Cisco Cisco Aironet 3700i Access Point White Paper

Página de 10
 
 
© 2015 Cisco and/or its affiliates. All rights reserved. This document is Cisco Public Information. 
Page 4 of 10 
As previously mentioned, a goal of Air Time Fairness is to establish airtime limits, defined as a percentage of total 
airtime, and to apply those limits on a per-SSID basis, where the SSID is used as parameter to define a group of 
clients. Other parameters can be used as well to define groups of clients, and will be discussed later in this paper. 
It is important to note that if a client or SSID exceeds the airtime limit, packets will be dropped. However, only 
packets in the downlink direction are dropped. This is because dropping downlink packets (access point to client) 
frees up airtime. Dropping uplink packets (client to access point) does not do anything to free up airtime since the 
packet has already been transmitted over the air by the client. Once the packet has been on the air, that airtime 
cannot be reclaimed. 
This can mean that a client that is transmitting a lot of uplink traffic can exhaust airtime. This is essentially a denial-
of-service attack that is outside the scope of what Cisco Air Time Fairness addresses. 
. 
Cisco Air Time Fairness Per SSID 
A key concept in allocating airtime is the idea of a time slice. A time slice typically represents a specified maximum, 
not to exceed a certain percentage of the available downstream airtime. 
Figure 2 describes the basic principles underlying the time slice. 
Figure 2.    Time Slice 
 
The configuration of Air Time Fairness 
from the network administrator’s perspective can be easily explained. 
Note:   The configuration procedures which follow are included as examples only. Refer to the configuration guide 
for supported configuration procedures. 
The first step in creating a time slice is to select a policy ID and policy name, and assign a numerical weight to the 
policy. Assigning a numerical weight to a policy corresponds to allocating a percentage of airtime to the policy. The 
greater the numerical weight, the greater the percentage of allocated airtime. Equal numerical weights achieve 
equal allocations of airtime. But the numerical weight is not literally equal to the actual percentage allocated.