ensuring that at least 10 calls
go through.
Failure notification- RSVP notifies that sender of
RSPEC why exactly the reservation was denied.
Cisco routers have demonstrated the ability to handle
more than 10,000 RSVP flows with minimal CPU
and Memory impact.
RSVP runs over IP, both IPv4 and IPv6.
Problems of InterServ
•
•
•
•
All applications need to make reservations.
Which means many states must be stored in each router.
As a result, IntServ works on a small-scale.
As we scale up to a system the size of the Internet, it is
difficult to keep track of all of the reservations.
– Many users might not even be able to make reservations
because bandwidth hasn’t increased.
• Hence IntServ is not very popular.
• Making reservation for best effort services or bursty
traffic.
– Waste of network resources
– Policing network, introducing more states in routers.
Differentiated Services (DiffServ)
• Is a method of trying to guarantee quality of service on large
networks such as the Internet.
• Is a layer 3 implementation and works over IP.
• DiffServ deals with bulk flows of data rather than single flows
and single reservations.
– A single negotiation will be made for all of the packets from a single
ISP or a single university.
• The contracts resulting from these negotiations are called
"service level agreements” or SLA.
• These service level agreements will specify
– classes of traffic will be provided
– guarantees that are needed for each class
– And how much data will be sent for each class.
• Example of SLA, we pay ISP monthly fees to get upto 5mbs
surfing, 600kbs downloading and 42 uploading for best effort.
How does DiffServ work?
• Sender marks the ToS octet of IPv4 or IPv6 which is determines a
particular forwarding treatment, or per-hop behavior, at each network
node.
– delay-bound
– jitter-bound
– bandwidth
• The DiffServ Code Point (DSCP) maps the packet to a particular
mapping behavior (PHB) for processing by DS-compliant router.
• Per-hop behaviors (PHBs) are applied to the traffic at a network
ingress point (network border entry) according to pre-determined
policy criteria.
• The PHB provides a particular service level (bandwidth, queueing,
and dropping decisions) in accordance with network policy.
• The traffic may be marked at this point, and routed according to the
marking, then unmarked at the network egress (network border exit).
DiffServ Code Point - DSCP
Each class specifies a buffer and bandwidth. So if class #2 has to be
dropped the routers will start dropping AF23 first then moving up to
AF22.
IPv4 ToS Byte
DSCP
DiffServ Network Router Queue
DiffServ Network
• "DiffServ cloud" is a collection of DiffServ routers.
• Individual routers in DiffServ cloud give the highest priority to
the packets with the highest value in ToS.
• The SLA establishes the policy criteria, and defines the traffic
profile.
• If there is so much traffic that it breaches the service level
agreement, then the sender may be liable for fines, according to
the details of the contract.
• There may also be a discard policy on the frequencies with
which each type of packet is discarded if the router runs out of
buffer space.
• The traffic may be split into Gold, Silver, and Bronze classes. In
each router, Gold traffic takes precedence over Silver traffic,
which takes precedence over Bronze.
DiffServ over Multiple Domains
Advantages of DiffServ
• One advantage of DiffServ, is that all the
policing and classifying is done at the
boundaries between DiffServ clouds.
• The core of the Internet, routers can get on with
doing the job of routing, and not care about the
complexities of collecting payment or enforcing
agreements.
DiffServ Disadvantages
• DiffServ is a simply mechanism for deciding which
packets to delay or drop at the expense of others.
• If DiffServ is working by dropping packets selectively,
traffic on the link is close to saturation.
• Any further increase in traffic will result in Bronze
services being taken out altogether.
• Since Internet traffic is highly bursty, this is almost
certain to happen on a regular basis if traffic on a link is
near the limit at which DiffServ becomes needed.
• For this reason, DiffServ will always be inferior to
adding sufficient network capacity to avoid packet loss
on all classes of traffic.
Disadvantages of DiffServ
• If a packet crosses two or more DiffServ clouds before reaching its
destination a Gold packet may be another's Bronze.
– Enforcing standardised policies across networks require complexity to their
already complex peering agreements.
– Also success of internet is highly credited to it’s lack of decision making
and simple routing.
• A glut of fibre capacity is far easier and cheaper to add than to
employ elaborate DiffServ policies as a way of increasing
customer satisfaction.
• DiffServ is for most ISPs only a way of rationing customer
network utilisation to allow greater overbooking of their existing
capacity.
– Use DiffServ tools to suppress peer-to-peer traffic, because of its ability to
saturate customer links indefinitely.
– ISP's business model relies on 1%-10% link utilization for most customers.
Multiprotocol Labeling Switching
(MPLS)
• Multiprotocol Label Switching (MPLS) is similar to DiffServ, as it
also marks traffic at ingress boundaries in a network, and unmarks
at egress points.
• MPLS-enabled routers are called a Label Switching Router (LSR)
• The MPLS technology can operate over various link-level
technologies, which includes packet-over-Sonet, frame relay,
ATM, Ethernet, and token ring.
• MPLS combines layer 2 switching technology with layer 3
network layer services while reducing the complexity and
operational costs.
• The percentage of service provider respondents indicating MPLS
deployment in some part of their network jumped from 47% in
2002 to 79% in 2003.
How does MPLS work?
• First an MPLS tunnel (also called LSP — Label Switched
Path) is set up between two routers.
– A software can be used to configure all routers on network.
• The entry point of an MPLS tunnel is called the ingress
router, the end point is called the egress router.
• Egress and ingress routers are called PE (Provider Edge).
• Devices that function only as transit routers are called P
(Provider) routers.
• The job of a P router is significantly easier than that of a
PE router, so they can be less complex and may be more
dependable because of this.
How does MPLS work?
How does MPLS work?
• Unlike DiffServ, MPLS markings are primarily designed to determine
the next router hop.
• In MPLS each router makes an independent forwarding decision for that
packet.
– Packet headers contain information to choose the next hop.
• The assignment of a particular packet is done just once, at the entrance.
• At the first hop, the router makes a forwarding decision based on the
destination address.
• Then router determines the appropriate label value, attaches the label to
the packet and forwards it to the next hop.
• At the next hop, the router uses the label value as an index into a predetermined table that specifies the next hop and a new label.
• The LSR attaches the new label, then forwards the packet to the next
hop.
• The route taken by an MPLS-labeled packet is called the Label Switched
Path (LSP).
Wireless: Some background
• Wired LAN is being replaced by Wireless due to
mobility.
• Wireless Is based on CSMA
– Listen before talk
• Wireless LAN has seen huge increases in
– Throughput about 125mbits
– Stability (interference management)
– Range (about 1 mile, more with WiMax)
• Wireless problems
– Collisions
• Hidden terminal even with RTS/CTS
– QoS
• 802.11e is designed to solve the QoS problem in
wireless network.
802.11e
• 802.11 defines two ways a wireless network can send
data
– Distributed Coordination Function (DCF)
– Point Coordination Function (PCF)
• 802.11e redefines DCF and PCF and enables QoS in
them by giving priority to real time applications.
• The new methods of sending data are
– Old DCF is now Enhanced Distributed Coordination Function
(EDCF)
– Old PCF is now Hybrid Coordination Function (HCF)
• Different methods give different performance benefits.
802.11 DCF Overview
• A station detects that the medium is free and starts
decrementing it’s back-off counter.
– The back off counter comes from a contention windows (CW)
• Contention window is maintained for collisions and increases if
transmission fails (binary exponential).
– Back-off counter starts to decrements if the medium has been
free for the DIFS period (DCF Inter-Frame Space).
• If another station has selected a lower number it starts
transmitting first.
– The station will hold its current back off count.
• When the medium clears clear again the station waits
for DIFS period and resumes counting.
• There are no transmit guarantee.
• All data is treated equally and burst data can choke
video/audio.
802.11 DCF simulation
802.11 DCF negatives
• In DCF mode, all the stations compete for the resources and channel with the
same priorities.
• There is no differentiation mechanism to guarantee bandwidth, packet delay and
jitter for high-priority stations or multimedia flows.
• SIMULATION
• STAs send three types of traffic (audio, video and background traffic) to each
other.
– Audio at 8KB/s flow.
– The video sending rate is 80KB/s with a packet size equal to 1280 bytes.
– Background traffic is 128 KB/s, using a 1600 bytes packet size.
• We vary the load rate from 9.6% to 90% by increasing the number of STAs
from 2 to 18.
• The number of STAs is up to 10.
– throughput of audio is about 7.8 KB/s;
– throughput of video is about 78KB/s;
– throughput of background is about 125KB/s; and delay is lower than 4ms.
• The number of STAs is larger than 10,
– They all experience the same delay.
• There is no way to guarantee the QoS requirements for high-priority audio and
video traffic unless admission control is used.
EDCF or EDCA overview
• The 802.11e provides two new mechanisms for resolving contention that
enables QoS:
– EDCA (Enhanced Distributed Channel Access) or EDCF
– and HCCA (Hybrid Controlled Channel Access) of HCF.
• EDCA improves on DCF by giving higher-priority traffic an advantage during
contention.
• Instead of waiting a DIFS period before transmitting after the back-off period
expires, higher-priority traffic can attempt to transmit only after a PIFS (point
coordination function interframe space) period and associated back-off time.
• Using the EDCA scheme, nodes that offer high-priority traffic, have a higher
probability of gaining channel access than the nodes offering lower-priority
traffic.
• Granularity within priority slots is possible by carrying this technique to
exponential back-off. Higher-priority devices will back off at a slower rate (i.e.,
fewer slot times) than lower-priority devices, giving higher-priority devices
another contention advantage.
• EDCA coexists with DCF-based devices because all DCF devices appear as
low-priority (DIFS) nodes. Using EDCA, an access point can support more
VoIP phones than DCF, for a given voice quality.
802.11e: EDCF
• The EDCF is designed for the contention-based prioritized QoS support.
• Each QoS-enhanced STA (QSTA) has 4 queues (ACs), to support 8 user
priorities (UPs)
– One or more UP is mapped to same queue.
– Is done to reduce overhead of maintaining queues.
• Each AC queue works as an independent DCF STA and uses its own backoff
parameters.
• Two main methods are introduced to support service differentiation:
– The first one is to use different InterFrame Space (IFS) sizes for different ACs.
– Arbitration IFS (AIFS) is used in EDCF, instead of DIFS in DCF.
– AIFS is also called CWOffset
• CWMin can be selected per TC.
• CWMax is the same for all TC.
– It provides priority mechanism for each TC
• High priority AC
• The AIFS [AC] is determined by AIFS [AC] = AIFSN [AC] · SlotTime + SIFS,
– Arbitration inter frame spacing number (AIFSN) is defined as either 1 or 2.
– When AIFSN = 1, high priority queues AC1, AC2 and AC3 have AIFS value equal to
PIFS.
– The low priority queue AC0 has AIFS value of DIFS.
802.11e: EDCF (cont..)
• When a frame arrives at an empty AC queue and the medium has
been idle longer than AIFS [AC] + SlotTime, the frame is
transmitted immediately.
• If the channel is busy, the arriving packet in each AC has to wait
until the medium becomes idle.
• So the AC with the smaller AIFS has the higher priority.
• The second method consists in allocating different CW sizes for
different ACs.
• Assigning a short CW size to a high priority AC ensures that in
most cases, high-priority AC is able to transmit packets ahead of
low-priority one.
• If the backoff counters of two or more parallel ACs in one QSTA
reach zero at the same time, a scheduler inside the QSTA will
avoid the virtual collision by granting the EDCF-TXOP to the
highest priority AC.
– The other colliding ACs will enter a backoff process and double the CW
sizes as if there is an external collision.
– This way EDCF is supposed to improve the performance of DCF under
congested conditions.
802.11e: EDCF
802.11e: EDCF burst
• To improve the throughput performance, EDCF packet bursting
can be used in 802.11e.
• That once a QSTA has gained an EDCF-TXOP, it can be allowed
to send more than one frame without contending for the medium.
• The QSTA can send multiple frames as long as the total access
time does not exceed the TXOPLimit bound determined by QAP.
• To ensure that no other QSTA interrupts the packet bursting, SIFS
is used between packet bursts.
• If a collision occurs, the EDCF bursting is terminated.
• This mechanism can reduce the network overhead and increase
throughput by multiple transmissions using SIFS and burst
acknowledgements.
• Bursting may also increase the jitter, so TXOPLimit should not be
longer than the time required to transmit the largest data frame.
802.11e: EDCF negatives
• Simulation results show that although internal
collision rates are low for EDCF.
• External collisions between the same priorities in
different QSTAs are still high.
• Also as network utilization increases low priority
data is choked by higher priority.
EDCF simulation
802.11 PCF
• Gives priority to Time-bounded services to get priority access to
the wireless medium.
• Is Coordinated by a station called Point Coordinator (PC).
• PCF uses a centralized polling scheme, which requires the AP as a
point coordinator (PC).
• If a network is set up with PCF-enabled, the channel access time
is divided into periodic intervals named beacon intervals.
• Beacon intervals is composed of
– Contention Free Period (CFP)
• the PCF is used for accessing the medium
– and a Contention Period (CP)
• the DCF is used for accessing the medium
• The time used by the PC to generate beacon frames is called
target beacon transmission time (TBTT).
• In the beacon, the PC denotes the next TBTT and broadcasts it to
all the other STAs in the BSS.
Beacon frame
802.11 PCF polling
• During the CFP, the PC maintains a list of registered STAs and polls
each STA according to its list.
• When an STA is polled, it gets the permission to transmit data frame.
• If the PC received no response from a polled station after waiting for
PIFS, it polls the next station.
• The PC continues with polling other stations until the CFP expires.
• Since every STA is permitted a maximum length of frame to transmit,
the maximum CFP duration for all the STAs can be known and decided
by the PC, which is called CFP_max_duration.
• In order to ensure that no DCF STAs are able to interrupt the operation
of the PCF, a PC waits for a PCF InterFrame Space (PIFS), which is
shorter than DIFS, to start the PCF.
• Then, all the other STAs set their NAVs to the values of
CFP_max_duration time, or the remaining duration of CFP in case of
delayed beacon.
• A specific control frame, called CF-End, is transmitted by the PC as the
last frame within the CFP to signal the end of the CFP.
• During the CP, the DCF scheme is used, and the beacon interval must
allow at least one DCF data frame to be transmitted.
802.11 PCF negatives
1. All the communications between two STAs have to go through the
AP, channel bandwidth is wasted.
2. The cooperation between CP and CFP modes may lead to
unpredictable beacon delays.
– The PC schedules the beacon at TBTT for the CFP interval, and then the
beacon can be transmitted when the medium has been found idle for an
interval of time longer than a PIFS.
– Depending on whether the wireless medium is idle or busy around the
TBTT, the beacon frame may be delayed.
– In the current 802.11 legacy standard, STAs are allowed to start their
transmissions even if the frame transmission cannot terminate before the
upcoming TBTT.
3. Third, the transmission time of a polled STA is difficult to control.
A polled STA is allowed to send a frame of any length between 0
and 2346 bytes, which introduce the variation of transmission
time.
1. The transmission time is can’t be predicted by Ap since PHY rate changes.
2. This makes a barrier for the AP to provide guaranteed QoS service for other
STAs in the polling list during the rest of the CFP.
HCCA or HCF Overview
• Nodes with priority traffic request to be added to a polling list
managed by the access point.
• The access point polls each QoS node and avoids contention with
DCF and EDCA nodes by transmitting polls before any of these
devices can begin to contend for channel access.
• HCCA is a hybrid mechanism, merging the best aspects of a
coordinated mechanism for QoS traffic and DCF for efficient
handling of bursty traffic.
• Since HCCA preempts the network for QoS traffic, it must
regularly go quiet to allow non-HCCA-enabled devices to transmit
as usual.
• During these quiet times, EDCA-based devices can gain channel
access, and transfer data.
• That is an important aspect of HCCA because it's likely that
networks will contain a mix of HCCA-, EDCA-, and DCF-based
devices that must coexist with each other.
802.11e HCF
• To support both IntServ and DiffServ 802.11e has defined a new
mechanism called Hybrid Coordination Function (HCF).
• HCF is composed of two access methods: contention-based
channel access (also called EDCF) and controlled channel access
mechanisms,
• One main feature of HCF is to introduce four access category
(AC) queues and eight traffic stream (TS) queues at MAC layer.
• When a frame arrives at MAC layer, it is tagged with a traffic
priority identifier (TID) according to its QoS requirement, which
can take the values from 0 to 15.
– The frames with TID values from 0 to 7 are mapped into four AC queues
using EDCF access rule.
– TID values from 8 to 15 are mapped into eight TS queues using HCF
controlled channel access rule.
• Another main feature of the HCF is the concept of transmission
opportunity (TXOP), which is the time interval permitted for a
particular STA to transmit packets.
802.11e HCF
• In HCF controlled channel access mechanism, QoS
guarantee is based on the (TSPEC) negotiation between
the QAP and the QSTA(s).
• Virtual connection called traffic stream (TS) is
established, before transmitting any frame requiring QoS
• A set of TSPEC parameters are exchanged between the
QAP and the corresponding QSTA(s).
• Based on TSPEC, the QAP scheduler computes the
duration of polled-TXOP for each QSTA, and allocates
the polled-TXOP to each QSTA.
• Then the scheduler in each QSTA allocates the TXOP for
different TS queue according to the priority order.
• If priority is the same, the scheduler will select the
minimum value of all maximum service interval.
802.11e HCF
• QAP is allowed to use an admission control algorithm to
determine whether or not to allow new TS into its BSS.
• When a TS is set up, the QAP attempts to provide QoS by
allocating the required bandwidth to the TS.
• During a CFP, the medium is fully controlled by the QAP.
• During a CP, it can also grab the medium whenever it wants
(after a PIFS idle time).
• After receiving a QoS CF-poll frame, a polled QSTA is allowed
to transmit multiple MAC frames denoted by contention-free
burst (CFB), with the total access time not exceeding the
TXOPLimit.
• All the other QSTAs set their NAVs with the TXOPLimit plus a
slot time when CFB is occuring.
HCF controlled channel access
• HCF can start the controlled channel access mechanism in both CFP and CP
intervals.
• During the CP, a new contention-free period named controlled access phase
(CAP) is introduced.
• CAPs are several intervals during which frames are transmitted using HCFcontrolled channel access mechanisms.
• HCF can start a CAP by sending downlink QoS-frames or QoS CF-Poll frames
to allocate polled-TXOP to different QSTAs after the medium remains idle for
at least PIFS interval.
• Then the remaining time of the CP can be used by EDCF.
• By using CAP, the HCF beacon interval size can be independent of targeted
delay bounds of multimedia applications.
– The HCF controlled channel access can increase the polling frequency by
initiating CAP at any time, thus guarantee the delay bound with any size of
beacon interval. So there is no need to reduce the beacon interval size that
increases the overheads.
– The problem of beacon delay in PCF is solved, because in HCF, a QSTA is
not allowed to transmit a frame if the transmission cannot be finished before
the next TBTT.
HCF
DCF
EDCF
HCF
DCF
EDCF
HCF
Problems with QoS
• The market has not yet favoured QoS services.
• Network that offers sufficient bandwidth for most
applications, most of the time, is already economically
stable, with little incentive to deploy non-standard
stateful QoS-based applications.
• If one needs more bandwidth, pay more and get more.
• Internet peering arrangements are already complex, not
many providers are for supporting QoS across peering
connections
• If dropping many packets on elastic low-QoS
connections, the connection is already overwhealmed
and will violate traffic contracts.