Dear all,
We're experimenting with Raspberry Pis and hostapd in order to quantify co-channel capacity and interference but we're encountering channel access issues.
The idea is to have multiple APs and clients on the same channel and verify that channel allocation is pretty much evenly split between all APs by creating traffic using iperf. Following are the 2 setups we've been using.
1 - Qualifying system, using tried and tested hardware, with validated co-channel access & allocation :
+-----------------+ +------------------+ +------------------+
| Debian PC | | Sagemcom | | Cell Phone |
| iPerf client | Ethernet | 802.11n 2.4Ghz | WLan | 802.11n 1x1 |
| UDP +-------------->+| Channel 6 +--------------->+ iPerf Server |
| Target 80M | | No QoS Policies | | UDP |
| | | | | |
+-----------------+ +------------------+ +------------------+
+-----------------+ +------------------+ +------------------+
| Debian PC | | Sagemcom | | Cell Phone |
| iPerf client | Ethernet | 802.11n 2.4GHz | WLan | 802.11n 1x1 |
| UDP +-------------->+| Channel 6 +--------------->+ iPerf Server |
| Target 80M | | No QoS Policies | | UDP |
| | | | | |
+-----------------+ +------------------+ +------------------+
Results :
- Using a single client & AP : 57 Mbps avg on both APs.
- Using 2 APs with a single client each :
- Total : 54 Mbps avg.
- Client 1 : 27.1Mbps avg.
- Client 2 : 26.9Mbps avg.
- Channel almost evenly split.
2 - Test setup, using mixed hardware :
+-----------------+ +------------------+ +------------------+
| Debian PC | | Sagemcom | | Cell Phone |
| iPerf client | Ethernet | 802.11n 2.4Ghz | WLan | 802.11n 1x1 |
| UDP +-------------->+| Channel 6 +--------------->+ iPerf Server |
| Target 80M | | No QoS Policies | | UDP |
| | | | | |
+-----------------+ +------------------+ +------------------+
+-----------------+ +------------------+ +------------------+
| Debian PC | | Raspi + hostapd | | Cell Phone |
| iPerf client | Ethernet | 802.11n 2.4GHz | WLan | 802.11n 1x1 |
| UDP +-------------->+| Channel 6 +--------------->+ iPerf Server |
| Target 80M | | No QoS Policies | | UDP |
| | | Linksys AE3000 | | |
+-----------------+ +------------------+ +------------------+
Results :
- Single client + Sagemcom : 56.1 Mbps avg.
- Single client + Raspi/hostapd : 33.5Mbps avg (expected considering the dongle used).
- Using 2 APs (1 Sagemcom & 1 Raspi/hostapd) with a single client each :
- Total : Inconsistent, between 27 and 33 Mbps.
- Client 1 (Sagemcom) : ~ 5 Mbps avg.
- Client 2 (Raspi/hostapd) : btw. 22 and 28 Mbps.
Seeing the results above made us believe the QoS settings were off in our hostapd.conf configuration file, especially since the results of channel sharing were an almost perfect 50/50 split with the 2 approved Sagemcom APs.
After setting up the wmm settings to the most relaxed settings in hostapd.conf (dubbed as "background" in the example file on w1.fi), we retried with the exact same results, as the problem really was the sharing of the radio channel.
Is there anything we might have missed ? Or is hostapd not backing off as much as it is supposed to do and take up all the radio channel ?
Thanks,
Maxime Renusson
Cellular Networks R&D Engineer
Orange Labs Networks / Wireless Engineering & Propagation
maxime.renusson@xxxxxxxxxx
_________________________________________________________________________________________________________________________
Ce message et ses pieces jointes peuvent contenir des informations confidentielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu ce message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages electroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deforme ou falsifie. Merci.
This message and its attachments may contain confidential or privileged information that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and delete this message and its attachments.
As emails may be altered, Orange is not liable for messages that have been modified, changed or falsified.
Thank you.
_______________________________________________
Hostap mailing list
Hostap@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/hostap
