From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: From: Sven Eckelmann Date: Mon, 05 Mar 2018 20:28:13 +0100 Message-ID: <2010709.fvC8HbtXRd@sven-edge> In-Reply-To: References: MIME-Version: 1.0 Content-Type: multipart/signed; boundary="nextPart4202079.LQ3ES6GMtR"; micalg="pgp-sha512"; protocol="application/pgp-signature" Subject: Re: [B.A.T.M.A.N.] Questions about hardware List-Id: The list for a Better Approach To Mobile Ad-hoc Networking List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paul Becue Cc: dries.berckmans@soundtalks.com, b.a.t.m.a.n@lists.open-mesh.org --nextPart4202079.LQ3ES6GMtR Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="iso-8859-1" [Please try to not contact me directly when it is about B.A.T.M.A.N. relate= d=20 things and at least Cc the official mailing list] On Montag, 5. M=E4rz 2018 19:04:29 CET Paul Becue wrote: > We are working on a device with a mesh network. We are doing tests with=20 > WiFi ad-hoc networks, and we tested last week BATMAN-adv and also babeld= =20 > on a Raspberry Pi 3 with a Ralink dongle attached to it. The results=20 > were very good. [..] > Could you give me some hints where I have to look for hardware and what=20 > kind of hardware I have to look for? I have the impression that there is= =20 > something that I don't see in the whole story. I hope ad-hoc WiFi is=20 > still a technology that is up to date. Or is there a technology that=20 > replaces it? Yes, IBSS is becoming a problem at the moment. You can try QCA ath10k hardw= are=20 with Ben Greears firmware [1] or Mediathek with the mt76 driver [2] when yo= u=20 want up-to-date HW. There is still some ath9k based stuff around when 802.1= 1n=20 is enough. Otherwise you might try to replace IBSS with meshpoint interfaces. Just set= up=20 it like 802.11s but with forwarding disabled [3]. This works at least with= =20 ath10k based HW and the official QCA ath10k firmware. But HW which uses ath10k and other things which uses special firmware which= =20 prevents you from using your "own" rate control algorith are a no-go when=20 you want to use throughput based metrics (as in B.A.T.M.A.N. V or olsrd2).= =20 Then you should stick either with ath9k or mt76 compatible HW. Btw. my tests of mt76 showed performance and stability problems in the past= =2E=20 This may have changed by recent fixes in the driver but I don't know the current state. Kind regards, Sven [1] https://www.candelatech.com/ath10k.php [2] https://github.com/openwrt/mt76 [3] http://openwrt-devel.openwrt.narkive.com/Bieu1aRR/ath10k-mesh-with-open= wrt-question --nextPart4202079.LQ3ES6GMtR Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part. Content-Transfer-Encoding: 7Bit -----BEGIN PGP SIGNATURE----- iQIzBAABCgAdFiEEF10rh2Elc9zjMuACXYcKB8Eme0YFAlqdmk0ACgkQXYcKB8Em e0bmbRAAiIph3bIc0rs3emC6O+9ZvEjBwg+j5R+Dv9qZpQc6dCPMAo4eaUWtTpJz R26VZazvPkaXEs455VNb3KaRXkb0VyfezPMegH8rpUjS4MMESbsRkAadDeLxTQU4 f/T5kFwj9WyPPquNGqOUY9ejAp0S6o9CNnDC9YXF8ai5iSbOVHaJrQFTsnuIu+p7 qsJGWywqr8n0DFYbIv3Cw86kLQvEFJbspHTjIxLgpnMgpsQYuJgFjBAwRGYpIeUu GWwc2QsT9rhvlVLs0d+2IuvoVp3ziSCUfsKdI0b7oTZlSGUDbiNPfHUe4izia+BC WxBPUiOcNO379MdnsKghptJs9gzGXabErIpVjAnDgBN8KrdnenhqaHJwsRlXrg0A RIw5QLU6XxzYL0oDR8Kh3qVO0H3DysXGBNnjurouY9mYnOmxaGIGLb9jLdmik80I v/uG8BEv1/5Tqn2KPfk+AlulaL/xGYKT3ADowRVG0zdwjNXV+fKyW5p3N+vTh9Bx Ox2D3F6Zn4b1/gNxgp5F9k1jEnitlJIF1bk3LAhK84FQlxOEfcM/Arowy2hEOsoq 6dGYlOPDQAMe5fhcWhtGi6WOPZLV8hYKT+CKzxKKmG7O8dNDg5bLYrMAiYiXU013 b6p3YM3cvbDB42FGYHJu5tXA/FdTG4gaOw1Rtqa/EX7PSnnxNfI= =Ib7O -----END PGP SIGNATURE----- --nextPart4202079.LQ3ES6GMtR--