We already use Wi-Fi more than cellular; Why not continue the trend? | Hacker News Reader

We already use Wi-Fi more than cellular; Why not continue the trend?

gigaom.com

35 pointsmtgx49 comments

Lol, carriers will never EVER willingly became "simple internet providers". EVER. They will scream and fight to the last dollar for their "right" to provide monopolized "services" - separate voice traffic, separate plaintext traffic, separate image traffic, and other "services". They'll cut and limit all generic internet connections, they'll filter unwanted packets (VoIP, iMessage etc.), they'll reroute traffic through their internal resources and inject ADs or spyware. They'll do lots of thing, except of cource becoming ISP.

But people might be able to create a wifi alternative without them. E.g. projects that blanket a city with wifi or that share people's routers with others in a secure manner. Then the carriers will be cut out of the picture.

You should check some of the MVNO providers in Germany.

These mobile operators compete on price and anything that smells like custom development of a blah-portal-blah is cut.

There was a time when ISPs used to think the could do value-add crap. Orange-dialup had a news service. BT had a portal and email. Like the ISPs, our carriers will become the fat pipe they fear. Just slower than we'd want.

The reason that wifi is better than cellular is because you don't have to fight with a carrier to use it. if AT&T starts pushing everything onto wifi, it won't make AT&T better, it'll just make wifi worse.

I almost never use wi-fi on my phone. It drains the battery too much.

edit: I'm not familiar with cellular hardware tech. My observation is true for my phone, necessarily, because my 3G reciever is active 100% of the time, and turning on wifi is just turning on another component. If this isn't the case for you, do you turn off your 3/4G while you're in a wifi spot? Does it manager this all automatically?

Do you have your wifi on while you're outside of a wifi spot? If so, how does that not cause a loss in battery life over never having it on?

Unfortunately I almost never use cellular data on my phone. It drains my bank account too much.

It seems to me to actually be the opposite. Having a working wifi connection will increase my battery-life by at least a factor of 2

That seems unlikely. WiFi should certainly drain your battery less than cellular. (I have yet to jump on the smartphone train but my iPad gets consistent 7 hours with 3G, 10 hours with WiFi.)

WiFi also kills my phone's battery.

It appears that for newer devices, wifi uses a lot less power.

There is, however, a cost in searching for WiFi networks every time you want to make a connection.

I imagine it becomes most apparent if your device is polling for email. On every interval it needs to wake up, turn on WiFi, look for any available networks, discover there are none and then finally revert to the cellular network. With WiFi disabled, it can go straight to cell.

The iPhones were famously bad for this in the early days, though they seem to improve later on; either through a hardware or software update, I'm not sure which.

I wish the new 802.11 ac standard would've gone the "battery efficiency" route, rather than doubling the performance. Technically, they could choose either doubling the performance or cutting the power consumption in half, but I'm pretty sure they will go the performance route, which really isn't necessary on a mobile phone right now (500+ Mbps).

802.11ac improves range and performance, but the new Bluetooth specs improve range and battery power.

If you mean Bluetooth 4.0, its bandwidth is like 50kbps IIRC. It can't be used for anything other than heartbeat monitors and stuff like that.

I know absolutely nothing about Bluetooth, but I watched a WWDC 2012 video about it just for fun. It was an extremely well-done introduction to this technology.

Session 703, Core Bluetooth 101 - https://developer.apple.com/videos/wwdc/2012/ (requires a free developer account)

Bluetooth runs at ~2 Mbps which is not too shabby. And there's a new (as of BT 3.0) "high-speed" mode which actually sends bluetooth traffic over 802.11, so you can keep your wifi radio off until you actually need to send a lot of data.

edit: I haven't watched your video yet, but thanks for the link :)

Sorry, I thought you were talking about Bluetooth 4.0, which is very low-power and very slow.

These two slides are from the aforementioned talk. Note that 'Classic' means 3.0 and 'Low power' means 4.0:

http://d.pr/i/cRvm

http://d.pr/i/qCHB

Well 4.0 includes "classic" and the new low-power mode. But I see on Wikipedia that BT has a 3 Mbps mode (aside from the 802.11 "high-speed" mode), do you know what that would refer to?

The cost of charging an iPhone 3G for AN ENTIRE YEAR is about 25 cents.

(Source: Electric Power Research Institute http://bit.ly/NV287P).

Granted, they charged it every other day but with even the heaviest Wi-Fi use, you couldn't more than quadruple it, which still leaves you at an inconsequential $2 a year.

My problem isn't the cost, its that I have a battery so I can be mobile--which means I can't recharge it just because the battery gets low.

I've been using Republic wireless (hyrbrid wifi phone) for about 6 months now. 99% of the time I forget it's anything other than a regular cell phone.

For me it works better than a regular cell phone, because it works in my basement workshop where I can't get reception from most carriers.

How does it work in the car?

My understanding (I signed up for their beta but opted not to sign a contract) is that their service switches over to cellular network when wifi isn't available. I seem to remember them saying that they could maintain a telephone conversation through that switch, but I've not used their service so have no real-world data. In a car, I'd assume you'll be using a cellular network (Sprint if I remember correctly).

They use some heuristic to determine when you're using the cellular network "too much", because their whole pitch is that traffic over a cellular network is more expensive than traffic over the existing ISPs. If you use their cellular network too much, they will terminate your contract - but again, I'm not a customer so I don't know where that line lives.

I'm pretty sure they switched their stance on "too much" termination.

Also, I believe while on a call you can transfer off of wifi, but not on to it.

"I'm pretty sure they switched their stance on "too much" termination."

Here's the most recent statement on that matter, I think: http://republicwireless.com/blog/unlimited

(Also, I think they made the wrong call, and they should have "stepped back from unlimited". It won't be unlimited, they shouldn't call it unlimited. But I'm not planning on abusing it either way.)

I'm sitting in their Wave G, patiently waiting. I refuse to pay for both a nice home connection and a cell data connection, so I still have a dumbphone. Buying a smart phone that primarily uses local Wifi (which I am in about 90% of the time) but can use cellular networks as a backup when there's no Wifi, and thus only charges for that "backup" contingency, is something I'm willing to pay for. I'll recover the cost of purchasing the phone in something like 6 months, even ignoring the fact it'll be a much better phone in the meantime.

> “For example a cellular picocell costs from $7,500 to $15,000 whereas a much higher capacity carrier-grade Wi-Fi access point costs around $2,000,” Thanki wrote. “The cost of a Wi-Fi chipset for a consumer device is around $5, whereas 3G cellular chipsets costs around $30.

Won't a cell tower cover at least a couple orders of magnitude more area than a wifi access point? Is this analysis silly or am I missing something?

Well he said a picocell which is around an order of magnitude greater range. I am not sure that range is the most important thing as towers become more and more congested I am pretty sure we are moving to smaller and smaller ranges.

>Well he said a picocell which is around an order of magnitude greater range [than a WiFi access point].

Ahh, yea, this is the key point. For those who aren't in the know: "A picocell is a small cellular basestation typically covering a small area, such as in-building (offices, shopping malls, train stations, stock exchanges, etc.), or more recently in-aircraft. In cellular networks, picocells are typically used to extend coverage to indoor areas where outdoor signals do not reach well, or to add network capacity in areas with very dense phone usage, such as train stations. Picocells provide coverage and capacity in areas difficult or expensive to reach using the more traditional Macrocell approach." [wikipedia]

> I am not sure that range is the most important thing as towers become more and more congested I am pretty sure we are moving to smaller and smaller ranges.

Not where I live in New Mexico :)

I think the argument is that a given amount of spectrum has a maximum capacity of data that can be transmitted. Using one cell tower to serve a large area (which means a large number of devices and a large amount of data) caps the amount of data that can be transmitted within that area. There are two solutions to that: use more spectrum or serve smaller areas. If you're serving smaller areas, there are cheaper technologies available than 3G, and they also don't require dedicated spectrum.

I see. So this argument is really only directed at urban areas where there is already total coverage but the bandwidth per user is limited because of congestion.

Yes, plus the cell network has dedicated bandwidth for each client whereas clients will fight for bandwidth on a wifi access point.

By that argument, you could also just build one extremely powerful tower and cover a vast region. But that would support far less data, because everyone in the region would be waiting for a turn to broadcast at the same frequency.

This is why cellular phones are called cellular phones. The key enabling technology was the ability to divide a large geographic area into many smaller cells. Each cell needs to use different frequencies than its neighboring cells, but you can reuse the same frequencies many times across the wider region.

The smaller you make the cells, the more data you can pack into the same set of frequencies, because non-neighboring cells can be using the same frequency concurrently.

Also, each tower has many directional antennas, so one tower acts as multiple cells.

With analog cellular, yes. With digital cellular, several connections can share a given frequency, somewhat expanding the number of connections per tower, although there is still a limit. However, I find that AT&T ignores the pressing bandwidth needs of the customers. During rush hour, no matter how perfect my signal strength, I get terrible data bandwidth. Its not the cell tower, its the connection from there onward.

Plus, with a single huge tower, your handsets are fighting the inverse square law as they struggle to pump out enough wattage to keep a connection.

Because Wi-Fi kinda sucks at roaming?

It's fine if you need a network connection at arbitrary end-points. But IME it's not so fine when you need a sustained connection while moving. (excluding on-train/on-plane wifi and the like , natch)

Agreed, it was very noticeable when listening to online radio in my University campus, since you can get the connection cut off for 20-30 seconds. For this particular use case it's fine as long as you adjust the buffer size, but for VoIP it wouldn't work.

You're right except that by 'kinda sucks' you actually mean that it doesn't support base station hand off at all. You can kind of simulate one ubiquitous wifi network by using the same SSID everywhere. Then, when a user moves out of range of one access point, they'll attempt to reconnect and pick up the new one that's in range.

Wifi and cellular radio are massively different technologies, and wifi does not fit the use case that we currently use cellular for from a technical point of view.

This article is misleading on several accounts. First, it completely ignores the incredible difference in range you can get with 4g vs. Wifi. This is directly related to the frequencies you are operating at.

Secondly, the reason that WiFi frequencies are not regulated the same was as cellular frequencies has to do directly with that limitation. If we suddenly had AT&T, Verizon, and other WiFi routers littering our streets every sixty feet, there would be a terrible amount of noise and it would interfere with people's home or business networks.

Suddenly, you would need to start licensing that spectrum. See how it works?

If you had a wifi-capable Blackberry and T-Mobile in the US, you could use what's called UMA calling- which offers normal uninterrupted voice service over wifi, with seamless handoff mid-call between Wifi hotspots and the normal cellular network (i.e. no dropped calls). The best part about it is the wifi calls don't count against your monthly minutes. It was an awesome service but never really caught on.

It's also available on Android. Unfortunately I've flashed mine to have a generic ROM, meaning that I've lost the ability. Hoping to see it ported back in.

Groove IP lets you make wifi (or dataplan) calls from your Android through Google Voice.

Is this really what we want?

Let's say AT&T decides to go the author's suggested route, and let's also pretend that I'm a paying AT&T customer. I travel to an area where AT&T is delivering their data through WiFi, but several shop owners in the same area have set up WiFi access points and have overrun the available WiFi spectrum in the area. The quality of my data connection suffers significantly, and I get upset at AT&T for providing poor service.

Does anyone really think this is in AT&T's best interest?

By using licensed spectrum, wireless providers gain tighter control over the engineering of their networks. Yes, it's a double-edged sword, but I don't see an opportunity for improvment by moving to unlicensed spectrum.

Carriers realize the value in WiFi - it's present everywhere, cheap to deploy at scale and already exists on most mobile devices. There's precedent for WiFi offload - particularly in dense urban areas - and the industry is moving to support that model even further with Hotspot 2.0 and Passpoint certification.

That said, WiFi's not a panacea. Technologically it's subject to interference, the protocols allow for NO control over the handset (at present), and its range is limited. Speaking from experience running several large muni WiFi networks, the protocol does not handle power asymmetry well - your phone's WiFi needs to be roughly as strong as the AP or performance is terrible. Compare this to licensed cellular technologies which can pull in a tiny weak handset signal from miles away.

From a business perspective, there's a breakpoint between licensed and unlicensed use - at the point that WiFi's shortcomings have less impact than the cost of deploying licensed spectrum, it gets used. There's plenty of talk about small cells in the industry and I think we'll see a hybrid deployment style emerge over the next few years, with WiFi handling the densest of deployments. Carriers want to offer plentiful data but they want to do it in a way they can control, for good reasons - people are paying them and expecting a service in exchange.

From a good-of-all-mankind perspective, what I'd like to see happen is a transition to all-IP on the handset, seamlessly decoupling every feature from the cellular network. Apple showed the potential here with iMessage, which works well if everyone's on an iPhone. SMS and voice themselves need to be carried over IP in a device-independent manner. When I can send an SMS from my Android phone to my iPhone over IP, and call a POTS number over IP, I'm truly carrier-independent and effectively paying them to be a WISP. That gives me (and the market) opportunities to build a service tailored to my needs.

The tech is not here yet:

Passpoint/Hotspot 2.0 is the first step of many that will eventually integrate Wi-Fi hotspots seamlessly into the carrier’s mobile networks. Networks will be able to pass subscribers from cell site to hotspot and back again as Wi-Fi gets incorporated into the increasingly diverse array of small cells that will make up future heterogeneous networks, or HetNets. That’s still a long time coming though.

Hotspot 2.0 only takes care of the discovery and authentication of carrier-owned or managed hotspots. If you’re an AT&T or an SK Telecom with tens of thousands of hotspots under your control, then Passpoint will be a boon, as certified devices will automatically connect to any carrier access point in range.

http://gigaom.com/broadband/wi-fi-alliance-begins-certifying...

When this happens, except "your" FIOS router to start serving data to Verizon wireless users.

Early in the CDMA days, there was talk of mounting small cellular antennas on utility poles to speed up rollouts, especially in suburban areas.

The problem is, managing tons of little leases is a nightmare, and the people who own telephone poles and streetlights (allowing you to have a few big leases) want an obscene amount of money to hang stuff on their pole.

As the article mentions, the carriers are focused on cellular because that's what they control and can get people to pay $30-80 / month to use their cellular data plans. The carriers' worst nightmare would be if more cities provided quality wifi everywhere. Few would pay for data plans anymore, and maybe not even voice plans either.

I'd love to see some statistics from Apple regarding Wi-Fi only iPad sales versus 3G/4G iPad sales.

Those may be skewed though by the amount of people that simply choose tethering over getting a 3g enabled tablet/device.

I'm one of those people as I have several devices (kindles and tablets) so I prefer paying for one data plan on my phone and tethering vs. individual data plans for each.

agreed

Us centric nonsense.

Sent via mobile internet on a bus.

Reply on news.ycombinator.com