Editing Wi-Fi (section)

=== Throughput ===
Various layer-2 variants of IEEE 802.11 have different characteristics. Across all flavours of 802.11, maximum achievable throughputs are either given based on measurements under ideal conditions or in the layer-2 data rates. This, however, does not apply to typical deployments in which data are transferred between two endpoints of which at least one is typically connected to a wired infrastructure, and the other is connected to an infrastructure via a wireless link.

This means that typically data frames pass an 802.11 (WLAN) medium and are being converted to 802.3 (Ethernet) or vice versa.

Due to the difference in the frame (header) lengths of these two media, the packet size of an application determines the speed of the data transfer. This means that an application that uses small packets (e.g. VoIP) creates a data flow with high overhead traffic (low [[goodput]]).

Other factors that contribute to the overall application data rate are the speed with which the application transmits the packets (i.e. the data rate) and the energy with which the wireless signal is received. The latter is determined by distance and by the configured output power of the communicating devices.<ref>{{cite conference |title=Towards Energy-Awareness in Managing Wireless LAN Applications |url=https://www.researchgate.net/publication/241631429 |publisher=IEEE/IFIP NOMS |date=2012 |conference=IEEE/IFIP Network Operations and Management Symposium |access-date=11 August 2014 |archive-date=13 August 2014 |archive-url=https://web.archive.org/web/20140813094612/http://www.researchgate.net/publication/241631429_Towards_energy-awareness_in_managing_wireless_LAN_applications?ev=prf_pub |url-status=live }}</ref><ref>{{cite web |title=Application Level Energy and Performance Measurements in a Wireless LAN |url=https://www.researchgate.net/publication/224264522 |publisher=The 2011 IEEE/ACM International Conference on Green Computing and Communications |access-date=11 August 2014 |archive-date=13 August 2014 |archive-url=https://web.archive.org/web/20140813113706/http://www.researchgate.net/publication/224264522_Application_Level_Energy_and_Performance_Measurements_in_a_Wireless_LAN?ev=prf_pub |url-status=live }}</ref>

The same references apply to the attached throughput graphs, which show measurements of [[User Datagram Protocol|UDP]] throughput measurements. Each represents an average throughput of 25 measurements (the error bars are there, but barely visible due to the small variation), is with specific packet size (small or large), and with a specific data rate (10&nbsp;kbit/s – 100&nbsp;Mbit/s). Markers for traffic profiles of common applications are included as well. This text and measurements do not cover packet errors but information about this can be found at the above references. The table below shows the maximum achievable (application-specific) UDP throughput in the same scenarios (same references again) with various WLAN (802.11) flavours. The measurement hosts have been 25 metres (yards) apart from each other; loss is again ignored.

{| style="margin: 0 auto;"
| [[File:Throughputenvelope80211g.png|thumb|upright=1.5|Graphical representation of Wi‑Fi application-specific performance envelope of 802.11g, which uses the {{nowrap|2.4 GHz}} band]]
| [[File:ThroughputEnvelope11n.png|thumb|upright=1.5|Graphical representation of Wi‑Fi application-specific performance envelope of 802.11n using a 40&nbsp;MHz channel width in the {{nowrap|2.4 GHz}} band]]
|}