provide privacy and security of user data. Figure shows the relationship between the OSI reference model and the IEEE 802.11 Standard. Figures and list amendments to the original IEEE 802.11 Standard affecting wireless LAN. This lesson examines 802.11e in detail.

Content 6.1 Implementing WLAN QoS 6.1.2 WLAN QoS Description The IEEE standardized 802.11e in 2005 as a set of technologies for prioritizing traffic and preventing packet collisions and delays to improve voice and video performance over WLANs. The 802.11e specification uses the same eight priority level structure that 802.11p uses. They are grouped into four access categories or traffic classes as follows:

Voice: Priority 7 or 6 for toll-quality VoIP calls requiring low latency
Video: Priority 5 or 4 for SDTV or HDTV video streams
Best Effort: Priority 3 or 0 for latency-insensitive, interactive applications
Background: Priority 2 or 1 for batch data transfer applications and indicates less than best effort treatment

Each access level provides a separate queue. To identify the class of each packet, the standard uses markers that are similar to markers used in wired Ethernet. Seeing those markers, an access point handles packets according to the packet’s assigned priority. To speed the adoption of QoS in the 802.11 marketplace, the Wi-Fi Multimedia (WMM) was implemented before the 802.11e standard was approved. WMM is a subset of the 802.11e standard using four access categories rather than the eight priority values of 802.11e. By default, these are voice, video, best effort, and background, but definitions of the four classes can be changed from the default definitions. WMM provides a group of features for wireless networks to improve audio, video, and voice application performances. The resulting QoS allows translation from 802.1p or differentiated services code point (DSCP) to the appropriate RF techniques giving high-priority traffic an increased probability of RF transmission over lower-priority traffic. WMM uses traffic-prioritization capabilities based on the four defined access categories. RF prioritization gives a higher access category the increased probability of being transmitted first. This allows the platinum level to obtain RF access for transmission before the gold, silver, or bronze levels. The access categories correspond to 802.1p or DSCP priorities to facilitate interoperability with QoS policy-management mechanisms. WMM priorities coexist with legacy devices that are not WMM-enabled. Packets that are not assigned to a specific access category are categorized by default as the best-effort access category. Figure depicts the WMM traffic prioritization scheme that the Wi-Fi Alliance has approved for assigning application data that is being sent to the client. WMM is an enhancement to the MAC sublayer that adds QoS functionality to Wi-Fi networks. The MAC sublayer handles access mechanisms, fragmentation, and encryption. WMM prioritization maps four independent transmit queues to eight 802.1e priority levels, as shown in Figure .

Content 6.1 Implementing WLAN QoS 6.1.3 WLAN QoS RF Backoff Timing The fundamental access method of 802.11 is CSMA/CA. CSMA/CA works by a "listen before talk scheme." WMM uses the CSMA/CA-based distributed coordination function (DCF) mechanism that gives all devices the same priority and is based on a best-effort, “listen-before-talk” algorithm. If the medium is not busy, the transmission may proceed. CSMA/CA uses a random backoff time to avoid collisions among stations sharing the medium, if the sending station’s physical or logical sensing mechanism indicates a busy medium. Each client waits a random backoff time and then transmits only if no other device is transmitting at that time. This collision-avoidance method gives all the devices the opportunity to transmit, but when traffic demand is high and networks can become overloaded, the performance of all devices will be affected. Using Enhanced DCF (EDCF), WMM stipulates different fixed and random wait times for the four access categories to provide more favorable network access for applications that are less tolerant of packet delays. Devices that have less time to wait have a better chance of being able to transmit than those that have a longer wait. The access categories—voice, video, best effort and background—map to Ethernet 802.1d prioritization tags to allow consistent QoS across wireless and wired network segments. WMM provides priority access to the RF medium in two ways:

First, the wireless access point must prioritize the data into four access categories (from highest to lowest: platinum, gold, silver, and bronze).
Second, the lower-priority traffic must use longer interframe wait timers to allow higher-priority traffic access to the wireless network first. The timers result from a summarization of the fixed short interframe space (SIFS), slot times (fixed-length time intervals based on priority), and a random slot timer (based on priority) shown in Figure for various QoS levels. The random backoff slot timers prevent media contention among traffic from within the same access category.

Content 6.1 Implementing WLAN QoS 6.1.4 Lightweight Access Point—Split MAC Architecture Two philosophically opposite approaches to providing wireless access exist. The first approach, which is becoming less popular, uses autonomous access points (AP) with a lot of intelligence. Intelligent access points can communicate with existing routers and thus support robust applications. This style of WLAN management, referred to as a distributed architecture, works well but is expensive and requires access points that can work within a specific vendor's particular infrastructure. The second approach takes the intelligence out of APs and puts the intelligence into switches or routers to allow networks to scale and to reduce the overall cost of WLAN deployment. This approach led to the development of the Lightweight Access Point Protocol (LWAPP), resulting in an entirely new paradigm for managing WLAN deployments. LWAPP uses the concept of a “split MAC”, which is the ability to separate the real-time aspects of the 802.11 protocol from most of the protocol’s management aspects. Cisco has designed a centralized, lightweight access point wireless architecture to address the unique RF management needs of enterprises. A core component is the split MAC architecture, which distributes the processing of 802.11 data and management protocols between a lightweight access point and a centralized WLAN controller. More specifically, the access point handles time-sensitive activities, such as beacon handling, handshakes with clients, MAC layer encryption, and RF monitoring. The WLAN controller handles all other functions and provides system-wide visibility. These functions include IEEE 802.11 management protocol, frame translation, and bridging functions, as well as system-wide policies for user mobility, security, QoS, and, perhaps most importantly, real-time RF management. Figure distinguishes real-time, or time-critical, MAC functions that the access point performs, and non-real-time, or network critical, MAC functions that the controller performs. The figure also introduces the concept of an LWAPP tunnel. Forthcoming topics explain that all traffic coming from wireless clients goes into LWAPP packets that are tunneled through the underlying network to the wireless LAN controller.

Content 6.1 Implementing WLAN QoS 6.1.5 WLAN QoS Challenges Providing QoS in a WLAN is a challenge due to the following :

End-to-end QoS uses DSCP for packet marking.
Wireless RF uses Layer 2 marking.
In the Cisco deployment model, traffic destined to access points does not contain 802.1p QoS tag information because the access points connect to a Cisco Catalyst switch port, which is not a trunk.
As a