So the Cellular industry is having seminars and investing big on Wi-Fi offload and some schools are considering LTE offload, what an irony.
At the end of the day the cost of providing 1 byte over LTE is much higher than the cost 1 byte over Wi-Fi. (DAS, Microcell, MacroTower, all more expensive than Wi-Fi…and much more complicated as far as contracting is concerned) Before doing anything I would first analyze the average monthly bandwidth need of a student and then do a comparison between Wi-Fi cost over 5-8 years VS LTE cost (with comparable quality of service). And BTW, why do we all need to upgrade to 802.11ac? (802.11n seems perfectly fine to me) Because premature EOL is coming upon us? On another note, having a low capacity/expensive Data Wireless in residence halls might have interesting side effects: -Students will watch their shows in class rather at night -Students will start reading books again in their rooms -Students will hang around campus late at night just to have Wi-Fi access (do you plan to shutdown campus Wi-Fi after hours ? ;-) -Students will meet in hallways and talk to each other rather using Social Media -No need to filter peer-to-peer in residence halls, no one can afford to download anything But here my main concern…how will you enable eduroam on Cellular? ;-) Philippe Philippe Hanset www.eduroam.us > On May 13, 2015, at 9:24 AM, Brian Helman <bhel...@salemstate.edu> wrote: > > I have a little more information to provide now. I absolutely appreciate > that it will be extremely tempting to respond with biased opinions. I don’t > think there is anything that can be said that I haven’t already expressed to > my team. However, that will not help me write up my recommendation. So that > being said, feel free to chime in with tangible reasons to do this or not… > > Apparently, our president heard that some schools are investigating > purchasing bulk data contracts with mobile (“cellular”) carriers for data. > The idea is, we would stop providing 802.11g/n/ac wireless in the residence > halls and instead provide students with the abilities to register their > devices with the mobile carrier to use 4G/LTE data. The University will pay > for this. > > Pros: > No wireless (802.11) to purchase, support > Reduced POE requirements on switches > No wireless driver/configuration mismatches problems to support > > Cons: > Is mobile wireless signal available everywhere inside the buildings? Costs > to improve signal. > What speeds are available (what range of speeds)? Is it by user or aggregate? > How is congestion handled? > What devices – mobile phones only? Hotspots to provide access to > non-cellular devices (e.g wifi-only tablets; laptops) > More Ethernet ports needed for devices that previously depended on wireless > What provider(s)? > Support shifted from “device to institutional wifi” to “device to myfi” or > “devide to 3rd party” > Cost per user, per GB? > > What else? > > If you know of any institutions who have attempted this (I have heard MIT is > looking at it, but we aren’t MIT), please let me know. > > By the way, the background here is .. we installed our 802.11n network ~5 > years ago and haven’t had any commitment to fund it since. So now we are > trying to deal with capacity (BYOD) issues that didn’t exist 5 years ago > while upgrading to 11ac. Of course, it’s not a 1:1 swap of equipment since > we’d be migrating from 2.4GHz to 2.4+5GHz. That puts the costs for forklift > upgrades pretty high (did I mention I’ve been unsuccessfully asking for > funding for 3 years?). > > I believe this can all best be summarized with a simple .. Oy. > > -Brian > > > > > > From: The EDUCAUSE Wireless Issues Constituent Group Listserv > [mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU > <mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU>]On Behalf Of Jerkan, Kristijan > Sent: Sunday, May 03, 2015 12:34 PM > To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU > <mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU> > Subject: [WIRELESS-LAN] AW: [WIRELESS-LAN] To provide (wireless) service, or > not to provide (wireless) service... > > As a public institution in the EDU sector we always had a byod policy in our > dorm network, specifically including „anything You want to connect to the > port in Your room“. > > Parameters: > -5k+ dorm rooms (1.8k the largest segment, 20 the smallest) > -120km radius > -at least one (mostly two) RJ45 port per room (cat5-7 to the switch, fiber > afterwards) > -10/100MBit ports (deliberatly did not go for 1GBit at the edge) > -no additional accounting, just dhcp with opt82 > -public ips behind reflexive acl (no shaping, etc.) > -uplink via the federal research network > -service neutral (whoever wants to can use a DSL provider also/instead and > may use the inhouse cable from their basement to their room for it) > -one service number (fixed number, forwarded to five cellphones – whoever > picks up first wins) > -managed by ~10 students (pro bono, but with a couple of incentives) > > That beeing said, here are a few points why this works for us and is not > generally applicable: > -people have to work together to archive common goals (state, local, > university and dorm administration – technical and administrative staff) > -it does not take much to put a service neutral CAT cable into every room > while they are beeing built/renovated instead of a cheaper telephone cable, > but it does take a joint effort and common goals > -to every dorm room there is a rent/contract, so we know who is „behind“ it > and can make one specific person liable (opt82) > -there are only single-bed rooms (this is a cultural thing and different than > in the US, I guess noone around here would even rent a shared room) > -almost no dorms are adjacent to the classrooms/labs (seamless wireless > coverage/services wouldn’t be possible anyway) > -in order to find enough students (5 for the core team) who will do the > occasionally needed actual work without payment, a balance between demands > and incentives is important > > Effect: > -very low capex and extremly low opex for the dorm network [numbers only off > list] > -very limited support calls (maybe 2/week; maybe 10-20 during the > move-in-phase, mostly students from the states asking about the non-existant > login/pw) > -no need to worry about deprication charges or every new feature (regarding > wireless: ABG to N to AC; MIMO, fequency analysis chipsets; 2.4ghz to 5ghz, > wave2) > -the least administrative overhead possible > -none of the students in our networking team had problems finding jobs after > they left (no trouble finding volunteers, very long participation period) > -scalabe system; got us from ~1.2k rooms (back in ’99) within a 1km radius to > 5k+ (today) in a 120km radius > -effective support answers („Yes, You can also attach every AP You want to > You port… No, don’t worry, if You are able to understand Your class reading, > You will also understand vendor X’s manual…) > -no secondary discussions (health, etc.) > -plug&play experience for students > -ability to consolidate our attention to more interesting projects; we still > provide wireless (eduroam), but only in common areas away from the rooms > (ALU/Aruba 6000, now 7210, anything between 124s and 270s except the cloud > based APs) > -over the years we had some (small and larger) dorms outsourced to different > (small and large) companies who provided full wireless-only coverage, > standard management as well as forbidden private wireless, but as our own > model proved technically resiliant and cost-effective time and again, our > external partners solutions didn‘t > > Basically our setup could be exactly what Your administrative staff/board is > aiming for. > My personal message to them would be to first and foremost take an honest > look at how and why things are the way they are. > If they just argue out of a mix of intuition and auserity, their good > intentions will cause a fail (probably utterly and completely, like many > others before). > It is possible to run a cost effective plug&play network, with a high > satisfaction rate amoung students (EDU did that long before the BYOD > marketing hype). But that requires a high level of cooperation (belivers, > ideally who themselves lived in dorms and remember how student life can be), > common goals, success in overcoming obstacles and also constant vigilance and > re-evaluation. > From an administrative and oversight point of view this is a lot more and > complex work than finding, distributing and approving funds. For various > reasons it is also not always something that can be implemented everywhere or > sustained for a meaningful period of time. Therefore it is often better to > honestly deal with the geographic/personal/political reality and to solve the > technical problem with money. > Even if Your board would want to, a change towards a system like ours takes > time. Your institution should definetly not run on an obsolete wireless > infrastructure during that periode (and wear out its staff and cause stir > among students in the process). > Hope this helps to balance the biased view. ;-) > > > Regards, > Kris > > > > > Von: The EDUCAUSE Wireless Issues Constituent Group Listserv > [mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU > <mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU>] Im Auftrag von Brian Helman > Gesendet: Freitag, 1. Mai 2015 17:23 > An: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU > <mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU> > Betreff: [WIRELESS-LAN] To provide (wireless) service, or not to provide > (wireless) service... > > A few weeks ago we made a pitch for funding to upgrade our res halls to > 802.11ac. This request for funding has had an unforeseen effect. I’m not > being asked to investigate NOT providing wireless networking in our res > halls. Here are the options, as it has been described to me: > > -No institutional wireless. Let the students bring in their own AP’s > -Some kind of managed service (wireless as a service) with 802.11 > -Some kind of institutionally owned/leased mobile wireless (e.g we provide > our own 4G) > -Hybrid > -Continue with 802.11n 2.4GHz and fill in holes as they pop up > > I’m not going to put my thoughts up here just yet. These are the > options/thoughts as presented by the levels above me. > > Let the discussion begin…. > > > > > ____________________________________ > Brian Helman, M.Ed | Director, ITS/Networking Services | (: 978.542.7272 > Salem State University, 352 Lafayette St., Salem Massachusetts 01970 > GPS: 42.502129, -70.894779 > > ********** Participation and subscription information for this EDUCAUSE > Constituent Group discussion list can be found at > http://www.educause.edu/groups/ <http://www.educause.edu/groups/>. ********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.
