One. The difference between online network and cable network
(1.) Transport medium is completely different
(1.) The transmission medium of the wireless network has neither an absolute boundary or is not easy to observe. (STA outside the boundary is not received)
(2.) Can't guarantee that the external signal is not interfered
(3.) Reliability is poor than wired transmission
(4.) Dynamic topology
(5.) There is a lack of complete connectivity, that is, the STAS may not be visible, so each STA can hear all other STAS assumptions are not established.
(6.) Asymmetric transmission characteristics
(2.) Address
two. Wireless network structure
IEEE 802.11 Defines the signal feature and modulation method of data transmission in the physical layer, defining two radio frequency (RF) transmission methods and an infrared transmission method. The RF transmission standard is direct sequence spread spectrum (DSSS) and frequency hopping spread (FHSS). Since conflict detection is difficult in wireless networks, the Media Access Control (MAC) layer adopts the Avoiding Conflict (CA) protocol, not conflict detection (CD), but can only reduce conflicts. IEEE 802.11 Physical Layer's wireless media (WM) determines that it is different from the MAC of the existing wired local area network, which has a unique media access control mechanism to share wireless media in CSMA / CA.
Basic Services (BSS) is the basic unit of WLAN, which features distributed coordination (DCF) and point coordination function (PCF). Coordination function is a station that determines working within BSS, allowing the logical function of the protocol unit through WM to transmit and possibly receive the protocol unit. DCF is the basic media access method of the IEEE802.11MAC protocol, which acts on basic service groups and basic network structures, which can be implemented in all stations, which supports competitive asynchronous services. PCF supports no competitive time limit business and no competitive asynchronous business. Wireless LAN MAC provides services with: secure services, MSDU reordering services and data services.
DCF is based on carrier listening multichannel access / collision prevention (CSMA / CA) without using carrier listening multi access / collision detection (CSMA / CD) multiple access / collision detection (CSMA / CD) multichannel access method in wired LAN. This is because the channel collision is not heard when the site is transmitted. Since the wireless channel is widely dynamic, it is difficult to adopt collision detection during effective bandwidth, so it can only take a random retraction method to reduce the probability of two frames.
PCF needs to coexist with DCF and logically on DCF. Non-Combined Demolition Interval (CFP Period) is used to determine the frequency generated by the PCF. A part of the time in the repetition interval is assigned to the competitive business stream, and the remaining business stream is provided. Non-competitive intervals are initialized by flag frame (B). The logo frame is transmitted by the access point, one of the main functions is synchronous and timing. The length of the non-competitive period is a modular parameter, which can vary with the input load. If the load is light, the access point will shorten the non-competitive period and provide more time to DCF. Figure 2 shows the case where the PCF and DCF coexistence within the non-competitive period.
IBSS, BSS, ESS, WIRED LAN, DS,
three. service
As a logical service interface called by the MAC sub-layer, the complete service set in the 802.11 architecture is:
(1.) Authentication
(2.) Association
(3.) DEAUTHENTICATION
(4.) Disassociation
(5.) Distribution
(6.) INTEGRATION
(7.) Privacy
(8.) REASSOCIATION
(9.) MSDU DELIVERY
Services are divided into two parts: STA and DS.
STA section SS includes:
(1.) Authentication
(2.) deAuthentication (3.) Privacy
(4.) MSDU Delivery
DS section DSS includes:
(1.) Association
(2.) Disassociation
(3.) Distribution
(4.) Integration
(5.) REASSOCIATION
In these services, the purpose of service is divided into two parts:
(1.) 6 of these services to support MSDU Delivery Between Stas
(2.) The remaining three services are used to control 802.11lan Access and Confidentiality
Distribution, integration is DS forwarded news
Association, Reassociation, DISASSOCIATION provides Support for Distribution Service
Authentication, Preauthentication, Deauthentication, Privacy is Access and Confidentiality Control Services
four. news
802.11 There are 3 kinds of message types: Data, Management and Control. There are also a variety of subtypes per message.
Data: DATA, DATA CF-ACK, DATA CF-POLL, DATA CF-ACK CF-POLL, CF-ACK, CF-POLL, CF-ACK CF-POLL,
Management: Association request, Association response, Reassociation request, Reassociation response, Probe request, Probe response, Beacon, Announcement traffic indication message (ATIM), Disassociation, Authentication, Deauthentication
Control: PS-POLL, RTS, CTS, ACK, CF-End, CF-End CF-ACK
Fives. Mac
The structure of the MAC layer is PCF and DCF, and PCF is based on DCF.
*********************************************************** ******************
DCF should be implemented in each communication station, regardless of IBSS or Infrastructure Network.
The DCF has a well-known conflict-prevention carrier listening multiplex (CSMA / CA) mechanism, which is implemented by random backoff interval. Conflict prevention is generally, when the network is idle, the site does not communicate immediately, but to choose an arbitrary rollback time, and the network is still free in this time. Because it is arbitrary, it greatly reduces the chances of conflict.
DCF Gets the control of the network through the RTS / CTS.
*********************************************************** *******************
The PCF is only in Infrastructure Network. Implemented at the access point AP of the BSS.
The PC is actually a selector, and is implemented by selecting the communication site in order from the polling list.
There is a virtual carrier listening mechanism in the PCF. Implemented by sending a Beacon management frame and setting a network assignment vector (NAV). And by using a small inter-interval interval (Interframe Space) than DCF, the PCF is transmitted higher than DCF.
===========================================================================00 It is a decrease in the reliability of wireless transmission. Each small piece length does not exceed Fragmentationthresh. By dividing, the reliability of communication can be improved, and the retransmission amount is reduced, and the network utilization is improved.
====================================
The data service of the MAC layer is primarily responsible: convert the MAC service request from the logical link layer (LLC) into an input signal that can be utilized by the MAC state machine; simultaneously converts the output signal of the MAC state machine to a logic chain. Road layer service index.
Five-1 DCF
All directed traffic should use immediate active reply (Immediate Positive Acknowledge), if the reply is confiscated, returning.
Conflict-prevented carrier listening multi-channel access in DCF is to reduce the possibility of conflicts in multiple sites transmission.
Carrier listening has both physical mechanisms that also have virtual mechanisms. The virtual mechanism is implemented by sending the reservation information.
The sending method of the appointment information is:
(1.) By switching the RTS / CTS frame. The Duration / ID domain inside defines the time required to send the data frame and the ACK frame. All sites that can receive RTS or CTS are available.
(2.) Duration / ID domain in all directional frames (Directed frame). An appointment time or to the next ACK, or the next piece of the next piece ACK.
*********************************************************** *********************
RTS / CTS frame removed as a rapid detection of conflicts or a detection of the transmission path. RTS / CTS can also resolve invisibleness between STAs (block is blocked by AP). However, when sending the Broadcast / Multicast frame, the RTS / CTS mechanism is not used, on the one hand, because of the conflict, on the other hand, it is an increase in the network burden.
The use of RTS / CTS is controlled by the dot11rtsthreshold property. But those sites that do not use the RTS / CTS mechanism, still applying the received RTS or CTS update NAV vector, and make CTSs to their own RTS accordingly. And the RTS / CTS transmission rate should be within the ABASicrateSet. =====================================
The most basic process is Backoff ProCedure during DCF. Figure 52 Backoff Procedure
After a task is completely over, first is a DIFS frame gap. If there is no SIFS and PIFS frames to transmit before this, then STAS will start the respective return process. If the task of other STAs is started (by receiving other STAs), they will update NAV, and wait until the task is completed, continue to remain retirement. The rollback process is reduced in aslottime.
After the fallback is completed, the process of transferring data is started. (Figure 53)
In the transmission data, the SIFS frame gap is used. The order of the media on the medium (from the sender and the receiver) is RTS-CTS- (Fragment 0-ACK0) - ... - (Fragment N-ACK N). Each frame has a Duration / ID domain, which is included in the next task (ie the next parentheses in the next parentheses). (Figure 55)
After the RTS is issued, the STA receiving the RTS should be after a SIFS, if the NAV indicates that the media is idle, then transmits the CTS; if NAV indicates that the media is busy, the RTS does not respond. The sender STA is starting with PHY-TXEND.CONFIRM after emitting RTS, waiting for a CTSTIMEOUT INTERVAL.
If there is no PHY-RXSTART.Indication in this time, the STA thinks that the RTS will send failed, and then make Backoff Procedure.
If the phy-rxstart.indication occurs within this time, the STA is waiting for the phy-rxend.indication to finalize whether RTS is successful.
When the CTS frame is active when the STA successfully identifies the frame issued by the RTS recipient, the RTS transmission is successfully transmitted, and the next frame sequence can be transmitted.
When the frame emitted by the RTS recipient identified by the STA is other frame, even if it is effective, it is also determined that the RTS transmission failed. Then after PHY-RXEND.Indication, Backoff Procedure.
If there is no PHY-RXSTART.Indication in this time, the STA thinks that the RTS will send failed, and then make Backoff Procedure.
If the phy-rxstart.indication occurs within this time, the STA is waiting for the phy-rxend.indication to finalize whether RTS is successful. When the CTS frame is active when the STA successfully identifies the frame issued by the RTS recipient, the RTS transmission is successfully transmitted, and the next frame sequence can be transmitted.
When the frame emitted by the RTS recipient identified by the STA is other frame, even if it is effective, it is also determined that the RTS transmission failed. Then after PHY-RXEND.Indication, Backoff Procedure.
*********************************************************** ***********************
Directed MPDU Transfer Procedure
When STA transmits Directed Frames, only RTS / CTS is transmitted with RTS / CTS when the MPDU length is greater than dot11rtstHResHold.
An object of MAC MIB when dot11rtSthreshold, can be set and restored via Mac LME. When 0, it means that all MPDUs should be transmitted via RTS / CTS. When it is greater than the maximum length of MSDU, it means that all MPDUs are not passed.
RTS / CTS transmission.
If transferred via RTS / CTS, it is transmitted after the CTS frame and a SIFS are performed, and the media state is not considered.
If it is transferred via RTS / CTS, it is performed after obtaining basic transmission permissions.
Regardless of the use of RTS / CTS, the receiving end STA must have ACK Procedure.
Broadcast and multicast MPDU Transfer Procedure
When there is no PCF, the Tods of Broadcast / Multicast MPDUS is empty and transports only basic Access Procedure. No RTS / CTS, no ACK returns.
If the TODS of the Broadcast / MultiCast MPDUS emitted by the STA is not empty, the transfer is required to follow the RTS / CTS because this MPDU is to the AP and then distributed to BSS.
All STA will filter out the frame containing your own address in Source Address.
For Broadcast / Muliticast Frame, there is no Mac-Level Recovery unless Tods is not empty. Results Reliability decreased.
Ack Procedure
If the frame is received by the AP, TODS is not empty, indicating that the ACKNOWLEDGMENT is required. This AP should send ACK after a SIFS instead of the network situation.
Source STA should be taken to start signing the start sign for the start flag at PHY-TXEND.CONFIRM before determining the MPDU sends failure.
If there is no PHY-RXSTART.Indication in this time, the MPDU will fail. (And make backoff procedure after Acktimeout Interval
If there is a phy-rxstart.indication, you will continue to wait for the phy-rxend.indication.
If the ACK frame is successfully identified, it is considered that the transmission is successful.
Otherwise, even if you identify other valid frames, it is also considered that the send failed. At this point, STA should perform backoff procedure after phy-rxend.indication. The only exception is: any valid frame issued by the PS-POLL recipient
It is considered to be a successful response to PS-POLL.
Duplicate Detection and Recovery
The filtering of repeating frames is done with the Sequence Control Field.
The Sequence Control Field consists of Sequence Number and Fragment Number. The same MSDU's MPSUS has the same Sequence Number, while different MSDUs have different sequencenumber. Sequence Number is an incremental integer generated by the sender STA.
The receiver STA should maintain a CACHE, store the recent received
Receiver STA should also reject the RESET BITs that are not empty and repeated frames in cache.
But the receiver STA still wants to make Ack Procedure for frames that require AcknowledgeGment, even if it is to be discarded.
DCF TIMING RELATIONS
Asifstime and Aslottime are fixed to each PHY.
PIFS = asifstime aslottime
Difs = asifstime 2 * aslottime
============================================================================================================================================================================================================= ================
Five-2 PCF
If the STA confiscates the ACK in the expected, it is not returned. To retransmit, you have to wait until the next selection, or retransmit during the CP.
If the PC confiscates the ACK in the expected, Backoff is performed before retransmission. To retransmit, wait until this AID is again selected again.
Time structure of CFP:
The PC generates CFP with CFPRATE frequency (ie, CFP Repetition Interval, which is determined by CFPperiod in the CF Parameter Set.
The longest duration of the CFP is determined by the CFPMAXDuration of the CF Parameter Set.
The remaining time of the CFP is determined by the cfpdurremaining of the CF Parameter Set.
*********************************************************** ******************
STAS Party NAV setting:
At each CFP begins (ie TBTT), each site is not a PC set its own NAV to cfpmaxduration. Since then, each of your NAV is updated to cfpdurremaining. At the end of the CFP, it will end with CF-END or CF-End ACK. When you receive these frames, the site resets your NAV.
*********************************************************** *******************
During PCF, the most basic is Beacon frame.
STA will determine the following status according to the received Beacon. (?? By Partial Virtual Bitmap and Bitmap Control, I decide whether I have been selected.) Decide whether to be a DTIM frame by DTIM Count, decide whether the BROADCAST / Multicast frame is received according to the 0th bit of Bitmap Control. At the same time, Beacon Interval decided to receive the period of the BEACON frame, which is especially important for sites in the PS state. At TBTT, each STA sets its own NAV to CFPMAXDURATION, while the PC first checks if the network is busy, wait until the network is idle (whether there is SIFS, etc.) immediately sends the Beacon frame. The STA received by Beacon will perform time synchronization, NAV update, determine Beacon Interval, (?? Determine if you are selected) and so on. After sending the SIFS time after sending the Beacon frame, the PC sequentially selects the site from the polling list according to the AID ascending principle, and sends a frame containing CF-POLL to the selected site. And wait for a response. If there is still not yet, then continue to select the next site after PIFS time. After receiving the site, you must respond. If it is noted to send data, send back NULL or CF-ACK. If there is, send back the DATA or DATA CF ACK. (There is also a situation that: if the site is selected, the DWELL TIME is not enough, and the MPDU will be sent and sent back to NULL or CF-ACK. After the AP is received, the site is not procedure based on the more framement and more data of the Frame Control field. If not, the PC sends a DAIA CF-ACK CF-POLL frame to another station in the polling list. (Where the CF-ACK portion of the frame is confirmed to the previous data frame. The polling and confirmation frames are combined between the point coordinator and the site to transmit the efficiency.) Until the cfpdurremaining time ends, or PC again There is no data to be sent and the Polling List is selected. At this time, the CFP will end in advance, and the PC will send CF-END frames to immediately end non-competitive period.
*********************************************************** *******************
Operation with overlapping point-coordinated BSSS
Since there is no CSMA / CA Contention Window Randomization and Backoff of the DCF, the overlapping, BSSS with PC, if CFP Rate is similar to the Beacon Interval, and on the same PHY channel,
Conflicts may occur.
So when the PC's Beacon is busy with the network, the PC should start CFP after Difs any Backoff delay. The PC may also use this backoff to rehabilitate the unrocked directed Data and Management frames.