Remote Macintosh conventions.


43 views
Uploaded on:
Category: People / Lifestyle
Description
2. Arrangement of remote MAC conventions. Remote MAC conventions. Settled task plans. Arbitrary access plans. Request task plans. Circuit-exchanged. CL parcel exchanged. CO parcel exchanged. 3. Layout. Requirement for remote MAC protocolsObtain task of assets per call ala circuit exchanging altered assignmentObtain task of assets per bundle ala parcel exchanging CL flavor:
Transcripts
Slide 1

Remote MAC conventions Prof. Malathi Veeraraghavan Elec. & Comp. Engg. Dept/CATT Polytechnic University mv@poly.edu EL604: Wireless & Mobile Networking

Slide 2

Classification of remote MAC conventions Wireless MAC conventions Fixed-task plans Random-access plans Demand task plans Circuit-exchanged CL bundle exchanged CO parcel exchanged

Slide 3

Outline Need for remote MAC conventions Obtain task of assets per call ala circuit exchanging settled task Obtain task of assets per parcel ala parcel exchanging CL flavor: irregular access CO flavor: request task

Slide 4

Random access MAC conventions Comparable to connectionless bundle exchanging No reservations are made; rather a remote endpoint essentially begins sending information bundles What can happen?

Slide 5

Answer Collision Need to maintain a strategic distance from crashes or identify impacts and retransmit What\'s the expense of being excessively cautious, making it impossible to dodge crashes? Usage will be yielded

Slide 6

Various arbitrary access MAC plans ALOHA Slotted ALOHA CSMA/CD CSMA/CA

Slide 7

ALOHA Simplest plan True free-for-all. At the point when a hub needs to send, it does as such. It listens for a measure of time equivalent to the most extreme round outing delay in addition to an altered augmentation. On the off chance that it hears an affirmation, fine; else it resends in the wake of holding up an irregular measure of time. After a few endeavors, it surrenders. Low defer if light load Max. usage: 18%

Slide 8

Analysis of ALOHA First transmission Retransmission t prop : most extreme one-way spread postponement between any two stations X = L/R, L: parcel length (steady) R: rate S: Throughput (likewise number of fresh introductions/X sec on the off chance that we expect that all bundles in the end make it) G: landing rate of new + retransmitted bundles/X sec; Poisson entry forms Probability of effective transmission is that there are no extra transmissions in the helpless time of 2X Y: arbitrary variable indicating number of aggregate entries in 2X seconds t-X t t+X t+X+2t prop t+X+2t prop +B backoff period powerless period

Slide 9

Analysis of ALOHA contd. The throughput S is the aggregate landing rate G times the likelihood of there being no crash The normal postponement relies on upon normal number of transmission endeavors per bundle The normal number of unsuccessful endeavors G/S-1 Average deferral is approximated by

Slide 10

ALOHA throughput Maximum throughput is 18% at G = 0.5

Slide 11

Slotted ALOHA Competition to send just happens toward the begin of every space (equivalent to X) Vulnerable period is X (not 2X as in ALOHA) What is greatest throughput?

Slide 12

CSMA Carrier Sense Multiple Access sense transporter if inert, send sit tight for ack If there isn\'t one, accept there was a crash, retransmit Vulnerable period: one t prop

Slide 13

Different systems 1-industrious: if occupied, always sense channel if inactive, send promptly if impact is identified, hold up an arbitrary measure of time before retransmitting Non-diligent: sense channel when station has a bundle to send if occupied, hold up an irregular measure of time before detecting once more; if inert, transmit when it is inert impacts decreased in light of the fact that detecting is not quickly rescheduled disadvantage: more postpone p-determined: consolidates 1-tireless objective of diminished inert channel time with the non-relentless objective of lessened impacts. sense continually if occupied and the station needs to send a parcel when the channel gets to be inert, transmit bundle with likelihood p with likelihood 1-p station holds up an extra t prop before detecting once more

Slide 14

CSMA/CD Ethernet (likewise 802.3) institutionalizes the 1-determined CSMA/CD multi-access control convention. Every station listens before it transmits. In the event that the channel is occupied, it holds up until the channel goes inert, and after that it transmits. On the off chance that the channel is inert it transmits quickly. Keep detecting. On the off chance that crash is identified, transmit a brief sticking sign, then stop transmission, sit tight for an irregular time, and retransmit. crash recognition is not by sitting tight for an ACK

Slide 15

Collisions in Ethernet The impact determination procedure of Ethernet requires that a crash is identified while a station is as yet transmitting. Accept: max. spread postponement on the transport is a.

Slide 16

Collisions in Ethernet Restrictions: Each edge ought to be at any rate twice the length of an ideal opportunity to recognize an impact (2 · most extreme engendering delay).

Slide 17

CSMA/CD CSMA/CD: In CSMA, if impact happens, need to hold up till harmed outlines have completely spread. For long casings contrasted with spread defer, this could prompt noteworthy misuse of limit. So include crash location. Guideline: Frames ought to be sufficiently long to permit impact recognition preceding the end of transmission

Slide 18

Exponential Backoff Algorithm If a station is included in a crash, it holds up an irregular measure of time before endeavoring a retransmission. The irregular time is controlled by the accompanying calculation: Set "opening time" to 2a. After first crash hold up 0 or 1 time unit. After i-th crash, hold up an irregular number between 0 and 2 i - 1 time openings. Try not to build arbitrary number extent if i=10. Surrender after 16 crashes.

Slide 19

Wireless 802.11 LAN Uses CSMA/CA Why CA and CD? Hard to identify impacts in a radio domain – why? Concealed station issue: Two commonly far away stations An and C need to send to B. At An and C, channel seems sit without moving But crash happens at B

Slide 20

Why is it hard to distinguish impacts in a radio domain? A transmitting station can\'t successfully recognize approaching feeble signs from commotion and the impacts of its own transmission; require a full duplex radio to listen and transmit on same recurrence (not valid in FDD frameworks)

Slide 21

Mechanisms for CA Use of Request-To-Send (RTS) and Confirm-to-Send (CTS) instrument When a station needs to send a parcel, it first sends a RTS. The getting station reacts with a CTS. Stations that can hear the RTS or the CTS then stamp that the medium will be occupied for the span of the solicitation (demonstrated by Duration ID in the RTS and CTS) Stations will modify their Network Allocation Vector (NAV): time that must pass before a station can test channel for inert status this is called virtual transporter detecting RTS/CTS are littler than long bundles that can impact Use of InterFrame Spaces (IFS)

Slide 22

extending CP CFP Frame CFP Super-outline Foreshortened CFP 802.11 MAC IEEE 802.11 consolidates an interest task MAC convention with irregular access PCF (Point Coordination Mode) – Polling CFP (Contention-Free Period) in which access point surveys has DCF (Distributed Coordination Mode) CP (Contention Period) in which CSMA/CA is utilized

Slide 23

DCF Distributed Coordination Function This method of 802.11 is an arbitrary access MAC When a hub needs to send information, it detects the medium. In the event that inert, sit tight for a time of DIFS and if the medium is still sit out of gear after DIFS, send instantly. In the event that when the medium is detected it is occupied; then sit tight for medium to be inactive for a DIFS (DCF IFS) period then decrement backoff clock until medium gets to be occupied once more; stop clock, OR clock achieves 0; transmit outline if two stations have their clocks achieve 0; crash will happen; for each retransmission endeavor, increment the dispute window (CW), inert period after a DIFS, exponentially; 2 i –1 beginning with CW min e.g., 7, 15, 31 .

Slide 24

DIFS SIFS CW Random backoff time DCF mode transmission without RTS/CTS Data source Ack goal NAV other Defer access Exercise: Show timing outline for DCF mode with RTS/CTS

Slide 25

DCF MAC Send quickly (after DIFS) if medium is inert If medium was occupied when detected, hold up a CW after it gets to be inactive (on the grounds that numerous stations might hold up when medium is occupied; in the event that they all send the moment the medium gets to be inactive, odds of crash are high)

Slide 26

PHY layer Three physical layer determinations are a piece of 802.11 Spread range Frequency bouncing (FH) Direct Sequence (DS) Infrared (IR)

Slide 27

FH What is FH? Balance the information flag to such an extent that it involves distinctive recurrence groups as transmission advances e.g., send a tune over numerous FM radio channels with some abide time per channel Why not FDMA? Multipath blurring influences tight recurrence groups so a few channels offer exceptionally poor transmission In FH, time spent in every channel is little

Slide 28

802.11 FH PHY 79 non-covering 1Mhz channels utilized 1Mbps signs transmitted over the 2.4Ghz band 2400 – 2483.5Mhz 83.5 Mhz of transfer speed (US: begins 2.402Ghz to 2.480 – so 79) A channel bounce happens each 224  s 78 jumping designs Divided into 3 sets of 26 examples each The sets are intended to maintain a strategic distance from delayed crash periods between bouncing successions in a set Hopping designs impact 3 times by and large, and 5 times in the most pessimistic scenario over a bouncing cycle; every bounce is a hop of at least 6 channels Each 802.11 LAN must utilize a specific jumping design The jumping designs take into consideration 26 systems to be gathered and still work at the same time

Slide 29

+1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 DS Modulate information signal by a sign that involves a much bigger data transfer capacity Chip rate: time to transmit a +1 or –1 To transmit an information bit, need 11 chip times 11 chip Barker arrangement To transmit +1, send To transmit - 1, send +1 - 1 - 1 - 1 - 1 11 image times 11 image times

Slide 30

802.11 DS Takes a 1Mbps information flag and changes over it into a 11 Mbps signal 11 directs in the 2.4Ghz band (5Mhz dispersing) Channels isolated by focus frequencies no less than 30Mhz separated can work without obstruction If all out transmission capacity is just 83.5 Mhz, just 3 802.11 LANs utilizing DS can have covering cells FCC just assigns somewhere around 2412 and 2462

Slide 31

Ad-hoc versus foundation based Ad-hoc No altered system base required A remote endpoint sends and all hubs inside reach can get flag Each bundle conveys

Recommended
View more...