Wimax 4

WIMAX

WIMAX

• WIMAX (Worldwide Interoperability for Microwave Access ):

– Protocol of communication network without wire, based on the standard IEEE 802.16

– Allows communications over long distances than WiFi, and a greater bandwidth. Cover approximately 40km.

• Field of application:– Better price points for both home and business

customers. – WIMAX allow competitors joint access to any subscriber in areas without preexisting physical cable or telephone

networks– would allow gamers access to ad hoc local networks of

other players with the same gear- without any internet access.

Introduction

• Goal of WIMAX: Provide high-speed Internet access to home and

business subscribers, without wires.

• Frequency range: 10-66 GHz and sub 11 GHz

• Supports:– Legacy voice systems– Voice over IP– TCP/IP– Applications with different QoS requirements.

Introduction

•802.16 consists of the access point, BS(Base Station) and

SSs (Subscriber Stations).

•All data traffic goes through the BS, and the BS control the allocation of bandwidth on the radio channel.

Introduction

• During a communication, all the information coming from a SS go to the BS and are retransmit to the right SS.

• Base stations (BS) can handle thousands of subscriber stations (SS).

• Two type of link are defined:– The downlink: From the BS to the SS.– The uplink: From the SS to the BS.

Introduction

Infrastructure of WIMAX

• A WIMAX tower: similar in concept to a cell-phone tower. A single WIMAX tower can provide coverage to a very large area (~8,000 km).

• A WIMAX receiver : The receiver and antenna could be a small box or PCMCIA card, or could be built into a laptop.

Introduction

A WIMAX tower An example of WIMAX receiver : PCMCIA card

PLAN

I Transmission of the data

II ARQ

III Scheduling

PLAN


II ARQ

III Scheduling


1 Mac and physical layers2 Structure of a SDU3 structure of a PDU4 fragmentation and packing

• The physical level:– Specify the frequencies– diagram of modulation– synchronizations– speeds,– techniques of cutting in the time (of type

TDMA: Time Division Multiple Access)– techniques of detection and correction of

error.

1 Mac and physical layers


The MAC level : Located at the top of the physical level, it

manages the allowance of the slots and uses the method Rammed-tdma.

The interface of communication with the applications :

this layer concentrates on the management of level IP and the encapsulation of packages IP in the screen adapted to the section of time.


• The MAC is comprised of three sublayers. • The Service Specific Convergence Sublayer (CS)

provides any transformation or mapping of external network data, received through the CS service access point (SAP), into MAC SDUs received by the MAC Common Part Sublayer (MAC CPS) through the MAC SAP.





2 structure of a SDU

• Higher-layer PDUs shall be encapsulated in the MAC SDU format. For some payload protocols, each payload consists of an 8-bit payload header suppression index (PHSI) field followed by actual payload.

• Other protocols map the higher layer PDU directly to the MAC SDU. A value of zero in the PHSI indicates no payload header suppression has been applied to the PDU.



3 Structure of a MPDU

MSB LSB

Generic MAC header Payload (optional) CRC (optional)

• The maximum length of the MAC PDU is 2048 bytes, including header, payload, and Cyclic Redundancy Check (CRC).

6 bytes Variable 4 bytes

• The size of the payload is variable, the payload can contain either data or management message.


• Two MAC header formats are defined:– Generic MAC Header that begins each

MAC PDU containing either MAC management messages or CS data.

– Bandwidth Request Header used by the SS to request additional bandwidth.

• The single-bit Header Type (HT) field distinguishes this two header formats: – HT=0 for Generic Header.– HT=1 Bandwidth Request Header.

• Two MAC header formats are defined:● Generic MAC Header that begins each

MAC PDU containing either MAC management messages or CS data.

● Bandwidth Request Header used by the SS to request additional bandwidth.


• The single-bit Header Type (HT) field distinguishes this two header formats: – HT=0 for Generic Header.– HT=1 Bandwidth Request Header.


1 Mac and physical layers2 Structure of a SDU3 structure of a PDU 4 fragmentation and packing

4 Fragmentation and packing

To allow efficient use of the airlink resource, two functions are included in the MAC layer:

• Packing: Several small MSDUs addressed to the same CID

may be concatenated by the transmitter to form a single MPDU. At the reception, the SDU is reassembled by the MAC layer.

• Fragmentation: MSDU might be fragmented by the transmitter to

form several MPDUs. At the reception the SDUs are separated by the MAC layer.


Reasons:

• Lack of the frame time when allocating the air time to the given MSDU

• High BER that requires employing integrity check for smaller data blocks


• The number of fragments can not be more than 16.

• When created, the MAC payload (MPDU) are assigned by:

- Fragment Serial Number (FSN) with possible value 0 to 15.

- Fragment Control code (FC) with the following meaning:

o 00 = non-fragmented MPDU. o 01 = last fragment. o 10 = first fragment. o 11 = continuing (middle) fragment.

• The FSN is always transmitted within the same MACmessage as the fragment data.


• The sequence number allows the SS to recreate the original payload and to detect the loss of any intermediate packets. • Upon loss, the SS shall discard all MAC PDUs on the connection until a new first fragment is detected or a non fragmented MAC PDU is detected.


• If packing is turned on for a connection, the MAC may pack multiple MAC SDUs into a single MAC PDU. Packing makes use of the connection attribute indicating whether the connection caries fixed-length or variable-length packets.

• The transmitting side has full discretion whether or not to pack a group of MAC SDUs in a single MAC PDU.


Decrease the MAC overhead


• Packing and fragmentation can occur in the same PDU.

Summarize

SDU 1 SDU 2 SDU 3 SDU 4

PDU 1 PDU 2 PDU 3

NO fragmentation& NO packing Packing Fragmentation

MAC SDUs

MAC PDUs

P

FEC 1 FEC 2 FEC 3Preamble

TCPDUs

BURST

FEC encoding

PDU which has started in the previous TC packet

First PDU which starts in this TC packet

Second PDU which starts in this TC packet

II ARQ A ARQ

• Stop and Wait• Sliding window technique• Feedback ( go back-N)• Selective repeat

B FEC

C H-ARQ

D Methods used by WIMAX

II ARQ

• Three methods are employed to makes the data transmission reliable in a unreliable connection ( airlink):

– ARQ ( automatic repeat request)

– FEC (Forward Error Correcting)

– H-ARQ (hybrid ARQ= ARQ+FEC)

II ARQ A ARQ

• Stop and Wait• Sliding window technique• Feedback ( go back-N)• Selective repeat

B FEC

C H-ARQ


II ARQ

• Three methods are employed for the ARQ wireless transmissions:

– Stop and Wait

– Feedback ( go back-N)

– Selective repeat

• Both feedback and selective algorithm are based on sliding window technique

II ARQ

A ARQ• Stop and Wait• Sliding window technique• Feedback ( go back-N)• Selective repeat

B FEC

C H-ARQ


Stop and Wait

• to each reception of a package, the receiver sends a particular message (ACK) to show reception.

• the transmitter preserves a copy of the emitted package and await the reception of the acknowledgement.

• after a certain time (time out), the package is retransmitted (and the transmitter waits again).

Stop and Wait

Transmitting Data

1 32 3Time

Received Data 1 2 3Time

AC

K

AC

K

NA

K

Output Data 1 2 3Time

Error

ACK: Acknowledge

NAK: Negative ACK

retransmission

Stop and Wait

Disadvantages of this method:

• Problem of acknowledgement of delivery transmission not very effective

• time between the emission of two packages

• Transmissions on the network in only one direction at the same time

=> use of the sliding window technique

II ARQ


B FEC

C H-ARQ


• One emits several packages before awaiting an acknowledgement.

• The number of packages is defined by the size of the window.

• In each acknowledgement, the window shifts (slips).

The sliding window technique:

II ARQ

The sliding window technique

Without window With a sliding window

size=0 (stop and wait) size=3

Tra

nsm

itte

r

Rec

epto

r

Tra

nsm

itte

r

Rec

epto

r

II ARQ


B FEC

C H-ARQ


Feedback ( go back-N)

• Based on the sliding window technique

• When an MPDU is lost, the transmitter is required to retransmit all the PDU starting from first MPDU was lost

Disadvantage of this method:

• Very bandwidth inefficient: some frames may be repeated several times while there are well received.

Feedback ( go back-N)

1Time

NA

K

Time

Error

Go-back 3

2 3 4 5 3 44 5 6 7 5

1 2 3 44 5

Error

NA

K

Go-back 5

1 2 3 44 5

II ARQ


B FEC

C H-ARQ


Selective repeat

• Based on the sliding window technique

• Only the lost MPDU is retransmitted

Selective repeat

1Time

NA

K

Error

Retransmission

2 3 4 5 3 6 7 8 9 7

1Time

2 4 3 6 8 7

Error

NA

K

Retransmission

5 9

1Time

2 4 3 6 8 75 9

1Time

2 4 3 6 8 75 9

II ARQ

A ARQ1. Stop and Wait2. Sliding window technique3. Feedback ( go back-N)4. Selective repeat

B FEC

C H-ARQ


FEC

• Error Detection Process:• Transmitter

– For a given frame, an error-detecting code (check bits) is calculated from data bits

– Check bits are appended to data bits

• Receiver– Separates incoming frame into data bits and check bits– Calculates check bits from received data bits– Compares calculated check bits against received check

bits– Detected error occurs if mismatch

FEC

• Transmitter– Forward error correction (FEC) encoder maps

each k-bit block into an n-bit block codeword.– Codeword is transmitted.

• Receiver– Incoming signal is demodulated– Block passed through an FEC decoder

FEC

• No errors present– Codeword produced by decoder matches original

codeword.

• Decoder detects and corrects bit errors.

• Decoder detects but cannot correct bit errors; reports uncorrectable error.

• Decoder detects no bit errors, though errors are present

FEC

II ARQ


B FEC

C H-ARQ


H-ARQ

– H-ARQ= FEC+ARQ

– FEC: turbo codes/ convolutional codes/ block codes/…

– ARQ: selective repeat (SR) / stop and wait (SAW)/ go-back-N (GBN)

II ARQ


B FEC

C H-ARQ


Methods used by WIMAX

• ARQ mechanism is an optional part of the

MAC layer in WIMAX.

• WIMAX can use ARQ ,FEC or H-ARQ.

1 ARQ :

• Uses the sliding window technique:

– Selective repeat is selected by default – Feedback algorithm in specified case

• More details:

– ARQ parameters shall be specified and negotiated during connection creation or change

– A connection can not have a mixture of ARQ and non-ARQ traffic

– The ARQ feedback information can be sent as a standalone MAC management message on the appropriate basic management connection or piggybacked on an existing connection

– ARQ feedback cannot be fragmented.


Transmitter state

1 ARQ :

Receiver state

1 ARQ :

2 FEC:


• 802.16 specifies the concatenation of a Reed-Solomon (RS) outer code and a rate-compatible convolutional inner code, on both uplink and downlink.

• The encoding is performed by first passing the data in block format through the RS encoder and then passing it through a zero-terminating convolutional encoder.

• Turbo convolutional codes (TC) and Turbo Block (TB) codes are specified as optional FEC schemes in the standard.

• Low density parity check (LDPC) codes are a new type of FEC codes that are gaining in popularity and might be specified as optional FEC scheme in 802.16e version

3 H-ARQ :• H-ARQ schema is basically a stop and wait

protocol:– Each H-ARQ packet is encoded and 4 subpackets are

generated from the encoded result

– The transmitter shall send the packet labeled ’00’ at he fist transmission

– Then the receiver attempts to decode the original encoder packet

• If it succeeds the receiver sends an ACK to the transmitter so that the transmitter stops sending additional subpackets.

• Otherwise the transmitter sends a NACK and the transmitter may send one among the fourth subpackets.



3 H-ARQ :

– These procedure go on until the SS successfully decodes the encode packet.

– The transmitter may send one among subpackets labeled ’00’,’01’,’10’,’11’ in any order.

– The transmitter can send more than a copy of any sub packet and can omit any subpacket except the subpacket labeled ’00’.

III Downlink/Uplink Scheduling

• Radio resources have to be scheduled according to the QoS (Quality of Service) parameters

• WIMAX Downlink scheduling: – the flows are simply multiplexed – the standard scheduling algorithms can be used :

• WRR (Weighted Round Robin)• VT (Virtual Time)• WFQ (Weighted Fair Queueing)• WFFQ (Worst-case Fair weighted Fair

Queueing)• DRR (Deficit Round Robin)• DDRR (Distributed Deficit Round Robin)


Plan

• A Downlink Scheduling

• B Uplink Scheduling


Plan



A Downlink Scheduling

Plan

3. RR 4. WRR5. VT6. WFQ7. WFFQ8. DRR9. DDRR

•

Plan

• RR • WRR• VT

• WFQ• WFFQ• DRR

• DDRR


RR

111VCC 1 (Source 1)

22VCC 2 (Source 2)

333VCC 3 (Source 3)

3 3

12 3112

WRR scheduler

Counter Reset Cycle

1 32 233 ….

• Round-Robin algorithm equitably distributes the load between each waiter whatever the current number of connections or the response times

RR

• This algorithm is adapted if the waiters of the cluster have the same processing capacities

• if not, certain waiters are likely to receive more requests than they can treat. Some while others will use only part of their resources.

• The WRR algorithm solves this problem.

Plan



• DDRR


WRR

• The WRR algorithm is based on the Round Robin algorithm but it takes into account the processing capacity of each waiter.

• The administrators manually assign a coefficient of performance to each waiter. ( 1, 2 and 3 in the example).

Counter Reset Cycle111VCC 1 (Source 1)

22 VCC 2 (Source 2)

333VCC 3 (Source 3)

3 3

2

1

3

123111 2 33333

WRR scheduler

Coefficients of performance

….

• A sequence of scheduling is generated automatically according to this value.

• The requests are then assigned to the various waiters according to a sequence of alternate repetition

VCC 2 (Source 2)

VCC 3 (Source 3)

111VCC 1 (Source 1)

22

3333 3

2

1

3

123111 2 33333

WRR scheduler

Counter Reset Cycle

Sequence of scheduling

…….

WRR

Plan

• RR • WRR• VT• WFQ• WFFQ• DRR• DDRR


VT

VT• VT : aims to emulate the TDM (Time Division

Multiplexing)

– connection 1 : reserves 50% of the link bandwidth– connection 2, 3 : reserves 20% of the link bandwidth

Connection 1 Average inter-arrival : 2 units



First-Come-First-Served service order

Virtual times

Virtual Clock service order


Plan


• It is not practical to have one queue for each

conversation so the WFQ employs a hashing algorithm which divides the traffic over a limited number of queues to be selected by the user or fixed by default.

• WFQ is like having several doors. When a packet arrives it is classified by the classifier and assigned to one of the doors. The door is the entry to a queue that is served together with some other in a weighted round-robin order. This way the service is 'fair' for every queue.

WFQ

• The packet arrives, then the classifier reads its header.

• Calculates a number between "1" and "number of queues“ by using information contained on the header (source address ,destination address, ip precedence, protocol, ...)

• Then, it locates the packet in the queue identify by this number.

WFQ

WFQScheduler

flow 1

flow 2

flow n

Classifier

Buffer management

WFQ

WFQ

• Each flow i given a weight (importance) wi

• WFQ guarantees a minimum service rate to flow i– ri = R * wi / (w1 + w2 + ... + wn)

– Implies isolation among flows (one cannot mess up another)

w1

w2

wn

R

Packet queues

WFQ

WRR algorithm

w1

water pipesw2

w3

t1

t2

w2 w3

water buckets

w1

WFQ

WFQ

• If flows can be served one bit at a time

• WFQ can be implemented using bit-by-bit weighted round robin

• During each round from each flow that has data to send, send a number of bits equal to the flow’s weight

WFQ

• FFQ (Fluid Fair Queue) : head-of-the line processor sharing service discipline– : guaranteed rate to connection i– C : the link speed– : the set of non-empty queue– The service rate for a non-empty queue i

• WFQ : picks the first packet that would complete service in the corresponding FFQ


Plan



• DDRR

• WFFQ is based on WFQ algorithm

• WFFQ : picks the first packet that would complete service among the set of packets that have started service in the corresponding FFQ

WFFQ

EXAMPLE (1)

– All packets have the same size 1 and link speed is 1Guaranteed rate for connection 1 : 0.5Guaranteed rate for connection 2-11 : 0.05

Connection 1 sends 11 back-to-back packets at time 0Connection 2-11 sends 1 packet at time 0

– The completion time of connection 1 :2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22

– The completion time of connection 2 – 11 : 20

WFFQ and WFQ

Connection 1

Connection 2

Connection 11

WFQ and WFFQ

WFQ Service Order

WFFQ Service Order

…… … … … …

EXAMPLE (2)

WFFQ and WFQ

VT and WFQ

• All packets are fixed size and require exactly one second to service

• Starting at time zero, 1000 packets from connection 1 arrive at a rate of 1 packet/second

• Starting at time 900, 450 packets from connection 2 arrive at a rate of 1 packet/second– The completion times of the 901, 902, 903, …

packets of connection 1 in FFQ system are 1802, 1904, 1806, …

– The completion times of the 1, 2, 3, … packets of connection 2 in FFQ system are 901, 902, 903, …

Plan



D R R

• Each connection is assigned a state variable called the DC (Deficit Counter).

• At the start of each round, DCi of queue i is incremented by a specific service share (quantum)

• If the length of the head of the line packet, Li, is less than or equal to DCi,, the scheduler allows the ith

queue to send a packet.

• Once the transmission is completed DCi is decremented by Li.

Qi DCi

3500

35002800 7800 2000

1500

5000

700

1400

2800 7800 2000

2800 7800 2000

2800 7800 2000

2800 7800 2000

initializing(1st round)

Serviced

Not serviced

Serviced

Serviced

(3rd round)

(4th round)

D R R

+3500

+3500

(2nd round)

+3500

-2000

-7800

-2800

Li

Plan



D D R R

• Each connection is assigned a state variable called the DC (Deficit Counter)

• If the value of the DCi is positive then the scheduler allows the ith queue to send a

packet

• Once the transmission is completed DCi is decremented by Li, the length of the transmitted packet

• At the start of the subsequent rounds, DCi is incremented by a specific service share (quantum)

D D R R

QiDCi

350035002800 7800 2000

1500

-6300

-2800

700

-2100

2800 7800 2000

2800 7800 2000

2800 7800 2000

2800 7800 2000

2800 7800 2000

initializing(1st round)

Serviced

Serviced

Not serviced

(2nd round)

Not serviced

(3rd round)

Serviced

-2000

-7800

-2800

+3500

+3500

+3500

+3500


Plan



B Uplink Scheduling

Uplink scheduling: – Responsible for the efficient and fair

allocation of the resources (time slots) in the uplink direction

– Uplink carrier : • Reserved slots• contention slots (random access slots)

– The standard scheduling algorithms can be used

Bibliography

• http://www.opalsoft.net/qos/WhyQos-2424.htm• http://www.math.tau.ac.il/~alx/courses/notes/Icc8_2.ppt• http://www.it.uu.se/edu/course/homepage/datakom/civinght04/schema/sliding_window.pps•

http://www.opalsoft.net/qos/WhyQos-2424.htm

http://www.math.tau.ac.il/~alx/courses/notes/Icc8_2.ppt

http://www.it.uu.se/edu/course/homepage/datakom/civinght04/schema/sliding_window.pps

Wimax 4

Technology

macand physicallayer

introduction goal of

s tructure

mac level

single wimax tower

example of wimax receiver

microwave access

gamers access